Self-diagnostic circuitry for emergency lighting fixtures

BACKGROUND OF THE INVENTION

1. Field of the Invention

The invention relates generally to emergency lighting systems normally operable from a mains power supply and configured to operate from an emergency power source on failure of mains power, the invention particularly relating to programmable self-test and/or self-diagnostic systems for emergency lighting fixtures and which are capable of manual or automatic performance of testing and diagnostic functions on the circuitry, power supply, charging system and lamping thereof.

2. Description of the Prior Art

Emergency lighting systems have become ubiquitous due to code requirements, power typically being provided to emergency lighting systems under normal operating conditions by mains power supplied both to a normal AC ballast for light generation and to an emergency ballast for maintaining a charge on a power supply typically carried by the emergency lighting system, the power supply usually taking the form of an internal battery. Loss of line voltage results in the switching of the emergency ballast to power the lamping of the emergency lighting system through the emergency power supply, power being typically provided to the lamping for operation of the emergency lighting system for a relatively short period of time. Since the emergency lighting system is effectively in stand-by mode during normal operation thereof, periodic testing of the system is desirable in order to ensure proper functioning during emergency conditions, that is, loss of mains power. Conventionally, emergency lighting systems are tested by the active intervention of building maintenance personnel by manual operation of a test switch to simulate a power outage and to monitor the operation of the emergency lighting system. Neglect of testing procedures often occurs and results in detection of the failure of the emergency lighting system only when an emergency situation arises. In relatively recent years, automatic self-test and/or diagnostic circuitry for emergency lighting systems has become available. In such circuitry, an internal control system such as a microprocessor automatically causes a number of different tests to be conducted in sequence, such tests including detection of lamp current flow, power transfer from charger to battery and battery voltage. Failure conditions are typically indicated in such circuitry through illumination of a visual indicator such as a light emitting diode to indicate that maintenance is required. Systems monitored at central locations would also produce an indication of test failure on a computer display terminal at a central monitoring location. Systems capable of automatic testing include the supervisory emergency lighting system disclosed by Balcom et al in U.S. Pat. No. 4,799,039, this patent also describing other emergency lighting systems having periodic testing and self-diagnostic capabilities. The supervisory system of Balcom et al continuously monitors selected parameters of an emergency lighting system and periodically tests the system under simulated full-load emergency conditions automatically. The Balcom et al system also includes a closed loop three-mode battery charging control circuit. Vosika et al, in U.S. Pat. No. 5,574,423, disclose a self-diagnostic circuit for an emergency lamp which includes a high-impedance circuit path connected in series with the lamp and including a visual indicator energized by battery current passing through the circuit path and lamp during standby mode operation. A second high impedance circuit path is connected in parallel with the lamp to energize a second visual indicator whenever proper electrical continuity does not exist through the lamp. In U.S. Pat. No. 5,666,029, McDonald automatically tests the emergency ballast of an emergency lighting system, testing functions being facilitated by providing the emergency ballast with a transistorized inverter cutoff unit.

While circuitry of the prior art provides improved testing and diagnosis of emergency lighting systems when compared to manual testing as has been standard in the art for many years, the art is improved by the present invention which provides a low power, low cost circuit board combining all functions of a self-test/self-diagnostic system with control of all diagnostic, charging and transfer functions through the use of software, a microprocessor so controlled not only monitoring operation of charger/transfer circuitry but also controlling the charger/transfer circuitry to enable alternate strategies for alleviation of a given failure. The present self-test and self-diagnostic circuitry is thus capable of optimal function at extraordinarily low cost, the present system thereby producing a valuable advance in the art.

SUMMARY OF THE INVENTION

The invention provides an emergency lighting fixture having both manual and automatic self-test/self-diagnostic capabilities and wherein all diagnostic, charging and transfer functions are controlled by a microprocessor operating with software instructions. Microprocessor control is accomplished without modification of standard charger, inverter and lamp topology such as are conventionally employed in emergency lighting systems now available in the marketplace. The present electronic self-test and/or self-diagnostic circuitry combines with an emergency lighting system to perform under manual or automatic control testing and diagnostic functions on the circuitry, power supply, charger and lamping of a fixture. The diagnostic circuitry of the invention not only monitors operation of charger/transfer circuitry but also controls such circuitry to enable implementation of alternate strategies for alleviation of a given failure. In an LED exit sign, for example, a two-stage inverter powers the light emitting diodes and is controlled by the microprocessor, the microprocessor being efficiently powered in inverter mode. Further, battery power supply to the microprocessor is controlled by the microprocessor and is discontinued after appropriate operation until primary power is restored. Configuration of the test and diagnostic system in an exit signage embodiment can occur through use of a two-wire serial link between modular elements of the system, allowing the system to be flexibly configured in a low power low cost circuit board, flexibility further being provided by separate treatment of lamping from charger and diagnostic electronics.

