Occupancy sensor and method of operating same

BACKGROUND OF THE INVENTION
The present invention relates to occupancy sensors.
A portion of the disclosure of this patent document contains material that is subject to copyright protection. The copyright owner has no objection to the facsimile reproduction by anyone of the patent document or patent disclosure, as it appears in the Patent and Trademark Office patent file or records, but otherwise reserves all copyright rights whatsoever.
An occupancy sensor is an energy conservation device designed to detect the presence of human occupant(s) in a given area. When occupancy is sensed, the various electrically-powered loads in that area controlled by the sensor (for example, lights, ventilation, and the like) are energized. When that same area has been unoccupied for a predetermined period of time, the sensor de-energizes the electrical loads that it controls. Occupancy sensors may therefore conserve a great deal of energy in areas where the occupants do not exercise diligence in de-energizing those electrical loads when they leave the area.
Over the last few decades, several events have led to the growth of a large consumer market for energy saving devices including occupancy sensors. First, there has been an increase in public awareness of energy conservation and its beneficial environmental consequences. In addition, there has been increased realization by both private and government-controlled power generation industries of the economic and environmental advantages of energy conservation as a means of meeting peak load power demands. Finally, there have been legislative mandates at the federal, state and local levels for the use of energy conserving devices, such as occupancy sensors, in government and other public buildings.
Occupancy sensors have been successfully designed and tested using a variety of technologies. A brief description of the most widely used occupancy sensor technologies along with the strengths and weaknesses of those technologies follows:
Active Ultrasonic Acoustic Doppler Occupancy Detection.
This technology allows continuous detection of moving objects that reflect ultrasonic acoustic energy. This method of occupancy detection is highly sensitive since it is based on an active source of ultrasonic acoustic energy. An apparatus and method of this type are disclosed in U.S. patent application Ser. No. 08/384,580, filed Feb. 6, 1995, now U.S. Pat. No. 5,640,14, assigned to the same assignee as the present invention and entitled: OCCUPANCY SENSOR AND METHOD FOR OPERATING SAME, the disclosure of which is incorporated herein by reference.
However, this method of occupancy detection has several limitations: first, it is insensitive to motion that is orthogonal to the direction toward the receiver; second it is insensitive to motion generally not in the line of sight of the receiver; third, it is subject to false tripping due to other sources of ultrasonic energy; fourth, it is subject to false tripping due to heating and air conditioning air flow; and finally, it has no means of range discrimination. Since occupancy sensors based on Doppler techniques have no means of range discrimination, a large-distant target moving at approximately the same speed as a smaller, nearby target might have similar target signatures.
Active Electromagnetic Doppler Occupancy Detection
This technology allows continuous detection of moving objects that reflect electromagnetic energy. This method of occupancy detection is highly sensitive since it is based on an active source of electromagnetic energy. However, this method of motion detection also has several limitations: first, it is insensitive to motion that is orthogonal to the direction toward the receiver; second, it is insensitive to motion generally not in the line of sight of the receiver; third, it is subject to false tripping due to other sources of electromagnetic energy; and finally, it has no means of range discrimination.
Passive Audio Acoustic Occupancy Detection
This technology allows continuous detection of objects that emit audio acoustic energy. This method of occupancy detection is quite sensitive but is subject to false tripping due to non-occupant sources of audio acoustic energy such as facsimile machine, telephone and security system tones, automobile and emergency vehicle horns, etc.
Passive Infrared Occupancy Detection
This technology allows continuous detection of moving objects that emit infrared energy. This method of occupancy detection is also quite sensitive even though it is based on passive sensing of moving sources of infrared energy. This method of occupancy detection also has several limitations: first, it is insensitive to sources generally not in the line of sight of the receiver; second, it is subject to being blinded by intense, stationary sources of infrared energy; third, it is subject to false tripping due to rapid fluctuations in the intensity of stationary infrared sources; and finally, it is subject to a detection coverage tradeoff involving the number of lens facets versus detection range.
Position Sensor Based Occupancy Detection
This technology uses one or more mercury switches to sense changes in the physical position of the sensor. This technology has several limitations: first, it is insensitive to minor motion that may be indicative of occupancy; and second, it is inherently a digital (off/on) device.
Piezoelectric Sensor Based Occupancy Detection
This technology senses the changes in the resistance of a piezoelectric sensor to sense occupancy. This technology is subject to false tripping due to changes in temperature.
Significant innovation in the design of occupancy sensors has occurred over the last few decades. The early occupancy sensors utilized primarily analog signal processing techniques. The large area motion sensor described in U.S. Pat. No. 3,967,283 by Clark et. al., issued Jun. 29, 1976, utilized electromagnetic motion detection and was based on analog signal processing techniques. The occupancy sensor described in U.S. Pat. No. 4,661,720 by Cameron, Jr. et. al., issued Apr. 28, 1987, and the low voltage motion sensor for activating a high voltage load described in U.S. Pat. No. 4,820,938 by Mix et. al., issued Apr. 11, 1989, utilized analog signal processing techniques. The variable gain amplifier used in these sensors required manual adjustment. The room occupancy sensor, lens and method of lens fabrication described in U.S. Pat. No. 5,221,919 by Hermans, issued Jun. 22, 1993, utilized passive infrared detection and was based on analog signal processing techniques. The motion detection sensor with computer interface described in U.S. Pat. No. 5,281,961 by Elwell, issued Jan. 25, 1994, utilized active ultrasonic motion detection and was based primarily on analog signal processing techniques. Although easy to design and relatively cheap to implement, the analog filters in these devices had filter response characteristics that drifted with temperature variations and that varied over the lifetime of the various analog filter components. The overall result of using a sensor based on analog signal processing techniques was an occupancy sensor whose performance was unpredictable.
Additionally, the majority of these early occupancy sensors were based on a single sensing technology. Since each technology has its own inherent limitations, these sensors were subject to false tripping due to a variety of sources. For example, ultrasonic Doppler sensors were subject to false trips due to air conditioning and heating system air flow. In addition, since these sensors had no means of range discrimination, they were subject to false trips due to motion outside the desired range of interest. Similarly, passive infrared (PIR) sensors were subject to being blinded by intense, stationary sources of infrared energy. The automatic lighting device described in U.S. Pat. No. 4,751,399 by Koehring et. al. issued Jun. 14, 1988 utilized only acoustic motion detection. This sensor was subject to false tripping due to non-occupant sources of audio acoustic energy such as facsimile machine, telephone and security system controller tones, emergency vehicle and automobile horns, etc. The selective illumination technique described in U.S. Pat. No. 4,225,808 by Saraceni issued Sep. 30, 1980 allowed the use of pressure, ultrasonic motion, microwave, photoelectric and audible sound sensors but failed to combine these technologies to achieve a more reliable sensor with a reduced probability of false tripping. In order to lessen the probability of false trips, the user was often forced to reduce the sensor's sensitivity. The overall result of using a sensor based on a single technology was an occupancy sensor with reduced sensitivity and reliability.
The next generation of occupancy sensors used two or more sensing technologies. These sensors typically required the user to specify a separate activation threshold for each detector technology in the sensor. The digital detector output of each sensor technology was then combined using classical digital logic to detect occupancy. The preset light controller including infrared sensor operable in multiple modes described in U.S. Pat. No. 5,128,654 by Griffin et. al., issued Jul. 7, 1992, used infrared and visible light sensors. The dual technology motion sensor described in U.S. Pat. No. 5,189,393 by Hu, issued Feb. 23, 1993, combined the outputs of its ultrasonic and infrared sensors using classical Boolean AND and OR hardware logic. In general, these multiple sensing technology sensors had better performance than their predecessors but still exhibited a sensitivity-false alarm tradeoff. For example, if the various detector signals were combined using the logical OR function, the overall sensitivity of the sensor increased at the expense of an increased incidence of false trips. On the other hand, if the various detector signals were combined using the logical AND function, the overall incidence of false trips decreased at the expense of decreased sensor sensitivity. Since each sensing technology has its own separate activation threshold, these sensors were often unable to reliably detect motion in marginal cases where one or more sensing technologies observed signal levels just below the user-defined threshold level. The overall result of using these early multiple sensing technology-based occupancy sensors was an improved performance occupancy sensor that was unable to sense occupancy in the more complex marginal sensor signal level situations.
In general, prior art occupancy sensors heretofore known suffer from a number of disadvantages, including:
1. Lack of a sophisticated multiple sensing technology sensor signal conditioning to more completely exploit the advantages of sensing technologies while minimizing disadvantages. The prior art failed to combine the various occupancy sensor detection technologies in a sophisticated fashion to increase the overall probability of occupancy detection while simultaneously lowering the overall probability of false tripping.
2. Lack of adaptive sensor behavior. The prior art failed to produce an occupancy sensor whose performance adapted over time to optimize the sensor's performance.
3. Lack of digital signal processing techniques. The prior art used analog signal processing techniques. The analog filters used in these sensors required manual tuning that was a costly, time consuming process. In addition, the performance of these analog filters was temperature dependent and drifted with time.
4. Lack of means to simply and efficiently communicate the status of the sensor to installation and maintenance personnel. An occupancy sensor, typically has a number of settings that determine its mode of operation, and that the person who installs or maintains the sensor may wish to review. The sensor is typically installed out of reach on a ceiling or wall such that its adjustment knobs or dials are not readily visible. The prior art does not incorporate a system to make such settings readily available and apparent to a person who wishes to query them.
5. Lack of means to check status of the controlled signal to determine if a load device is connected, or if the controlled output is misconnected or shorted.
6. Lack of permanent storage of sensor variables. The prior art failed to permanently store various sensor settings. In the event of a power failure, these sensors had no means of recovering their previous settings.
7. Lack of no means to recognize an excessively reverberant controlled space with excessive ultrasonic return signal amplitude, and lack of means to compensate by adjusting the ultrasonic transmitter amplitude.
8. Lack of ultrasonic receiver preamplifier and demodulator performance monitoring means. The prior art did not monitor ultrasonic receiver preamplifier and demodulator performance and did not have means for making adjustments to accommodate a poorly executed installation or highly acoustically reflective space. A sophisticated ultrasonic sensor incorporates a high gain receiver preamplifier that may become saturated due to excessive acoustic reflections from room walls and other hard structures within the space. Furthermore, the sensor may be installed incorrectly too close to a fixed acoustic reflector such as a wall, exit sign, or other architectural feature. Saturation of the receiver preamplifier causes the motion signal to be lost, and the sensor to be effectively blinded by the excessive signal level. It is desirable that the sensor may be installed by unskilled personnel, and that the sensor be able to accommodate non-ideal situations created either by improper installation or difficult acoustic environments. The prior art has no means to determine saturation of the receiver preamplifier, nor any means to correct for such saturation.
9. Lack of occupancy cycle detection and utilization. The prior art did not detect the typical daily and weekly occupancy cycle of the sensor's environment and use that information to make occupancy decisions. A workspace is typically occupied according to a cycle that varies predictably throughout the day, and also according to a set pattern through the work week. Heretofore, sensors have not taken into account this pattern, and the prior art has no means to survey and record the typical daily and weekly occupancy patterns, nor to store that information, nor to act on the basis of that information.
10. Lack of adaptive PIR sensitivity adjustment to accommodate different and time-varying levels of ambient PIR noise
SUMMARY OF THE INVENTION
The present invention solves the above-noted failings in the prior art by providing an occupancy based load controller, comprising least one occupancy sensor for producing at least one occupancy estimator signal indicative of motion within a space, a programmable microprocessor, connected to the at least one occupancy sensor, for calculating an occupancy signal from the at least one occupancy estimator signal, for comparing the occupancy signal to an activation threshold, and for adjusting a sensitivity of the at least one occupancy sensor as a function of time-varying noise that corrupts the at least one occupancy estimator signal; and a controllable load energizing device responsive to the programmable microprocessor, operable to automatically energize an electrical load when the microprocessor determines that the occupancy signal is greater than the activation threshold.
The at least one occupancy sensor may be a PIR sensor, in which case the programmable microprocessor adjusts the sensitivity of the PIR sensor as a function of seasonal infrared noise that corrupts the output of the PIR sensor.
The invention also includes a method of operating an occupancy based load controller, including: at least one occupancy sensor that produces at least one occupancy estimator signal indicative of motion within a space, a programmable microprocessor, connected to the at least one occupancy sensor, for comparing the occupancy estimator signal to a predetermined threshold; and a controllable load energizing device responsive to the programmable microprocessor, operable to automatically energize an electrical load when the microprocessor determines that the occupancy estimator signal is greater than the predetermined threshold; the method comprising, estimating noise that corrupts the at least one occupancy estimator signal; and adjusting a sensitivity of the at least one occupancy sensor as a function of the estimated noise. The estimating step may include estimating noise by filtering the occupancy estimator signal using a statistical order filter, by digitally filtering the occupancy estimator signal by time integrating the occupancy estimator, by averaging the occupancy estimator signal over time, or by detecting an envelope of the occupancy estimator signal. Sensitivity adjustment is preferably done based upon samples of the occupancy estimator signal taken while the electrical load is de-energized. Further, sensitivity adjustment may be accomplished by adjusting the predetermined threshold as a function of the estimated noise.
Other features and advantages of the invention will become apparent from a consideration of the drawings and ensuing detailed description.

BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a simplified functional block diagram of the present invention.
FIG. 2 is a detailed functional block diagram of the present invention.
FIG. 3 is the user controls and switches circuit of the present invention.
FIG. 4 is the ultrasonic transmitter circuit of the present invention.
FIG. 5 is the ultrasonic receiver circuit of the present invention.
FIG. 6 is the ultrasonic transmitter signal conditioning circuit of the present invention.
FIG. 7(a) is the ultrasonic receiver signal conditioning circuit of the present invention.
FIG. 7(b) is an alternative sampling control circuit useable in the circuit of FIG. 7(a)
FIG. 8 is the infrared sensor circuit of the present invention.
FIG. 9 is the infrared signal conditioning circuit of the present invention.
FIG. 10 is the acoustic sensor circuit of the present invention.
FIG. 11 is the acoustic signal conditioning circuit of the present invention.
FIG. 12 is the microwave transmitter/receiver circuit of the present invention.
FIG. 13 is the microwave signal conditioning circuit of the present invention.
FIG. 14 is the photo cell sensor circuit of the present invention.
FIG. 15 is the software time division multiplexing scheme used for task allocation of the present invention.
FIGS. 16(a)-(f) are graphs illustrating prior art dual technology occupancy detection using the Boolean AND function.
FIG. 17 is a decision surface for the arithmetic sum formula in accordance with the present invention.
FIGS. 18(a)-(f) are graphs illustrating dual technology occupancy detection in accordance with the present invention.
FIG. 19 is a decision surface for the Yager Union function (K=1, p=2) in accordance with the present invention.
FIG. 20 is a decision surface for the Yager Union function (K=1, p=0.5) in accordance with the present invention.
FIG. 21 is a decision surface for a table look-up version of the Yager Union function in FIG. 19.
FIG. 22 is a decision surface for a table look-up version of the Yager Union function in FIG. 20.
FIG. 23 is a time line depicting the adaptation of sensor sensitivity for a false-on event in accordance with the present invention.
FIG. 24 is a time line depicting the adaptation of sensor sensitivity for a hallway strike false-on-stay-on event in accordance with the present invention.
FIG. 25 is a time line depicting the adaptation of sensor sensitivity for a false-off event in accordance with the present invention.
FIG. 26 is a time line depicting a stay on with no major activity.
FIG. 27 is a time line depicting a stay-on-multiple-activation event
FIG. 28 is a time line depicting an expanded occupancy estimator signal for a stay-on-multiple-activation event.
FIG. 29 is the stay-on-multiple-activation event and strong airflow analysis task flowchart for the present invention.
FIG. 30 is a diagram showing sensor installations that may result in adaptation of sensor sensitivity.
FIG. 31 is the occupancy cycle detection and utilization measurement method for the present invention.
FIG. 32 is the occupancy cycle detection and utilization task flowchart for the present invention.
FIG. 33 is a typical learning curve describing the adaptive behavior of the present invention.
FIG. 34 is a "lights off" task flowchart when using an acoustic sensor to prevent "false off" for the present invention.
FIG. 35 is a PIR noise level adjustment task flowchart for the present invention.
FIG. 36 is a time line illustrating the adjustment of ultrasonic sensitivity based on acoustic signal detection.
FIG. 37 is a timer adjustment task flowchart for the present invention.
FIG. 38 is a minimize energy usage at power-up task flowchart for the present invention.
FIG. 39 is the main (foreground) routine flowchart of the present invention.
FIG. 40 is the hardware and system variable initialization routine flowchart of the present invention.
FIG. 41 is the interrogation status report routine flowchart of the present invention.
FIG. 42 is the two minute status report routine flowchart of the present invention.
FIG. 43 is the false-on correction task flowchart of the present invention.
FIG. 44 is the false-off correction task flowchart of the present invention.
FIG. 45 is the interrupt (background) routine flowchart for the present invention.
FIG. 46 is the sensor or task flowchart for the present invention.
FIG. 47 is the ultrasonic DSP task block diagram for the present invention.
FIG. 48 is the infrared DSP task block diagram for the present invention.
FIG. 49 is a time line of two unstretched occupancy signals
FIG. 50 is a time line of two stretched occupancy signals
FIG. 51 is a time line of two peak stretched occupancy signals
FIG. 52 is the HIGH CONFIDENCE task flowchart for the present invention.
FIG. 53 is the HIGH SENSITIVITY task flowchart for the present invention.
FIG. 54 is the average time between motion routine flowchart for the present invention.
FIG. 55 is the update counters task flowchart for the present invention.
FIG. 56 is the adjust timer to minimize energy usage task flowchart for the present invention.
FIG. 57 is the MUX task flowchart for the present invention.
FIG. 58 is the adapt ultrasonic sampling routine flowchart for the present invention.
FIG. 59 is a flowchart of the ultrasonic transmitter duty cycle adjustment method of the present invention.
FIG. 60 is s time line of input receiver saturation
FIG. 61 is the LED task flowchart for the present invention.
FIG. 62(a) is the LED status task flowchart for the present invention.
FIG. 62(b) is an example of a status report according to the present invention.
FIG. 63 is the relay suppress task flowchart for the present invention.
FIG. 64 is the second task flowchart for the present invention.
FIG. 65 is the relay task flowchart for the present invention.
FIG. 66 is the minute task flowchart for the present invention.
FIG. 67 is the initialize I/O port data direction registers task flowchart for the present invention.
FIG. 68 is the interrogate task flowchart for the present invention.
FIG. 69 is the EEPROM task flowchart for the present invention.
FIG. 70 is a graph showing seasonal variation of PIR noise.
FIG. 71 is a graph of PIR thresholds.
FIG. 72 is a PIR threshold adjustment flowchart of the present invention.