In the LED exit signage mode, the present diagnostic circuitry employs the two-stage inverter indicated herein as being controlled by a microprocessor to convert voltage from one or more cells to a higher voltage for driving a string of light emitting diodes such as form the lamping of the exit sign or the like, driving of the light emitting diodes being accomplished in a manner similar to that described in U.S. Pat. No. 5,739,639, the disclosure of which is incorporated hereinto by reference. An EEPROM memory is used in this embodiment of the present system to set factory determined configuration parameters, thereby allowing accurate calibration without the need to use unreliable and costly potentiometers and further providing a field diagnostic log containing information relating to testing and diagnostic history. The LED exit signage embodiment of the present system further utilizes all solid state transfer switches for switching between emergency charging and diagnostic modes, this embodiment of the invention not utilizing relays and therefore using less power with more reliability than prior systems so configured. Since the microprocessor used according to the invention controls its own power supply and turns itself off once battery voltage has reached a low voltage disconnect threshold, very low power is consumed once the required emergency discharge is complete, a further benefit which allows the system to be shipped in this first embodiment without disconnecting the battery. Emergency lighting fixtures configured according to this first embodiment of the invention therefore do not require connection of the battery on site during installation, thereby reducing on-site installation time and possible error. System flexibility in the LED exit sign embodiment of the invention is maximized through use of a two-wire serial data interface providing linkage between elements of the system which can be modular in nature and include optional circuit boards to allow optional features to be readily added to the lighting system. Software controlling the microprocessor of the invention in the LED exit sign embodiment of the invention, drives a low cost battery maintenance algorithm through use of a simple shunt circuit to reduce battery trickle charge current once the battery is charged, battery life thus being prolonged.

A second embodiment of the invention capable of self-test and/or self-diagnosis of emergency unit lighting fixtures is necessitated by differences in the-basic load and battery chemistries which exist between normally incandescent emergency unit fixtures and emergency exit signage illuminated by an array of light emitting diodes. While the first embodiment of the invention finds primary use with exit signage illuminated by light emitting diodes, the second embodiment of the invention is used in emergency unit fixtures wherein sealed lead-acid batteries are normally used to provide power for direct current lamps when AC mains power fails. Typical drain rates in the emergency mode of operation in such emergency unit fixtures range from three to thirty amps depending upon the number of direct current lamps operated within the unit fixture. By comparison, battery drain rates required from an efficient light emitting diode exit sign is less than approximately 650 milligrams. Accordingly, LED exit signs can be operated in the emergency mode by one or more NiCd batteries using a relatively simple current control charger. The use of lead-acid batteries in emergency unit fixtures requires use of a voltage-control charger in order to maintain the batteries properly in the normal, non-emergency mode of operation. The need for sealed lead-acid batteries in emergency unit fixtures is accompanied by the need for temperature compensation of charger output voltage in order to maximize battery life over a range of operating temperatures.

As a second consideration for the need for the two embodiments of the invention, a large difference exists in output mode requirements between emergency exit signage and unit emergency fixtures. In the situation involving an exit sign illuminated by light emitting diodes, in particular, the load is small enough to be driven with a small boost converter circuit having a battery-drain rate of less than 650 milliamps. In emergency unit fixtures utilizing incandescent lamping in the emergency mode, output load of greater than three amps necessitate use of a relatively simple relay transfer circuit for the output load. A further complication of the increased output load in emergency unit fixtures dictates that diagnostic circuitry for unit fixtures requires a more complicated current sense circuit as opposed to a relatively simple shunt element, that is, a resistor, as is used in the exit sign which is capable of functioning with the first embodiment of the circuitry of the invention. The self-diagnostic aspect of the second embodiment of the invention requires the diagnostic circuitry to measure both charge and discharge currents, such diagnostic circuitry therefore requiring a large dynamic range and being capable of measuring both positive and negative currents.

The second embodiment of the invention which is particularly useful with emergency unit fixtures having incandescent lamping is capable of operation without the use of an inverter. Further, the second embodiment of the invention does not provide a zero power consumption feature as occurs with the first embodiment of the invention, it being necessary to ship emergency unit fixtures so configured with the battery disconnected. Still further, the emergency unit fixtures having the second embodiment of the diagnostic circuitry forming a portion thereof utilizes a relay to connect direct current lamps to lead-acid batteries during emergency mode operation. A relay can preferably be used when the system is configured with lead-acid batteries. The battery-charging algorithm used in the emergency unit fixture is also more complex than the algorithm employed in the first embodiment of the invention which is used with LED emergency exit signage. The more complex algorithm is necessary due to the differing charger topology, the emergency unit fixtures requiring a voltage-control charger while the LED exit signs utilize a relatively current-control charger using a shunt regulation topology.

Accordingly, it is an object of the invention to provide electronic self-test and/or self-diagnostic systems particularly useful with emergency lighting fixtures, the systems performing testing and diagnostic functions on the circuitry, power supply, charger and lamping of a fixture either by manual or automatic initiation.

It is another object of the invention to provide diagnostic circuitry for emergency lighting systems which not only monitor operation of charger/transfer circuitry but also control the charger/transfer circuitry to enable implementation of alternate strategies for alleviation of a given failure.

It is a further object of the invention to provide self-test/self-diagnostic emergency lighting fixtures configured in a low power, low cost circuit board with control of all diagnostic, charging and transfer functions by means of software, the system including a standard charger, an inverter or relay and standard lamp topology.

It is a still further object of the invention to provide in a first embodiment of self-test/self-diagnostic emergency lighting fixtures an inverter controlled by a microprocessor and converting voltage from one or more cells to a higher voltage to drive lamping such as a light emitting diode array as in an exit sign, inverter operation being fully microprocessor controlled.

It is yet another object of the invention to provide in a second embodiment of self-test/self-diagnostic emergency lighting fixtures a relay transfer circuit controlled by a microprocessor and a voltage-controlled charger for operation of emergency unit fixtures utilizing incandescent lamping, operation of the fixtures being microprocessor controlled.

Further objects and advantages of the invention will become more readily apparent in light of the following detailed description of the preferred embodiments.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1

is a block diagram of a first embodiment of the emergency lighting system of the invention including self-test/self-diagnostic circuitry;

FIG. 2

is a simplified schematic illustrating central features of the first embodiment of the invention;

FIG. 3

is a diagram illustrating the relationship of

FIGS. 3A

,

3

B and

3

C;

FIGS. 3A

,

3

B and

3

C in combination form a schematic of a charger circuit utilized in the first embodiment of the invention;

FIG. 4

is a schematic of a self-diagnostic option board;

FIG. 5

is a diagram illustrating the relationship of

FIGS. 5A

,

5

B and

5

C; and,

FIGS. 5A

,

5

B, and

5

C in combination form a schematic of a circuit utilized in the second embodiment of the invention.

DESCRIPTION OF THE PREFERRED EMBODIMENTS.

Referring now to the drawings,

FIGS. 1 through 3

illustrate a first embodiment of the invention as comprising an emergency lighting system seen generally at

10

. In the system

10

, lamping

12

preferably takes the form of an array of light emitting diodes arranged in series. The lamping

12

is operated under normal, non-emergency conditions through the power supply

14

which includes normal line voltage circuitry

16

for sensing the presence of AC line voltage being fed to the lamping

12

as an input signal fed to the circuitry

16

from the power supply

14

, the circuitry

16

then feeding a signal to microprocessor

18

. The microprocessor

18

controls boost converter

20

to supply power of reduced current to the lamping

12

on discontinuation of line voltage, voltage being supplied to the boost converter by means of battery

22

. The microprocessor

18

also controls battery charging circuit

24

which shuts down on control by the microprocessor

18

. The power supply

14

supplies power to regulator circuit

26

which supplies power to the microprocessor

18

with control exerted by the boost converter

20

.