DESCRIPTION OF ILLUSTRATIVE EMBODIMENTS
Glossary of Acronyms
A/D analog-to-digital converter
ASA automatic sensitivity adjustment
ATA automatic timer adjustment
CCP capture/compare/PWM
CMOS complimentary metal oxide semiconductor
DSP digital signal processing
EEPROM electrically erasable programmable read only memory
GSB gain setback
HVAC heating, ventilation, air conditioning
I/O input/output
LED light emitting diode
MUX multiplexer
OP AMP operational amplifier
PIR passive infrared
PWM pulse width modulation
RAM random access memory
ROM read only memory
TTL transistor-transistor logic
VDC volts direct current
A simplified, functional block diagram of the multi-technology-based occupancy sensor 100 of the present invention is shown in FIG. 1. The occupancy sensor inputs are on the left side of FIG. 1. The digital microcontroller 101 and various signal conditioning blocks are in the center of FIG. 1. The various outputs are on the right hand side of FIG. 1. FIG. 1 illustrates the use of five different sensor technologies: ultrasonic 102, infrared 103, acoustic 104, microwave 105 and visible light 106. The analog output signal of the ultrasonic sensor 102 is digitized using an analog-to-digital converter (A/D) 107. The digital ultrasonic sensor signal is then processed using an ultrasonic DSP algorithm 108 on the digital microcontroller. The resulting ultrasonic output signal, referred to as an occupancy estimator signal, is then fed to a digital sensor fusion detector 109. In a similar fashion, occupancy estimator signals are generated for the analog signals from the infrared, acoustic, microwave and visible light sensors 103, 104, 105, 106 by digitizing the analog signal using A/Ds 110, 111, 112, 113 and processing the signal using the appropriate DSP algorithms 114, 115, 116, 117. The resulting occupancy estimator signals are fed to the digital sensor fusion detector 109. The digital sensor fusion occupancy detector 109 then combines the different processed sensor signals to produce a composite occupancy estimator that is compared to a composite sensor fusion threshold 118 (either a composite activation or composite maintenance threshold) to determine occupancy. The output of the sensor fusion detector 109 is used to actuate a relay that energizes the electrical load(s) controlled by the occupancy sensor.
If the ultrasonic DSP 108 output signal exceeds the user-defined ultrasonic activation threshold 119 as determined by comparator 121, the ultrasonic motion indicators will be turned on. Similarly, if the infrared DSP 114 output signal exceeds the user-defined infrared activation threshold 122 as determined by comparator 123, the infrared motion indicators will be turned on. It should be noted that the individual sensor technology activation thresholds 119 and 122 are preferably used only to drive the various sensor technology motion indicators and are not used alone to sense occupancy.
A more detailed functional block diagram of the present invention is shown in FIG. 2. The occupancy sensor inputs are on the left hand side of the diagram. The digital microcontroller and its related blocks are in the center of the diagram. The various outputs are on the right hand side of the diagram. The occupancy sensor has the following major functional blocks: power supplies circuit 201; clock generator circuit 202; power reset circuit 203; user controls and switches 204; EEPROM circuit 206; ultrasonic transmitter circuit 207; ultrasonic receiver circuit 102; ultrasonic transmitter signal conditioning circuit 209; ultrasonic receiver signal conditioning circuit 211; infrared sensor circuit 103; infrared signal conditioning circuit 214; acoustic sensor circuit 104; acoustic signal conditioning circuit 217; microwave sensor circuit 105, microwave signal conditioning circuit 215, microwave transmitter circuit 225, microwave transmitter signal conditioning circuit 210, photo cell sensor circuit 106; photo cell signal conditioning circuit 219; ultrasonic motion indicator circuit 223; infrared motion indicator circuit 224; electrical load(s) switching and dimming control interface 226; and digital microcontroller circuit 229.
Each of these electrical hardware blocks is now described in detail. An exemplary embodiment of the power supplies circuit 201 accepts an unregulated input voltage, Vin, and produces one or more regulated out voltages that are used to power the various components shown in FIG. 2. An exemplary embodiment has an input voltage is 24 VDC and produces nominal regulated output voltages of 13.1, 5.1 and 2.5 VDC, although other input and output voltages would also be acceptable. The 2.5 VDC supply is also referred to as VREF.
Clock generator circuit 202 may be of conventional crystal oscillator design, and in the exemplary embodiment, includes a crystal and a clock driver circuit internal to the digital microcontroller 229. The clock generator circuit 202 has no input and the output of the clock generator circuit is the digital oscillator clock signal for the digital microcontroller 229. An 8.0 MHz or 9.8 MHz parallel cut crystal is used depending on the desired ultrasonic transmitter frequency.
Power reset circuit 203 is of conventional design. The input to power reset circuit 203 is the 5.1 VDC power supply. The output of power reset circuit 203 is an exponentially rising analog signal used to reset the digital microcontroller 229 upon application of power.
An exemplary embodiment of the user controls and switches block 204 is shown in FIG. 3 and comprises two banks of switches 301, 302, four potentiometers 303 and a CMOS 4 channel differential analog multiplexer 304. The inputs to the user controls and switches block 204 are the user-specified settings of these switches and potentiometers. The outputs of the user controls and switches block 204 are the analog voltages corresponding to the user-specified settings.
The first bank of switches 301 includes four switches 306, 307, 308, 309 that determine the basic operational mode of the sensor. Specifically, these switches include manual override switch 306, dual technology mode switch 307, LED override switch 308 and adaptation reset switch 309.
Manual override switch 306 is used to override the normal operation of the sensor. In the ON state, the electrical load switched by the sensor will be energized at all times. In the OFF state, the electrical load switched by the sensor will be energized whenever occupancy is sensed.
Dual technology mode switch 307 determines the thresholds to be used for comparing to the composite occupancy estimator. The system can be configured with any combination of ultrasonic, PIR and acoustic detection technologies. In the OFF state (referred to as the HIGH CONFIDENCE THRESHOLD mode), the sensor's composite activation threshold is set at a high level. In this mode, the sensor is less likely to false trigger, but the sensor is not as sensitive and therefore less likely to detect marginal detection cases. In the ON state (referred to as the HIGH SENSITIVITY THRESHOLD mode), the sensor's composite activation threshold is set at a lower level than the HIGH CONFIDENCE THRESHOLD mode. In this mode, the sensor is more sensitive and will more accurately detect marginal cases at the risk of a greater likelihood of false triggering.
LED override switch 308 is used to override the normal operation of the sensor's LEDs. In the OFF state, all sensor LEDs will be disabled. In the ON state, all sensor LEDs will be enabled.
Adaptation reset switch 309 is used to reset the sensor's adaptive variables. If the sensor's environment changes significantly a long time after the power on reset condition, the sensor's adaptive variables may need to be reset so that the sensor can begin adapting to its new environment. In the OFF state, the sensor's adaptive variables will not be reset. In the ON state, the sensor's adaptive variables will be reset.
The second bank of switches 302 comprises four switches 311, 312, 313, 314 that select the more advanced features of the sensor. Specifically, these switches include: strong airflow algorithm switch 311, gain setback switch 312, automatic timer adjustment switch 313, and automatic sensitivity adjustment switch 314.
Strong airflow algorithm switch 311 is used to modify the ultrasonic sensor's response in order to better compensate for the presence of air flow. In the OFF state, the sensor uses its standard airflow ultrasonic DSP algorithm. In the ON state, the sensor uses its strong airflow ultrasonic DSP algorithm.
Gain setback switch 312 is used to determine the sensor's gain as a function of time since motion was last sensed. In the OFF state, the sensor's gain does not vary as a function of time. In the ON state, the sensor's gain will vary as a function of time.
Automatic timer adjustment switch 313 is used to determine the sensor's timer delay setting as a function of time. In the OFF state, the sensor's timer delay setting does not automatically vary as a function of time. In the ON state, the sensor's timer delay setting varies automatically as a function of time.
Automatic sensitivity adjustment switch 314 is used to determine the sensor's sensitivity setting as a function of time. In the OFF state, the sensor's sensitivity setting does not automatically vary as a function of time. In the ON state, the sensor's sensitivity setting varies automatically as a function of time.
The user controls and switches block 204 also contains four potentiometers: ultrasonic sensitivity potentiometer 316, PIR sensitivity potentiometer 317, daylight control sensitivity potentiometer 318, and timer delay potentiometer 319.
Ultrasonic sensitivity potentiometer 316 is used to increase or decrease the sensor's ultrasonic sensitivity. The ultrasonic sensitivity potentiometer setting maps linearly to sensor range. PIR sensitivity potentiometer 317 is used to increase or decrease the sensor's PIR sensitivity. Daylight control sensitivity potentiometer 318 is used to adjust the sensor's daylight control sensitivity by increasing or decreasing the daylight control ambient light threshold. Timer delay potentiometer 319 is used to adjust the sensor's timer delay by increasing or decreasing the length of time the sensor's electrical load(s) are energized following occupancy detection. The timer delay potentiometer setting maps linearly to sensor time delay.
The user controls and switches block 204 also contains a CMOS 4 channel differential analog multiplexer 304. Multiplexer 304 may be a 4052 type multiplexer, available from National Semiconductor. Multiplexer 304 allows the user to simultaneously read two analog input signals. The inputs to multiplexer 304 are two banks of four analog signals 302, 303 and two digital logic control signals A and B generated by digital microcontroller 229 (FIG. 2). The outputs of multiplexer 304 are the two analog signals selected using the multiplexer 304 for application to digital microcontroller 229.
In the exemplary embodiment, EEPROM circuit 206 comprises a CMOS 256.times.8 serial EEPROM for non-volatile storage of the various occupancy sensor registers although other memories may be used. EEPROM circuit 206 has two inputs, a serial clock and serial data line. The output of the EEPROM circuit is a serial data line. In the exemplary embodiment, EEPROM 206 uses a standard serial bus protocol for saving and retrieving the various occupancy sensor register values.
An exemplary embodiment of the ultrasonic transmitter circuit 207 is shown in FIG. 4 and is comprised of one or two 16 mm narrowband air ultrasonic ceramic transducers 401, 402. The input 403 to these transducers 401, 402 is either a 0 to 15 volt or -15 to 15 volt variable duty cycle square wave, although other driving signal may also be acceptable. The output of transducers 401, 402 is an ultrasonic continuous-wave tone at the desired frequency of interest.
An exemplary embodiment supports ultrasonic transmitter frequencies of 25 KHz, 32 KHz and 40 KHz, but other frequencies would also be acceptable. A variety of commercial ultrasonic transmitting transducers are available. For example, acceptable 16 mm transmitting transducers available from S-Square are shown in Table 1.
TABLE 1______________________________________Exemplary S-Square Ultrasonic Transmitting Transducers Sound InputFrequency Model Level Bandwidth Capacitance Power(KHz) Number (dB) (KHz) (pF) (watts)______________________________________25 250ST1 112 2 2400 0.2 6032 328ST1 115 2 2400 0.2 6040 400ST1 119 2 2400 0.2 60______________________________________
An exemplary embodiment of ultrasonic receiver circuit 102 is shown in FIG. 5 and includes one or two 16 mm narrowband air ultrasonic ceramic transducers 501, 502. The input to these transducers 501, 502 is an ultrasonic continuous-wave tone centered at the desired frequency of interest with an additional Doppler-shifted signal corresponding to motion in the sensor's field of view. The output 503 of ultrasonic receiver circuit 102 is an electrical signal corresponding to the received acoustic signal. An exemplary embodiment of the present invention supports ultrasonic receiver frequencies of 25 KHz, 32 KHz and 40 KHz, and variety of commercial ultrasonic receiving transducers are available. For example, acceptable 16 mm receiving transducers available from S-Square are shown in Table 2.
TABLE 2______________________________________Exemplary S-Square Ultrasonic Receiving Transducers InputFrequency Model Sensitivity Bandwidth Capacitance Power(KHz) Number (dB) (KHz) (pF) (watts)______________________________________25 250SR1 -65 2 2400 0.2 6032 328SR1 -67 2.5 2400 0.2 6040 400SR1 -65 2.5 2400 0.2 60______________________________________
An exemplary ultrasonic transmitter signal conditioning circuit 209 is shown in FIG. 6 and includes an NPN transistor circuit 601, a CMOS buffer circuit 602, a CMOS inverter circuit 603 and two CMOS driver circuits 604, 605. Input 606 to circuit 209 is a variable duty cycle TTL-level square wave generated by the digital microcontroller 229. Output 403 of circuit 209 is applied to ultrasonic transmitter circuit 207 (FIG. 4) and is either a single-ended or double-ended ultrasonic transmitter signal at the desired frequency of interest.
The variable duty cycle TTL-level square wave input signal 606 generated by the digital microcontroller 229 is generated by a latch (not shown) that is set at a rate determined by a carrier period counter value, and that is cleared within each carrier period at a point determined by the carrier pulse width counter value. Unequal turn-on and turn-off times of the transmitter signal conditioning circuit 209 that drives the carrier send transducer may result in an asymmetric signal at the amplifier output, and thus less than full power being dissipated in transmitter. The sensor algorithm may compensate for the unequal turn-on and turn-off times by setting the carrier pulse width counter to a predetermined value intended to yield a symmetric square wave at the amplifier output. In the event that it is desired to decrease the transmitter output power level, the carrier pulse width counter may be varied to deviate from a 50% duty cycle square wave, lowering the effective AC drive level to the transmitting transducer, and thus the output level.
NPN transistor circuit 601 is used as a level shifter. The input to NPN transistor circuit 601 is a TTL-level square wave. The output of NPN transistor circuit 601 is a 0 to 15 volt level square wave. CMOS buffer circuit 602 is composed of a single CMOS inverter with Schmitt trigger. This CMOS inverter 602 is used to increase the output drive capability of the NPN transistor circuit 601. The input to the CMOS buffer 602 is a 0 to 15 volt level square wave. The output of the CMOS buffer 602 is a 0 to 15 volt level square wave. CMOS inverter circuit 603 is comprised of a single CMOS inverter with Schmitt trigger. CMOS inverter circuit 603 inverts the ultrasonic transmitter signal to allow double-ended transmitter drive capability. The two CMOS driver circuits 604,605 are each comprised of two CMOS inverters with Schmitt triggers, in parallel. These two driver circuits 604, 605 allow the ultrasonic transducer(s) to be driven in either single ended or differential mode. For single ended mode the transducer(s) is connected between driver circuit 604 via capacitor 607, and the effective drive applied to the transducer(s) is a square wave with amplitude +7.5V to -7.5V. For differential drive, the transducer(s) is connected between driver circuit 604, coupled via capacitor 607, and driver circuit 605, and the effective drive applied to the transducer(s) is a square wave of amplitude +15V to -15V.
The ultrasonic receiver signal conditioning circuit 211 is shown in FIG. 7(a) and includes a variable gain op amp circuit 701, a PNP transistor circuit 702, an NPN transistor circuit 703, three RC circuits 704, 706, 707, an op amp buffer circuit 708 and a two stage op amp circuit 709. The inputs to ultrasonic receiver signal conditioning circuit 211 are the Doppler-shifted analog ultrasonic receiver signal 503 and a digital sampling point signal 711. The outputs of ultrasonic receiver signal conditioning circuit 211 are the demodulated, filtered analog ultrasonic receiver signal 712 and an analog ultrasonic sampling point signal 713.
Variable gain op amp circuit 701 is an ultrasonic receiver preamplifier circuit. The input to this circuit is the modulated analog ultrasonic receiver signal 503. The output of this circuit is an amplified, modulated analog ultrasonic receiver signal. Circuit 701 uses diodes in the negative feedback path to switch in parallel with a resistor, thus decreasing the overall gain for large signal excursions, and preventing hard-limiting of the amplifier in the event of excessive continuous wave receiver signals. The resulting nonlinear transfer characteristic is advantageous when the sensor is installed in a confined space where wall reflections cause a large amount of acoustic energy to be directed into the ultrasonic receiver.
PNP transistor circuit 702 is a zero crossing phase lock loop sampling point circuit. The input to circuit 702 is a digital sampling point signal 711 generated by the digital microcontroller 229. The output of circuit 702 is an analog sampling signal used to drive the synchronous demodulator circuit 703. Circuit 702 varies the position (or phase) of the sampling point on the ultrasonic receiver waveform under control of digital microcontroller 229. This prevents the loss of motion information due to large signal levels. For optimum sensitivity, the synchronous sample point on the ultrasonic receiver waveform should lie as close to the zero crossing as possible.
NPN transistor circuit 703 is a synchronous demodulator circuit. The two inputs to circuit 703 are the amplified modulated analog ultrasonic receiver signal and the analog sampling point signal. The output of circuit 703 is the demodulated Doppler-shifted analog ultrasonic receiver signal. The NPN transistor in circuit 703 is arranged in a series pass configuration and is driven by the analog sampling point signal generated by PNP transistor circuit 702.
RC circuit 704 is an envelope detector circuit. The input to RC circuit 704 is the Doppler-shifted analog ultrasonic receiver signal. The output of RC circuit 704 is the filtered, demodulated Doppler-shifted analog ultrasonic receiver signal. RC circuit 704 acts as an envelope detector and filters out the ultrasonic carrier while preserving the low frequency Doppler shift signal information.
Op amp buffer circuit 708 serves to increase the drive capability of the synchronous demodulator circuit 703. The input to circuit 708 is the filtered, demodulated Doppler-shifted analog ultrasonic receiver signal. The output of circuit 708 is the increased drive, filtered, demodulated Doppler-shifted analog ultrasonic receiver signal.
RC circuit 706 is a lowpass filter circuit. The input to RC circuit 706 is the increased drive, filtered, demodulated Doppler-shifted analog ultrasonic receiver signal. The output of RC circuit 706 is the increased drive, lowpass filtered, demodulated Doppler-shifted analog ultrasonic receiver signal. The output of circuit 708 is lowpass filtered to remove contributions due to motion in the environment, leaving a DC signal that represents the receiver carrier amplitude at the ultrasonic sample point. This signal is sampled by the digital microcontroller 229 through signal 711 to yield the analog ultrasonic sampling point signal 713.
Two stage op amp circuit 709 is a bandpass filter circuit. The input to circuit 709 is the increased drive, filtered, demodulated Doppler-shifted analog ultrasonic receiver signal. The output of circuit 709 is the increased drive, bandpass filtered, demodulated Doppler-shifted analog ultrasonic receiver signal. The pass band of circuit 709 is designed to pass the Doppler-shifted signal for motions of interest.
RC circuit 707 is an anti-aliasing filter circuit. The input to RC circuit 707 is the increased drive, bandpass filtered, demodulated Doppler-shifted analog ultrasonic receiver signal. The output of RC circuit 707 is the anti-aliased, increased drive, bandpass filtered, demodulated Doppler-shifted analog ultrasonic receiver signal. The output signal 712 is sampled by the digital microcontroller 229 A/D circuitry and processed using a variety of digital signal processing techniques as discussed in detail below.
FIG. 7(b) is an exemplary embodiment of a circuit that may be used in combination with the circuit of FIG. 7(a) to affect sampling of the Doppler shifted ultrasonic signal without the use of microcontroller 229 to generate the sampling signal. The circuit of FIG. 7(b) is based on a commercially available 555 timer circuit 714 and peripheral circuitry. The input to the timer circuit 714 is the same as the input 606 to the transmitter signal conditioning circuit shown in FIG. 6, and the output of the timer circuit 714 is connected to the sampling point signal input 711 to the ultrasonic receiver signal conditioning circuit of FIG. 7(a).
An exemplary embodiment of the infrared sensor circuit 103 is shown in FIG. 8 and includes a dual element pyroelectric infrared motion sensor circuit. The input to infrared sensor circuit 103 is infrared electromagnetic radiation. The output of infrared sensor circuit 103 is an electrical signal indicative of motion.
A variety of commercial infrared motion sensors are available. An acceptable example is the Heimann LHi 878 dual element pyroelectric infrared motion sensor. This sensor is designed to detect motion of human body radiation in the range of electromagnetic wavelengths of 5 to 14 micrometers. Other infrared sensors would also be acceptable.
An exemplary embodiment of the infrared signal conditioning circuit 214 is shown in FIG. 9 and includes two cascaded op amp bandpass circuits 901, 902. The input 903 to infrared signal conditioning circuit 214 is the electrical PIR sensor signal indicative of motion. The output 904 of infrared signal conditioning circuit 214 is a bandpass filtered infrared sensor signal indicative of motion. The output signal 904 of this circuit is sampled by A/D circuitry within the digital microcontroller 229 and processed using a variety of digital signal processing techniques discussed below.
An exemplary embodiment of the acoustic sensor circuit 104 is shown in FIG. 10 and includes an audio acoustic sensor 1002 and surrounding circuitry. The input to acoustic sensor circuit 104 is sound energy. The output 1001 is an electrical signal indicative of occupancy such as human speech and sounds generated by human occupants.
A variety of commercial acoustic sensors 1002 are available. An acceptable example is the Panasonic WM-52BM electret condenser microphone cartridge. It includes a high voltage internal electret membrane, metal electrode and a field effect transistor (FET). Frequency range extends from 20-16,000 Hz.
An exemplary embodiment of the acoustic signal conditioning circuit 217 is shown in FIG. 11 and includes two cascaded op amp band pass circuits 1102 and 1103 followed by a fast attack, slow decay peak detector 1104. The input 1101 to acoustic signal conditioning circuit 217 is the electrical acoustic sensor signal indicative of occupancy 1001. The output 1105 of the acoustic signal conditioning circuit 217 represents the magnitude of acoustic energy. The output signal of this circuit is sampled by A/D circuitry within the digital micro controller 229 and processed using a variety of digital signal processing techniques discussed in detail below.
An exemplary embodiment of the microwave sensor circuit 105 is shown in FIG. 12, and includes a microwave antenna 1202 and an oscillator circuit 1203. Circuit 105 combines the transmitting and receiving elements into one antenna 1202. The antenna 1202, inductor 1204, capacitor 1205 determine oscillator frequency and with NPN transistor circuit 1206 form the oscillator and transmitter. The output impedance of the transmitter is high enough not to significantly reduce reflected signal reception. The input to microwave sensor circuit 105 is electromagnetic energy reflected from the observed space. The output 1201 is an amplitude modulated radio frequency signal indicative of motion. A variety of constructions for antenna 1202 are available from wave guides to simple wire.
An exemplary embodiment of the microwave signal conditioning circuit 215 is shown in FIG. 13 and includes two cascaded op amp band pass circuits 1302 and 1303, followed by a fast attack, slow decay peak detector 1304. The input 1201 to microwave signal conditioning circuit 215 is the electrical microwave sensor signal indicative of motion. The output 1305 of microwave signal conditioning circuit 215 represents the magnitude of received Doppler microwave energy. The output signal 1305 is sampled by A/D circuitry within the digital micro controller 229 and processed using a variety of digital signal processing techniques discussed in detail below.
An exemplary embodiment of the photo cell sensor circuit 106 is shown in FIG. 14 and is comprised of a visible light photo sensor 1401 and supporting circuitry. The input to photo cell sensor circuit 106 is visible electromagnetic radiation. The output of photo cell sensor circuit 106 is an electrical signal indicative of ambient light. A variety of commercially available photo cells are available for use as light photo sensor 1401, including, for example, the Vactec VT90N4 photoconductive cell sensor.
Photo cell signal conditioning circuit 219 may be, for example, an RC lowpass filter. The input to this RC lowpass filter is the electrical photo cell sensor signal indicative of ambient light. The output of this RC lowpass filter is a lowpass filtered photo cell sensor signal indicative of ambient light. The output signal of photo cell signal conditioning circuit 219 is sampled by A/D circuitry within the digital microcontroller 229 and processed using a variety of digital signal processing techniques as described in detail below.
Ultrasonic motion indicator circuit 223 and infrared motion indicator circuit 224 may each be comprised of LEDs driven by digital microcontroller 229. The outputs of these indicators 223 and 224 are lights indicating that the ultrasonic or infrared portion of the sensor has detected motion above user-specified thresholds. In the preferred embodiment, the ultrasonic activation threshold and the infrared activation threshold are used only for the purpose of driving the ultrasonic motion indicator and infrared motion indicator and are not used by the digital sensor fusion detection algorithm to determine occupancy.
The electrical load switching and dimming control interface circuit 226 may take many forms and typically includes a conventional transistor-based relay driver circuit and relay that interfaces digital microcontroller 229 with an electrical load that is operated at a higher voltage than the power supplied to microcontroller 229. Provisions may be made to control any number of electrical loads. Incandescent light loads may be controlled by a simple relay, while fluorescent lighting may be controlled using dimming ballasts or other electrical load dimming devices in a known manner.
In accordance with one embodiment of the present invention, the digital microcontroller circuit 229 may be an 8-bit CMOS microcontroller with A/D converter for sophisticated communication and control of the sensor. An acceptable commercially available microcontroller is an 8-bit low-cost, high-performance Microchip PIC 16C73 A fully-static, EEPROM-based 28-pin CMOS microcontroller, however, other microcontrollers or microprocessors may also be acceptable. The microcontroller clock frequency in an exemplary embodiment is either 8.0 MHz or 9.83 MHz depending on the desired ultrasonic frequency. The preferred device has an advanced RISC-like Harvard architecture and includes a large register set, an eight-level deep hardware stack and supports multiple internal and external interrupt sources. The device has 192 bytes of RAM and 22 I/O pins. It also has a variety of on-chip peripherals including five 8-bit A/D channels, three timer/counters modules, two capture/compare/pulse width modulation (PWM) modules and two serial ports. The synchronous serial port of the device can be configured as either a 3-wire serial peripheral interface device or a two-wire Inter-Integrated Circuit (I.sup.2 C) bus device. The serial communications interface can be configured as either a synchronous or an asynchronous device. The device offers four oscillator options and includes a highly reliable watchdog timer to provide protection against software malfunctions.
Although the hardware aspects of the present invention have been described with reference to a particular exemplary embodiment, it will be understood that addition, deletions and changes may be made to the exemplary embodiment without departing from the scope of the present invention
The software used to control the operation of digital microcontroller 229 in accordance with one embodiment of the invention includes two primary routines, the main (foreground) routine and the interrupt (background) routine. The main foreground routine initializes the sensor and then enters the main loop that periodically displays the status of the sensor and resets the microcontroller's watch dog timer. The background interrupt routine is executed on a regular, periodic basis and performs most of the sensor's major functions.
The various operational functions of the sensor are separated into tasks such as the sensor task, MUX task, LED task, etc. Each of these high-level tasks may be further broken down into a sequence of smaller tasks. For example, the sensor task may be broken down into the ultrasonic sensor task, the PIR sensor task, the acoustic sensor task, and the photo sensor task Each task may be enabled by setting that task's enable flag. Similarly, each task may be disabled by clearing that task's enable flag.
One of the timer/counter peripherals within digital microcontroller 229 is used to interrupt microcontroller 229 on a regular, periodic basis. This scheme allows the instruction bandwidth of microcontroller 229 to be evenly divided into a number of execution time slots as shown in FIG. 15. This time-division multiplexing scheme is used to run the various sensor tasks at different periodic intervals. For example, the sensor's display status task is run once every two minutes while the sensor's second task is run once a second. By counting interrupts (time slots) and balancing the load of each task, the various sensor tasks can be multiplexed in time to keep the microcontroller evenly loaded. For simplicity, the present implementation uses 256 time slots, each time slot having a duration of 1 millisecond.
The early multi-technology occupancy sensors utilized a separate activation threshold for each sensor technology in order to generate a separate digital occupancy signal for each sensor technology. Typically, whenever a given sensor signal was greater than or equal to the activation threshold value for that technology, the digital occupancy estimator signal for that sensor technology was a logical one. Whenever that sensor signal was below the activation threshold value for that technology, the sensor's digital occupancy estimator signal for that sensor technology was a logical zero. The various digital occupancy estimator signals of each sensor technology were then combined using classical Boolean (digital) logic in order to generate a digital composite occupancy estimator signal.
For example, consider a dual technology sensor that used ultrasonic and infrared sensors and combined their digital outputs using the AND logic function. Expressed symbolically, the digital output of each of the sensors and the overall digital output of the dual technology occupancy sensor would be: ##EQU1##
In general, these early multi-technology sensors exhibited a sensitivity-false alarm tradeoff depending on which Boolean logic function was used to combine the individual digital occupancy estimator signals. If the various individual digital occupancy estimator signals were combined using the Boolean OR function, the overall sensitivity of the sensor increased at the expense of an increased incidence of false trips. If the various individual digital occupancy estimator signals were combined using the Boolean AND function, the overall incidence of false trips decreased at the expense of decreased sensor sensitivity. Since each technology had its own separate activation threshold, these prior art sensors were often unable to reliably detect motion in complex marginal cases where one or more technologies observed signal levels just below the user-defined threshold level.
An example of this situation for a prior art dual technology sensor is shown in FIGS. 16 (a)-(f). FIG. 16(a) depicts a typical ultrasonic sensor signal that occasionally exceeds the ultrasonic activation level threshold. FIG. 16(b) depicts a typical passive infrared sensor signal that also occasionally exceeds the passive infrared activation level threshold. FIG. 16(c) depicts the corresponding individual digital occupancy estimator signals. When the two individual digital occupancy estimator signals are combined using the Boolean AND function, the sensor's digital composite occupancy estimator signal (not shown) indicates that occupancy has been detected. In contrast, FIGS. 16(d)-(f) depict a more complex situation. FIG. 16(d) depicts an ultrasonic sensor signal that is close to the ultrasonic activation level threshold but never exceeds the ultrasonic activation level threshold. FIG. 16(e) depicts a typical passive infrared sensor signal that occasionally exceeds the passive infrared activation level threshold. FIG. 16(f) depicts the corresponding individual digital occupancy estimator signals. When the two individual digital occupancy estimator signals are combined using the Boolean AND function, the sensor's digital composite occupancy estimator signal (not shown) indicates that occupancy has not been detected.