The microprocessor

18

further controls an indicator light emitting diode shown at

28

and can readily be configured to control options shown in phantom at

30

, the system

10

therefore being extremely flexible. EEPROM

32

maintains a basic timing cycle and maintains an activity log useful for field return fault diagnosis. The EEPROM stores the number of hours of battery charge and discharge, the number of times a system went into emergency mode, the number of power failures, lamp failures, battery failures as well as other information which can conveniently be kept in a record by use of the EEPROM

32

. The memory of the EEPROM

32

is also used to set predetermined configuration parameters during manufacture of the system

10

, thereby guaranteeing accurate calibration without use of unreliable and costly potentiometers or the like. As will be described hereinafter, the EEPROM

32

provides other functions and advantages. It is to be understood that the electrically erasable and rewriteable read-only memory, that is, EEPROM

32

, is central to the functioning of the present emergency lighting system

10

. While EEPROM devices are well known in the prior art, the EEPROM

32

is used in a novel manner in the system

10

to reduce cost and to provide advanced diagnostic capabilities. Since it is an object of the invention to provide an emergency system capable of performing testing cycles at specified intervals, it is desirable for the system

10

to maintain testing cycles accurately in spite of power interruptions. EEPROM devices have previously been employed in emergency lighting systems for recording the state of the timing cycle should a power interruption occur. The present invention utilizes the EEPROM

32

for a variety of functions in addition to maintenance of the basic timing cycle.

Testing and diagnostics within the system

10

are carried out through the microprocessor

18

with test points being shown at

34

to include temperature, battery voltage, charger/load current and manual test switch inter alia as will be described in detail hereinafter.

Referring now to

FIG. 2

, it is seen that the boost converter

20

of the system

10

of

FIG. 1

is comprised in

FIG. 2

of transistor

36

, inductances

38

and

40

, diode

42

and resistor

44

. The circuit elements

36

through

44

, coupled with batteries

46

and

48

which are preferably NiCd batteries, act to convert a low battery voltage to a higher voltage and are thus capable of driving a light emitting diode lamp

50

. Such circuitry is disclosed in U.S. Pat. No. 5,739,639, referred to hereinabove. The light emitting diode lamp

50

comprises an array of light emitting diodes

52

configured in any desired manner including as configured in the aforesaid patent.

The transistor

36

is controlled in response to a software program contained in the microprocessor

18

, this control of the transistor

36

being accomplished through the use of a signal at

54

from the microprocessor

18

, the microprocessor

18

not being seen in FIG.

2

. When signal

54

is driven above 0.7 volts, the transistor

36

turns on and current flows from the batteries

46

and

48

through the inductances

38

and

40

, via the transistor

36

and through the resistor

44

, thus completing the circuitry to ground. The voltage at

56

therefore rises and is monitored by amplifier

58

with the gain of the amplifier

58

being set by resistors

60

and

62

so as to produce a maximum output, the output being approximately 4 volts, when no current is flowing. As the voltage at

56

rises, the voltage on a signal at

64

falls. At a certain point, as defined by the gain of the amplifier

58

, the voltage at

66

set by a resistor divider formed by resistors

68

and

70

(the resistor divider being identified at

72

), with that certain point being further defined by the threshold of the Schmitt trigger interrupt input (not shown) of the microprocessor

18

of

FIG. 1

, the certain point being connected to the signal at

64

, the microprocessor

18

being interrupted. The microprocessor

18

then turns off the transistor

36

for a predetermined “offtime” by setting the signal at

54

to ground allowing energy stored in the inductances

38

and

40

to be released via the diode

42

into capacitor

74

and, via diodes

78

and

80

, into capacitor

76

. The microprocessor

18

then turns the transistor

36

back on again, repeating the energy transfer cycle.

Operation of the present system as described hereinabove has advantages over prior art systems employing a simple oscillator in that the peak current through the inductances

38

and

40

is a fixed value regardless of the value of the inductor or load, the only significant variation being caused by the tolerances of the resistors

44

,

68

and

70

and a five volt power supply voltage which is regulated, the voltage being indicated at

82

. Total variation is therefore around ten percent. By regulating the peak current, the maximum lamp current is regulated and a more consistent light output is maintained.

Using the signal at

54

, the microprocessor

18

can therefore determine whether to operate an inverter, that is, the boost converter circuit

20

as required by software operating the microprocessor

18

. Further, the microprocessor

18

determines a desired “OFF time”. By varying the OFF time of the inverter, average lamp brightness can be controlled to best advantage in the operation of differing lamps and in view of differing local regulations or differing product variations where lamp brightness can be traded for battery life depending upon relative importance in a given installation. In the present system, the “OFF time” and the variation of “OFF time” during a battery discharge are set by parameters recorded on a product-by-product basis in the memory chip (not shown) of the EEPROM

32

during factory calibration. A software-controlled battery discharge profile can thus be set as necessary, such variation in product capability being achieved essentially at no cost.

Microprocessor control of the inverter involves use of the two inductances

38

and

40

rather than a single inductor as is common in the prior art. In the present system

10

, the inverter also powers the microprocessor

18

and its support circuitry when in emergency mode. The two inductances

38

and

40

form a “split inductor” which allows a lower voltage to be provided to the input of regulator

84

than the voltage which is provided to the lamp

50

. As can also be appreciated, the regulator

84

acts to generate the precise five volts required by the microprocessor

18

. Accordingly, addition of two low cost components, that is, the inductance

38

and the diode

78

, results in an increase of up to ten percent in overall power conversion efficiency when the inverter is operating, thereby allowing a greater proportion of battery energy to be provided to the lamp

50

to meet lamp brightness requirements as necessary.

Once the inverter of the system

10

ceases to operate as determined by the software of the microprocessor

18

, no power is supplied to the microprocessor

18

or its associated support chips (not shown in FIGS.

1

and

2

). The diode

80

, which comprises a Zener diode, blocks the current path from the batteries

46

and

48

via the inductance

38

to the regulator

84

which is the five volt regulator. If a voltage of less than 3.3 volts exists across the Zener diode

80

, no current flows. Since the batteries

46

and

48

have a nominal voltage of 1.2 volts each and a realized maximum terminal voltage of approximately 1.4 volts, the Zener diode

80

will never conduct unless the inverter is operating. If the microprocessor

18

is not powered, then the microprocessor does not provide the signals at

54

and at

88

and total power consumption is essentially zero.

Accordingly, no need exists to isolate the terminals of the batteries

46

and

48

during shipping to prevent complete discharge or damage caused by a battery polarity reversal, conditions which can develop with two series cells when one cell becomes completely discharged. Prior art lighting systems require battery isolation by disconnection or addition of a temporary insulator. Control of the inverter through the microprocessor

18

further reduces cost and reduces the possibility of an installation mistake such as by forgetting to remove an insulator (not shown) from an isolation position relative to the batteries

46

and

48

.

Advantages derived through use of the EEPROM

32

have previously been alluded to. In addition to maintaining a basic timing cycle, the EEPROM

32

provides additional advantages which include elimination of the use of potentiometers and the manual process of setting such potentiometers in a factory environment. Emergency systems require accurate calibration to ensure that charge voltages and currents are correct, that the system goes into emergency mode at the required mains voltage (80% of nominal) and begins charging at the required voltage (85% of nominal). Calibration software can further be provided to the microprocessor

18

, the calibration software interacting with software in automated factory test equipment to store required thresholds in the EEPROM

32

. Storage in the EEPROM

32

of configuration parameters allows control of system operation and system features. In addition to the features described above relative to the configuration of “OFF time”, certain features are made available only on certain products and it is desirable to be able to easily change the features at the time of manufacture without using differing versions of software running in the microprocessor

18

. This ability is particularly desirable when the microprocessor

18

is “masked” with predefined code during manufacture. The system

10

configured according to the present invention contains 52 bytes of configuration information which the microprocessor software uses to control operation. The invention uses 40 bytes of EEPROM storage to count events such as the number of brownouts and specified tests for later evaluation should a product be returned from the field. Precise evaluation is thus allowed of the performance of the system

10

under actual conditions and allows improvement using such knowledge.