While these prior art multi-technology-based occupancy sensors showed improved performance over single technology occupancy sensors, they were still unable to sense occupancy in the more complex marginal sensor signal level situations.
In contrast, the multi-technology sensor-fusion-based occupancy sensor of the present invention does not generate a separate digital occupancy signal for each sensor technology. Instead, the multi-technology sensor-fusion method of the present invention takes the various sensor output signals and forms a composite occupancy estimator signal. This composite occupancy estimator signal is then compared with a composite activation threshold value. Whenever the composite occupancy estimator signal is greater than or equal to the composite activation threshold value, the sensor's digital composite occupancy signal is a logical one. Whenever the composite occupancy estimator signal is below the composite activation threshold value, the sensor's digital composite occupancy signal is a logical zero.
For example, consider a dual technology sensor-fusion-based occupancy sensor in accordance with the present invention that uses ultrasonic and infrared sensors and combines their outputs using an arbitrary generalized non-linear function. Expressed symbolically, the composite occupancy estimator signal and the overall digital composite occupancy signal of the dual technology sensor-fusion-based occupancy sensor is: ##EQU2## where, f(, , , . . . , ) denotes an arbitrary generalized function of N variables (i.e., one variable for each sensor technology). Note that for the dual technology sensor of the present invention with (N=2), the function f(, ) may be an arithmetic sum, a weighted arithmetic sums, or an arbitrary non-linear function of the variables expressed either algorithmically or in tabulated form. The algorithmic form is more amenable to mathematical analysis whereas the tabulated form has distinct advantages from a microprocessor implementation point of view.
One embodiment of the present invention utilizes an arithmetic sum based occupancy detection algorithm to implement a sophisticated multi-technology sensor fusion detection scheme. This algorithm increases the overall probability of occupancy detection while simultaneously lowering the overall probability of false tripping. In a dual technology sensor using ultrasonic and PIR sensors, the invention combines the outputs of the ultrasonic and PIR sensors to produce a composite occupancy estimator that is compared to a composite threshold to determine occupancy. This produces a highly sensitive yet highly reliable occupancy sensor that can detect minor motion in more complex marginal sensor signal level situations.
A plot of the two-dimensional arithmetic sum surface is shown in FIG. 17. When the arithmetic sum is greater than the activation threshold shown in FIG. 17, a load is activated. Once activated, the load remains activated when the arithmetic sum is greater than the maintenance threshold shown in FIG. 17. By combining these signals prior to thresholding, sub-threshold signal levels are able to contribute useful information to the decision process. This allows the sensor to effectively utilize sensor signal values that are high enough to be indicative of occupancy but too low to indicate certain occupancy according to a given single sensing technology. In an exemplary embodiment of the invention, the computation and evaluation of the composite occupancy estimator signal is performed separately for HIGH CONFIDENCE and HIGH SENSITIVITY modes using fixed-point arithmetic.
An example of this situation for a dual technology occupancy sensor according to the present invention is shown in FIGS. 18 (a)-(f). FIG. 18(a) depicts a typical ultrasonic sensor signal. FIG. 18(b) depicts a typical passive infrared sensor signal. Note that neither of the individual sensor technologies have a separate activation threshold. FIG. 18(c) depicts the sensor's composite occupancy estimator signal and the sensor's composite activation threshold level. When the two individual sensor signals are combined using an arithmetic sum function, the sensor's composite occupancy estimator signal suggests that occupancy has been detected whenever its value exceeds the sensor's composite activation threshold level. FIGS. 18(d)-(f) depict a more complex situation. FIG. 18(d) depicts a weak ultrasonic sensor signal. FIG. 18(e) depicts a typical passive infrared sensor signal. FIG. 18(f) depicts the sensor's composite occupancy estimator signal and the sensor's composite activation threshold level. When the two individual sensor signals are combined using an arithmetic sum function, the sensor's composite occupancy estimator signal suggests that occupancy has again been detected. It should be noted that the prior art multi-technology occupancy sensor was unable to detect occupancy in this situation (FIGS. 16 (d)-(f)).
An example of a more complex occupancy function of two variables is the function
f(x,y)=min (1, (K.sub.x *xP+K.sub.y *yP).sup.1/p)
This function produces what is known as the Yager Union decision surface. A condition may occur where either sensing technology alone assures reliable detection of occupancy when its associated occupancy estimator signal occurs at a relatively high level. Additionally, the occurrence of moderate levels of both the occupancy estimators together may be a reliable indication of occupancy. In this case, a composite occupancy estimator is required that indicates occupancy when the component occupancy estimators occur together at relatively low levels, or individually at relatively high levels. An example of the given function with K.sub.x =K.sub.y =1 and p=2 is shown in FIG. 19. This function combines the component occupancy estimators to form a composite occupancy estimator as desired.
A condition may occur that requires a relatively high degree of confidence in the occupancy estimate, and where the component occupancy estimators are generally reliable when taken together, but where the component occupancy estimators are generally unreliable taken alone. In this case, a composite occupancy estimator is required that indicates occupancy with a moderate level of indication from both component occupancy estimators, but rejects even large signals from the component occupancy estimators individually. An example of the given function with K.sub.x =K.sub.y =1 and p=0.5 is shown in FIG. 20. This function combines the component occupancy estimators to form the desired composite occupancy estimator.
Other functions may also be acceptable to combine the individual component occupancy estimators to produce a composite estimator signal, and the specification of theses exemplary functions does not limit the scope of the invention. In addition, although the exemplary functions are described using two occupancy estimators produced by two occupancy sensing technologies, this is due primarily to the ease of illustration using two estimators. However, it will be understood that a composite occupancy estimator may be derived from any number of individual occupancy estimators. For example, PIR, ultrasonic and microwave sensing technologies may be used to produce three individual component occupancy estimators that are then combined using a desired function (for example, algebraic sum, weighted sum, Yager Union, and the like) to produce a composite occupancy estimator.
The Yager Union function requires a significant computational resource to evaluate analytically, more than may be available in real time from a cost effective microprocessor. FIGS. 21 and 22 depict the same functional relationships as in FIGS. 19 and 20, but expressed in a discrete tabular form for fast efficient table lookup. In order to achieve different functional characteristics, the microprocessor may have multiple such tables to choose from, corresponding to different occupancy environment situations. These tables may then be selected adaptively or before hand. Additionally, the tables may be dynamic, and periodically recalculated according to the adaptation rules of the sensor automatic adjustment algorithm. This is possible, even with a limited computational resource, because the adaptation is an occasional low speed process, unlike the function evaluation, that must be executed at a high repetition rate in order to achieve fast response to occupancy signals. The Yager Union function is offered as an exemplary embodiment of the table look up aspect of the present invention, it will be understood that other mathematical functions used to calculate the composite signal may also be stored as a table.
The present invention may also include a number of adaptive features. When enabled, the following sensor variables are adapted as a function of time: ultrasonic sensor lights ON sensitivity setting; ultrasonic sensor lights OFF sensitivity setting; infrared sensor sensitivity setting; timer delay setting, infrared noise level estimate, smallest ultrasonic motion level estimate and a time between motion estimate.
The adaptation of the sensor's various sensitivity settings is generally referred to automatic sensitivity adjustment (ASA). The adaptation of the sensor's timer delay setting is generally referred to as automatic timer adjustment (ATA).
At power on reset, the various ultrasonic and infrared sensitivity settings and the timer delay setting are initialized according to the user-specified potentiometer settings (FIG. 3, 316, 317, 319). Automatic adjustments to the ultrasonic and infrared sensitivity settings and the sensor's timer delay setting are implemented as deviations from the user-specified potentiometer settings. As a result, any change to a potentiometer value results in an immediate change to the sensor adapted characteristic. The sensor continues to adapt based on both the potentiometer setting and the adapted adjustment value.
To accomplish ASA, a variety of situations automatically modify the sensor's ultrasonic, PIR and acoustic sensitivity settings, including: false-on event, false-on-stay-on event, false-off event, lights stay on-with no major occupancy event, lights stay on-with major occupancy event, occupancy cycle detection and utilization adjustment, infrared noise level adjustment, acoustic background noise level adjustment and smallest ultrasonic motion level adjustment. ASA can be accomplished in single technology or multiple technology sensors.
A false-on event occurs when the sensor's load(s) are initially de-energized, energized for a short period of time (for example, 0-8 seconds longer than timer delay setting), and then de-energized again. FIG. 23 describes a false-on adaptation event. The timer setting in this example is 15 minutes, occupancy is detected at 2301 only for a very short period of time (less than 15 seconds) and the load is energized. Because occupancy was not detected for 15 minutes after load was initially energized, lights turn off at 2302 and sensitivity is reduced (2303). A false-on event causes the sensor's ultrasonic, PIR, microwave, or acoustic sensitivity to be reduced.
A false-on-stay-on event occurs when sensor's load(s) are initially de-energized then energized for multiple timer settings without the occupancy signal ever exceeding activation threshold. FIG. 24 describes a false-on-stay-on adaptation event for a sensor that is sensing occupancy incorrectly through an open door into a hallway. Every time someone passes by the door, occupancy is detected. In this example, the timer is set at 15 minutes, occupancy is detected 2405, timer is reset and the load energized. Before the timer expires, occupancy is again detected 2406, 2407, 2408 and 2411 as someone passes by the door. After each occurrence, the timer is reset. This pattern of detecting occupancy without the signal exceeding the activation threshold 2402 continues for more than three timeout periods or 45 minutes 2404 and the sensitivity 2403 is reduced. Now when persons pass by the open door 2409 the sensor does not detect occupancy and the load is then de-energized 2410, 15 minutes after occupancy was last detected.
A false-off event occurs when the sensor's load(s) are initially energized when the occupancy signal is greater than the activation threshold 2801, then de-energized for a short period of time (for example, 0 to 15 seconds), and then energized again. FIG. 25 describes a false-off adaptation event for a sensor controlling room lights. The sensor detects occupancy at 2502 and the is timer is reset to 15 minutes. At 2511, the timer expires and load is de-energized. The occupant then reactivates the lights 2507 with occupancy signal 2503 within 15 seconds of load(s) de-energizing. At 2507, the sensitivity and timer are increased to 2512 and 2506 respectively. The timer is temporarily increased to 80 (5+60+15=80) minutes at 2506 (for example, 5 is the increase in timer value after false-off event, 60 is the temporary additional increase in timer value for the next eight hours and 15 minutes was the timer value before false-off event). Occupancy detection 2504 resets the timer and after 80 minutes, the load(s) de-energize at 2508. Eight hours after false-off event occurred 2509, the timer is reduced to 20 minutes 2513. When occupancy 2505 is next detected at 2510, the timer is reset to 20 minutes and load(s) energized.
Lights stay on-no activation event occurs when load(s) are energized constantly for multiple timer settings and occupancy signals never exceed activation threshold. FIG. 26 describes a lights stay on-no activation adaptation event. The sensor detects major occupancy 2603 when the occupancy signal exceeds the activation threshold 2602 (for example, a person just stood up and left the area), the timer is reset and the load(s) remain energized. The occupancy threshold is constantly exceeded for the next 120 minutes by small disturbances 2604 (for example, air noise, sensor electrical noise, etc . . . ) with no signals exceeding 2602 activation threshold. Sensitivity 2605 is reduced and now these small disturbances don't reset the timer and the load de-energizes at 2607.
Lights stay on-with major occupancy event occurs when load(s) are energized constantly for multiple timer settings and occupancy signal exceeds activation threshold. Several conditions can cause this event for example, the sensor is mounted close to a powerful air register, sensor mounted close to door and sensor pointed in the wrong direction and sensor mounted in correct application but sensitivity is way out of adjustment. FIG. 27 describes a lights stay on- with major occupancy adaptation event. Occupancy signal 2703 exceeds occupancy threshold 2701 and load(s) are energized at 2705. In quick succession, occupancy signals 2704, 2707, 2708 and 2709 occur with 2704, 2707 and 2708 crossing the activation threshold 2702. Quick succession of occupancy signals constantly reset timer 2706. Even in very active occupancy areas where sensor is adjusted and applied correctly, signal 2710 will settle down for short periods of time.
Using only a single technology sensor, for example ultrasonic, it is difficult to differentiate a very active room with lots of people verses a sensor mounted close to a powerful air register. However if multiple technologies are used, a much better opportunity is provided to adapt correctly, identify if the sensor is mounted close to a powerful air register or identify incorrect sensor application. Two examples are presented showing a dual technology ultrasonic and PIR sensors. In both cases the room is unoccupied. One sensor is mounted correctly with a very high initial ultrasonic sensitivity setting and in the other example the sensor is mounted close to a powerful air register.
FIG. 28 shows an expanded composite occupancy signal 2803 for ultrasonic and PIR detectors. Ultrasonic signal 2804 is very large and provides most of the signal contribution if an algebraic sum is used to combine the two signals. PIR signal 2805 is very quiet with very little signal variance. Since PIR is much more immune to air turbulence than ultrasonic it can be concluded that either the sensor is mounted close to an air register or the ultrasonic sensitivity is adjusted very high.
FIG. 29 shows an example algorithm to reduce the sensitivity of either PIR or ultrasonic signals. In 2901 it is determined if the composite signal exceeds the activation threshold for multiple timer settings. If this condition is satisfied, it is determined if the PIR and US signals are occasionally correlated (2910). If the sensor detects strong occupancy by both technologies at the same time, the area could be occupied. If signals do not correlate, it is determined which signal contributes most to composite occupancy signal (2902). In this example, ultrasonic sensitivity is reduced (2903). If ultrasonic sensitivity is reduced by more than 15% full scale (2905), the sensor is too close to a powerful air register and the sensor sends strong air flow status (2906). This is true because at low ultrasonic sensitivity levels it's coverage pattern is much smaller than the PIR pattern. Also, the PIR pattern is very dense and strong PIR occupancy signals would be received if the ultrasonic pattern area was occupied.
In situations where air flow is moderate or non-existent, after a couple of ultrasonic sensitivity reductions the composite signal fails (2901). The first time decision block 2907 is entered, the sensor is most likely in a "Lights stay on-no major motion detected" event. In block 2909, it is determined that ultrasonics are the major signal contributor and ultrasonic sensitivity is reduced in 2912. If ultrasonic sensitivity is less than 15% full scale (2913), the sensor is close to a powerful air register and the sensor sends strong air flow status 2906. If ultrasonics sensitivity is greater than 15% of full scale (2913), the loop is resumed until an error occurs (2906) or the sensor is properly adapted (2908).
It can also be determined if the sensor is mounted incorrectly or put into an unacceptable application and flash the appropriate status information. FIGS. 30(a) and (b) show two applications where sensors 3001 and 3004 are improperly applied.
Sensor 3001 is mounted looking out the door in room 3002. As people walk past the door in hallway 3003 the lights will be tripped on. A variety of sensitivity reduction techniques, as previously discussed with reference to FIGS. 23-29, will significantly reduce sensitivity. Working at desk 3006 will not maintain electrical load(s) and will periodically de-energized. The sensitivity will increase from "false-off" adaptation event. When the space is unoccupied, the sensitivity will decrease. This cycling will be detected and an application error status message will be displayed.
Sensor 3004 is mounted in a very large room 3007. This room is much larger than the coverage pattern sensor 3004 can provide. Working at desk 3005 will increase sensitivity from "false-off" adaptation events. Eventually, sensor 3004 sensitivity will be increased through a variety of sensitivity reduction techniques previously discussed for example some of these could be "false-on", "false-on-stay-on" or "lights stay on-no major motion" adaptation events will significantly reduce sensitivity. This cycling will be detected and an application error status message will be displayed.
The occupancy cycle detection and utilization adjustment modifies the sensor's ultrasonic and infrared sensitivity settings based on the historical usage patterns of the area controlled by the sensor. FIG. 31 is a histogram that depicts the occupancy cycle detection and utilization measurement scheme of the present invention. The various sensitivity settings are slightly increased during time periods when the area is normally occupied. Similarly, the various sensitivity settings are slightly decreased during time periods when the area is normally unoccupied. FIG. 32 is a simplified occupancy cycle detection and utilization task for the present invention in flowchart form. In operation, when it is time to run the task (3201), the activity level for the immediately preceding time period is computed (3202). Then, the activity level is saved (3203). This preferably takes the form of an accumulation of newly computed activity with activity that has been computed and accumulated previously for the same time period. Thus, a histogram like that shown in FIG. 31 is built. Then, the sensor sensitivity is adjusted (3204) for the upcoming time period based upon the level of activity stored in the histogram for the next time period.
In accordance with one aspect of the present invention, a learning curve is used to adapt the sensor sensitivity and timer settings as a function of the elapsed time since the initial power on reset state, toggling of the sensor's Adaptation Reset switch 309 (FIG. 3) or continuous cycling in one of the previously discussed adaptation events. In general, each sensor technology has its own separate (and potentially different) learning curve. A typical learning curve plot is shown in FIG. 33. The shape of the learning curve implies that greatest incremental sensitivity and timer changes occur at the beginning of the learning curve cycle. As the elapsed time increases in the learning curve cycle, the adaptive behavior will gradually decrease until very small changes in sensitivity and timer settings will occur.
If the sensor's environment experiences a significant change long after power on reset state, the sensor may require some adjustment to operate properly in its new environment. In this unusual situation, the use may toggle the sensor's Adaptation Reset switch 309 (FIG. 3). This will reset all of the sensor's adaptive algorithms and reset to the beginning of the learning curve plot in FIG. 24. The sensor will then begin to adapt to its new environment according to its various learning curves.
Controlling very large spaces with a minimum number of sensors or controlling areas that have unusual shapes can easily be accomplished by adding an acoustic detector. This new application method only enables the acoustic detector just before and after the load(s) are de-energized. This eliminates many of the disadvantages that dual technology PIR/Acoustic and Ultrasonic/Acoustic sensors have. When acoustic occupancy detectors are used alone or in dual technology configurations, load(s) are energized for much longer time periods because of incorrect occupancy detection due to external noises such as radios, fans, passing automobiles etc. In these devices, the acoustic detector is enabled continuously. In accordance with the present invention, as illustrated in, FIG. 34, the acoustic detector is only enabled 15 seconds before load(s) are de-energized 3402 and 15 seconds after. During the first 15 seconds, the acoustic sensor establishes an average acoustic occupancy threshold. After the load(s) are de-energized 3403 the acoustic detector is enabled for an additional 15 seconds 3404 and listens for acoustic occupancy such as human speech. If the acoustic occupancy threshold is exceeded, the load(s) are energized 3405 and start "false off" event adaptation timer begins 3406. The preferred embodiment of this acoustic detection method is to complement a single technology ultrasonic or PIR sensor or dual technology ultrasonic and PIR sensor.
Traditional analog PIR sensor manufactures provide a manual PIR gain adjustment or fix the gain during the manufacturing process. Providing a manual adjustment may lead to misadjustment. For example, the gain may be erroneously set so large that the sensor occasionally false trips on internally generated noise. Manufactures that provide fixed gain are not utilizing full detector sensitivity to allow for device tolerances in the PIR detector, filter and amplifier stages. FIG. 35 provides a method of recording PIR signal levels 3501 when load(s) are de-energized and areas unoccupied. New peak PIR values are stored 3503 and a new occupancy threshold offset can be calculated. Using this method, maximum PIR sensitivity settings can be calculated.
In accordance with one aspect of the present invention, the process of recording PIR signal levels while the area is unoccupied facilitates the estimation of the noise portion of the PIR signal. By determining the level of noise present in the PIR signal in an unoccupied space, an adaptation algorithm determines an optimal setting for the PIR sensitivity that provides maximum possible sensitivity while providing ample immunity against false trips due to the noise.
In some situations, the sensitivity setting of the PIR sensor may be either too high or too low. Improper settings can result from either installer error or over-adaptation. Seasonal temperature changes can also affect the sensitivity of a PIR sensors. FIG. 70 demonstrates the seasonal variation of a PIR sensor. The sensor in this example has a fixed threshold 7001. During the winter months, the sensor has a low noise level 7002 well away from the threshold 7001. However, the noise level 7002 approaches the threshold 7001 with the onset of summer until the noise 7002 is so close to the threshold 7002, as shown in area 7003, it may begin to cause false trips.
A noise level too close to the threshold is undesirable. The composite occupancy estimator technique of the present invention complicates the choice of the best choice of threshold. Not only is it desirable to keep the PIR noise level a distance away from the threshold, the PIR noise should also be small enough so that the noise does not contribute to the composite occupancy estimator signal. Alternatively, the PIR threshold should not be set so high that the only PIR motion hits that contribute to the composite occupancy estimator are the extremely large ones. FIG. 71 demonstrates this point graphically. If the threshold is set too low, then the noise will be in the region of marginal motion 7101. It is preferable that the PIR noise level is located beneath the region of minor motion 7102 and above the region of reduced PIR sensitivity 7104.
In one exemplary embodiment of the invention, the microcontroller employs a PIR noise estimation algorithm that uses a statistical order filter (SOF) also referred to as a rank order filter to perform the estimation of the PIR noise level of the PIR output (output of block 4803, FIG. 48). In order to assure estimation of only PIR noise and not real PIR signals, noise estimation is performed only when the occupied flag is not set (determined by block 5201, FIG. 52). The SOF algorithm operates on a fixed number of samples of a signal. The algorithm sorts the array of samples in ascending or descending order and then outputs a weighted sum of the samples in the sorted array.
In the exemplary embodiment, the output of the SOF is chosen to be the maximum value of the signal because such a SOF can be efficiently implemented in a microcontroller. However, other choices would also be acceptable. Furthermore, filtering algorithms other than a SOF may be used. For example, noise estimation may be accomplished through the use of digital filters, integration, envelope detection, or simple averaging.
FIG. 72 is a flow chart of an exemplary algorithm that updates the PIR threshold in such a way that the noise level will move closer to the proper noise region 7103. In the exemplary embodiment, the microcontroller executes this routine every time occupancy begins. In particular, the estimation algorithm is invoked when the load is energized in step 5204 (FIG. 52) or in step 5304 (FIG. 53), and operates on signal samples taken before the load is energized. Estimated noise value is checked at 7201. If the estimated noise value is below the proper noise region 7103, then the algorithm increases the PIR sensitivity 7202. Alternatively, if the estimated noise value is above the proper noise region 7103, then the algorithm reduces the PIR sensitivity 7203. Eventually, the PIR sensitivity will be set to a value, that will place the noise in the proper region.
Ultrasonic detection technologies are most susceptible to strong air flow when turning loads on, as compared to steady state (maintaining loads on) or turning lights off. Overcoming false occupancy detection when air flow turns on or creating a "Air Puff" detector would be a major improvement in ultrasonic sensing technology. The most common way of improving ultrasonic performance is adding PIR or microwave detectors. While use of these additional sensing technologies improved overall sensor performance, they did not eliminate the deleterious effects of air flow is on the ultrasonic detector.
Sensors mounted next to an HVAC register can acoustically detect several things when air flow begins. First, the compressor turns on, second the air flow will typically exhibit a unique sound signature during startup and continuous operation. This unique sound signature may be a function of the building configuration, and the structures of the air diffuser and air duct.
The closer the sensor is to the HVAC register, the greater the affect of air flow. An acoustic detector designed to monitor air flow sounds can be used to complement ultrasonic detector systems. FIG. 36 illustrates signal levels for a room that is empty, quiet and unoccupied when air flow turns on 3601, the acoustic sound energy 3602 is detected by the acoustic detector before the ultrasonic detector is affected. The ultrasonic detector detects signal 3603 at 3604. This air flow on transition from 3604 to 3605 is difficult for ultrasonic detectors to reject. The Doppler shift spectrum in this region is broad and robust. Using the acoustic detector to temporarily inhibit or reduce ultrasonic sensitivity from 3601 to 3606 eliminates the ultrasonic detector from creating a "false-on" event.
The sensor's timer supports a number of special modes intended to address a variety of user needs such as to allow the sensor's installer to force a short time-out period for purposes of sensitivity adjustment, to avoid short timer settings in normal operation, and to avoid a long load-energized period with no motion if the sensor's circuit breaker is cycled. If the sensor is powered up with the timer adjustment potentiometer (319, FIG. 3) in the fully-CCW position, or if the timer adjustment potentiometer is adjusted to the fully CCW position after being at another setting, the sensor enters an 8-second timeout test mode. If the timer is adjusted from the fully-CCW position to another setting, the timer exits the 8-second timeout test mode and returns to normal operation. As an aid to the installer, the flowchart of FIG. 37 illustrates what happens if the sensor's timer is left in minimum timeout (8-second) or maximum timeout (32 minutes) setting for more than an hour. If the timer is at minimum (full CCW) the timer is set at 7 minutes (3702, 3704 and 3705). If the timer is at maximum (full CW) the timer is set at 15 minutes (3703, 3706 and 3707).
To accomplish ATA, a variety of situations modify the sensor's timer delay setting.
A false off event shown, for example, in FIG. 25, will slightly increase the timer delay setting. In addition, as shown in FIG. 38, the sensor also incorporates a minimize energy usage algorithm that checks for occupancy within the first two minutes after power up 3802. If motion is detected, keep lights on and use normal timer setting 3804. If motion is not detected, turn lights off 3803. This feature solves the problem if power is interrupted by utility failure or a general area shut down during the evening and power is subsequently restored. Conventional sensors power on and begin a complete time-out cycle, typically leaving the lights on needlessly for twenty minutes.
A flowchart of the main software routine of an exemplary embodiment of the present invention is shown in FIG. 39. In practice, the flowcharts depicted in FIGS. 32 and 39-69 are encoded in an appropriate programming language and loaded into the program memory of digital microcontroller 229 (FIG. 2), to cause digital microcontroller to control the operation of the apparatus of the present invention to perform the methods of the present invention.
Referring to FIG. 39, upon power on reset (3901), the digital microcontroller's various timer, counter, serial port and high speed 8-bit A/D converter peripherals are set up and the various sensor variables are initialized (3902). The microcontroller 229 then enters the occupancy sensor's main loop (3903). The sensor's main loop initially checks to see if either the manual interrogate status flag or the two minute status flag is enabled (3904, 3905). If so, the appropriate status report task is executed. The sensor then checks to see if a sensor false-on event or false-off event has occurred (3906, 3907). If either event has occurred, the appropriate correction task is executed. Finally, at the bottom of the main loop, the microcontroller's watch dog timer (WDT) is reset (3908). This timer is used to reset the microcontroller 229 to the power on reset state in the event of a software malfunction.
Thus, the main routine shown in FIG. 39 includes the following major software tasks that are described in detail in the figures indicated: hardware and system variable initialization task (FIG. 40); interrogation status report task (FIG. 41); two minute status report task (FIG. 42); false-on correction task (FIG. 43); and false-off correction task (FIG. 44).
A flowchart of the hardware and system variable initialization task for the present invention is shown in FIG. 40. The hardware and system variable initialization task is executed on power up to configure the sensor for normal operation.
Initially the system on-chip and off-chip hardware is configured (4001). Specifically, the power on reset and brown out reset bits are reset. Global interrupts are disabled. The A, B and C I/O ports are cleared. The option register is cleared. The timer prescalar is set for 1:16 and is then assigned to the real time clock/counter (RTCC) peripheral. The various I/O port data direction registers are cleared and then set as required for normal operation. The RTCC timer count down value is set. All peripheral interrupts are disabled. The second CCP timer is set up. Finally, the serial EEPROM is initialized.