Prior art self-test/self-diagnostic systems typically use a relay to connect an emergency lamp to a battery when power fails and/or to disconnect input power to a charger during diagnostic cycles to check transfer switch operation. Due to the fact that relays are typically expensive, it is therefore desirable to eliminate the use of relays in the first embodiment of the invention if at all possible due in part to use of NiCd batteries as the batteries

46

and

48

. Further according to the present invention, and specifically the first embodiment thereof, a number of diodes are used to separate the power to various circuit elements, thereby eliminating the need for a relay. The basic circuit topology of the system

10

is similar to the non-diagnostic emergency system described in the aforesaid patent. In this basic circuitry, an input of 277 volts or 120 volts to either capacitor

90

or capacitor

92

respectively reduces the AC mains voltage, a bridge rectifier

94

converting the mains voltage from AC to DC, the DC current passing through light emitting diodes

52

of the lamp

50

through current measurement resistor

96

and the batteries

46

and

48

before returning to the rectifier

94

. It should be understood that a string of the light emitting diodes

52

can be taken to be any desirable number and not be limited to the four light emitting diodes shown in FIG.

2

.

The boost converter

20

described previously herein includes the diode

42

in its topology and, by adding diode

98

to the path from the normal power supply, normal power and emergency power can be combined to drive the same lamping, that is, the lamp

50

. Power for the electronics of the microprocessor

18

is also derived from the normal power supply via diode

100

and is combined with the output of the inverter, that is, the emergency power, via the diode

78

. Accordingly, both the lamp

50

and microprocessor electronics are effectively powered by either source without the need for a relay. Additionally, microprocessor signal

102

is used to drive a transistor

104

to shunt the AC power source to allow software to perform diagnostic tests. By using the series battery/light emitting diode lamp technology of the aforesaid patent, the total power available from the AC power source can be kept small with easy bypass being possible through use of the small low cost Darlington transistor

104

shown in the circuit. Zener diode

106

is disposed across the transistor

104

for transient voltage protection.

FIG. 3

is a block diagram showing the relationship of

FIGS. 3A

,

3

B and

3

C, which

FIGS. 3A

,

3

B and

3

C illustrate in combination certain circuitry utilized in a first embodiment of the invention, the space available on a sheet of drawings being:insufficient to allow said circuit to be reproduced on a single sheet. Reference to

FIG. 3

herein constitutes a reference to the circuitry shown in the combined illustrations of

FIGS. 3A

,

3

B and

3

C. As is seen in

FIG. 3

, a ten-pin connector

108

can be employed for providing all essential signals and power from the microprocessor

18

to one or more optional circuit boards such as seen in

FIG. 4

to provide optional features. As is indicated at

30

in FIG.

1

and in relation to

FIG. 3

, the connector

108

connects to reset circuitry

120

to provide unregulated DC power for subsystems shown on options board

122

which regulate their own five volt or higher power supplies. The connector

108

further provides a reset signal, regulated five volts for circuits which require five volts, battery voltage for allowing other circuits to monitor battery voltage, a buzzer output seen in

FIG. 4

for systems requiring audible alarms and a one-wire serial data connection to a low power radio receiver (not shown) thus allowing remote control devices to be used to trigger self-test or other modes. Various daughter boards providing options can be plugged into the converter

108

. Still further, a pair of wires carrying industry standard I

2

C data signals from the EEPROM

32

and the microprocessor

18

allow intelligent subsystems to communicate with the self-diagnostic microprocessor. Full charging of a rechargeable battery results in dissipation as heat all the energy taking the form of electric current flowing through the battery cell. Increasing the temperature of a battery cell reduces cell life. Microprocessor control of a shunt battery regulator configured as shown herein in

FIG. 2

has as its purpose the reduction in charge current to the minimum required to overcome the effects of self-discharge once the battery is to be charged. The circuitry shown in

FIG. 2

is similar to the shunt regulated battery charging system of U.S. Pat. No. 5,646,502, the disclosure thereof being incorporated hereinto by reference. The system of the patent is essentially identical in most respects to the circuitry of

FIG. 2

with the exception that the microprocessor

18

controls the process.

Referring again to

FIG. 2

, it is to be understood that a percentage of the battery charge current through the lamp light emitting diodes

52

via the resistor

96

will flow through the resistor

110

and the transistor

112

when the microprocessor generates a positive voltage above 0.7 volts at signal

114

. A circuit is thus provided which allows the microprocessor

18

to reduce the charge current once the batteries

46

and

48

have been charged for a given time or have reached a given voltage as measured at

116

. Battery life can be extended substantially by providing such circuitry.