Next, the system variables of the occupancy sensor are initialized (4002 ). Specifically, a stack pointer is set up. The various ultrasonic gains, thresholds and offset values are initialized. The various infrared offset values are initialized. The minute and second counters are initialized. The various motion flags and motion counters are initialized. The primary sensor tasks are enabled except for the sensor task. The occupied bit is set. The various sensor LED registers are initialized. The RTCC interrupts are enabled. A variable delay is executed until the serial EEPROM initialization has been completed. A single EEPROM location is read to determine if the EEPROM has had sensor information previously stored in it. If it has, the contents of the EEPROM are read and the various sensor variables are restored. If the EEPROM has not had sensor information previously stored in it, the current sensor variables are stored in the EEPROM. Finally, the sensor task is enabled.
A flowchart of the interrogation status report task for the present invention is shown in FIG. 41. The interrogation status report task displays the status of the sensor in response to a user's interrogation of the sensor. If a status task is not currently in progress (4101), and if the LED task is not busy (4102), ultrasonic sensor 102 (FIG. 2) is checked to determine if wind noise is present (4103). Control then passes to block 4104, and then on to block 4105.
A flowchart of the two minute status report task for the present invention is shown in FIG. 42. The two minute status report task displays the status of the sensor at periodic two minute intervals. If a status task is not currently in progress 4201), and if the LED task is not busy (4202), control passes to block 4203 where the sensor's status is stored in the sensor status message word and the sensor's status digit counter is initialized. The LED blink preamble time is stored in the LED counter and the LED is turned off. The LED task is enabled and the LED status state initialized. The LED status flash mode is set and the LED status in process flag is set. Finally, the TwoMinStatTaskEnable flag is set.
A flowchart of the false-on correction task for the present invention is shown in FIG. 43. The false-on correction task attempts to correct sensor false-on events by increasing the threshold required for the detection of motion. Initially a false-on event counter is incremented (4301). Then the sensor determines which technology dominated the false-on event (4302). A new adaptive learning coefficient is computed and the appropriate threshold offset is adjusted (4303, 4304).
A flowchart of the false-off correction task for the present invention is shown in FIG. 44. The false-off correction task attempts to correct sensor false-off events by decreasing the ultrasonic threshold required for the detection of motion. A new adaptive learning coefficient is computed and the appropriate threshold offset is adjusted.
A flowchart of the occupancy sensor's interrupt software routine for the present invention is shown in FIG. 45. Upon interrupt, the contents of the 8-bit W accumulator register and Status register in microcontroller 229 are saved (4501). The timer/counter peripheral that generated the interrupt is reset and the microcontroller's interrupt flag is cleared (4501). The sensor task is executed if it is enabled (4502). The various occupancy sensor counters are updated (4503). The sensor's timer delay is adjusted to minimize energy consumption (4504). The multiplexer (MUX) task is run if it is enabled (4505). The ultrasonic sensor's sampling point is adapted (4506). Next, the LED task is run if it is enabled (4507). The relay suppress task is then run if it is enabled (4508). The second and minute tasks are then executed if enabled (4509, 4510). If the interrogate task is enabled it is executed (4511). The sensor's EEPROM task is then executed (4512). Finally, the microcontroller's watch dog timer (WDT) is cleared and the W register and the status register contents are restored (4513).
Thus, the interrupt routine shown in FIG. 45 includes the following major software tasks that are described in detail in the figures indicated: sensor task (FIGS. 46-54); update counters task (FIG. 55); adjust timer to minimize energy usage task (FIG. 56); MUX task (FIG. 57); adapt ultrasonic sampling point task (FIG. 58); LED task (FIG. 61); relay suppress task (FIG. 63); second task (FIG. 64); minute task (FIG. 66); interrogate task (FIG. 68); and EEPROM task (FIG. 69). Each of these interrupt routine software tasks is described in detail below.
The sensor task inputs and processes data from each of the various sensor technologies. A flowchart of the sensor task of the present invention is shown in FIG. 46. First the ultrasonic task is executed (4601, shown in detail in FIG. 47). Next, the PIR task is executed if it is the proper time to run it (4602, shown in detail in FIG. 48). If the dual tech mode flag is set (4603) according to the state of user switch 307 (FIG. 3), the dual technology HIGH CONFIDENCE task is run (4604, shown in detail in FIG. 52). Otherwise, the dual technology HIGH SENSITIVITY task is run (4605, shown in detail in FIG. 53). Finally if the dual technology sense flag is set (4606), the average time between motion task is executed (4607, shown in detail in FIG. 54).
The ultrasonic task inputs data from the ultrasonic sensor and processes that data using an ultrasonic DSP algorithm. An exemplary ultrasonic DSP algorithm is shown schematically in FIG. 47. Although the ultrasonic DSP algorithm is represented as hardware components in FIG. 47, the algorithm is actually embodied in software. Referring to FIG. 47, first, the ultrasonic data is sampled using a sampling frequency f.sub.S of approximately 1,000 Hz. The ultrasonic data is initially passed through a low order digital highpass filter 4701. The digital ultrasonic data is then passed through a low order digital bandpass filter 4702. Next, the absolute value of the ultrasonic data signal is computed (4703) and passed through a variable gain stage 4704. The gain of stage 4704 is derived from the setting of user-specified ultrasonic sensitivity potentiometer 316 (FIG. 3). The output of the variable gain stage 4704 is then decimated by factor of 64 (f.sub.S =.about.16 Hz) 4705, and passed through a low order digital lowpass filter 4706. Low pass filter 4706 has a cutoff frequency of approximately 2.5 Hz. The output of Low pass filter 4706 is multiplied by a variable factor, 4707, depending on the presence of strong air flow. If the strong air flow option switch 311 (FIG. 3) is ON indicating the presence of strong air flow, the signal is multiplied by a factor of 1.0. Otherwise the signal is multiplied by a factor of 0.75. A fixed offset value 4708 is added to the result that is then fed into the negative (-) input of a clipping asymmetric ramp detector 4709. The positive (+) input of ramp detector 4709 is connected to the output of the variable gain stage 4704. Whenever the positive input of ramp detector 4709 is greater than the negative input, the output of ramp detector 4709 increases by 512. Conversely, whenever the negative input of ramp detector 4709 is greater than the positive input, the output of ramp detector 4709 decreases by 128. The output of the ultrasonic DSP block is passed through a digital pulse stretcher to increase the time duration of the peaks in the ultrasonic signal. This technique slightly increases the probability of detection of small time duration events by stretching the ultrasonic signal peaks in time until the infrared sensor responds to the corresponding event, and is described in detail below with reference to FIGS. 49, 50 and 51. After pulse stretching, the output of ramp detector 4709 is passed to the sensor's fusion detection occupancy algorithm, as shown in more detail in the flow chart of FIGS. 52 and 53. Finally, the output of ramp detector 4709 is compared (4710) to a user-specified ultrasonic sensor activation threshold (4711). If the signal is above that threshold, the ultrasonic motion indicator(s) will be activated.
The PIR task inputs data from the infrared sensor and processes that data using a passive infrared DSP algorithm. An exemplary PIR DSP algorithm is shown schematically in FIG. 48. Although the PIR DSP algorithm is represented as hardware components in FIG. 48, the algorithm is actually embodied in software. Referring to FIG. 48. The PIR data is sampled using a sampling frequency f.sub.S of approximately 32 Hz. The PIR data is initially passed through a low order digital highpass filter 4801. The digital PIR data is then passed through a low order digital lowpass filter 4802. Next, the PIR difference signal is computed by subtracting the output of the PIR low pass filter 4802 from the output of the PIR high pass filter 4801. Next, the absolute value of the PIR difference signal is computed 4803. The absolute value of the PIR difference signal is then pulse stretched in a manner similar to that done with the ultrasonic signal, and passed to the sensor's multi-technology sensor fusion detector (FIGS. 52 and 53) to determine occupancy. Finally, the absolute value of the PIR difference signal is compared 4804) to a user-specified PIR activation threshold (4805). If this signal is above the threshold, the infrared motion indicator(s) will be activated.
As mentioned above with reference to FIGS. 47 and 48, pulse stretching is applied in both the ultrasonic and PIR DSP algorithms. This is desirable because valid occupancy signals from multiple technologies may occur nearly, but not exactly simultaneously. For instance, ultrasonic motion detectors are preferentially sensitive to motion along the direction to the sensor, while infrared are more sensitive to motion perpendicular to the direction to the sensor. Under normal circumstances, the movements of an occupant of the controlled space may vary between the directions toward and perpendicular to the sensor, resulting in occupancy estimator signals that peak at different times.
The need for pulse stretching is illustrated in FIG. 49. Without pulse stretching, motion signals from detection technology 1 and 2 occur at different times and are generally not coincident in time. In this case a composite occupancy estimator will not indicate occupancy, despite the near coincidence of signals from the individual detection technologies. The result of stretching these signals is shown in FIG. 50. In this case the stretch of signal 1 (5001) brings it into temporal coincidence (5005) with signal 2 in the first case, and the stretch of signal 2 (5002) brings it into temporal coincidence (5006) with signal 1 in the following case. Now the signals overlap, and a composite occupancy estimator will indicate occupancy.
Pulse stretching as shown in FIG. 50 generally requires a circular buffer large enough to store a number of samples equal to the stretch time period divided by the sample rate. It also requires a search algorithm to find the maximum value within the buffer, that becomes the current stretched signal value. An alternative implementation is to recognize the local maximum of a signal, and stretch only the maximum value. This implementation requires a lesser computational resource, and the storage of only two values, the signal maximum value, and its stretch ending time. The result of this pulse stretching implementation is shown in FIG. 51.
The HIGH CONFIDENCE task implements the sensor's HIGH CONFIDENCE occupancy detection mode. A flowchart of the sensor's HIGH CONFIDENCE task is shown in FIG. 52. If the area covered by the occupancy sensor is not currently occupied (5201), a composite occupancy estimator is computed using an arithmetic sum formula (5202). If the resulting value is above the predetermined composite activation threshold (5203), the sensor's electrical load(s) are energized and the DualTechSense flag is set (5204).
If the area covered by the occupancy sensor is currently occupied, a composite occupancy estimator is computed using an arithmetic sum formula (5205). If the resulting composite occupancy estimator value is above the predetermined composite maintenance threshold (5206), the sensor's relay countdown timer is reset and the DualTechSense flag is set (5207).
The HIGH SENSITIVITY task implements the sensor's HIGH SENSITIVITY occupancy detection mode. A flowchart of the sensor's HIGH SENSITIVITY task is shown in FIG. 53. Initially, an composite occupancy estimator is computed using an arithmetic sum formula (5301). If the resulting value is above the predetermined composite maintenance threshold (5302), the sensor checks to see if the area covered by the sensor is currently occupied (5303). If the area covered by the occupancy sensor is not currently occupied, the sensor's electrical load(s) are energized and the DualTechSense flag is set (5304). If the area covered by the occupancy sensor is currently occupied, the sensor's relay countdown timer is reset and the DualTechSense flag is set (5305).
The average time between motion task estimates the average time between motion. A flowchart of the average time between motion task of the present invention is shown in FIG. 54. If it is time to run this task (5401), the previous estimate of the average time between motion is updated and saved (5402). This information is used as a cue for the various adaptive sensor algorithms.
The update counters task updates the primary three byte interrupt counter used in the occupancy sensor. A flowchart of the update counters task of the present invention is shown in FIG. 55. The low byte of the counter is first incremented (5501). If zero (5502), the middle counter byte is incremented (5503). If zero (5504), the high byte of the counter is incremented (5505). If the high byte is zero (5506), a counter roll over flag is set (5507).
The adjust timer to minimize energy usage task implements the power up timer adjust algorithm. A flowchart of the adjust timer to minimize energy usage task of the present invention is shown in FIG. 56. The adjust timer to minimize energy usage task acts to reduce the time out on an ongoing basis. The amount of reduction depends upon the current value of the time delay potentiometer 319 (FIG. 3). In the illustrated embodiment, the amount of timer reduction is done according to the following formulas:
If time out.gtoreq.60 minutes, reduce by 30 minutes/4.66 hours,
If time out.gtoreq.30 minutes, reduce by 5 minutes/4.66 hours,
If time out.gtoreq.15 minutes, reduce by 1 minutes/4.66 hours,
If time out.gtoreq.10 minutes, reduce by 30 seconds/4.66 hours,
Otherwise, reduce timer by 2 seconds/4.66 hours
Initially, the timer delay adjustment value is reduced (5601). The timer delay adjustment value is then clipped at zero (5602).
The MUX task reads the status of the various user-selected mode option switches (discussed in detail above with reference to FIG. 3), and saves the results in the various sensor variables and flags. A flowchart of the MUX task is shown in FIG. 57. The MUX task initially sets the A and B MUX controls to cause MUX 304 (FIG. 3) to read the ultrasonic sensitivity potentiometer setting and the ASA switch setting (MUX=00). On the next pass, the MUX 304 reads the PIR sensitivity potentiometer setting and the ATA switch setting (MUX=01). On the next pass, the MUX 304 reads the timer pot setting and GSB switch setting (MUX=10). Finally, on the fourth pass, the MUX 304 reads the photo cell potentiometer setting and the strong air flow switch setting (MUX=11). This sequence of four MUX states is then sequentially repeated on a periodic basis.
The adapt ultrasonic sampling point task adjusts the ultrasonic sampling point to prevent the loss of motion information due to ultrasonic analog front end receiver signal saturation. A flowchart of the adapt ultrasonic sampling point task of the present invention is shown in FIG. 58. The sample point setting algorithm adjusts the sample point to achieve a receive carrier amplitude level at the sample point that is as close to zero crossing (VREF voltage) as possible. The sample point is adjusted by varying the timer value that sets the sample point within the carrier period.
When the sensor is initially powered, the sample point is adjusted to a default initialization value that is correct for the sensor in a nominal situation. The receive carrier sample point value is tested (5801), and if found to lie within a window about zero crossing (5802), no change to the sample point is made. If the receive carrier sample point value is found to be outside the desired window (5802), the sample point is precessed in discrete steps (5803), until the receive carrier sample point value is within the window(5804). The sample point is further perturbed to search out the value that is closest to zero crossing (5805). This becomes the synchronous demodulator sample point.
If it is impossible to find a sample point that results in a carrier value within the window (5806), the allowable window size is increased (5807), and the sample point search is made again. This process is repeated until the window has been increased to the point of receiver preamplifier saturation. In the event that the sensor is installed so close to a reflecting wall that there is no sample point that may be found that does not saturate the receiver preamplifier, the sample point search has failed, and the sensor reports this via the status reporting feature. Alternatively, the sensor may adjust the carrier send output level.
Specifically, when an ultrasonic sensor is installed within a small space, reverberant energy from the close walls may cause the sensor receive preamplifier to saturate. This saturation may be extreme beyond the ability of the sample point search algorithm FIG. 58 to find a satisfactory sample point within the granularity of the sample point timer adjustment. The correct response to this situation is to decrease the transmitter power, since the small space will be adequately covered. In another situation an ultrasonic sensor may be installed in a large space, but too close to an acoustic reflecting surface, resulting in receiver preamplifier saturation extreme beyond the ability of the sample point search algorithm to find a satisfactory sample point. The correct response in this situation is to decrease the transmitter power, unless this causes the sensor to emit insufficient acoustical energy to effectively survey the space. If this is the case, the sensor is unable to achieve a satisfactory drive level that both avoids preamplifier saturation and effectively surveys the space, so the sensor signals its inability to perform in a satisfactory manner by reporting a malfuction. This prompts service personnel to review the sensor installation and move it away from the acoustic reflector that is causing the saturation problem. This process is detailed in FIG. 59.
The receive carrier sample point value is periodically retested while the sensor is in operation. In the event that the sample point is recognized to be outside the window, and is found to lie consistently outside the window, indicating a change in the sensor mounting, the sample point search is reinitiated. The search results in a new smaple point.
FIG. 60 depicts the situation of waveform saturation. Time is shown on horizontal axis, and the waveform at the preamplifier output is shown as a function of time. Point 6001 on the waveform indicates the preamplifier output at a voltage below the positive saturatio limit. Point 6002 is the preamplifier output in saturation, where it cannot achieve the ideal waveform value of point 6003. Point 6004 is the ideal sample point at the zero crossing of the preamplifier signal. The region bounded by 6005 and 6006 is a voltage window around zero crossing within which the sample point should ideally lie to avoid the preamplifier saturation characteristic. The region bounded by 6007 and 6008 is the corrsponding window in time within which the sample point must lie in order to achieve a sample point voltage within region 6005.
The LED task determines the LED flash mode and then flashes the various sensor LEDs in the appropriate fashion. A flowchart of the LED task is shown in FIG. 61. In the exemplary embodiment there are four different LED flash modes: single flash mode, continuous flash mode, status flash mode, and null flash mode. The LED single flash mode flashes the LED on and off a single time. The LED continuous flash mode flashes the LED on and off continuously. The LED status flash mode flashes the current status of the sensor, and is discussed in more detail with reference to FIG. 62(a). The LED null flash mode may be used for testing.
The LED status task manages the overall operation of the LEDs in order to output the status of the various sensor registers and its overall status. A flowchart of the LED status task of an example of the present invention is shown in FIG. 62(a).
Each status register value and the overall status of the sensor are converted into a sequence of Morse code dots and dashes, or other coding scheme. The LED status routine keeps track of the various dot and dash characters and makes sure the complete status information is sent. An exemplary status report is shown with reference to FIG. 62(b). The status report takes the form of a series of short and long LED flashes that, in the exemplary embodiment are in Morse code. The report is preceded by a preamble 6201 and then by one or more data fields. For example, after the preamble, the general status (6202) of the sensor may be reported. General status my be, for example, normal, windy (for a windy installation), failed or any other status that is useful. In the example of FIG. 62(b), the general status is followed by the delay timer setting in minutes (6203) and the ultrasonic sensitivity setting (6204). Thus, in the example given, the status of the sensor is normal ("S"), the delay timer is set for 16 minutes, the ultrasonic sensitivity setting is 10 and the PIR sensitivity is set to 8. Other types of information may also be displayed, including by way of example, PIR sensitivity setting and microwave sensitivity setting.
As described below in detail with reference to FIGS. 66 and 68, status may be reported periodically, without user intervention, or it may be reported in response to user interrogation. In the former case, a shorted status report may be given (for example, only the general sensor status), whereas a more complete status report may be given in response to user interrogation. FIGS. 61 and 62 (a) and (b) illustrate various possibilities for using LEDs to provide status information on the sensor. The specific information provided may be of any useful nature, and is not limited to any specific type of sensor status information.
The sensor's relay suppress task disables the sensor's relay for a predetermined period of time. A flowchart of the relay suppress task of the present invention is shown in FIG. 63. This routine is used to prevent sensor false triggering by large voltage spikes from the sensor's relay.
Once the relay's predetermined suppression time has elapsed (6301), the relay task enable flag is set (6302),and the relay suppress task enable flag is cleared (6303).
The sensor's second task is run once a second. A flowchart of the sensor's second task is shown in FIG. 64. This task initially increments the seconds counter (6401). If the relay is currently on (6402), the relay task is executed (6403). If any user-specified mode options have changed (6404), the various auto-sensitivity adjust (ASA) variables are reset (6405). Finally, if the sensor's electrical load(s) are energized (6406), the energized load counter is incremented (6407). If the sensor's electrical load(s) are off, the de-energized load counter is incremented (6408).
The relay task decrements the sensor's relay counter. A flowchart of the relay task of the present invention is shown in FIG. 65.
After decrementing the sensor's relay counter (6501), the relay task turns off the sensor's relay if the relay counter has reached zero (6302). If the relay counter has not reached zero, the relay suppression task is ended. If the ASA task is enabled (6503), and the lights were previously on for about the same time as the sensor's timer delay setting (6504), the ultrasonic and PIR sensitivities are reduced and the false on correction task is enabled (6505).
The sensor's minute task is run once a minute. A flowchart of the minute task of the present invention is shown in FIG. 66. This task initially increments the minutes counter (6601). If it is time to run the two minute status task (6602), the two minute status task enable flag is set (6603). Finally, the initialize I/O port data direction registers task is executed (6604).
The initialize I/O port data direction registers task periodically resets the I/O ports on the digital microcontroller 229 (FIG. 2). A flowchart of the initialize I/O port data direction registers task is shown in FIG. 67. The Port A (6701), Port B (6702), and Port C (6703) data direction reregisters are periodically reset to ensure proper operation.
The sensor's interrogate task is used to command the sensor into a special mode of operation (for example, a status reporting mode) by using the motion sensing capabilities of the sensor and without the need for adding a mode selection switch. In the exemplary embodiment, the special mode is invoked by the detection of five successive hand waves by the user. A flowchart of the interrogate task of the present invention is shown in FIG. 68. This task is implemented as an eighteen state machine. The state counter (NGATEstate) is monitored at a rate determined by the interrogation counter STATcnt reaches zero (6801). If the state counter ever reaches a value of eighteen, the interrogation has been successfully requested (6802), so the bit allowing the sensor status to be displayed is set and the state counter is set to zero (6803). If the state counter is less than eighteen, the state of the USSense is used to modify the state counter. The two's place bit of the state counter is XOR'ed with the USSense bit (6804) to determine if the state counter should be incremented to the next even state. If the USSense bit contains the desired value, the state is incremented twice and the STATcnt is set equal to FF (6805). If the USSense bit doesn't contain the desired value, the state is evaluated (6806). If NGATEstate is even it is incremented once, and the STATcnt is set equal to FF(6807). This branch of program flow creates all of the odd states. If NGATEstate is odd, NGATEstate is set to zero (6808) By this method, the sensor may be switched from normal mode to status reporting mode using the motion detecting functions of the sensor, and without the need for adding a status mode selection switch.
The sensor's EEPROM task reads and writes data to the sensor's EEPROM. A flowchart of the EEPROM task is shown in FIG. 69. If the EEPROM is not busy (6901), and an EEPROM write task is scheduled (6902), the EEPROM task writes the various sensor variables to the sensor's EEPROM (6903), and then clears the EEPROM write task flag. If the EEPROM is not busy (6901), an EEPROM write task is not scheduled (6902), and an EEPROM read task is scheduled (6904), the EEPROM task reads the various sensor variables from the sensor's EEPROM and then clears the EEPROM read task flag (6905).
Although the description above contains many specifics, these should not be construed as limiting the scope of the invention but as merely providing illustrations of some of the presently preferred embodiments of this invention. For example, the sensor might use one or more subsets of the various detection technologies described above. Similarly, the sensor might use a different digital microcontroller, a different ultrasonic transducer or a different ultrasonic transducer frequency from the exemplary embodiment. Thus the scope of this invention and the method of operating the same should be determined by the appended claims and their legal equivalents and should not be limited to the example embodiments described above.
Table 3 provides a correlation between the attached source code appendix, and the flowcharts of FIGS. 39-69 and 72. The source code appendix is written in the assembly language, and is an example of programming for the digital microcontroller used in the exemplary embodiment to practice the present invention. Other programs may be used without departing from the scope of the invention.
TABLE 3______________________________________Source Code Label References for Primary Sensor TasksFigure Number Description of Routine Code Label______________________________________39 Main (Foreground) Routine main40 Hardware and Variable 02init Initialization Task41 Interrogation Status Report statussend Task42 Two Minute Status Report twominstatus Task43 False On Correction Task fixfalson28 False Off Correction Task fixfalsoff45 Interrupt (Background) inta Routine47 Ultrasonic DSP Task ultrasonictask48 Infrared DSP Task runPIR52 HIGH CONFIDENCE Task dtdecidenorm53 HIGH SENSITIVITY Task dtHSscheme55 Update Counters Task cntrupdt56 Adjust Timer to Minimize reducetimer Energy Usage Task57 MUX Task runMUXT61 LED Task LEDTASK62 LED Status Task statusled63 Relay Suppress Task suppress64 Second Task SECONDtask65 Relay Task relaytask66 Minute Task mcnt67 Initialize I/O Port Data setTrisRegs Direction Registers68 Interrogate handleSTAT69 EEPROM Task eepromwait72 IR Noise Estimate Adaptation INEMonitorPIRDetector______________________________________
__________________________________________________________________________; o2dt32.asm; COPYRIGHT (C) MYTECH CORP. 1996, 1997;carrier set 1;---------------------------------------------------------------------------------------------------------------------------------------; The 32Khz carrier;---------------------------------------------------------------------------------------------------------------------------------------timvala set H`067` ; 0.times.100-0.times.67=0.times.99 ; 999.837hz --> ; (fosc/4)/(12 instr+0.times.98*16) ; the factor of 16 comes from ; the use of the prescalertmr2val set H`04a` ; 75 decimal, this gives 32Khz carrier ; 32.768Khz=(fosc/4)/75include "o2dt.asm"LIST C=160,N=80,F=INHX8M,P=16C74,R=HEXTITLE "Omni II Dual Technology Sensor"SUBTITL "COPYRIGHT (C) MYTECH CORP. 1996, 1997";-----------------------------------------------------------; o2dt.asm;; Source code for the Multi-Tech sensor (N=2); Copyright (C) MYTECH Corp. 1996, 1997; 7-25-96 JJF;;-----------------------------------------------------------include "c:.backslash.dev.backslash.mytech.backslash.pic16c74.inc"include "c:.backslash.dev.backslash.mytech.backslash.support.inc";; FLOATOFFSET causes the threshold to be; offset by 1/4 of the average PD output.;FLOATOFFSET set 1ADAPT set 1AutoTimerAdj set 0useGSB set 1useUS set 1useIR set 1carrier set 1 ; 0:25Khz;1:32Khz;2:40KhzuseLCD set 0 ; includes the code for the lcd debuguseDA set 0 ; includes the code for the da debug;--------------------------------------------------------------;--------------------------------------------; we begin page 0 allocations here;--------------------------------------------include "c:.backslash.dev.backslash.mytech.backslash.o2ctlreg.inc"include "c:.backslash.dev.backslash.mytech.backslash.o2dtregs.inc"include "c:.backslash.dev.backslash.eeprom.backslash.o2eeprom.inc"CBLOCK FONREF ; false-on correction wait counter CHAR ; used by LCDENDC;--------------------------------------------; we begin page 1 allocations here;--------------------------------------------include "c:.backslash.dev.backslash.mytech.backslash.adaptreg.inc"CBLOCK STATCNTL STATWORD usPot pirPot FONCNT FOFFCNT lcdcntr ; used by lcdENDC;*********************************if useDA==1include "c:.backslash.dev.backslash.mytech.backslash.doutd.inc"endiforg 0.times.1c0include "c:.backslash.dev.backslash.mathlib.backslash.mathlib.asm"include "filter.asm"include "c:.backslash.dev.backslash.mytech.backslash.ledtask.asm" ; LED taskinclude "c:.backslash.dev.backslash.mytech.backslash.relaytsk.asm" ; relay countdown taskinclude "c:.backslash.dev.backslash.eeprom.backslash.o2e2prom.asm"include "c:.backslash.dev.backslash.mytech.backslash.muxtask.asm"include "o2dtinit.asm"include "dtustask.asm" ; ultrasonic sensor taskinclude "dtirtask.asm" ; pir sensor task; include "c:.backslash.dev.backslash.mytech.backslash.o2lodplr.asm"include "c:.backslash.dev.backslash.mytech.backslash.o2stat.asm"include "o2adapt.asm"include "o2sermon.asm"if useLCD==1#define TEMP MTEMP5include "c:.backslash.dev.backslash.lcd.backslash.omni2lcd.asm"endifo2start;*********************************; The hardware setup is done here. The; timers are setup and started.;*********************************o2initmainmovlw high noserrptmovwf PCLATHbtfss CBITS1,7goto noserrptbcf CBITS1,7movlw high serrptmovwf PCLATHcall serrptnoserrptif useLCD==1movlw high nolcdreportmovwf PCLATHbtfss CBITS2,3goto nolcdreportbcf CBITS2,3; movlw high lcdreport; movwf PCLATH; call lcdreport; lcdreportnolcdreportendif;---------------------------------------------------------------; send the interrogation status report if necessary;---------------------------------------------------------------movlw high noireportmovwf PCLATHbtfss ReportTaskEnablegoto noireportmovlw high statussendmovwf PCLATHcall statussend; to prevent the seldom, but possible, event; of an interrogation and a 2 min report from; occurring at the same timemovlw high no2minreportmovwf PCLATHgoto no2minreportnoireport;---------------------------------------------------------; send the 2 minute status report if it is time;---------------------------------------------------------movlw high no2minreportmovwf PCLATHbtfss TwoMinStatTaskEnablegoto no2minreportmovlw high twominstatusmovwf PCLATHcall twominstatusno2minreportif ADAPT==1;------------------------------------------------------; run the false-on correction if necessary;------------------------------------------------------movlw high nofalseonmovwf PCLATHbtfss SBITS,6goto nofalseonmovlw high fixfalseonmovwf PCLATHcall fixfalseonusePage1incf FONCNT,FusePage0nofalseon;------------------------------------------------------; run the false-off correction if necessary;------------------------------------------------------movlw high nofalseoffmovwf PCLATHbtfss SBITS,5goto nofalseoffmovlw high fixfalseoffmovwf PCLATHcall fixfalseoffusePage1incf FOFFCNT,FusePage0nofalseoffendifmovlw high novarstoremovwf PCLATHbtfss SBITS,1 ; is it time to store var's?goto novarstore ; nobcf SBITS,1 ; disable taskStoreVarsE2 ; store the var'snovarstoremendclrwdtmovlw high mainmovwf PCLATHgoto main;-------------------------------------------------------------------------org 0movlw high o2startmovwf PCLATHgoto o2start ; reset vector;----Interrupt Routine (comes here at 1 KHz)org 4 ; interrupt vector locationintamovwf TEMPW ; save w in TEMPWswapf STATUS,W ; get status (swapf doesn't alter status bits)usePage0movwf TEMPS ; save status in TEMPSmovf PCLATH,Wmovwf TMPPCHmovf FSR,Wmovwf TMPFSRmovlw timvala ; reset RTCCmovwf RTCCbcf INTCON,T0IF ; reset RTCC interrupt flagclrf PCLATH ; set up page;-------------------------------------------------------------------------; check the drvrstat pin (PORTC 5) if currently occupied;-------------------------------------------------------------------------movlw high drvrstatchkendmovwf PCLATHbtfss Occupiedgoto drvrstatchkendbtfsc DriverStatusgoto drvrstatchkend ; since drvrstat is high, then relay driver OK; we have a problembcf PORTB,2 ; turn off the relaybsf DriverFailurebcf Occupieddrvrstatchkend;------------------------------------------------; run the sensor task if it is enabled;------------------------------------------------btfss SBITS,0goto skipsnsrmovlw high ultrasonictaskmovwf PCLATHcall ultrasonictaskclrf PCLATH ; reset page after call;------------------------------------------; do the pir routine every 32ms;------------------------------------------dopirmovf CCNT0,Waddlw 1 ; time slice offsetandlw H`1f`btfss STATUS,Z ; pir sensor runs at 1Khz/32goto dtdecisionmovlw high runPIRmovwf PCLATHcall runPIRclrf PCLATH ; reset page after calldtdecision;-------------------------------------------------------; This is where we do the composition of; individual technologies to determine; occupancy. We'll handle this differently; depending on the state of the lights and; feature select switches;-------------------------------------------------------btfsc Occupied ; if occupied use the normal decision algorithmgoto dtdecidenorm; when not occupied for less than .about.15sec, always use the HSalgorithmmovf LOFFTIMEHI,Wbtfss STATUS,Z ; if lofftimehi!=0, then >15sec for sure.goto dtdecidenormmovf LOFFTIMELO,Wandlw H`f0` ; if any of upper 4 bits, then >15sec for surebtfss STATUS,Zgoto dtdecidenormmovlw high dtHSschememovwf PCLATHcall dtHSschemeclrf PCLATHgoto enddtdecidedtdecidenormbcf DualTechSensebtfsc options,0goto dtHSmotionmovlw high dtHCschememovwf PCLATHcall dtHCschemeclrf PCLATHgoto enddtdecidedtHSmotionmovlw high dtHSschememovwf PCLATHcall dtHSschemeclrf PCLATH ; reset pageenddtdecide;----------------------------------------------------; we need to compute an average of the; time between motion occurences. This; quantity is then used as a cue for the; ASA algorithm.;----------------------------------------------------btfss DualTechSensegoto skipsnsrusePage1movf timesincemothi,Wiorwf timesincemotlo,WusePage0btfsc STATUS,Z ; ignore this motion occurrencegoto cntrupdt ; if it is less than 1sec since the ; previous one.;---------------------------------------------------------------------------; we'll add code here for looking at the time since motion;---------------------------------------------------------------------------usePage1movf timesincemothi,WusePage0movwf TEMP7usePage1movf timesincemotlo,Wsubwf maxtimesincemotlo,WusePage0movwf TEMP6btfss STATUS,C ; borrow?incf TEMP7,F ; yesmovf TEMP7,WusePage1subwf maxtimesincemothi,Wbtfsc STATUS,C ; borrow?goto cntrupdt ; no;------------------------------------------------------------------------------; the most recent timesincemot is greater than previous max; so we need to update the max and manipulate timeradj(?);------------------------------------------------------------------------------movf timesincemothi,Wmovwf maxtimesincemothimovf timesincemotlo,Wmovwf maxtimesincemotlo; no timer manipulation for nowskipsnsrcntrupdtusePage0;-----------Update Counters-----------incf CCNT0,F ; increment counter low bits (1ms)btfss STATUS,Z ; zero?goto t1klbl1 ; no...incf CCNT1,F ; increment counter low mid bits (every 256ms)btfss STATUS,Z ; zero?goto t1klbl1 ; no...incf CCNT2,F ; increment counter high mid bits (every 65.536sec)btfss STATUS,Z ; zero?goto t1klbl1 ; no...incf CCNT3,F ; yes, increment counter high bits (every 4.66hrs)btfsc STATUS,Z ; if CCNT3 rolls over,bsf CntrRollOver ; set the flag;-------------------------------------------------------------------; here we add the hooks for manipulating the timer;-------------------------------------------------------------------movlw high reducetimermovwf PCLATHcall reducetimerclrf PCLATH;-------------------------------------------------------------------------------; we also need to store the adaptation variables in EEPROM;-------------------------------------------------------------------------------bsf SBITS,1t1klbl1 nopbtfsc FourMinOnBitgoto morethan4movlw H`fc` ; mask off upper 6 bitsandwf CCNT2,Wbtfss STATUS,Z ; if any of lower 2 bits set, then power on forbsf FourMinOnBit ; for more than 4min, set the appropriate bitmorethan4;------------------------------------------------------; The muxtask gets called at a 16ms rate; The effective sample rate of the A/D; readings coming through is 15hz.;------------------------------------------------------movf CCNT0,Waddlw 3 ; time slice offsetandlw H`0f`btfss STATUS,Zgoto nomuxmovlw high runMUXTmovwf PCLATHcall runMUXTclrf PCLATH ; reset the pagenomux; movlw high lowfdoppler; movwf PCLATH; call lowfdoppler; clrf PCLATHmovlw 99h ; 1001 1001 enable channel3 a/d againmovwf ADCON0 ; (This is the receiver a/d.) ; We switch it now so the mux will settle ; in time for the reading.; We will clock the LED every 4ms; movlw high skipled; movwf PCLATH ; don't forget to setup the goto's!!! btfss SBITS,2 ; is LED scheduled to run? goto skipled ; no. so skip task.movlw high LEDTASKmovwf PCLATHcall LEDTASKclrf PCLATHskipledbtfss CBITS3,7 ; is green LED scheduled to run?goto skipgled ; no. so skip task.if useDA==0movlw high gLEDTASKmovwf PCLATHcall gLEDTASKclrf PCLATHendifskipgledbtfss Suppressgoto nosprsmovlw high suppressmovwf PCLATHcall suppressclrf PCLATH ; reset pagenosprsSECONDtaskmovlw high endSECONDmovwf PCLATHmovf SECONDreflo,Wsubwf CCNT0,W ; CCNT0-SECONDreflobtfss STATUS,Zgoto endSECONDmovf SECONDrefhi,W ; movf only affects the Z bitbtfss STATUS,C ; did the subtract borrow?incf SECONDrefhi,W ; yes.(SECONDrefhi borrows from CCNT1)subwf CCNT1,W ; CCNT1-SECONDrefhibtfss STATUS,Zgoto endSECOND; setup for the next secondmovlw H`e8`addwf SECONDreflo,Fbtfsc STATUS,Cincf SECONDrefhi,Fmovlw H`03`addwf SECONDrefhi,F ; 03e8h is 1000dincf SECONDLO,Fbtfsc STATUS,Cincf SECONDHI,F ; SECONDHI:SECONDLO increments every 1.0sbtfss Occupiedgoto secrelayusePage1incf timesincemotlo,Fbtfsc STATUS,Zincf timesincemothi,FusePage0secrelaymovlw high relaytaskmovwf PCLATHbtfsc SBITS,3 ; is the relay on?call relaytask ; yes, so service the countdown task.clrf PCLATH; read the option bitsbcf options,0btfss DTSwitch ; invert switch logicbsf options,0bcf options,1btfss LEDOverride ; invert switch logicbsf options,1bcf options,2btfss AdaptReset ; invert switch logicbsf options,2; determine if the options have been changedusePage1movf lastoption,Wandlw H`7f` ; mask off bit 7usePage0movwf TEMPLOmovlw H`7f` ; mask off bit 7andwf options,Wsubwf TEMPLO,Wbtfsc STATUS,Zgoto no1secasareset;----------------------------------------------------------; options changed, so reset the asa variables; need to verify this feature;----------------------------------------------------------clrf usonadjclrf usoffadjclrf iroffadjclrf timeradjloclrf timeradjhiclrf CCNT3bcf CntrRollOver;!!!!!!!!!; should we add code to reset e2prom var's????;!!!!!!!!!no1secasaresetmovf options,WusePage1movwf lastoptionusePage0;--------------------------------------------------------------------; when a driver failure has been declared, we should; periodically clear the bit, this prevents the; sensor from having permanent failure.;--------------------------------------------------------------------movlw high secdrvrchkendmovwf PCLATHmovlw H`03` ; check every 4secandwf SECONDLO,Wbtfss STATUS,Zgoto secdrvrchkendbtfsc DriverFailurebcf DriverFailuresecdrvrchkend;-----------------------------------------; this is used for the lcd report;-----------------------------------------btfss SECONDLO,0bsf CBITS2,3;-----------------------------------------; this is for the serial report;-----------------------------------------movlw H`07` ; every 8secandwf SECONDLO,Wbtfsc STATUS,Zbsf CBITS1,7movlw high ltsoffmovwf PCLATHbtfss CBITS1,2 ; are the lights on?goto ltsoff ; no. goto lights off SECONDltsonincf LONTIMELO,Fbtfsc STATUS,Zincf LONTIMEHI,Fmovf LONTIMELO,W ; have the lights been on 8+sec?andlw H`f8` ; only if any of upper 5 bits setbtfss STATUS,Zbsf CBITS2,6 ; yes, so set the 8sec lights on flaggoto endSECONDltsoffincf LOFFTIMELO,Fbtfsc STATUS,Zincf LOFFTIMEHI,Fif useGSB==1 movf LOFFTIMELO,W andlw H`e0` ; if any of upper 3 bits set btfsc STATUS,Z ; then lights off > 30 sec goto ltsoffnogsb ; when the gsb switch is turned on, we need ; to set usgain to the offgain value which ; is .75 of the ongain value corresponding ; to the square law mapped pot setting movf usoffgain,W btfsc GSBEnable ; is gsb turned on? movwf usgain ; yes.ltsoffnogsbendifendSECONDmcnt movlw high endmcntmovwf PCLATHmovf mcntref,Wsubwf SECONDLO,Wbtfss STATUS,Zgoto endmcnt; setup for the next minutemovlw D`60`addwf mcntref,Fincf mcntlo,F ; mcntlo increments once every minutebtfsc STATUS,Cincf mcnthi,F;----------------------------------------------------------------------------; no two minute status when the LED override switch is on;----------------------------------------------------------------------------btfsc NoLEDFlaggoto skip2minstatusbtfss mcntlo,0 ; do status every other minutebsf TwoMinStatTaskEnableskip2minstatus;---------------------------------------------------------------------; Per Microchip advice, we need to periodically reset; the TRIS registers;---------------------------------------------------------------------movlw high setTrisRegsmovwf PCLATHcall setTrisRegsclrf PCLATH;----------------------------------------------------------------------; When in installer test mode, we need to come out of; it after an hour. Increment and Check ITMClk.;----------------------------------------------------------------------movlw high min.sub.-- noitmmovwf PCLATHbtfss InstallerTestModegoto min.sub.-- noitmincf ITMClk,Fmovlw D`60`subwf ITMClk,W ; W = ITMClk - 60btfss STATUS,C ; if borrow, then <60 min in test modegoto min.sub.-- noitm ; the subtract borrowed; get out of test modebcf InstallerTestModemin.sub.-- noitm;----------------------------------------------------------------------; in order to get a better estimator of the ir noise; level, we maintain a running max of the noise. Once; every minute this running max is compared with the; current ir noise estimate. The largest is the new; noise estimate of the ir circuit.;----------------------------------------------------------------------usePage1movf irRecentMax,WusePage0movwf TEMP8usePage1movf irNoiseEst,WusePage0movwf TEMP7movlw high maxmovwf PCLATHcall maxclrf PCLATHusePage1movwf irNoiseEstmovlw H`80`movwf irRecentMaxusePage0endmcnt;------------------------------------------------------; service the interrogate task if necessary;------------------------------------------------------movlw high nostatmovwf PCLATHbtfss Interrogatinggoto nostatstatcntrdecfsz NTRGATEcntgoto nostatdostatmovlw high handleNTRGATEmovwf PCLATHcall handleNTRGATEnostat;---------------------------------------------------------------------------------; The eeprom code is placed in the interrupt routine so; that it will not be disturbed by interrupts, otherwise; interrupts may have to be disabled during eeprom read/write.;---------------------------------------------------------------------------------movlw high eepromwritemovwf PCLATHeepromwaitbtfss E2FLAG,1goto eepromwritemovlw high eepromdlyendmovwf PCLATHmovf CCNT0,Wsubwf E2DLY,Wbtfsc STATUS,Z ; is it exactly 11ms?goto eepromdlyend ; yesbtfss STATUS,C ; if not, did the subtract borrow (more than 11ms)eepromdlyendbcf E2FLAG,1 ; yeseepromwritemovlw high eepromreadmovwf PCLATHbtfss EEPROMWTASK ; is an eeprom write scheduled?goto eepromread ; no...movlw high endeeprommovwf PCLATHbtfsc E2FLAG,1 ; is the EEPROM still writing?goto endeeprom ; yesmovlw high WRBYTEmovwf PCLATHcall WRBYTEbcf EEPROMWTASKmovlw high endeeprommovwf PCLATHgoto endeeprom ; if we have just written, then we can't ; read for .about.11ms so jump over read taskeepromreadmovlw high endeeprommovwf PCLATHbtfss EEPROMRTASKgoto endeeprombtfsc E2FLAG,1 ; is the EEPROM still writing?goto endeeprom ; yesmovlw high RDBYTEmovwf PCLATHcall RDBYTEbcf EEPROMRTASKendeepromendi clrwdtmovf TMPFSR,Wmovwf FSRmovf TMPPCH,Wmovwf PCLATHswapf TEMPS,Wmovwf STATUS ; restore statusswapf TEMPW,F ; swap value in TEMPWswapf TEMPW,W ; restore w registerretfi retfie ; return from interruptdtHC ;--------------------------------------------------------; To do the high confidence operation, we; use a modified bounded sum operator.;--------------------------------------------------------movlw high dtHC0movwf PCLATHmovf ustruth,Waddwf pirtruth,Wmovwf TEMP6 ; TEMP6 = truthmovlw H`60` ; 0.times.60 is the lower truth limitsubwf TEMP6,Fbtfss STATUS,C ; borrow?goto dtHC0 ; yes, so no motion for suremovf TEMP6,Wbtfsc TEMP6,7movlw H`7f`returndtHC0clrwreturndtHS movlw high dtHSmaxmovwf PCLATHmovf ustruth,Waddwf pirtruth,Wmovwf TEMP9btfsc TEMP9,7goto dtHSmax ; if overflow return 7fh; movlw 07fhreturndtHSmaxmovlw H`07f`returndtHCschemebtfsc Occupied ; are we currently occupied?goto stilloccup ; yes...see if lights on timer should be resetmovlw high dtHCmovwf PCLATHcall dtHC ; w = HC(ustruth,pirtruth)movwf TEMP9;------------------------------------------------------; We want the truth to be >= limit;------------------------------------------------------amoccuptstmovlw high amoccup.sub.-- yesmovwf PCLATHmovf TEMP9,Wsublw H`040` ; 40h - Wbtfsc STATUS,Z ; is the truth exactly what we want?goto amoccup.sub.-- yes ; yes...space is occupiedbtfsc STATUS,C ; is the truth > limitreturn ; no...space not occupied;----------------------------------------------------------; We have determined that the lights should; be turned on.;----------------------------------------------------------amoccup.sub.-- yesmovlw high turnOnLightsmovwf PCLATHcall turnOnLights ; turn on lights + bookkeepingclrf PCLATHbsf DualTechSensereturnstilloccup;-------------------------------------------------------; Since we are currently occupied we use; the high sensitivity operation;-------------------------------------------------------movlw high dtHSmovwf PCLATHcall dtHSmovwf TEMP9stilloccuptst;-------------------------------------------------------; We want the truth to be >= limit;-------------------------------------------------------movlw high dt1occup.sub.-- yesmovwf PCLATHmovf TEMP9,Wsublw H`07f` ; limit - Wbtfsc STATUS,Z ; is the truth exactly what we want?goto dt1occup.sub.-- yes ; yes...reset the relaybtfsc STATUS,C ; is the truth > than limitreturn ; no...;----------------------------------------------------------; We have determined that the lights should; be kept on.;----------------------------------------------------------dt1occup.sub.-- yesmovlw high setrelaymovwf PCLATHcall setrelay ; this will simply reset the ; relay countdown timerbsf DualTechSensedt1end returndtHSschememovlw high dtHSmovwf PCLATHcall dtHSmovwf TEMP9;-----------------------------------------------------; We want the truth to be >= limit;-----------------------------------------------------movlw high dt2occup.sub.-- yesmovwf PCLATHmovf TEMP9,Wsublw H`07f` ; limit - Wbtfsc STATUS,Z ; is the truth exactly what we want?goto dt2occup.sub.-- yes ; yes...reset the relaybtfsc STATUS,C ; is the truth > than limitreturn ; no...;-----------------------------------------------------------; We have determined that motion is present;-----------------------------------------------------------dt2occup.sub.-- yesbtfsc Occupied ; are we currently occupied?goto dt2stilloccup ; yes...lights on timer should be resetmovlw high turnOnLightsmovwf PCLATHcall turnOnLightsclrf PCLATHbsf DualTechSensereturndt2stilloccupmovlw high setrelaymovwf PCLATHcall setrelay ; this will simply reset the ; relay countdown timerbsf DualTechSensedt2end return END; pic16c74.inc; COPYRIGHT (C) MYTECH CORP. 1996, 1997;----- STATUS Bits ---------------------------------------------------------------------------------------------------------------IRP EQU H`0007`RP1 EQU H`0006`RP0 EQU H`0005`NOT.sub.-- TO EQU H`0004`NOT.sub.-- PD EQU H`0003`Z EQU H`0002`DC EQU H`0001`C EQU H`0000`W EQU H`0000`F EQU H`0001`INDF equ 00hRTCC equ 01hPCL equ 02hSTATUS equ 03hFSR equ 04hPORTA equ 05hPORTB equ 06hPORTC equ 07hPORTD equ 08hPORTE equ 09hTRISA equ 05hTRISB equ 06hTRISC equ 07hTRISD equ 08hTRISE equ 09hPCLATH equ 0ahADCON0 equ 1fhADCON1 equ 1fh ; must be in page 1ADRES equ 1ehINTCON equ 0bhOPTREG equ 01h ; must be in page 1PIR1 equ 0chPIE1 equ 0ch ; must be in page 1PIR2 equ 0dhPIE2 equ 0dh ; must be in page 1TMR1L equ 0ehPCON equ 0eh ; must be in page 1TMR1H equ 0fhT1CON equ 10hTMR2 equ 11hT2CON equ 12hPR2 equ 12h ; must be in page 1SSPBUF equ 13hSSPADD equ 13h ; must be in page 1SSPCON equ 14hSSPSTAT equ 14h ; must be in page 1CCPR1L equ 15hCCPR1H equ 16hCCP1CON equ 17hRCSTA equ 18hTXSTA equ 18h ; must be in page 1TXREG equ 19hSPBRG equ 19h ; must be in page 1RCREG equ 1ahCCPR2L equ 1bhCCPR2H equ 1chCCP2CON equ 1dh;----- INTCON Bits (except ADC/Periph) -------------------------------------------------------------------------------------GIE EQU H`0007`PEIE equ H`06`T0IE EQU H`0005`INTE EQU H`0004`RBIE EQU H`0003`T0IF EQU H`0002`INTF EQU H`0001`RBIF EQU H`0000`;----- OPTION Bits -----------------------------------------------------------------------------------------------------------------NOT.sub.-- RBPU EQU H`0007`INTEDG EQU H`0006`T0CS EQU H`0005`T0SE EQU H`0004`PSA EQU H`0003`PS2 EQU H`0002`PS1 EQU H`0001`PS0 EQU H`0000`;----- ADCON0 Bits --------------------------------------------------------------------------------------------------------------ADCS1 EQU H`0007`ADCS0 EQU H`0006`CHS2 EQU H`0005`CHS1 EQU H`0004`CHS0 EQU H`0003`GO EQU H`0002`GO.sub.-- DONE EQU H`0002`ADON EQU H`0000`;----- ADCON1 Bits --------------------------------------------------------------------------------------------------------------PCFG2 EQU H`0002`PCFG1 EQU H`0001`PCFG0 EQU H`0000`;----- PIE1 and PIR1 ADC Bits ---------------------------------------------------------------------------------------------------ADIE EQU H`0006`ADIF EQU H`0006`TXIF EQU H`04`RCIF EQU H`05`;----- I2C Bits for PIR1 and PIE1-----------------------------------SSPIE equ H`03` ;PIE1SSPIF equ H`03` ;PIR1 Z; support.inc; COPYRIGHT (C) MYTECH CORP. 1996, 1997;LSB equ 0#define.sub.-- C STATUS,C;**********************; John's macro's;**********************usePage0 macrobcf STATUS,5 ; select page 0endmusePage1 macrobsf STATUS,5 ; select page 1endm; all uses of this macro depend on; no changes to WdisableInts macrobcf INTCON,GIE ; disable global interruptsendm; all uses of this macro depend on; no changes to WenableInts macrobsf INTCON,GIE ; enable global interruptsendm;;;;;; all uses of this macro depend on; no changes to WenterCritical macrodisableIntsendm; all uses of this macro depend on; no changes to WexitCritical macroenableIntsendm; push -- push the contents of W onto; the stack and dec the stack pointer.; Thw W register is not affected.pushw macromovwf INDFdecf FSR,Fendm; pop -- get the value pointed to by the; stack and inc the stack pointer. The; value is returned in W.popw macromovf INDF,Wincf FSR,Fendmi16lsrw macro hireg,loreg rlf hireg,W rrf hireg,F rrf loreg,Wendmi16lsr macro hireg,loreg rlf hireg,W rrf hireg,F rrf loreg,Fendmi16lsl macro hireg,loreg bcf STATUS,C rlf loreg,F rlf hireg,Fendmchkof macrobtfsc STATUS,Cincf overflowendmchkuf macro btfss STATUS,C incf underflowendmjmpl macro addr movlw high addrmovwf PCLATHgoto addrendmicall macro addr movwf INDF movlw high addrmovwf PCLATHmovf INDF,Wcall addrendm;; o2ctlreg.inc; COPYRIGHT (C) MYTECH CORP. 1996, 1997; control registers for the Omni II;; 7-22-96 JJF;---------------------------------------------------------------------------------------------------------------------------------------;---------------------------------------------------------------------------------------------------------------------------------------CBLOCK 0.times.20 TEMPW ; this exists at both 20h and a0h TEMPS TMPPCH TMPFSR CCNT0 ; 1ms ticks CCNT1 ; 256 ms ticks CCNT2 ; 65.536sec ticks CCNT3 ; 16777.216sec ticks ; 279.62min=4.66hr ticks ; max ctr is 1188hours(49.5days) CBITS1 ; bit 0: 1-US sense motion ; 1: 1-PIR sense motion ; 2: 1-occupied bit ; 3: 1-US ON ; 4: 1-PIR ON ; 5: 1-power on for 60+ minutes ; 6: 1-Installer Test Mode enable ; 7: **used with Serial Debug Report** CBITS2 ; bit 0: 1-US dominant;0-IR dominant ; 1: 1-lights off 1+sec flag ; 2: 1-Counter has rolled over ; 3: **used with LCD Debug** ; 4: 1-GSB enable flag ; 5: 1-False-Off Timer return state ; 6: 1-False-Off timer return enabled ; 7: 1-DRVRSTAT Failure!!! CBITS3 ; bit 0: 1-sensitivity too high ; 1: 1-Two minute status report enabled ; 2: 1-EEPROM write task enabled ; 3: 1-EEPROM read task enabled ; 4: 1-ASA reset signal ; 5: 1-interrogate task enabled ; 6: 1-interrogation report task enabled ; 7: 1-Green LED task enabled SBITS ; bit 0: 1-sensor task enabled ; 1: 1-Store Vars task enabled ; 2: 1-Red LED task enabled ; 3: 1-lights-on countdown ; 4: 1-relay suppress enable ; 5: 1-false off correction ; 6: 1-false on correction ; 7: 1-periodic ASA TEMPLO ; These temp registers are for use TEMPHI ; in the interrupt routine ONLY! TEMP6 ; ` ` TEMP7 ; ` ` TEMP8 ; ` ` TEMP9 ; ` ` MTEMP ; These temp registers are for use MTEMP1 ; in the main loop ONLY! MTEMP2 ; ` ` MTEMP3 ; ` ` MTEMP4 MTEMP5 TMRPOTHI ; the lights on timer setting in seconds TMRPOTLO timeradjhi timeradjlo SECONDHI ; this is the 1 second counter SECONDLO SECONDrefhi ; the reference for the 1sec counter SECONDreflo ; relative to the 1ms counter, when ; SECONDHI:SECONDLO increments, ie once/sec, ; SECONDrefhi:SECONDreflo is set to 1000 + ; CCNT1:CCNT0.(CCNT0 increments every ms) relaylo ; this is the relay task counter. it is relayhi ; decremented once every second. sprscnt ; this is the relay suppress counter. mcnthi ; the minute counter mcntlo mcntref ; the minute counter reference, relative ; to the second counter. LOFFTIMEHI LOFFTIMELO LONTIMEHI LONTIMELO options ; bits 0-2: dual tech mode ; bit 3: strong air compensation ; bit 4: ASA enable ; bit 5: ATA enable ; bit 6: GSB enable ; bit 7: dt motion sense NTRGATEcnt NTRGATEstate FCOUNT ; generic lowpass corner freq. control ITMClk ; installer test mode clock (minutes) lastTimer ; for determining test mode entry/exitENDC;; convenient macros for bit addresses;#define Occupied CBITS1,2#define RelayEnable CBITS2,1#define Relay PORTB,2#define DLCRelay PORTA,2#define USSense CBITS1,0#define IRSense CBITS1,1#define GreenLEDEnable CBITS3,7 ; the task enable bit#define GreenLED PORTA,4#define TwoMinStatTaskEnable CBITS3,1#define EEPROMWTASK CBITS3,2#define EEPROMRTASK CBITS3,3#define Interrogating CBITS3,5#define ReportTaskEnable CBITS3,6#define Suppress SBITS,4#define StrongAirCompensate options,3#define ASAEnable options,4#define ATAEnable options,5#define GSBEnable options,6#define NoLEDFlag options,1#define FourMinOnBit CBITS1,5#define InstallerTestMode CBITS1,6#define ASATimerReturn CBITS2,6#define ASATimerReturnState CBITS2,5#define CntrRollOver CBITS2,2#define DTSwitch PORTB,0;#define DTSwitch2 PORTB,6;#define DTSwitch3 PORTB,7#define LEDOverride PORTB,6#define AdaptReset PORTB,7#define DriverStatus PORTC,5#define DriverFailure CBITS2,7;; o2dtregs.inc;; register allocations for the dual tech sensor; COPYRIGHT (C) MYTECH CORP. 1996, 1997include "c:.backslash.dev.backslash.mytech.backslash.o2usregs.inc"include "c:.backslash.dev.backslash.mytech.backslash.o2irregs.inc"CBLOCK pirtruth ustruth LightLevel ; the photocell A/D readingENDC#define DualTechSense options,7;; o2irregs.inc; COPYRIGHT (C) MYTECH CORP. 1996, 1997; register allocations for the Omni II PIR Sensor;;--------------- globals ---------------CBLOCK; PIRadin ; PIR PIRLPLO ; PIR PIRLPHI ; PIR PIRthresh iroffadj PIRAVGLO PIRAVGHI pirdecision irpscnt gLEDST ; Green LED state gLEDSELECT ; Green LED function select gLEDCNTR ; Green LED TASK counter ENDC;; o2usregs.inc; COPYRIGHT (C) MYTECH CORP. 1996, 1997; register allocations for the Omni II US sensor;CBLOCK w2hi ; filter state memory w2lo w1hi w1lo AD1LPLO AD1LPHI; lfdlplo; lfdlphi; lfdtemp usgain ; the currently used gain setting, usongain usonadj usoffadj ; the ASA algorithm alters this variable. ; it is added to usoffgain. usoffgain ; the gain when the lights are off decimlo decimhi RECTAVGlo RECTAVGhi integhi ; the 16bit integrator output integlo usdecision uspscnt ; pulse stretch cnt LEDST ; Red LED state LEDSELECT ; Red LED function select LEDCNTR ; Red LED TASK counter ENDCVC1hi set w1hiVC1lo set w1loIL1hi set w2hiIL1lo set w2lo;; o2asareg.inc; COPYRIGHT (C) MYTECH CORP. 1996, 1997; register allocations for the Omni II; ASA algorithm.;CBLOCK 0.times.b0 usLearnCoef irLearnCoef timerLearnCoef irNoiseEst irRecentMax avgtimesincemot timesincemotlo timesincemothi maxtimesincemotlo maxtimesincemothi lastoption dominantTech uspulsecnt0 uspulsecnt1 uspulsemax usminpulse usminpw0 usminpw1ENDCCBLOCK E2FLAG ; Common flag bits register E2BUF ; Bit buffer E2ERR ; Error code (to indicate bus status) E2ADDR ; Address register DATAI ; Stored data input register DATAO ; Stored data output register E2WBUF ; Write buffer E2RBUF ; Read buffer E2CNT ; Bit counter E2DLY ; Delay variable; SLAVEENDCFLAG set E2FLAGEEPROM set E2BUFERCODE set E2ERRADDR set E2ADDRTXBUF set E2WBUFRXBUF set E2RBUFCOUNT set E2CNT;#define SLAVE B`10101110` ; eeprom address in I2C space;#define SLAVE B`10100000`;************************************************; Convenient Stuff for addressing bits;************************************************; FLAG Bits;ERRORFLG EQU 0 ; Error flag;; EEPROM Bits;DI EQU 7 ; EEPROM inputDO EQU 6 ; EEPROM output;*******************************************; PIC I2C Pin Assignments;*******************************************SDA EQU 4 ; RC4, data in/outSCL EQU 3 ; RC3, serial clock;SDA EQU 6;SCL EQU 7;; mathlib.asm; COPYRIGHT (C) MYTECH CORP. 1996, 1997; 10-3-96 JJF;;----------------------------------------------------------------; Bin2BCD -- This routine converts the 8bit; binary value in W to a 3 digit BCD value; in the registers MTEMP3:MTEMP2:MTEMP1.; MTEMP1 is the least significant digit.;----------------------------------------------------------------Bin2BCDclrf MTEMP2clrf MTEMP3movwf MTEMP1bin2bcdlbl1movlw D`10`subwf MTEMP1,W ; MTEMP1 - 10btfss STATUS,Cgoto bin2bcdlbl2movwf MTEMP1incf MTEMP2goto bin2bcdlbl1bin2bcdlbl2movlw D`10`subwf MTEMP2,W ; MTEMP2 - 10btfss STATUS,Cgoto bin2bcdlbl3movwf MTEMP2incf MTEMP3goto bin2bcdlbl2bin2bcdlbl3retlw 0;--------------------------------------------------------------; filter.asm -; COPYRIGHT (C) MYTECH CORP. 1996, 1997;--------------------------------------------------------------;----------------------------------------------------------------------; genlp -- generic lowpass filter; Inputs: FSR -- points to low byte of; filter state; FCOUNT -- divide count which; sets the corner freq; values are 1-7.; TEMP9:TEMP7 -- 16bit input sample;; corner frequency is approximately:; (fs/2pi)*ln(2 FCOUNT/(2 FCOUNT-1));----------------------------------------------------------------------genlpmovf INDF,Wmovwf TEMPLOincf FSR,F ; point to high bytemovf INDF,Wdecf FSR,F ; return pointer to low bytemovwf TEMPHIcomf TEMPLO,Fincf TEMPLO,Fbtfsc STATUS,Zdecf TEMPHI,Fcomf TEMPHI,F ; -lpstatemovf TEMP7,Waddwf TEMPLO,Fbtfsc STATUS,Cincf TEMPHI,Fmovf TEMP9,Waddwf TEMPHI,F ; TEMPHI:TEMPLO has input - lpstatemovlw high genlploopmovwf PCLATHgenlplooprlf TEMPHI,Wrrf TEMPHI,Frrf TEMPLO,Fdecfsz FCOUNT,Fgoto genlploopmovf TEMPLO,Waddwf INDF,Fincf FSR,F ; point to high bytebtfsc STATUS,Cincf INDF,Fmovf TEMPHI,Waddwf INDF,F ; lpstate is updated to ; lps[n] = x[n] + (x[n] - lps[n-1]/(FCOUNT+1))return;-----------------------------------------------------; highpass -; Input: ADVAL - 8bit input sample; Output: TEMPHI:TEMPLO 16 output;-----------------------------------------------------; This filter is implemented by first; computing the lowpass of the input; sample and then computing the output; via input-lowpass(input); we remove the dither term as well;-----------------------------------------------------highpass macro; TEMP9 already contains the raw us doppler signal; just place the dither term in TEMP7movlw H`a0`movwf TEMP7movlw high genlpmovwf PCLATHmovf FSR,Wmovwf TMPFSRmovlw 2movwf FCOUNTmovlw AD1LPLOmovwf FSRcall genlp ; lowpass(ADVAL) in AD1LPHI:AD1LPLOmovf TMPFSR,Wmovwf FSR;-----------------------------------------------------; calculate ADVAL - lowpass(ADVAL);-----------------------------------------------------movf AD1LPHI,Wmovwf TEMPHImovf AD1LPLO,Wmovwf TEMPLOcomf TEMPLO,F ; negate ad1lpincf TEMPLO,Fbtfsc STATUS,Z ; zero bitdecf TEMPHI,Fcomf TEMPHI,F ; -ad1lp in TEMPHI, TEMPLOmovf TEMP9,Waddwf TEMPHI,F ; ad1-ad1lp in TEMPHI, TEMPLOendm; return;-----------------------------------------------------------------------; bpf -- the appropriate filter is chosen based; on the value of the `carrier` flag, which is; set at compile time.;; Inputs: TEMP8:TEMP6 -- 16bit input sample; Outputs: TEMPHI:TEMPLO -- 16bit output sample;-----------------------------------------------------------------------if carrier==0;--------------------------------------------; fc .about. 100; Direct Form II Implementation;--------------------------------------------bpfrlf TEMP8,Wrrf TEMP8,Frrf TEMP6,Frlf TEMP8,Wrrf TEMP8,Frrf TEMP6,Frlf TEMP8,Wrrf TEMP8,Frrf TEMP6,Fcomf w2hi,Fcomf w2lo,Fincf w2lo,Fbtfsc STATUS,Zincf w2hi,F ; -w(n-2)movf w2lo,Wmovwf TEMPLOmovf w2hi,Wmovwf TEMPHI ; -w(n-2) in TEMPi16lsr w2hi,w2lo ; 1/2movf w2lo,Wmovwf TEMP7movf w2hi,Wmovwf TEMP9 ; T9-7 = -1/2w(n-2)i16lsr w2hi,w2lo ; 1/4i16lsr w2hi,w2lo ; 1/8movf w2lo,Waddwf TEMP7,Fbtfsc STATUS,Cincf TEMP9,Fmovf w2hi,Waddwf TEMP9,F ; T9-7 = -w(n-2) (1/2+1/8)i16lsr w2hi,w2lo ; 1/16movf w2lo,Waddwf TEMP7,Fbtfsc STATUS,Cincf TEMP9,Fmovf w2hi,Waddwf TEMP9,F ; T9-7 = -w(n-2) (1/2+1/8+1/16)movf w1hi,W ; shift state memorymovwf w2himovf w1lo,Wmovwf w2lomovf w1lo,Waddwf TEMP7,Fbtfsc STATUS,Cincf TEMP9,Fmovf w1hi,Waddwf TEMP9,F ; T9-7 = (1)w(n-1)-w(n-2) (1/2+1/8+1/16)i16lsr w1hi,w1lo ; 1/2w(n-1)i16lsr w1hi,w1lo ; 1/4movf w1lo,Waddwf TEMP7,Fbtfsc STATUS,Cincf TEMP9,Fmovf w1hi,Waddwf TEMP9,F ; w(n-1) (1+1/4)-w(n-2) (...)i16lsr w1hi,w1lo ; 1/8movf w1lo,Waddwf TEMP7,Fbtfsc STATUS,Cincf TEMP9,Fmovf w1hi,Waddwf TEMP9,F ; w(n-1) (1+1/4+1/8)-w(n-2) (...)movf TEMP6,Waddwf TEMP7,Fbtfsc STATUS,Cincf TEMP9,Fmovf TEMP8,Waddwf TEMP9,W ; w(n)=x(n)+w(n-1) (...)-w(n-2) (...)movwf w1himovf TEMP7,Wmovwf w1lomovf w1lo,Waddwf TEMPLO,Fbtfsc STATUS,Cincf TEMPHI,Fmovf w1hi,Waddwf TEMPHI,F ; TEMPbcf STATUS,Crlf TEMPLO,Frlf TEMPHI,Freturnendif ;25Khz bpfif carrier==1;--------------------------------------------; fc .about. 128; Direct Form II Implementation;--------------------------------------------bpfrlf TEMP8,Wrrf TEMP8,Frrf TEMP6,Frlf TEMP8,Wrrf TEMP8,Frrf TEMP6,Frlf TEMP8,Wrrf TEMP8,Frrf TEMP6,Fcomf w2hi,Fcomf w2lo,Fincf w2lo,Fbtfsc STATUS,Zincf w2hi,F ; -w(n-2)movf w2lo,Wmovwf TEMPLOmovf w2hi,Wmovwf TEMPHI ; -w(n-2) in TEMPi16lsr w2hi,w2lo ; 1/2movf w2lo,Wmovwf TEMP7movf w2hi,Wmovwf TEMP9 ; T9-7 = -1/2w(n-2)i16lsr w2hi,w2lo ; 1/4i16lsr w2hi,w2lo ; 1/8i16lsr w2hi,w2lo ; 1/16i16lsr w2hi,w2lo ; 1/32movf w2lo,Waddwf TEMP7,Fbtfsc STATUS,Cincf TEMP9,Fmovf w2hi,Waddwf TEMP9,F ; T9-7 = -w(n-2) (1/2+1/32)i16lsr w2hi,w2lo ; 1/64movf w2lo,Waddwf TEMP7,Fbtfsc STATUS,Cincf TEMP9,Fmovf w2hi,Waddwf TEMP9,F ; T9-7 = -w(n-2) (1/2+1/32+1/64)movf w1hi,W ; shift state memorymovwf w2himovf w1lo,Wmovwf w2lomovf w1lo,Waddwf TEMP7,Fbtfsc STATUS,Cincf TEMP9,Fmovf w1hi,Waddwf TEMP9,F ; T9-7 = (1)w(n-1)-w(n-2) (1/2+1/32+1/64)i16lsr w1hi,w1lo ; 1/2w(n-1)movf w1lo,Waddwf TEMPLO,Fbtfsc STATUS,Cincf TEMPHI,Fmovf w1hi,Waddwf TEMPHI,F ; TEMP = w(n-1) (1/2)-w(n-2)i16lsr w1hi,w1lo ; 1/4i16lsr w1hi,w1lo ; 1/8movf w1lo,Waddwf TEMP7,Fbtfsc STATUS,Cincf TEMP9,Fmovf w1hi,Waddwf TEMP9,F ; w(n-1) (1+1/8)-w(n-2) (1/2+1/32+1/64)movf TEMP6,Waddwf TEMP7,Fbtfsc STATUS,Cincf TEMP9,Fmovf TEMP8,Waddwf TEMP9,W ; w(n)=x(n)+w(n-1) (1+1/8)-w(n-2) (1/2+1/32+1/64)movwf w1himovwf TEMP9movf TEMP7,Wmovwf w1loi16lsr TEMP9,TEMP7 ; w(n) (1/2)movf TEMP7,Waddwf TEMPLO,Fbtfsc STATUS,Cincf TEMPHI,Fmovf TEMP9,Waddwf TEMPHI,F ; TEMP = w(n) (1/2)+w(n-1) (1/2)-w(n-2)bcf STATUS,Crlf TEMPLO,Frlf TEMPHI,Freturnendif ; 32Khz bpfif carrier==2;--------------------------------------------; fc .about. 