The invention is further understood by reference again to

FIG. 3

, which includes the circuitry of

FIG. 2

therein. As noted above,

FIG. 2

essentially illustrates a mains input power circuit section, a boost converter essentially comprising the inverter

20

and lamp and battery configurations. The entire circuitry of

FIG. 2

is incorporated into the circuit of FIG.

3

. As has been indicated previously, the reset circuitry

120

effectively looks at a five volt line and remains inactive unless a drop occurs on the five volt line. If a drop occurs, the reset circuitry

120

outputs a low pulse which resets the microprocessor

18

, the microprocessor

18

being brought back to a given point in the program controlling the self-diagnostic function. Transistors

124

and

126

function with Zener diode

128

inter alia to provide the reset function within the circuitry

120

, manual reset at

130

also being provided.

As carryover from the description of the circuitry of

FIG. 2

, it is to be understood that the transistor

104

is utilized to simulate power outages through turning said transistor on. The transistor

36

essentially comprises the heart of the boost converter

20

.

Continuing on with the circuitry of

FIG. 3

, capacitor

132

is seen to function as a filter capacitor and to provide DC offset at

134

, the resistors comprising the DC offset

134

dividing the network to monitor the power line. Operational amplifier

136

effectively detects the level of AC line voltage and connects to DC level by comparison to a reference voltage. At

138

, DC level is one/infinity to AC input. Integrator

140

comprised of capacitor

142

, resistor

144

and diode

146

is disposed between the operational amplifier

136

and the microprocessor

18

in order to provide appropriate function. Pull up resistors

148

and

150

are provided between EEPROM

32

and the microprocessor

18

. DC level filtering is provided at

152

by capacitors

154

,

156

and resistor

158

,

160

. One of the capacitors looking at battery voltage while the other looks at load current to provide a leveling filter function. In essence, the operational amplifier

136

comprises the heart of a difference amplifier

137

. The output of the difference amplifier

137

is DC level which is inversely proportional to AC output.

The EEPROM

32

essentially comprises a memory device having a non-volatile program which stores configuration variables used to make decisions. Log variables are also stored by the EEPROM

32

to provide a record of failures and the like. Even if power is discontinued to the EEPROM

32

, information is kept even in the absence of battery power or AC mains power. Capacitors

162

and

164

essentially comprise decoupling capacitors which suppress noise. The serial connection between the EEPROM

32

and the microprocessor

18

allows expansion of functionality with other plug-in modules.

Remaining portions of the circuitry of

FIG. 3

including a crystal clock

166

comprising a two megahertz resonator providing clocking for microprocessor frequency standards. Timing functions for the analog to digital converter in microprocessor

18

are provided by capacitor

168

. Switch

170

is provided for test purposes particularly for testing, rescheduling tests or cancelling tests. Capacitors

172

and

174

stabilize the five volt power supply and bypass high frequencies. Diode arrangement

176

provides bicolor light emitting diodes functioning with resistor

178

as an indicator circuit.

In essence, the microprocessor

18

comprises integrated circuitry having various control functions. In particular, if the microprocessor

18

detects loss of power, data is stored to the EEPROM

32

in order to save operational history.

Considering now the second embodiment of the invention as is particularly shown in

FIG. 5

, it is to be noted that the differences between the circuitry of FIG.

5

and

FIG. 3

are related as has been described hereinabove to basic load and battery chemistry considerations as exist between emergency exit signage utilizing light emitting diodes, requiring the circuitry of

FIG. 3

, and emergency unit equipment utilizing incandescent lamping, requiring the circuitry of FIG.

5

. In the circuitry of

FIG. 3

, nickel/cadmium batteries are utilized as the emergency power source while lead-acid batteries are utilized with the circuitry of FIG.

5

.

Much of the circuitry of

FIG. 3

is incorporated into the circuitry of

FIG. 5

, particularly the microprocessor

18

and the EEPROM

32

. Further, reset circuitry

120

is essentially identical as are the crystal clock

106

and the connector

108

. D.C. local filtering at

152

is produced in a similar fashion and the pull-up resistors operate in a similar manner. The difference amplification function centering on the operational amplifier

136

and associated circuitry is also provided in the same manner as is the D.C. offset function at

134

. Other similarities exist between the circuits of FIG.

3

and FIG.

5

.

The following description of

FIG. 5

is based in part upon differences between the self-diagnostic circuits of FIG.

3

and

FIG. 5

respectively.