160; Direct Form II Implementation;--------------------------------------------bpfrlf TEMP8,Wrrf TEMP8,Frrf TEMP6,Frlf TEMP8,Wrrf TEMP8,Frrf TEMP6,Frlf TEMP8,Wrrf TEMP8,Frrf TEMP6,Fcomf w2hi,Fcomf w2lo,Fincf w2lo,Fbtfsc STATUS,Zincf w2hi,F ; -w(n-2)movf w2lo,Wmovwf TEMPLOmovf w2hi,Wmovwf TEMPHI ; -w(n-2) in TEMPi16lsr w2hi,w2lo ; 1/2movf w2lo,Wmovwf TEMP7movf w2hi,Wmovwf TEMP9 ; T9-7 = -1/2w(n-2)i16lsr w2hi,w2lo ; 1/4movf w2lo,Waddwf TEMP7,Fbtfsc STATUS,Cincf TEMP9,Fmovf w2hi,Waddwf TEMP9,F ; T9-7 = -w(n-2) (1/2+1/4)i16lsr w2hi,w2lo ; 1/8i16lsr w2hi,w2lo ; 1/16movf w2lo,Waddwf TEMP7,Fbtfsc STATUS,Cincf TEMP9,Fmovf w2hi,Waddwf TEMP9,F ; T9-7 = -w(n-2) (1/2+1/4+1/16)movf w1hi,W ; shift state memorymovwf w2himovf w1lo,Wmovwf w2lomovf w1lo,Waddwf TEMP7,Fbtfsc STATUS,Cincf TEMP9,Fmovf w1hi,Waddwf TEMP9,F ; T9-7 = (1)w(n-1)-w(n-2) (1/2+1/4+1/16)i16lsr w1hi,w1lo ; 1/2w(n-1)movf w1lo,Waddwf TEMPLO,Fbtfsc STATUS,Cincf TEMPHI,Fmovf w1hi,Waddwf TEMPHI,F ; TEMP = w(n-1) (1/2)-w(n-2)movf TEMP6,Waddwf TEMP7,Fbtfsc STATUS,Cincf TEMP9,Fmovf TEMP8,Waddwf TEMP9,W ; w(n)=x(n)+w(n-1) (1)-w(n-2) (1/2+1/4+1/16)movwf w1himovwf TEMP9movf TEMP7,Wmovwf w1loi16lsr TEMP9,TEMP7 ; w(n) (1/2)movf TEMP7,Waddwf TEMPLO,Fbtfsc STATUS,Cincf TEMPHI,Fmovf TEMP9,Waddwf TEMPHI,F ; TEMP = w(n) (1/2)+w(n-1) (1/2)-w(n-2)bcf STATUS,Crlf TEMPLO,Frlf TEMPHI,Freturnendif ; 40Khz bpf;; ledtask.asm; Copyright (C) 1996, 1997 Mytech;; the omni2 led task and flasher routines; orig: jjf#if useUS; When the counter hits 0; LEDTBL is "call'ed". LEDTBL uses the LEDST 4-7; register to determine which led function to; jump to. LEDST 0-3 are used by the led functions; as state information.LEDTASKmovlw high endledmovwf PCLATHmovf CCNT0,Wandlw H`03`btfss STATUS,Z ; every 4msgoto endleddecf LEDCNTR,Fbtfss STATUS,Z ; has task counter reached zero?goto endled ; no, skip the rest.movlw high LEDTBLmovwf PCLATHcall LEDTBLendledreturnLEDTBL; bits 0-3 of LEDSELECT are LED `function ptr`movlw high LEDTBLJMPmovwf PCLATHmovlw 003handwf LEDSELECT,Wmovwf TEMPLObcf STATUS,Crlf TEMPLO,Fbcf STATUS,Crlf TEMPLO,Waddlw low LEDTBLJMPbtfsc STATUS,Cincf PCLATH,Fmovwf PCLLEDTBLJMPmovlw high sglflashmovwf PCLATHgoto sglflashnopmovlw high nullflashmovwf PCLATHgoto nullflashnopmovlw high statusledmovwf PCLATHgoto statuslednop; movlw high contflash1; movwf PCLATH; goto contflash1; nop retlw 0ffhinitnullflashmovlw H`40`movwf LEDCNTRmovlw H`01`movwf LEDSELECTclrf LEDSTbsf SBITS,2 ; schedule LED taskbcf PORTB,1 ; turn LED offreturnnullflashclrf LEDSTclrf LEDSELECTclrf LEDCNTRbcf SBITS,2bcf PORTB,1returnsglflash; the state of the single flash led is held in bit 0 of LEDST.; 0: initial dead spot for relay; 1: onbtfsc LEDST,0 ; what state are we in?goto sgfstop ; just finished state 1, so stopmovlw 080h ; just finished state 0, need tomovwf LEDCNTR ; set `on` timebsf PORTB,1 ; turn LED onbsf LEDST,0returnsgfstopbcf SBITS,2 ; unschedule LED taskclrf LEDSTclrf LEDCNTRclrf LEDSELECTbcf PORTB,1 ; turn LED offreturninitsgfbtfsc SBITS,2 ; if led is already running just returnreturnmovlw H`00` ; set LED task to state 0(off)movwf LEDSTclrf LEDSELECT ; set function 0 (single flash)movlw 32h ; set initial 100ms dead timemovwf LEDCNTRbsf SBITS,2 ; schedule LED taskbcf PORTB,1 ; turn LED offreturnif 0contflash1; continuous on-off using 512ms ticks; 0: led off; 1: led onbtfsc LEDST,0 ; what state are we in?goto cf1down ; just finished state 1, so back to 0movlw 0ffh ; just finished state 0, need tomovwf LEDCNTR ; set `on` timebsf PORTB,1 ; turn LED onbsf LEDST,0 ; set state to `1`returncf1downmovlw 0ffh ;movwf LEDCNTRbcf PORTB,1 ; turn LED offclrf LEDST ; set state to `0`returninitcf1movlw 0ffh ; set initial timemovwf LEDCNTRbsf SBITS,2 ; schedule LED taskbcf PORTB,1 ; turn LED offmovlw H`00` ; set LED task to state 0(off)movwf LEDSTmovlw H`03`movwf LEDSELECT ; set `func ptr`: 1 (continuous 512ms flash)returnendifstatusledmovlw high statusledjmpmovwf PCLATHmovlw H`03`andwf LEDST,Wmovwf TEMP6bcf STATUS,Crlf TEMP6,Fbcf STATUS,Crlf TEMP6,Waddlw statusledjmpbtfsc STATUS,Cincf PCLATH,Fmovwf PCLstatusledjmpmovlw high statusled0movwf PCLATHgoto statusled0nopmovlw high statusled1movwf PCLATHgoto statusled1nopmovlw high statusled2movwf PCLATHgoto statusled2nopmovlw high statusled3movwf PCLATHgoto statusled3statusled0bsf PORTB,1 ; turn on ledbsf LEDST,0 ; set state to 1movlw H`80` ; load the led taskusePage1btfss STATWORD,0 ; counter based onmovlw H`20` ; the value of the bitusePage0movwf LEDCNTR ; being displayedreturnstatusled1bcf PORTB,1 ; turn off ledbcf LEDST,0bsf LEDST,1 ; set state to 2movlw H`40`movwf LEDCNTR ; load cntr with postamble timereturnstatusled2;usePage1movlw H`07`; ; mask off upper 5 bitsandwf STATCNTL,W ; to see if more bitsusePage0movwf TEMP6movlw high statusledmoremovwf PCLATHdecfsz TEMP6 ; did we reach zerogoto statusledmore ; no; we reached zero,usePage0clrf LEDSTclrf LEDCNTRclrf LEDSELECTbcf SBITS,2bcf PORTB,1usePage1bcf STATCNTL,7usePage0returnstatusled3; this is a dummy state, we'll handle large digit count msg's hereusePage0bsf LEDST,0 ; set state to 3returnstatusledmoreusePage1movlw H`f8` ; mask off low 3 bitsandwf STATCNTL,FusePage0movlw H`7` ; don't forget to put updated count in cntl regandwf TEMP6,WusePage1iorwf STATCNTL,F ; `or` in the cnt...preserve upper 5 bitsusePage0bcf STATUS,CusePage1rrf STATWORD,F ; roll the new bitusePage0statusledinitmovlw H`80`movwf LEDCNTR ; load cntr with preamble timebcf PORTB,1bsf SBITS,2clrf LEDST ; set state to 0movlw H`02`movwf LEDSELECTusePage1bsf STATCNTL,7 ; set status-in-process flagusePage0return#endif#if useIRgLEDTASKusePage0movlw high endgledmovwf PCLATHmovf CCNT0,Waddlw 2 ; time slice offsetandlw H`03`btfss STATUS,Z ; every 4msgoto endgleddecf gLEDCNTR,Fbtfss STATUS,Z ; has task counter reached zero?goto endgled ; no, skip the rest.movlw high gLEDTBLmovwf PCLATHcall gLEDTBLendgledreturngLEDTBL; bits 0-3 of gLEDSELECT are LED `function ptr`movlw 00fhandwf gLEDSELECT,Waddwf PCLgoto gmfflash; goto contflash1; goto pflash1; goto pflash2gmfflash; the state of the single flash led is held in bit 0 of LEDST.; 0: initial dead spot for relay; 1: onbtfsc gLEDST,0 ; what state are we in?goto gmfstop ; just finished state 1, so stopmovlw 0ffh ; just finished state 0, need tomovwf gLEDCNTR ; set `on` timemovlw H`0ef`andwf PORTA,F; bsf PORTA,4 ; turn LED on bsf gLEDST,0returngmfstopbcf CBITS3,7 ; unschedule LED taskclrf gLEDSTclrf gLEDCNTRclrf gLEDSELECTmovlw H`010`iorwf PORTA,F; bcf PORTA,4 ; turn LED offreturninitgmfbtfsc CBITS3,7 ; if led is already running just returnreturnmovlw H`00` ; set LED task to state 0(off)movwf gLEDSTclrf gLEDSELECT ; set function 0 (single flash)movlw 02h ; set initial 100ms dead timemovwf gLEDCNTRbsf CBITS3,7 ; schedule LED taskmovlw H`010`iorwf PORTA,F; bcf PORTA,4 ; turn LED offreturn#endif;; relaytsk.asm; COPYRIGHT (C) MYTECH CORP. 1996, 1997; relaytask -; This task is called once every second. It handles; the countdown timer for the relay.relaytaskmovlw high endrelaymovwf PCLATHmovlw 01subwf relaylo,Fbtfss STATUS,Cdecf relayhi,Fmovf relaylo,Wiorwf relayhi,W ; if relayhi&relaylo are both zerobtfss STATUS,Z ; then the counter has just timed out.goto endrelay; The relay counter has just reached zero; turn off the lights and set appropriate flags.reloff bcf Occupied ; occupied bits off bcf RelayEnable ; lights off for less than x sec bcf SBITS,3 ; disable this taskclrf LOFFTIMELOclrf LOFFTIMEHIbsf Suppress ; enable suppress task; we need a mechanism to switch between 2 valuesmovlw H`08` ; this gives .about.3/8 sec US blindspot;movlw H`17` ; this gives .about.3sec IR blindspotmovwf sprscntbcf Relay ; relay offbcf DLCRelay ; dlc relay offif ADAPT==1movlw high endrelaymovwf PCLATHbtfss ASAEnablegoto endrelaymovlw high endrelayfalseonmovwf PCLATH; check to see if lights were only on for around the amount; of time the pot is set for. by definition, LONTIME>=TMRPOTmovf LONTIMELO,Wmovwf TEMP7movf LONTIMEHI,Wmovwf TEMP9movf TMRPOTLO,Wsubwf TEMP7,F ; lobyte[LONTIMELO - TMRPOTLO]btfss STATUS,C ; did the subtract borrow?decf TEMP9,F ; yesmovf TMRPOTHI,Wsubwf TEMP9,F ; hibyte[LONTIMEHI [- 1] - TMRPOTHI]btfss STATUS,Z ; if the high byte isn't zerogoto endrelayfalseon ; then LONTIME>TMRPOT+8 for sure.movlw H`f8`andwf TEMP7,W ; is the difference>=8?btfss STATUS,Zgoto endrelayfalseon ; yes?; The lights were on for less than 8 seconds more than the; timer pot setting. This is very likely indicative of a; false trip. The sensitivity should be reduced.btfsc InstallerTestMode ; if in test mode, no false-ongoto endrelayfalseonmovf mcntlo,Waddlw 2movwf FONREFbsf SBITS,6 ; schedule false trip correctionmovlw high endrelaymovwf PCLATHgoto endrelayendrelayfalseon;-----------------------------------------------------------------------; this was not a false on, so we need to look at the min; motion pulse while the lights were on. If it was; sufficiently large we can back off the sensitivity.;-----------------------------------------------------------------------usePage1movlw high endrelaymovwf PCLATHbtfss usminpulse,6 ; is it greater than 40h?goto endrelay ; no, so don't adjust.usePage0decf usonadj,F ; just down by 1 for now.endifendrelayusePage0movf usoffgain,Wmovwf usgainreturnturnOnLightsbtfss RelayEnable ; is the relay enabled?return ; nope...movlw high setrelaymovwf PCLATHcall setrelay ; note that if the lights are already ; on, this will simply reset the timer.bsf Occupied ; set occupied bitclrf LONTIMELOclrf LONTIMEHIclrf timesincemotloclrf timesincemothiif ADAPT==1 ; we need to determine which technology was most ; dominant on the motion hit. in a tie, choose ; us since it is naturally dominant. One or both ; of irdecision and usdecision will be >0. bsf CBITS2,0 ; choose us dominant btfsc pirdecision,7 ; is irdecision >0? goto turnonlbl1 ; no, so it must be us bcf CBITS2,0 ; choose ir dominant btfsc usdecision,7 ; is usdecision>0? goto turnonlbl1 ; no, so it must be ir ; they're both>0 so compare to find greatest movf usdecision,W subwf pirdecision,W ; ir - us btfsc STATUS,Z ; are they equal? goto turnonlbl1 ; yes. btfss STATUS,C ; is us>ir bsf CBITS2,0 ; yes.turnonlbl1 ; now save the appropriate quantity from the ; dominant technology sensor movf usdecision,W btfss CBITS2,0 movf pirdecision,W usePage1 movwf dominantTech usePage0 ; do the check for a possible false-off movlw high fshut movwf PCLATH btfss ASAEnable goto fshut movlw high nofshut movwf PCLATH movf LOFFTIMEHI,F ; the high byte must be zero btfss STATUS,Z goto nofshut movlw H`0f` ; 15sec subwf LOFFTIMELO,W ; w = lofftimelo - 15s btfsc STATUS,C ; if borrow, run false shut-off correction goto nofshut ; no borrow. ; we now need to handle the situation of the ; lights going out and false-on correction getting ; scheduled, then the lights being turned on ; immediately and false-off correction being scheduled ; To handle this, false-on correction is delayed for ; up to 2 minutes. If a false-off is detected AND ; false-on correction is currently scheduled, false-off ; correction wins the contradiction. movlw high fshut movwf PCLATH btfsc InstallerTestMode ; if in test mode, no false-off goto fshut bsf SBITS,5 btfss SBITS,6 ; is false-on correction scheduled? goto fshut ; no ; false-on correction is scheduled...unschedule it. clrf FONREF bcf SBITS,6 goto fshutnofshut ;--------------------------------------------------------------- ; false shut-off correction is not necessary. We ; need to determine if an adjustment to the ir ; threshold is necessary. We look at the ir noise ; estimate. If it is sufficiently small then we can ; lower the threshold(increase sensitivity). ;--------------------------------------------------------------- movlw high fshut movwf PCLATH usePage1 comf irNoiseEst,W addlw 1 ; -noise estimate in W andlw H`f0` ; is it greater than 15 below threshold? btfsc STATUS,Z ; if `and` is 0, then noise is too large goto fshut ; noise estimate too large! usePage0 decf iroffadj,F ; reduce threshold=increase sensitivityfshutendifusePage1movlw H`7f`movwf usminpulse ; init to max valueusePage0return;; setrelay will set the relay task counter to count; down from the proper number of 1s ticks. The; counter is derived from secondcnt which increments; every 1s (CCNT0 increments every 1ms).setrelaybtfss RelayEnable ; if the relay is enabled, turn on lightsreturn ; otherwise forget about it.srdoit ; (re)load the relay countdown timer; If less than 4 minutes since power on, timeout; is 2 minutes, unless the timer pot is set at; min in which case the timeout is 8 seconds; for 1hr OR until the pot is moved away from min.; Otherwise the min timeout is 8 minutes.; When 8 second min timeout is being used, no timer; adjustment is applied. This setting is usually; indicative of some kind of test mode.btfss InstallerTestModegoto srnotitm; we are in installer test mode, use 8 sec timeoutsruse8secminmovlw H`08` ; 8 second min timeoutmovwf relayloclrf relayhimovlw high setrelaylbl2movwf PCLATHgoto setrelaylbl2srnotitmbtfsc FourMinOnBitgoto sruse8minutemin ; use normal 8 minute timeoutmovlw H`78` ; setup for 2 min timeoutmovwf relayloclrf relayhimovlw high sraddadjmovwf PCLATHgoto sraddadjsruse8minuteminmovlw H`e0` ; 8 minute min timeoutmovwf relaylomovlw H`01`movwf relayhisraddadjmovf timeradjlo,W ; add any timer adjustmentaddwf relaylo,Fbtfsc STATUS,Cincf relayhi,Fmovf timeradjhi,Waddwf relayhi,Fsetrelaylbl2; turn on the relay if it isn't already onbtfss Relaybsf Relay; we need to add the check for daylight control; should check LightLevelbtfss DLCRelaybsf DLCRelaymovf usongain,Wmovwf usgain; if the lights are just being turned on, then; there is some bookkeeping we want to dobtfsc Occupied ; are lights already on?return ; yes, so skip the restbsf Occupiedbsf SBITS,3 ; enable the relay countdown taskreturnif 0;; setrelay will set the relay task counter to count; down from the proper number of 1s ticks. The; counter is derived from secondcnt which increments; every 1s (CCNT0 increments every 1ms).setrelaybtfss RelayEnable ; if the relay is enabled, turn on lightsreturn ; otherwise forget about it.srdoit; (re)load the relay countdown timer; if less than 60 minutes since power on, timeout; is 2 minutes, unless the timer pot is set at min in; which case the min timeout is 8 seconds; otherwise; the min timeout is 8 minutes.; When 8 second min timeout is being used, no timer; adjustment is applied. This setting is usually; indicative of some kind of test mode.btfsc SixtyMinOnBitgoto sruse8minutemin ; use normal 8 minute timeoutmovf TMRPOTHI,Wiorwf TMRPOTLO,Wbtfsc STATUS,Z ; if both zero, then `or` is zerogoto sruse8secminmovlw H`78` ; setup for 2 min timeoutmovwf relayloclrf relayhimovlw high sraddadjmovwf PCLATHgoto sraddadjsruse8secminmovlw H`08` ; 8 second min timeoutmovwf relayloclrf relayhimovlw high setrelaylbl2movwf PCLATHgoto setrelaylbl2sruse8minuteminmovlw H`e0` ; 8 minute min timeoutmovwf relaylomovlw H`01`movwf relayhisraddadjmovf timeradjlo,W ; add any timer adjustmentaddwf relaylo,Fbtfsc STATUS,Cincf relayhi,Fmovf timeradjhi,Waddwf relayhi,Fsetrelaylbl2; turn on the relay if it isn't already onbtfss PORTB,2bsf PORTB,2movf usongain,Wmovwf usgain; if the lights are just being turned on, then; there is some bookkeeping we want to dobtfsc Occupied ; are lights already on?return ; yes, so skip the restbsf Occupiedbsf SBITS,3 ; enable the relay countdown taskreturnendif; suppress -; This task is called at a 1ms rate. It handles the timed; process of disabling the relay for a given period. When; the timer reaches zero, the relay is enabled once again.suppressmovlw high endsuppressmovwf PCLATHmovf CCNT0,Wandlw H`7f` ; do it every 128msbtfss STATUS,Zgoto endsuppressdecfsz sprscnt,Fgoto endsuppressbsf RelayEnable ; un-suppress relaybcf Suppress ; disable this taskendsuppressreturn;**********************************************; o2eeprom.asm;; Omni II E2BUF routines;; 8-16-96 JJF;; Portions from Microchip Example code;**********************************************;**********************************************; in our application the slave address is always the same; so we can save that register.#define SLAVE B`10100000`;**********************************************E2Init macroclrf E2FLAGclrf E2ERRmovlw B`00111011` ; I2C master mode enabledmovwf SSPCONusePage1movf PIE1,Wandlw B`11000111` ; disable SSP interrupt & SPI INT'smovwf PIE1usePage0movf PORTC,Wandlw B`11100111`; andlw B`00111111`movwf PORTC ; Set SCL & SDA to low when not tri-statedusePage1movf TRISC,Wiorlw B`11011000`; iorlw B`11000000` ; tristate the SCL and SDA movwf TRISCusePage0endmE2WriteValue macro address,valuemovlw addressmovwf E2ADDRmovlw valuemovwf DATAOclrf E2ERRbsf EEPROMWTASKbtfsc EEPROMWTASKgoto $ - 1endmE2WriteReg macro addr,regmovlw addrmovwf E2ADDRmovf reg,Wmovwf DATAOclrf E2ERRbsf EEPROMWTASKbtfsc EEPROMWTASKgoto $ - 1endmE2WriteRegIndirect macro addrreg,regmovf addrreg,Wmovwf E2ADDRmovf reg,Wmovwf DATAOclrf E2ERRbsf EEPROMWTASKbtfsc EEPROMWTASKgoto $ - 1endmE2Read macro addressmovlw addressmovwf E2ADDRclrf E2ERRbsf EEPROMRTASKbtfsc EEPROMRTASKgoto $ - 1endmE2ReadIndirect macro addressmovf address,Wmovwf E2ADDRclrf E2ERRbsf EEPROMRTASKbtfsc EEPROMRTASKgoto $ - 1endmStoreVarsE2 macrousePage0E2WriteReg 1,usonadjE2WriteReg 2,usoffadjE2WriteReg 3,iroffadjE2WriteReg 4,CCNT3clrwbtfsc CntrRollOveriorlw H`01` ; bit 0 is cntrrollover bitbtfsc ASATimerReturniorlw H`02` ; bit 1 is asatimerreturn enable bitbtfsc ASATimerReturnStateiorlw H`04` ; bit 2 is asatimerreturnstate bitmovwf MTEMPE2WriteReg 5,MTEMPE2WriteReg 6,timeradjloE2WriteReg 7,timeradjhiusePage1movf timerLearnCoef,WusePage0movwf MTEMPE2WriteReg 8,MTEMPendm;---------------------------------------------------------------------------------------------------------------------------------------; BYTE-WRITE, write one byte to EEPROM device;---------------------------------------------------------------------------------------------------------------------------------------; Input : DATAO = data to be written; ADDR = EEPROM data address; SLAVE = device address (1010xxx0); Output : Data written to EEPROM device;---------------------------------------------------------------------------------------------------------------------------------------WRBYTEbcf STATUS,RP0; movf SLAVE,W ; Put SLAVE addressmovlw SLAVEmovwf TXBUF ; in xmit buffercall BSTART ; Generate START bitcall TX ; Output SLAVE addressbcf STATUS,RP0movf ADDR,W ; Put WORD addressmovwf TXBUF ; in xmit buffercall TX ; Output WORD addressbcf STATUS,RP0movf DATAO,W ; Move DATAmovwf TXBUF ; into buffercall TX ; Output DATA and detect acknowledgementcall BSTOP ; Generate STOP bitusePage0movlw D`11` ; setup 11ms delay after writeaddwf CCNT0,Wmovwf E2DLYbsf E2FLAG,1 ; enable eeprom wait taskreturn;---------------------------------------------------------------------------------------------------------------------------------------; TRANSMIT 8 data bits subroutine;---------------------------------------------------------------------------------------------------------------------------------------; Input : TXBUF; Output : Data transmitted to EEPROM device;---------------------------------------------------------------------------------------------------------------------------------------TXbcf STATUS,RP0movlw .8 ; Set counter for eight bitsmovwf COUNTTXLPbcf EEPROM,DO ; Default 0 bit outbtfsc TXBUF,7 ; If shifted bit = 0, data bit = 0bsf EEPROM,DO ; otherwise data bit = 1movlw high BITOUTmovwf PCLATHcall BITOUT ; Send bitbcf STATUS,RP0rlf TXBUF ; Rotate TXBUF leftskpc ; f(6) ------> f(7)bcf TXBUF,0 ; f(7) ------> carryskpnc ; carry ------> f(0)bsf TXBUF,0movlw high TXLPmovwf PCLATHdecfsz COUNT ; 8 bits done?goto TXLP ; No.movlw high BITINmovwf PCLATHcall BITIN ; Read acknowledge bitbcf STATUS,RP0movlw high ERRmovwf PCLATHmovlw 3btfsc EEPROM,DI ; Check for acknowledgementcall ERR ; No acknowledge from devicebcf STATUS,RP0retlw 0;---------------------------------------------------------------------------------------; Single bit data transmit from PIC to serial EEPROM;---------------------------------------------------------------------------------------; Input : EEPROM register, bit DO; Output : Bit transmitted over I2C; Error bits set as necessary;---------------------------------------------------------------------------------------BITOUTmovlw high BIT0movwf PCLATHbtfss EEPROM,DOgoto BIT0bsf STATUS,RP0bsf TRISC,SDA ; Output bit 0movlw high CLK1movwf PCLATHmovlw 2bcf STATUS,RP0btfsc PORTC,SDA ; Check for error code 2goto CLK1movlw high ERRmovwf PCLATHcall ERRmovlw high CLK1movwf PCLATHgoto CLK1 ; SDA locked low by deviceBIT0bsf STATUS,RP0bcf TRISC,SDA ; Output bit 0nop ; DelaynopnopCLK1bsf STATUS,RP0bsf TRISC,SCL ; Attempt to set SCL highmovlw high BIT2movwf PCLATHmovlw 1 ; Error code 1bcf STATUS,RP0btfsc PORTC,SCL ; SCL locked low?goto BIT2 ; No.movlw high ERRmovwf PCLATHcall ERR ; Yes, set errorBIT2nop ; Timing delaynopnopnopbsf STATUS,RP0bcf TRISC,SCL ; Return SCL to lowbcf STATUS,RP0retlw 0;---------------------------------------------------------------------------------------------------------------------------------------; BYTE-READ, read one byte from serial EEPROM device;---------------------------------------------------------------------------------------------------------------------------------------; Input : ADDR = source address; SLAVE = device address (1010xxx0); Output : DATAI = data read from serial EEPROM;---------------------------------------------------------------------------------------------------------------------------------------RDBYTEbcf STATUS,RP0; movf SLAVE,W ; Move SLAVE addressmovlw high BSTARTmovwf PCLATHmovlw SLAVEmovwf TXBUF ; into xmit buffer (R/W = 0)call BSTART ; Generate START bitmovlw high TXmovwf PCLATHcall TX ; Output SLAVE address. Check ACK.bcf STATUS,RP0movf ADDR,W ; Move WORD addressmovwf TXBUF ; into xmit buffermovlw high TXmovwf PCLATHcall TX ; Output WORD address. Check ACK.movlw high BSTARTmovwf PCLATHcall BSTART ; START READ (if only one device isbcf STATUS,RP0 ; connected to the I2C bus); movf SLAVE,Wmovlw SLAVEmovwf TXBUFbsf TXBUF,0 ; Specify READ mode (R/W = 1)movlw high TXmovwf PCLATHcall TX ; Output SLAVE addressmovlw high RXmovwf PCLATHcall RX ; READ in data and acknowledgemovlw high BSTOPmovwf PCLATHcall BSTOP ; Generate STOP bitbcf STATUS,RP0movf RXBUF,W ; Save data from buffermovwf DATAI ; to DATAI file register.return;---------------------------------------------------------------------------------------------------------------------------------------; RECEIVE eight data bits subroutine;---------------------------------------------------------------------------------------------------------------------------------------; Input : None; Output : RXBUF = 8-bit data received;---------------------------------------------------------------------------------------------------------------------------------------RXbcf STATUS,RP0movlw .8 ; 8 bits of datamovwf COUNTclrf RXBUF;RXLPrlf RXBUF ; Shift data to bufferskpcbcf RXBUF,0 ; carry ------> f(0)skpncbsf RXBUF,0call BITINbcf STATUS,RP0btfsc EEPROM,DIbsf RXBUF,0 ; Input bit =1decfsz COUNT ; 8 bits?goto RXLPbsf EEPROM,DO ; Set acknowledge bit = 1call BITOUT ; to STOP further inputretlw 0;---------------------------------------------------------------------------------------------------------------------------------------; Single bit receive from serial EEPROM to PIC;---------------------------------------------------------------------------------------------------------------------------------------; Input : None; Output : Data bit received;---------------------------------------------------------------------------------------------------------------------------------------BITINbsf STATUS,RP0bsf TRISC,SDA ; Set SDA for inputbcf STATUS,RP0bcf EEPROM,DIbsf STATUS,RP0bsf TRISC,SCL ; Clock highmovlw high BIT1movwf PCLATHmovlw 1bcf STATUS,RP0btfsc PORTC,SCL ; Skip if SCL is highgoto BIT1movlw high ERRmovwf PCLATHcall ERRBIT1movlw high ACKOKmovwf PCLATHbcf STATUS,RP0btfss PORTC,SDA ; Read SDA pin, for ACK lowgoto ACKOKbsf EEPROM,DI ; DI = 1ACKOKbsf STATUS,RP0nop ; Delaynopnopbcf TRISC,SCL ; Return SCL to lowbcf STATUS,RP0retlw 0;---------------------------------------------------------------------------------------------------------------------------------------; DELAY, Provide a 10.78mS delay;---------------------------------------------------------------------------------------------------------------------------------------; Input : None; Output : None;---------------------------------------------------------------------------------------------------------------------------------------Delaybcf STATUS,RP0movlw 7movwf TEMP6clrf TEMP7 ;clear last locationdly1nopmovlw high dly1movwf PCLATHdecfsz TEMP7 ;reduce countgoto dly1 ;Inner loop time = 1.54mSdecfsz TEMP6goto dly1 ;Total time = 10.78mSretlw 0;*****************************************************; DELAY, Provide a 10.78mS delay...11ms;*****************************************************E2Delay10msmovlw D`11`addwf CCNT0,Wmovwf E2DLYe2dly1movf CCNT0,Wsubwf E2DLY,Wbtfsc STATUS,Z ; is it exactly 11ms?goto e2dlyend ; yesbtfsc STATUS,C ; did the subtract borrow (more than 11ms)goto e2dly1 ; no...wait longere2dlyendreturn;---------------------------------------------------------------------------------------------------------------------------------------; START bit generation routine;---------------------------------------------------------------------------------------------------------------------------------------; input : none; output : initialize bus communication;---------------------------------------------------------------------------------------------------------------------------------------;;Generate START bit (SCL is high while SDA goes from high to lowtransition);and check status of the serial clock.BSTARTbsf STATUS,RP0bsf TRISC,SDA ; Make sure SDA is highbsf TRISC,SCL ; Set clock highmovlw 1 ; Ready error status code 1bcf STATUS,RP0btfss PORTC,SCL ; Locked?call ERR ; SCL locked low by device, flag errorbsf STATUS,RP0bcf TRISC,SDA ; SDA goes low during SCL highnop ; Timing adjustment, 1uS @4MHznopnopbcf TRISC,SCL ; Start clock trainbcf STATUS,RP0RETLW 0;---------------------------------------------------------------------------------------------------------------------------------------; STOP bit generation routine;---------------------------------------------------------------------------------------------------------------------------------------; Input : None; Output : Bus communication, STOP condition;---------------------------------------------------------------------------------------------------------------------------------------;;Generate STOP bit (SDA goes from low to high during SCL high state);and check bus conditions.BSTOPbsf STATUS,RP0bcf TRISC,SDA ; Return SDA to lowbsf TRISC,SCL ; Set SCL highnopnopnopmovlw high ERRmovwf PCLATHmovlw 1 ; Ready error code 1bcf STATUS,RP0btfss PORTC,SCL ; High?call ERR ; No, SCL locked low by devicebsf STATUS,RP0bsf TRISC,SDA ; SDA goes from low to high during SCL highmovlw high ERRmovwf PCLATHmovlw 4 ; Ready error code 4btfss TRISC,SDA ; High?call ERR ; No, SDA bus not release for STOPbcf STATUS,RP0retlw 0;---------------------------------------------------------------------------------------------------------------------------------------; Two wire/I2C -- CPU communication error status table and subroutine;---------------------------------------------------------------------------------------------------------------------------------------; input : W-reg = error code; output : ERCODE = error code; FLAG(ERROR) = 1;; code error status mode;; 1 : SCL locked low by device (bus is still busy); 2 : SDA locked low by device (bus is still busy); 3 : No acknowledge from device (no handshake); 4 : SDA bus not released for master to generate STOP bit;---------------------------------------------------------------------------------------------------------------------------------------;;Subroutine to identify the status of the serial clock (SCL) and serialdata;(SDA) condition according to the error status table. Codes generatedare;useful for bus/device diagnosis.ERRbcf STATUS,RP0btfss FLAG,ERRORFLG ; If not first error, do not change codemovwf ERCODE ; Save error codebsf FLAG,ERRORFLG ; Set error flagretlw 0;; MUXtsk.asm; Copyright(C) 1996, 1997 Mytech;; test code for the external mux; modified by:jdw;;reads the PORTB latch and uses RB3 and RB4 to select;a location to record an a/d conversion from RA1;and a bit read from RC0;RB4 .backslash. RB3 .backslash. a/d loc. .backslash. bit loc.; 0 .backslash. 0 .backslash. MUXpot0 .backslash. MUXdip,0; 0 .backslash. 1 .backslash. MUXpot1 .backslash. MUXdip,1; 1 .backslash. 0 .backslash. MUXpot2 .backslash. MUXdip,2; 1 .backslash. 1 .backslash. MUXpot3 .backslash. MUXdip,3; remember that the switches are electrically negative logic.; we'll invert the logic when we read the switches.runMUXT movlw H`09` ;Fosc(1/32)movwf ADCON0 ;next reading will be RA1clrf TEMP8movf PORTB,0 ;get RB3 and RB4andlw H`18`movwf TEMP8bcf STATUS,Crrf TEMP8,fbcf STATUS,Crrf TEMP8,f ;look at currentbcf STATUS,Crrf TEMP8,f ;state of PortB latchconvM bsf ADCON0,2 ;start A/D conversionmovlw high waitMmovf PCLATHwaitM btfsc ADCON0,2goto waitM ; wait on GO/DONE bitadM movf ADRES,Waddlw H`80`movwf TEMP6movlw high muxtblstartmovwf PCLATHbcf STATUS,Crlf TEMP8,Fbcf STATUS,Crlf TEMP8,W ; shift MUXclk up to jump 4 lines worthaddlw low muxtblstartbtfsc STATUS,Cincf PCLATH,Fmovwf PCLmuxtblstartmovlw high mux00movwf PCLATHgoto mux00nopmovlw high mux01movwf PCLATHgoto mux01nopmovlw high mux10movwf PCLATHgoto mux10nopmovlw high mux11movwf PCLATHgoto mux11nopMUXendmux00movf TEMP6,W ; record case 0 info.movwf TEMP7 ; us sensivity potbcf ASAEnablebtfss PORTC,0 ; invert switch logicbsf ASAEnablebsf PORTB,3 ; set input 01 on MUXbcf PORTB,4; we need to apply the square law adjustment; to the us sensitivity setting.movlw high gainlookupmovwf PCLATHmovlw H`80`addwf TEMP7,F ; make pot reading unipolarbcf STATUS,Crrf TEMP7,Fbcf STATUS,Crrf TEMP7,Fbcf STATUS,Crrf TEMP7,Fbcf STATUS,Crrf TEMP7,Wmovwf usongainusePage1movwf usPotusePage0call gainlookupmovwf usongainbtfsc Occupiedmovwf usgainif useGSB==0 movwf usoffgainendifif useGSB==1 movlw high muxnogsb movwf PCLATH btfss GSBEnable ; if gsb switch is turned off, goto muxnogsb ; then just set offgain to ongain ;----------------------------------------------------------- ; otherwise we set offgain to .75 of ongain ; the motivation is to prevent hallway falsing ;----------------------------------------------------------- bcf STATUS,C rrf usongain,W ; 1/2 usongain in W movwf usoffgain ; 1/2 usongain in usoffgain bcf STATUS,C rrf usoffgain,W ; 1/4 usongain in W addwf usoffgain,F ; 3/4 usongain in usoffgain movlw high muxgsbend movwf PCLATH goto muxgsbendmuxnogsb ; gsb is disabled movf usongain,W movwf usoffgainmuxgsbend ; be sure to set usgain properly movf usongain,W btfsc Occupied movwf usgain movf usoffgain,W btfss Occupied movwf usgainendifreturnmux01movf TEMP6,W ; record case 1 info.movwf TEMP7 ; ir sensitivity potbcf ATAEnablebtfss PORTC,0 ; invert switch logicbsf ATAEnablebcf PORTB,3 ; set input 10 on MUXbsf PORTB,4; turn the pir sensitivity setting into; a threshold.movlw H`80`addwf TEMP7,F ; make pot reading unipolarmovf TEMP7,WusePage1movwf pirPotrrf pirPot,Frrf pirPot,Frrf pirPot,Frrf pirPot,Wandlw H`0f` ; mask off bits shifted inmovwf pirPotusePage0;@@@@@ temp for pll testing; movf TEMP7,W; movwf lfdtemp;@@@@@comf TEMP7,F ; since the pot is a sensitivity ; setting, it is intuitive for a ; large setting to give greater ; range. However, a greater range ; is more consistent with a smaller ; threshold, so we use the complement ; of the pot setting as the threshold.bcf STATUS,Crrf TEMP7,W ; /2movwf PIRthreshmovwf TEMP7bcf STATUS,Crrf TEMP7,F ; /4bcf STATUS,Crrf TEMP7,F ; /8bcf STATUS,Crrf TEMP7,W ; /16subwf PIRthresh,Fmovlw H`0f` ; we add an offset to the thresholdaddwf PIRthresh,F ; to get past the noise floor.returnmux10movf TEMP6,W ; record case 2 info.movwf TEMP7 ; timer potbcf GSBEnablebtfss PORTC,0 ; invert switch logicbsf GSBEnablebsf PORTB,3 ; set input 11 on MUXbsf PORTB,4; we need to scale the timer pot reading to; 1 sec units. The raw reading is in 8sec units; so shifting by 3 will be necessary.movlw H`80`addwf TEMP7,F ; make pot reading unipolarmovf TEMP7,Wmovwf TMRPOTLOclrf TMRPOTHIbcf STATUS,Crlf TMRPOTLO,Frlf TMRPOTHI,Fbcf STATUS,Crlf TMRPOTLO,Frlf TMRPOTHI,Fbcf STATUS,Crlf TMRPOTLO,Frlf TMRPOTHI,F; movlw H`E0` ; 8 minute(480 sec.) min.; addwf TMRPOTLO,F; btfsc STATUS,C; incf TMRPOTHI,F; movlw H`01`; addwf TMRPOTHI,F;-----------------------------------------------------------; This is where we determine test mode entry; via timer pot setting.;-----------------------------------------------------------movlw high tmrmux.sub.-- notitmmovwf PCLATHbtfss InstallerTestMode ; are we in installer test mode?goto tmrmux.sub.-- notitm ; no; we are in installer test modemovlw H`fe`andwf TEMP7,W ; is new pot reading > 1?btfsc STATUS,Z ; not if `and` is zeroreturn; the timer pot has been increased to > 1bcf InstallerTestModeclrf ITMClktmrmux.sub.-- notitm; we are not in installer test mode, should we bemovlw high tmrmuxendmovwf PCLATHmovlw H`fe`andwf lastTimer,W ; is last timer > 1?btfsc STATUS,Z ; not if `and` = 0goto tmrmuxend ; lastTimer <= 1movlw H`fe`andwf TEMP7,W ; is new timer < 2btfss STATUS,Z ; yes if `and` = 0goto tmrmuxend ; new timer >= 2bsf InstallerTestModeclrf ITMClktmrmuxendmovf TEMP7,W ; new timermovwf lastTimerreturnmux11movf TEMP6,W ; record case 3 info.movwf LightLevel ; muxpot3bcf StrongAirCompensatebtfss PORTC,0 ; invert switch logicbsf StrongAirCompensatebcf PORTB,3 ; set input 00 on MUXbcf PORTB,4return; o2dtinit.asm; COPYRIGHT (C) MYTECH CORP. 1996, 1997; initialization and setup code for Omni IIsetTrisRegsusePage1movlw B`11111001`movwf PORTC ; set C1 & C2 to outputs; C3, C4, C1, C5, ; C6 & C7 are set as inputs, setting C6 & 7 to outputs ; causes USART to screw up on transmissionmovlw 0ebh ; per mfg instructions, reset PORTA to allmovwf TRISAmovlw B`11000001`movwf TRISBusePage0returno2init macrousePage1bsf H`8e`,0 ; reset power on reset bitbsf H`8e`,1 ; reset brown out reset bitusePage0clrf INTCONbcf INTCON,GIE ; disable global interruptsclrf PORTA ; clear ports prior to setting directionclrf PORTBclrf PORTCusePage1clrf OPTREG ; clear OPTION register, tmr0src=fosc/4; The prescaler has to be set according to the divisor needed; to get the 1Khz sample rate.; all carrier versions use a prescale of 16bsf OPTREG,0 ; prescale set to 1:16bsf OPTREG,1bcf OPTREG,3 ; assign prescaler to RTCCbsf OPTREG,7 ; enable the weak pullupsclrf TRISB ; clear register prior to setting directionclrf TRISCclrf TRISAmovlw high setTrisRegsmovwf PCLATHcall setTrisRegsusePage0clrf INTCONmovlw timvala ; initialize RTCCmovwf RTCCmovlw 99h ; 1001 1001 enable ain3 a/dmovwf ADCON0usePage1movlw 02h ; 0000 0000 set ra0-ra3 analog, vref=VDDmovwf ADCON1usePage0bcf INTCON,GIE ; disable global interruptsusePage1clrf PIE1 ; disable peripheral interruptsclrf PIE2 ; disable peripheral interrupts; CCP setupusePage0movlw H`04`movwf T2CONusePage1movlw tmr2valmovwf PR2usePage0movlw H`0c`movwf CCP1CONmovlw tmr2valmovwf TEMPLObcf STATUS,Crrf TEMPLO,Wmovwf CCPR1Lclrf CCPR1Hmovlw H`0c`movwf CCP2CON; movlw H`02` ; this is the falling edge of themovlw H`15`movwf CCPR2L ; sample pulse. (was 02h for omni 1)clrf CCPR2H ; 0.times.37 was determined by experimentation;---------------------------------E2Init ; eeprom init; #if 0usePage1; serial port setup stuff; movlw H`3f` ; 9600 baud @9.8304Mhz movlw H`0f`movwf SPBRG; movlw b`10100100` ;Async, High baud rate movlw B`10100000` movwf TXSTAusePage0movlw b`10000000` ;Enable serial portmovwf RCSTA; end serial port setup; endif;stck; movlw 0ffh ; initialize FSR to end of page 1 SRAM; movwf FSR ; This will be our stack. Stay above B0h.; usePage0; movlw H`5f`; movwf H`7f` movlw H`20` movwf FSRmovlw high initloopmovwf PCLATHinitloopclrf INDFincf FSR,Fbtfss FSR,7goto initloopmovlw 01hmovwf PORTBmovlw 80hmovwf ADRESmovlw 0ffhmovwf usgainmovwf usongainmovwf usoffgain; clrf usoffadj; clrf usonadj; clrf iroffadj; clrf uspscnt; clrf irpscnt; clrf LONTIMEHI; clrf LONTIMELO; clrf LOFFTIMEHI; clrf LOFFTIMELOmovlw H`03`movwf SECONDrefhimovlw H`e8`movwf SECONDreflo ; 3e8h is 1000movlw D`60`movwf mcntref; clrf mcntlo; clrf mcnthi; clrf SBITS; bcf SBITS,2 ; disable LED (for test/debug); bcf SBITS,4; bcf SBITS,5; bcf SBITS,6; clrf CCNT0; clrf CCNT1; clrf CCNT2; clrf CCNT3; clrf CBITS1; clrf CBITS2; clrf CBITS3; bsf CBITS2,7 ; high gain mode bsf CBITS2,0 ; set US dominant by default; clrf avgturnon; movlw H`d0`usePage1; movlw H`ff`; movwf avgtimesincemot ; this inits the avgtime to over 4min; clrf timesincemotlo; clrf timesincemothi clrf FONCNT clrf FOFFCNT clrf uspulsecnt0 clrf uspulsecnt1 clrf uspulsemax movlw H`7f`movwf usminpulse ; max valueusePage0movlw H`f0` ; 1sec ticks pwr-on timeout -- 4minmovwf TMRPOTLO; clrf TMRPOTHI; clrf timeradjlo; clrf timeradjhi; movlw H`0`; movwf optionsbsf RelayEnablemovlw high setrelaymovwf PCLATHcall setrelaybsf Occupiedmovlw high initsgf ; initialize the LEDmovwf PCLATHcall initsgfif useDA==0movlw high initgmf ; initialize the LEDmovwf PCLATHcall initgmfendifusePage1clrf STATCNTLclrf STATWORDusePage0clrf NTRGATEstate; bcf Interrogating; movlw H`0a2`; movwf FSRenableIntsbsf INTCON,5 ; enable RTCC interrupt;*****************************************************; for development only;*****************************************************if useDA==1 doutd.sub.-- sendifif useLCD==1 movlw high LCDInit movwf PCLATH call LCDInitendif;*****************************************************;********************************************; This initial delay is for the EEPROM to initialize *;********************************************movlw D`11`addwf CCNT0,Wmovwf MTEMPinitdly1movf CCNT0,Wsubwf MTEMP,Wmovlw high initdlyendmovwf PCLATHbtfsc STATUS,Z ; is it exactly 11ms?goto initdlyend ; yesmovlw high initdly1movwf PCLATHbtfsc STATUS,C ; did the subtract borrow (more than 11ms)goto initdly1 ; no...wait longerinitdlyend; remember the sensor task is disabled until nowmovlw high eepromnotinitmovwf PCLATHE2Read 0 ; read eeprom address 0movf DATAI,Wmovwf MTEMPsublw H`5a`btfss STATUS,Zgoto eepromnotinit; we read the eeprom here and initialize the variablesE2Read 1 ; read eeprom address 1movf DATAI,Wmovwf usonadjE2Read 2 ; read eeprom address 2movf DATAI,Wmovwf usoffadjE2Read 3 ; read eeprom address 3movf DATAI,Wmovwf iroffadjE2Read 4 ; read eeprom address 4movf DATAI,Wmovwf CCNT3E2Read 5 ; read eeprom address 5bcf CntrRollOverbtfsc DATAI,0 ; the stored cntrrollover bit just readbsf CntrRollOverbcf ASATimerReturnbtfsc DATAI,1bsf ASATimerReturn ; store ASATimerReturn enable bitbcf ASATimerReturnStatebtfsc DATAI,2bsf ASATimerReturnState ; store ASATimerReturnState bitE2Read 6 ; read eeprom address 6movf DATAI,Wmovwf timeradjloE2Read 7 ; read eeprom address 7movf DATAI,Wmovwf timeradjhiE2Read 8 ; read eeprom address 8movf DATAI,WusePage1movwf timerLearnCoefusePage0movlw high endeeprominitmovwf PCLATHgoto endeeprominiteepromnotinitE2WriteValue 0,H`5a`E2WriteValue 1,0 ; usonadjE2WriteValue 2,0 ; usoffadjE2WriteValue 3,0 ; iroffadjE2WriteValue 4,0 ; CCNT3E2WriteValue 5,0 ; bit fields, bit 0: cntrrolloverE2WriteValue 6,0 ; timeradjloE2WriteValue 7,0 ; timeradjhiE2WriteValue 8,0 ; timerLearnCoefendeeprominitbsf SBITS,0 ; enable sensor task; return endm;; ustask.asm; source code for the ultrasonic sensor task; COPYRIGHT (C) MYTECH CORP. 1996, 1997; JJFultrasonictask ; level 1 on the pic stackbsf ADCON0,2 ; start A/Dbcf USSense ; clear the us sense bitbsf PORTB,5btfss CCNT0,0bcf PORTB,5 ; external dithermovlw high snsrad1movwf PCLATHsnsrad1 btfsc ADCON0,2 ; wait til conversion completegoto snsrad1movf ADRES,0 ; get a/d inputaddlw 80hmovwf TEMP9 ; save first a/d input; if useDA==1; movf TEMP9,W; addlw H`80`; doutd 0 ; 000; endifhighpass; if useDA==1; movf TEMPHI,W; addlw H`80`; doutd 1 ; 001; endifmovf TEMPHI,Wmovwf TEMP8movf TEMPLO,Wmovwf TEMP6cfilt movlw high bpfmovwf PCLATHcall bpfsnsrlbl3movlw high snsrlbl3movwf PCLATH; if useDA==1 ;@@@@@@; movf TEMPHI,W; addlw H`80`; doutd 2 ; 010;@@@@@@; endifabs btfss TEMPHI,7 ; is it negative? goto abspos ; noabsneg comf TEMPHI,Fcomf TEMPLO,Waddlw 1btfsc STATUS,Cincf TEMPHI,Fmovwf TEMPLOabspos nop ; abs(BPF) in TEMPHI:TEMPLOxgain ;--------------------------------------------------------; Gain stage: this is an 8 by 16 mult; We'll only use the hi 16 bits of the; result.; 8 by 16 mult: (abcd) * (ef); ab*ef*256 + cd*ef; We'll throw away the low byte of cd*ef;--------------------------------------------------------if ADAPT==1 movlw high usnogainasa movwf PCLATH btfss ASAEnable goto usnogainasa movlw high gainadjust movwf PCLATH call gainadjust ; adjusted gain in W movwf TEMP7 movlw high usgainasaend movwf PCLATH goto usgainasaendusnogainasa movf usoffgain,W btfsc Occupied movf usongain,W movwf TEMP7usgainasaendelse movf usoffgain,W btfsc Occupied movf usongain,W movwf TEMP7endifmovlw high mult8.times.8movwf PCLATHmovf TEMPHI,Wmovwf TEMP8 ; store hi byte of .vertline.BPF.vertline.movf TEMPLO,Wmovwf TEMP9 ; lo byte of .vertline.BPF.vertline. in TEMP9movf TEMP7,W ; we now have the gain term in Wcall mult8.times.8 ; TEMPHI:TEMPLO = W * LO(.vertline.BPF.vertline.)movf TEMPHI,W ; we only want the hi byte of this mult,movwf TEMP6 ; since we are doing an 8.times.16 and taking ; only the hi 16 bits of the result.movf TEMP8,Wmovwf TEMP9 ; hi byte of .vertline.BPF.vertline. in TEMP9movf TEMP7,W ; we now have the gain term in Wcall mult8.times.8 ; TEMPHI:TEMPLO = W * HI(.vertline.BPF.vertline.)movf TEMP6,Waddwf TEMPLO,Fbtfsc STATUS,Cincf TEMPHI,Fbcf STATUS,Crlf TEMPLO,Frlf TEMPHI,Fbcf STATUS,Crlf TEMPLO,Frlf TEMPHI,Fbcf STATUS,Crlf TEMPLO,Frlf TEMPHI,Fmovlw high nogainsatmovwf PCLATHbtfss TEMPHI,7goto nogainsatmovlw H`7f`movwf TEMPHImovlw H`ff`movwf TEMPLOnogainsat; if useDA==1 ;@@@@@@; movf TEMPHI,W; addlw H`80`; doutd 3 ; 011;@@@@@@; endifmovf TEMPHI,Wmovwf TEMP9movf TEMPLO,Wmovwf TEMP7 ; rect*gain in TEMP9:TEMP7;------------------------------------------; do the decimation filter here; This gives a cutoff of .about.2.5hz;------------------------------------------movlw high genlpmovwf PCLATHmovf FSR,Wmovwf TMPFSRmovlw 6movwf FCOUNTmovlw decimlomovwf FSRcall genlpmovf TMPFSR,Wmovwf FSR;--------------------------------------------; do the decimation here, we will; subsample by 64 which gives a; new fs of .about.16hz.;--------------------------------------------movlw H`3f`andwf CCNT0,Wbtfss STATUS,Z ; if zero, then samplegoto nodecimmovf decimlo,Wmovwf TEMP7movf decimhi,Wmovwf TEMP9movf FSR,Wmovwf TMPFSRmovlw high genlpmovwf PCLATHmovlw 4movwf FCOUNTmovlw RECTAVGlomovwf FSRcall genlpmovf TMPFSR,Wmovwf FSRnodecim; if useDA==1 ;@@@@@@; movf RECTAVGhi,W; addlw H`80`; doutd 4;@@@@@@; endif;------------------------------------------------------------; compute the new threshold; rectavg + offset + fudgefactor(=f(rectavg));------------------------------------------------------------movf RECTAVGhi,Wmovwf TEMPHImovf RECTAVGlo,Wmovwf TEMPLOrlf TEMPHI,Wrrf TEMPHI,Frrf TEMPLO,F ; 1/2movf TEMPLO,Wmovwf TEMP6movf TEMPHI,Wmovwf TEMP8btfsc StrongAirCompensate ; if strong air compensation is enabled,goto strongaircomp ; then add 1*avg instead of .75*avgrlf TEMPHI,Wrrf TEMPHI,Frrf TEMPLO,F ; 1/4strongaircompmovf TEMPLO,Waddwf TEMP6,Fbtfsc STATUS,Cincf TEMP8,Fmovf TEMPHI,Waddwf TEMP8,F; rlf TEMPHI,W; rrf TEMPHI,F; rrf TEMPLO,F ; 1/8; rlf TEMPHI,W; rrf TEMPHI,F; rrf TEMPLO,F ; 1/16; rlf TEMPHI,W; rrf TEMPHI,F; rrf TEMPLO,F ; 1/32 rectavg in TEMPHI:TEMPLOloioff set H`40`hiioff set H`01`movlw loioffaddwf TEMP6,Fbtfsc STATUS,Cincf TEMP8,Fmovlw hiioffaddwf TEMP8,F ; 3/4rectavg + H`0140` in TEMPHI:TEMPLOmovf TEMP6,Wmovwf TEMPLOmovf TEMP8,Wmovwf TEMPHImovf RECTAVGlo,Waddwf TEMPLO,Fbtfsc STATUS,Cincf TEMPHI,Fmovf RECTAVGhi,Waddwf TEMPHI,F ; 1.75rectavg + H`0140` in TEMPHI:TEMPLObtfss TEMPHI,7goto nothreshsatmovlw H`7f`movwf TEMPHImovlw H`ff`movwf TEMPLOnothreshsat; if useDA==1;@@@@@@; movf TEMPHI,W; addlw H`80`; doutd 5;@@@@@@; endifinteg ;--------------------------------------; the integ update algorithm;--------------------------------------lointegup set H`00`hiintegup set H`02`lointegdn set H`80`hiintegdn set H`00`movlw high integdownmovwf PCLATHmovf TEMPLO,Wsubwf TEMP7,Fbtfss STATUS,C ; borrow?decf TEMP9,F ; yesmovf TEMPHI,Wsubwf TEMP9,F ; TEMP7:TEMP9 = rect*gain - threshbtfss STATUS,C ; borrow?goto integdown ; yesbtfsc TEMP9,7 ; result negative?goto integdown ; yesmovlw high integmaxmovwf PCLATHmovlw lointegupaddwf integlo,Fbtfsc STATUS,C ; carry?incf integhi,F ; yesbtfsc integhi,7 ; overflow?goto integmax ; yesmovlw hiintegupaddwf integhi,Fbtfsc STATUS,C ; carry?goto integmax ; yesmovlw high endintegmovwf PCLATHbtfss integhi,7 ; overflow?goto endinteg ; nointegmaxmovlw H`7f` ; saturate at 0.times.7fffmovwf integhimovlw H`ff`movwf integlomovlw high endintegmovwf PCLATHgoto endintegintegdownmovlw high integminmovwf PCLATHmovlw lointegdnsubwf integlo,Fbtfss STATUS,C ; borrow?decf integhi,F ; yesbtfsc integhi,7 ; negative?goto integmin ; yesmovlw hiintegdnsubwf integhi,Fbtfss STATUS,C ; borrow?goto integmin ; yesmovlw high endintegmovwf PCLATHbtfss integhi,7 ; negative?goto endinteg ; nointegminclrf integhi ; saturate at 0.times.0000clrf integloendintegnop; if useDA==1;@@@@@@; movf integhi,W; addlw H`80`; doutd 6;@@@@@@; endifmovlw high uszerotrthmovwf PCLATHmovf usgain,0andlw 0fehbtfsc STATUS,Z ; zero bit set?goto uszerotrth ; don't sense if sens gain pot <==1; ------Comparator----------------------lothresh set H`00`hithresh set H`30`comp movf integlo,Wmovwf TEMP7movf integhi,Wmovwf TEMP9movlw lothreshsubwf TEMP7,Fbtfss STATUS,C ; borrow?decf TEMP9,F ; yesmovlw hithreshsubwf TEMP9,F ; TEMP9:TEMP7 = integ - threshmovf TEMP9,Wmovwf usdecision; if useDA==1; movf usdecision,W; addlw H`80`; doutd 7 ; 111; endifustlu ;---------------------------------------------------------; do the lookup for the us decision here;---------------------------------------------------------; first check to see if the decision is >=0, ie; the sensor has detected motion.;---------------------------------------------------------movlw high suremotionmovwf PCLATHcomf usdecision,W ; if decision is positive, then negatingaddlw 1 ; will make it <0,movwf TEMP9btfsc STATUS,C ; if -decision=0, then motion for sure.goto suremotionbtfsc TEMP9,7 ; if -decision<0, then motion for sure.goto suremotion;--------------------------------------------------------------; the decision value is below the motion threshold; TEMP9 now holds -(integ - thresh), if; greater than 16 then no motion for sure.;--------------------------------------------------------------movlw high uszerotrthmovwf PCLATHmovf TEMP9,Wandlw H`f0` ; if any hi bit set, then -decision>16btfss STATUS,Z ; if zero, then motion is marginalgoto uszerotrth ; no movement...;------------------------------------------------------------; The motion is marginal, do the truth lookup;------------------------------------------------------------movlw high ustruthlookupmovwf PCLATHmovf usdecision,Waddlw H`10` ; limit to 0-15, should be anyway...andlw H`0f`call ustruthlookupmovwf TEMP8movlw high occupmovwf PCLATHgoto occupsuremotionmovlw high occupmovwf PCLATH; just saw motionmovlw H`07f`movwf TEMP8bsf USSense ; set us sense bit, for us only mode;-----------------------------------------------------------------------------------; we enable interrogation when the we know we have definately; seen motion and we are not currently interrogating.;-----------------------------------------------------------------------------------btfsc Interrogating ; if interrogation already enabled,goto occup ; skip over enable step;-----------------------------------------------------------------------------------; no interrogation when the LED override switch is on;-----------------------------------------------------------------------------------btfsc NoLEDFlaggoto occupbsf Interrogating ; otherwise enable it andmovlw H`02` ; set up next step of interrogationsmovwf NTRGATEstatemovlw H`ff`movwf NTRGATEcntoccup; this is where we do the pulse stretching; we only do pulse stretching for valid motion(truth level)movlw high endusstretchmovwf PCLATHdecf uspscnt,F ; reduce stretch countbtfsc STATUS,Z ; did count reach zero?clrf ustruth ; yesmovf ustruth,Wsubwf TEMP8,W ; newtruth - oldtruthbtfss STATUS,C ; did it borrow?goto endusstretch ; yes, so keep oldmovf TEMP8,Wmovwf ustruthmovlw H`5f`movwf uspscntendusstretch; Turn on LED;-----------------------------------------------------------------; if we are currently reporting status with the; LED's then don't send any indication of motion;-----------------------------------------------------------------movlw high aro1movwf PCLATHusePage1btfsc STATCNTL,7goto aro1;-----------------------------------------------------------------; if LED is currently in use, then leave it alone;-----------------------------------------------------------------usePage0btfsc SBITS,2 ; if led task on, then skipgoto aro1;-----------------------------------------------------------------; if led override switch is on, leave LED alone;-----------------------------------------------------------------btfsc NoLEDFlag ; is led override switch turned ongoto aro1 ; yes, so don't turn on LED;-----------------------------------------------------------------; we now need to determine if the LED should be; turned on based on the dual tech mode and the; `truth` of our motion signal.;-----------------------------------------------------------------movlw high usHSLEDmovwf PCLATHbtfsc options,0 ; which dual tech modegoto usHSLED ; we are in HS mode;-----------------------------------------------------------------; we are in HC mode, so we need to find the; line `a+b-60h=40h` (the `and` equation). We; choose the point where a=b, which is 50h.;-----------------------------------------------------------------movlw high usHCLEDmovwf PCLATHmovlw H`50`subwf ustruth,W ; ustruth - 50hbtfsc STATUS,Zgoto usHCLED ; ustruth = 50hmovlw high aro1movwf PCLATHbtfss STATUS,C ; status is still from subwfgoto aro1 ; ustruth < 50husHCLEDbsf PORTB,1movlw high aro1movwf PCLATHgoto aro1usHSLED movlw H`7f`subwf ustruth,W ; ustruth - 7fhbtfsc STATUS,Z ; are they equal?bsf PORTB,1 ; yes, so turn on ledmovlw high aro1movwf PCLATHgoto aro1uszerotrth; there's no motion, unless we are currently stretching; a previous pulsemovlw high uszeroyesmovwf PCLATHmovf uspscnt,Wbtfsc STATUS,Zgoto uszeroyes; we are stretching a previous pulsedecf uspscnt,Fmovlw high aro1movwf PCLATHgoto aro1uszeroyesclrf ustruth; leave led alone if it is in usemovlw high aro1movwf PCLATHbtfsc SBITS,2goto aro1; make sure led is offbtfsc PORTB,1bcf PORTB,1aro1snsrendusePage0; bcf PORTB,7 ; how much compute left?; if useDA==1;@@@@@@; movf ustruth,W; addlw H`80`; doutd 6 ; 110;@@@@@@; endif; this is where we determine the characteristics of any; motion currently happening. the objective is to find; the smallest motion pulse while the lights are on.; this will give a measure of how much the lights-on; sensitivity can be lowered.btfss Occupiedreturn; if currently not inside a pulse, check for the; beginning of a pulse, otherwise check for the; end of the pulse.movlw high usfindpulsemovwf PCLATHusePage1movf uspulsecnt1,Wiorwf uspulsecnt0,W ; if both are 0, the `or` is 0btfsc STATUS,Z ; if 0, then not in pulsegoto usfindpulse ; not inside pulse, find oneusePage0; inside pulsemovlw high usendpulsemovwf PCLATHbtfsc usdecision,7 ; is usdecision negative?goto usendpulse ; yes, so end of pulse. is it the smallest?; not at end of pulse, see if level is max of the pulseusePage1incf uspulsecnt0,Fbtfsc STATUS,Zincf uspulsecnt1,F ; increment pulse widthusePage0movf usdecision,WusePage1subwf uspulsemax,W ; max - decisionusePage0btfsc STATUS,C ; did it borrow?return ; no, so decision not max of pulse; save the new max of the pulsemovf usdecision,WusePage1movwf uspulsemaxusePage0returnusendpulse; we are at the end of the pulse, check to see if the; recent pulse is the smallest so far.movlw high usendpulseendmovwf PCLATHusePage1movf uspulsemax,Wsubwf usminpulse,W ; min pulse - pulse maxbtfss STATUS,C ; did it borrow?goto usendpulseend ; yes, so pulse is not smallestusePage1movf uspulsemax,Wmovwf usminpulsemovf uspulsecnt0,Wmovwf usminpw0movf uspulsecnt1,Wmovwf usminpw1usendpulseendusePage1clrf uspulsecnt0 ; clear pulse countclrf uspulsecnt1usePage0returnusfindpulse; we are not currently looking at a pulse, so we; need to look for a new oneusePage0btfsc usdecision,7return ; not a new pulse; we have found a new pulsemovf usdecision,WusePage1movwf uspulsemaxmovlw 1movwf uspulsecnt0clrf uspulsecnt1usePage0return ; end sensor taskustruthlookup; W has the motion level [0,15]movwf TEMP7movlw high ustrthlumovwf PCLATHmovf TEMP7,Waddlw low ustrthlubtfsc STATUS,Cincf PCLATH,Fmovwf PCLustrthluretlw 0.times.0 ;0retlw 0.times.02 ;1retlw 0.times.04 ;2retlw 0.times.08 ;3retlw 0.times.0c ;4retlw 0.times.10 ;5retlw 0.times.18 ;6retlw 0.times.28 ;7retlw 0.times.40 ;8retlw 0.times.58 ;9retlw 0.times.68 ;10retlw 0.times.6f ;11retlw 0.times.74 ;12retlw 0.times.78 ;13retlw 0.times.7c ;14retlw 0.times.7f ;15;-----------------------------------------------------------------; mult8.times.8; 8 by 8 multiply; Inputs:; W, TEMP9; Outputs:; 16bit result in TEMPHI:TEMPLO;-----------------------------------------------------------------mult8.times.8usePage0clrf TEMPHIclrf TEMPLObcf STATUS,C ; clear carry bitbtfsc TEMP9,0addwf TEMPHI,1rrf TEMPHI,1rrf TEMPLO,1btfsc TEMP9,1addwf TEMPHI,1rrf TEMPHI,1rrf TEMPLO,1btfsc TEMP9,2addwf TEMPHI,1rrf TEMPHI,1rrf TEMPLO,1btfsc TEMP9,3addwf TEMPHI,1rrf TEMPHI,1rrf TEMPLO,1btfsc TEMP9,4addwf TEMPHI,1rrf TEMPHI,1rrf TEMPLO,1btfsc TEMP9,5addwf TEMPHI,1rrf TEMPHI,1rrf TEMPLO,1btfsc TEMP9,6addwf TEMPHI,1rrf TEMPHI,1rrf TEMPLO,1btfsc TEMP9,7addwf TEMPHI,1rrf TEMPHI,1rrf TEMPLO,1return;----------------------------------------------------------------; gainadjust --; the adjustment is a two's complement num; the gain is an unsigned number.; The adjustment should be bounded well within; +/-127.; return the adjusted gain in W register;----------------------------------------------------------------gainadjustmovlw high gainaddposmovwf PCLATHmovf usoffgain,Wbtfsc Occupiedmovf usongain,Wmovwf TEMP7movf usoffadj,Wbtfsc Occupiedmovf usonadj,Wmovwf TEMP8btfss TEMP8,7 ; is it negative or positive?goto gainaddpos; adjust term is negative, so check for negative overflowaddwf TEMP7,Wbtfss STATUS,Cmovlw H`1` ; minimalreturngainaddpos; adjust term is positive, so check for overflowaddwf TEMP7,Wbtfsc STATUS,Cmovlw H`ff` ; maxreturngainlookupusePage0andlw H`0f` ; limit to 16movwf TEMP7movlw high gainlustartmovwf PCLATHmovf TEMP7,Waddlw low gainlustartbtfsc STATUS,Cincf PCLATH,Fmovwf PCLgainlustartretlw D`1`retlw D`4`retlw D`9`retlw D`16`retlw D`25`retlw D`36`retlw D`49`retlw D`64`retlw D`81`retlw D`100`retlw D`121`retlw D`144`retlw D`169`retlw D`196`retlw D`225`retlw D`255`;; PIRtask.asm; COPYRIGHT (C) MYTECH CORP. 1996, 1997;; the PIR task; orig: erw; composition mechanism 8-15-96 JJF; tighter filters 8-20-96 JJF;irtruthlookupmovwf TEMP9movlw high pirlumovwf PCLATHmovf TEMP9,Waddlw low pirlubtfsc STATUS,Cincf PCLATH,Fmovwf PCLpirlu retlw 0.times.0 ;0retlw 0.times.2 ;1retlw 0.times.4 ;2retlw 0.times.8 ;3retlw 0.times.0c ;4retlw 0.times.10 ;5retlw 0.times.18 ;6retlw 0.times.28 ;7retlw 0.times.40 ;8retlw 0.times.58 ;9retlw 0.times.68 ;10retlw 0.times.70 ;11retlw 0.times.74 ;12retlw 0.times.78 ;13retlw 0.times.7c ;14retlw 0.times.7f ;15PIRhpfilt macro; --- first calculate pir low pass ----------; TEMP9 already contains raw pir signalclrf TEMP7movlw high genlpmovwf PCLATHmovf FSR,Wmovwf TMPFSRmovlw 6movwf FCOUNTmovlw PIRLPLOmovwf FSRcall genlp ; lowpass(PIR) in PIRLPHI:PIRLPLOmovf TMPFSR,Wmovwf FSR;---- calculate pir high pass -------------; yhp(n) = x(n) - ylp(n);---------------------------------------------comf PIRLPHI,Wmovwf TEMP8comf PIRLPLO,W ; negateaddlw 1btfsc STATUS,C ; carry bitincf TEMP8,Fmovwf TEMP6 ; -pirlp in TEMP8, TEMP6movf TEMP9,Waddwf TEMP8,F ; TEMP8:TEMP6 = pir - pirlpendm; returnPIRavgfilt macromovf TEMP8,Wmovwf TEMP9movf TEMP6,Wmovwf TEMP7movlw high genlpmovwf PCLATHmovf FSR,Wmovwf TMPFSRmovlw 8movwf FCOUNTmovlw PIRAVGLOmovwf FSRcall genlp ; avg(PIR) in PIRAVGLO:PIRAVGLOmovf TMPFSR,Wmovwf FSRendm; return;; max routine; returns the max of registers TEMP7 and TEMP8 in W;max movlw high max2movwf PCLATHmovf TEMP8,Wsubwf TEMP7,W ; TEMP7-TEMP8, if borrow, TEMP8>TEMP7btfsc STATUS,Cgoto max2movf TEMP8,Wreturnmax2 movf TEMP7,WreturnrunPIR nopmovlw 0a1h ; 1010 0001 enable an4 a/dmovwf ADCON0 ; next reading will be PIR input (ch4 pin 7)bcf IRSense ; clear the pir sense bit.movlw high pirwait1movwf PCLATHmovlw D`24`movwf TEMP6pirwait1decfsz TEMP6,Fgoto pirwait1conv7 bsf ADCON0,2 ; start a/d conversionmovlw high waitPIRmovwf PCLATHwaitPIR btfsc ADCON0,2 ; wait til conversion complete goto waitPIR ; skip if conversion is completeadPIR movf ADRES,0 ; get a/d inputaddlw 080h ; center a/d output around 0movwf TEMP9 ; save PIR input; if useDA==1; movf TEMP9,W; addlw H`80`; doutd 0 ; 000; endifPIRhpfilt; if useDA==1; movf TEMP8,W; addlw H`80`; doutd 1 ; 001; endifPIRavgfilt; if useDA==1; movf PIRAVGHI,W; addlw H`80`; doutd 2 ; 010; endif; ------PIR Window Comparator-----------------------------------pirwindmovf PIRAVGLO,Wsubwf TEMP6,F ; lobyte(hp-avghp)btfss STATUS,C ; borrow?decf TEMP8,F ; yesmovf PIRAVGHI,Wsubwf TEMP8,F ; hibyte(hp-avghp)movlw high PIRposmovwf PCLATHbtfss TEMP8,7 ; skip next if sign is 1goto PIRposPIRneg comf TEMP8,F ; get neg of (hp-avghp)comf TEMP6,Fincf TEMP6,Fbtfsc STATUS,Zincf TEMP8,FPIRpos nop ; abs(PIRavg-TEMP8) in TEMP7; if useDA==1; movf TEMP8,W; addlw H`80`; doutd 3 ; 011; endifmovlw high PIRnomotionmovwf PCLATHmovlw H`7f` ; 127subwf PIRthresh,Wbtfsc STATUS,Z ; if PIRthresh is 127 (max.) thengoto PIRnomotion ; do not try to sense motionPIRcomp ; Compare abs(hp-hpavg) to thresholdmovlw high irnoadjustmovwf PCLATHmovf PIRthresh,Wbtfss ASAEnable ; apply adjustment only if asa enabledgoto irnoadjustmovf iroffadj,Fbtfsc STATUS,Z ; no adjustment presentgoto irnoadjustmovlw high iradjposmovwf PCLATHmovf PIRthresh,Wbtfss iroffadj,7 ; is adjustment negative?goto iradjposaddwf iroffadj,Wbtfss STATUS,Cmovlw H`16` ; min threshmovwf TEMP7movlw high irnoadjustmovwf PCLATHmovf TEMP7,Wgoto irnoadjustiradjposaddwf iroffadj,Wbtfsc STATUS,C ; make sure adjustment did not rollovermovlw H`7f` ; if it did, saturate at most positive #irnoadjustmovwf TEMP9subwf TEMP8,W ; .vertline.(hp-hpavg).vertline. - thresholdmovwf pirdecisionif useDA==1movf pirdecision,Waddlw H`80`doutd 7 ; 100endif; if useDA==1; movf TEMP9,W; addlw H`80`; doutd 7 ; 110; endifmovlw high pirlbl3movwf PCLATHbtfss pirdecision,7 ; if motion detected, forget noise estimategoto pirlbl3btfsc Occupied ; if occupied, forget noise estimategoto pirlbl3;--------------------------------------------------------; compute an estimate of the ir noise floor; when the space is not occupied.;--------------------------------------------------------movf pirdecision,Wmovwf TEMP8usePage1movf irRecentMax,WusePage0movwf TEMP7movlw high maxmovwf PCLATHcall maxusePage1movwf irRecentMaxusePage0pirlbl3movlw high pirlbl2movwf PCLATH; here we do the lookup of the; truth of the pir sensor value.btfsc pirdecision,7 ; is result positive?goto pirlbl2 ; No. motion, if any, is marginal.pir1truthmovlw H`7f` ; Compare was positive, definite motion.; movwf pirtruth ; the truth is 0.times.7f for sure.movwf TEMP8bsf IRSense ; set ir sense bit for ir only modemovlw high pirmotionmovwf PCLATHgoto pirmotionpirlbl2comf pirdecision,Waddlw 1btfsc STATUS,C ; if add carried, then motion for suregoto pir1truthmovwf TEMP7 ; TEMP7 = -pirdecisionandlw H`f0`btfss STATUS,Z ; if any high bits, then pirdecision wasgoto PIRnomotion ; large negative, so no motion.movlw high irtruthlookupmovwf PCLATHmovf TEMP7,W ; remember TEMP7=-(.vertline.PIRHPHI-PIRHPAVG.vertline. -threshold)andlw H`0f` ; limit to 0-15 rangesublw H`0f` ; reverse the limitcall irtruthlookup; movwf pirtruthmovwf TEMP8pirlbl1movlw high PIRnomotionmovwf PCLATHmovf pirtruth,Fbtfsc STATUS,Z ; is pirtruth = 0?goto PIRnomotion ; yes, so no motion sensed;pirmotionledpirmotion; this is where we do the pulse stretching; we only do pulse stretching for valid motion(truth level)movlw high endirstretchmovwf PCLATHdecf irpscnt,F ; reduce stretch countbtfsc STATUS,Z ; did count reach zero?clrf pirtruth ; yesmovf pirtruth,Wsubwf TEMP8,W ; newtruth - oldtruthbtfss STATUS,C ; did it borrow?goto endirstretch ; yes, so keep oldmovf TEMP8,Wmovwf pirtruthmovlw H`04`movwf irpscntendirstretch;--------------------------; we have motion;--------------------------movlw high PIRendmovwf PCLATHbtfsc NoLEDFlag ; is led override switch turned ongoto PIRend;----------------------------------------------------------------; we now need to determine if the LED should be; turned on based on the dual tech mode and the; `truth` of our motion signal.;----------------------------------------------------------------movlw high irHSLEDmovwf PCLATHbtfsc options,0 ; which dual tech modegoto irHSLED ; we are in HS mode;----------------------------------------------------------------; we are in HC mode, so we need to find the; line `a+b-60h=40h` (the HC equation). We; choose the point where a=b, which is 50h.;----------------------------------------------------------------movlw high irHCLEDmovwf PCLATHmovlw H`50`subwf pirtruth,W ; pirtruth - 50hbtfsc STATUS,Zgoto irHCLED ; pirtruth=50hmovlw high PIRendmovwf PCLATHbtfss STATUS,C ; STATUS is still result of subwfgoto PIRend ; pirtruth<50hirHCLEDbcf GreenLED; if useDA==0; movlw high initgmf; movwf PCLATH; call initgmf ; turn on the LED, remember this; ; will delay for 100ms.; endifmovlw high PIRendmovwf PCLATHgoto PIRendirHSLEDmovlw high PIRendmovwf PCLATHmovlw H`7f`subwf pirtruth,W ; pirtruth - 7fhbtfss STATUS,Z ; are they equal?goto PIRend ; no, so don't turn on ledbcf GreenLED; if useDA==0; movlw high initgmf; movwf PCLATH; call initgmf ; turn on the LED, remember this; ; will delay for 100ms.; endifmovlw high PIRendmovwf PCLATHgoto PIRendPIRnomotion; there's no motion unless we are stretching; a previous pulsemovlw high irzeroyesmovwf PCLATHmovf irpscnt,Wbtfsc STATUS,Zgoto irzeroyes; we are stretching a previous pulsedecf irpscnt,Fmovlw high PIRendmovwf PCLATHgoto PIRendirzeroyesclrf pirtruthbsf GreenLEDPIRend; if useDA==1; movf pirtruth,W; addlw H`80`; doutd 5; endifreturn;; o2sermon.asm; COPYRIGHT (C) MYTECH CORP. 1996, 1997; jjfserrpt; start by sending the header: preamble, version,; # of data bytes, and checksum.; [5ah][1-0-A-0-5][numbytes][checksum]usePage0movlw high serw1movwf PCLATHserw1 btfss PIR1,TXIFgoto serw1movlw H`5a` ; preamble charmovwf TXREGmovwf MTEMP1 ; start checksummovlw high serw2movwf PCLATHserw2 btfss PIR1,TXIFgoto serw2movlw H`31` ; ascii `1`movwf TXREGaddwf MTEMP1 ; add to checksummovlw high serw3movwf PCLATHserw3 btfss PIR1,TXIFgoto serw3movlw H`30` ; ascii `0`movwf TXREGaddwf MTEMP1 ; add to checksummovlw high serw4movwf PCLATHserw4 btfss PIR1,TXIFgoto serw4movlw H`41` ; ascii `A`movwf TXREGaddwf MTEMP1 ; add to checksummovlw high serw5movwf PCLATHserw5 btfss PIR1,TXIFgoto serw5movlw H`30` ; ascii `0`movwf TXREGaddwf MTEMP1 ; add to checksummovlw high serw6movwf PCLATHserw6 btfss PIR1,TXIFgoto serw6movlw H`36` ; ascii `6`movwf TXREGaddwf MTEMP1 ; add to checksummovlw high serw7movwf PCLATHserw7 btfss PIR1,TXIFgoto serw7movlw H`c0` ; number of data wordsmovwf TXREGaddwf MTEMP1 ; add to checksummovlw high serw8movwf PCLATHserw8 btfss PIR1,TXIFgoto serw8movf MTEMP1,W ; now send the checksummovwf TXREGmovlw H`20`movwf FSRmovlw high serw9movwf PCLATHserw9 btfss PIR1,TXIFgoto serw9movf INDF,Wmovwf TXREGincf FSR,Fbtfss FSR,7 ; overflowgoto serw9movlw H`a0`movwf FSRmovlw high serw10movwf PCLATHserw10 btfss PIR1,TXIFgoto serw10movf INDF,Wmovwf TXREGincf FSR,Fbtfsc FSR,7 ; overflowgoto serw10return; irnoise.asm; Copyright (C) 1996, 1997 Mytech Corp.;; John J. Fowler;;; This file should be included in the main sensor; file when using the IR Noise Estimate Adaptation; feature. The macros contained in this file should; then be placed in the appropriate places as; defined below.;; INEMonitorPIRDetector -- This macro should be placed; at the end of the PIR sensor DSP code. It monitors; the output of the threshold comparison to determine; if motion has occurred. The routine assumes that there; is a variable named `pirdecision` which contains the; result of the comparison. There must be a bit variable; called `Occupied` which is high the the space is occupied.; The routine also uses a 16bit variable called; LOFFTIMEHI:LOFFTIMELO which holds the amount of time that; the space has been unoccupied. The routine places the new; peak noise estimate in the variable named `irRecentMax`.;; INEBeginLightsOff -- This macro simply resets the 4 minute; running maximum of the IR signal which is held in the; variable called `irRecentMax`. This routine should be run; each time the load is de-energized.;; INEMinuteTask -- This macro uses the short time running maximum; of the IR signal, irRecentMax, to compute the long term; running max of the IR noise. Effectively, this routine; is performing the statistical order filter during each; unoccupied period maintaining the result in the variable; called `irNoiseEst`. This routine also handles the task of; double buffering `irNoiseEst` by moving the most recent; value of `irNoiseEst` into the variable called `irNoiseEstPrev`.; This routine should be run every minute. It uses a variable; called `mcntlo` to determine when it should perform the; actions mentioned above.;; INEFiniteMemory -- This macro is used to give the IR Noise; estimate a finite life span. Without this routine, the; IR Noise Estimate will never decay once it reaches the; highest possible value. This routine guarantees that once; the IR Noise naturally begins to decline (perhaps due to; seasonal change), the IR Noise Estimate will follow.; The routine should be run each time the load is energized.;; IRNoiseExecute -- This routine performs the actual check of; the IR Noise Estimate against the boundaries of the desired; noise region and makes adjustmentsthe variable `iroffadj` as; necessary. This routine should be placed in the `main`; routine. It is scheduled by setting the bit variable; called `IRNoiseCheckEnable`. The routine should be; scheduled each time the load is energized.;INEMonitorPIRDetector macromovlw high dontdoirmaxmovwf PCLATHbtfss pirdecision,7 ; if motion detected, forget noise estimategoto dontdoirmaxbtfsc Occupied ; if occupied, forget noise estimategoto dontdoirmax;--------------------------------------------------------; compute an estimate of the ir noise floor; when the space is not occupied.;--------------------------------------------------------movlw high doirmaxmovwf PCLATHmovf LOFFTIMEHI,Wbtfss STATUS,Z ; if lofftimehi!=0, then >=8sec for suregoto doirmaxmovlw high dontdoirmaxmovwf PCLATHmovf LOFFTIMELO,Wandlw H`F8`btfsc STATUS,Z ; if `and`=0, then <8sec since lights offgoto dontdoirmaxdoirmaxmovf pirdecision,Wmovwf TEMP8usePage1movf irRecentMax,WusePage0movwf TEMP7movlw high i8maxmovwf PCLATHcall i8maxusePage1movwf irRecentMaxusePage0dontdoirmaxendmINEBeginLightsOff macrousePage1movlw H`80`movwf irRecentMax ; set ir max to most negative numberusePage0endmINEMinuteTask macro;----------------------------------------------------------------------; in order to get a better estimator of the ir noise; level, we maintain a running max of the noise. Once; every 4 minutes while the space is not occupied this; running max is compared with the current ir noise; estimate. The largest is the new noise estimate of; the ir circuit.;----------------------------------------------------------------------; we only want to check the noise estimate once every; four minutes. If double buffered, the noise est.; will always be at least 4 min prior to the turn-on.;----------------------------------------------------------------------; execute at proper timeslicemovlw high minnomaxresetmovwf PCLATHmovlw H`03`andwf mcntlo,Wbtfss STATUS,Zgoto minnomaxresetmovlw high endminirestmovwf PCLATH; do only when vacantbtfsc Occupiedgoto endminirest; only when vacant for at least .about.29 minmovlw H`07`subwf LOFFTIMEHI,W ; lofftimehi - 7btfss STATUS,C ; if borrow then < .about.29.87mingoto endminirest ; borrowusePage1movf irRecentMax,WusePage0movwf TEMP8usePage1movf irNoiseEst,Wmovwf irNoiseEstPrev ; we double buffer the noise estimatorusePage0movwf TEMP7 ; INE --> TEMP7movlw high maxmovwf PCLATHcall maxusePage1movwf irNoiseEstendminirestusePage1movlw H`80`movwf irRecentMaxminnomaxresetusePage0endmINEFiniteMemory macro;-------------------------------------------------------------------; we need to remove the `infinite memory` trait of; the ir noise estimate. We'll reduce the estimate; by 25% every time an occupancy cycle begins.; We really want to reduce the estimate of the; noise floor. Since the estimate is usually < 0,; a mathematical reduction by 25% actually is an; increase in the estimate. We solve this problem; by making the estimate unsigned, applying the; mathematical reduction, and going back to signed.;-------------------------------------------------------------------usePage1movlw H`80` ; roll the number around foraddwf irNoiseEst,F ; convenient mathbcf STATUS,Crrf irNoiseEst,F ; 1/2 INE --> INEbcf STATUS,Crrf irNoiseEst,W ; 1/4 INE --> Waddwf irNoiseEst,F ; 3/4 INE --> INEmovlw H`80`addwf irNoiseEst,F ; don't forget to roll it back!usePage0endmIRNoiseExecute macromovlw high noircheckmovwf PCLATHbtfss IRNoiseCheckEnablegoto noircheckbcf IRNoiseCheckEnable;--------------------------------------------------------------; We look at the ir noise; estimate. If it is sufficiently small then we can; lower the threshold(increase sensitivity).;--------------------------------------------------------------movlw high irtoolrgmovwf PCLATHusePage1comf irNoiseEstPrev,W ; use the older noise estimateaddlw 1 ; -noise estimate in WusePage0movwf MTEMP ; store -noise estimateandlw H`e0` ; is it greater than 31 below threshold?btfsc STATUS,Z ; if `and` is 0, then -noise is too largegoto irtoolrg ; -(noise estimate) might be too large!; noise level really low, maybe threshold too high...movlw B`01100000` ; 3/8 weighting factormovwf MTEMP4movlw high o2AdjustIR.sub.-- upmovwf PCLATHcall o2AdjustIR.sub.-- upusePage1incf INECount,Fmovlw H`ff`btfsc STATUS,Zmovwf INECountusePage0movlw high endirnoisefixmovwf PCLATHgoto endirnoisefixirtoolrgmovlw high endirnoisefixmovwf PCLATHmovf MTEMP,Wandlw H`f0` ; is it >= 15btfss STATUS,Z ; if `and` is 0, then -noise is too smallgoto endirnoisefix ; noise is within range [-15,-32], no change; noise too largemovlw B`01100000` ; 3/8 weighting factormovwf MTEMP4movlw high o2AdjustIR.sub.-- dnmovwf PCLATHcall o2AdjustIR.sub.-- dnusePage1incf INECount,Fmovlw H`ff`btfsc STATUS,Zmovwf INECountusePage0endirnoisefixnoircheckendm__________________________________________________________________________

Number	Name	Date
3074053	McDonough et al.	Jan 1963
3383678	Palmer	May 1968
3721972	Hermans	Mar 1973
3725888	Solomon	Apr 1973
3777139	Peel	Dec 1973
3781859	Hermans	Dec 1973
3801978	Gershberg et al.	Apr 1974
3838408	McMaster	Sep 1974
3958118	Schwarz	May 1976
3967283	Clark et al.	Jun 1976
4021679	Bolle et al.	May 1977
4051472	Albanese et al.	Sep 1977
4172253	Hermans	Oct 1979
4207559	Meyer	Jun 1980
4225808	Saraceni	Sep 1980
4243979	Kleinschmidt	Jan 1981
4275390	Heywang et al.	Jun 1981
4315596	Johnson, Jr. et al.	Feb 1982
4331952	Galvin et al.	May 1982
4336464	Weber	Jun 1982
4337402	Nowakowski	Jun 1982
4361767	Pelka et al.	Nov 1982
4383288	Hess, II et al.	May 1983
4391406	Fried	Jul 1983
4401976	Stadelmayr	Aug 1983
4461977	Pierpoint et al.	Jul 1984
4468658	Rossin	Aug 1984
4521843	Pezzolo et al.	Jun 1985
4551654	Barnum	Nov 1985
4575659	Pezzolo et al.	Mar 1986
4593234	Yang	Jun 1986
4636774	Galvin et al.	Jan 1987
4660024	McMaster	Apr 1987
4661720	Cameron, Jr. et al.	Apr 1987
4678985	Moskin	Jul 1987
4703171	Kahl et al.	Oct 1987
4710750	Johnson	Dec 1987
4713598	Smith	Dec 1987
4751339	Koehring et al.	Jun 1988
4751399	Koehring et al.	Jun 1988
4757303	Scheidweiler	Jul 1988
4769765	Green	Sep 1988
4787222	Claytor	Nov 1988
4815046	Dorr	Mar 1989
4820938	Mix et al.	Apr 1989
4833450	Buccola et al.	May 1989
4843283	Chen	Jun 1989
4857912	Everett, Jr. et al.	Aug 1989
4866425	Lindmark	Sep 1989
4874962	Rowland	Oct 1989
4882567	Johnson	Nov 1989
4890093	Allison et al.	Dec 1989
4929833	Smith	May 1990
5015994	Hoberman et al.	May 1991
5077548	Dipoala	Dec 1991
5083025	Blomberg	Jan 1992
5084696	Guscott et al.	Jan 1992
5101194	Sheffer	Mar 1992
5124546	Hu	Jun 1992
5124566	Hu	Jun 1992
5128654	Griffen et al.	Jul 1992
5142199	Elwell	Aug 1992
5181010	Chick	Jan 1993
5189393	Hu	Feb 1993
5194848	Kerr	Mar 1993
5196826	Whiting	Mar 1993
5221919	Hermans	Jun 1993
5243327	Bentz et al.	Sep 1993
5258743	Nelson et al.	Nov 1993
5258775	Casey et al.	Nov 1993
5258899	Chen	Nov 1993
5276427	Peterson	Jan 1994
5281961	Elwell	Jan 1994
5282118	Lee	Jan 1994
5283551	Guscott	Feb 1994
5293097	Elwell	Mar 1994
5307054	Concannon, Sr. et al.	Apr 1994
5317304	Choi	May 1994
5317620	Smith	May 1994
5319283	Elwell	Jun 1994
5349330	Diong	Sep 1994
5357170	Luchaco	Oct 1994
5374854	Chen	Dec 1994
5386210	Lee	Jan 1995
5406176	Sudgen	Apr 1995
5471194	Guscott	Nov 1995
5473202	Mudge et al.	Dec 1995
5482314	Corrado et al.	Jan 1996
5489827	Xia	Feb 1996
5504476	Cecic et al.	Apr 1996
5543777	Vane et al.	Aug 1996
5578988	Hoseit et al.	Nov 1996
5581237	DiPoala	Dec 1996
5586048	Coveley	Dec 1996
5640143	Myron et al.	Jun 1997
5673022	Patel	Sep 1997
5701117	Platner et al.	Dec 1997

Number	Date	Country
0 066 691	Dec 1982	EPX
40 05 169	Aug 1991	DEX
4302836	Jul 1994	DEX
WO 9210074	Jun 1992	WOX
WO 9305627	Mar 1993	WOX
WO 9420937	Sep 1994	WOX
WO 9625021	Aug 1996	WOX

Occupancy sensor and method of operating same

Information

Patent Number

Date Filed

Date Issued

Inventors

Original Assignees

Examiners

Agents

CPC

US Classifications

Field of Search

US

International Classifications

Abstract

Description

Claims

Parent Case Info

US Referenced Citations (97)

Foreign Referenced Citations (7)

Non-Patent Literature Citations (5)

Continuation in Parts (1)

Entry
International Search Report dated May 18, 1998 (PCT/US98/02010) (MYTE:023P).
Imbrecht et al., In: Building Energy Efficiency Standards, California Energy Commision, Jul. 1988.
O'Mara, "Dual-technology sensors solve installation problems," Security Distributing & Marketing, pp. 81-83, Aug. 1990.
International Search Report dated Jun. 7, 1996.
Hubbell, "Mass 1500A and 750A Motion Switching Sustem," 1993.