It is to be understood that

FIG. 5

is a block diagram showing the relationship of

FIGS. 5A

,

5

B and

5

C, which

FIGS. 5A

,

5

B and

5

C illustrate in combination certain circuitry utilized in a second embodiment of the invention, the space available on a sheet of drawings being insufficient to allow said circuitry to be reproduced on a single sheet. Reference to

FIG. 5

herein constitutes a reference to the circuitry shown in the combined illustrations of

FIGS. 5A

,

5

B and

5

C. For those portions of the circuitry of

FIG. 5

which differ from the circuitry of

FIG. 3

, a more detailed discussion will be provided hereinafter.

The D.C. input capacitors

90

and

92

of

FIGS. 2 and 3

comprise an impedance divider network responsible for reducing input voltage and limiting input current to the circuitry of FIG.

3

. In

FIG. 5

, a step-down transformer

200

is employed to produce this function. The transformer

200

combines with bridge rectifier diodes

202

,

204

,

206

and

208

and voltage regulator

210

to comprise the main current carrying elements of the charger circuitry of FIG.

5

. Transistor

212

controls the voltage regulator

210

which is used as a pass element for the charge current in the circuitry of FIG.

5

. The microprocessor

18

of

FIG. 5

provides a control signal from pin

9

to turn charge current on and off to battery

216

via the transistor

212

and the voltage regulator

210

. The software program contained in the microprocessor

18

determines whether this “switch” is on or off. In essence, the software program in the microprocessor

18

is responsible for regulating charge voltage, via pin

9

, to a temperature-compensated voltage level, based on charge current and battery voltage that the microprocessor

18

is monitoring, via pins

2

and

3

, of the microprocessor

18

.

Software control in the circuits of the invention for implementation of charger control is advantageous relative to chargers utilizing hardware for control of voltage regulation set points and temperature compensation adjustment factors. In particular, these advantages include the fact that software control-allows a system to take advantage of the monitoring of battery voltage and current inherent in the self-diagnostic system and utilizes this function for the dual purpose of charge control. Further, such a system is easily configurable to different battery voltage levels and battery chemistries by simply changing configuration variables used by the software program contained in the memory of the EEPROM

32

, these configuration variables containing charge voltage level setpoints and temperature compensation factors. Still further, the microprocessor

18

can make use of an internal temperature sensing diode to provide a low cost method of measuring the temperature inside of the unit so that the charger output voltage can be temperature compensated for ambient temperature to maximize battery life at temperature extremes.

Referring back to

FIG. 5

in particular, a lamp output section of the circuitry of

FIG. 5

consists of a simple relay transfer circuit comprising resistor

218

, transistor

220

, diode

222

and relay

224

. This relay transfer circuit essentially replaces the inverter and peak current detection components of the circuitry of FIG.

3

. Pin

8

of the microprocessor

18

is used to turn on and off the relay

224

via the transistor

220

when microprocessor software detects AC power failure or for scheduled self-diagnostics testing. Since the incandescent lamping

226

of

FIG. 5

provides a larger DC load in the emergency mode, it is necessary to utilize the relay transfer circuit shown in FIG.

5

.

In the circuitry of

FIG. 3

, a simple current sensing element is provided in the form of the resistor

96

. In the circuitry of

FIG. 5

, a current sensing circuit is seen to be provided by processor

18

, amplifier

230

, resistors

232

,

234

,

236

,

238

,

240

and

258

; capacitors

242

,

246

,

248

and

250

; and diodes

252

,

254

and

256

. A voltage converter circuit is provided by the voltage converter integrated circuit

228

and the capacitors

246

,

248

and

250

and diode

256

, this voltage converter circuitry

228

providing a −5V rail for the amplifier circuit which is formed around the amplifier

230

. The amplifier circuit formed around the operational amplifier

230

is a variation of an active non-saturating, full-wave precision rectifier circuit such as is referred to as an absolute value circuit. The left-hand portion of the circuit is an active-wave rectifier circuit while the right-hand portion of the circuit is an inverting summing amplifier.

In operation, the circuitry of

FIG. 5

should first be considered to be in the emergency mode. In this state, the coil of the relay

224

is energized to provide a path for battery current to flow to ground through the lamping

226

which comprise DC lamps. This lamp current by design must flow back into the negative terminal of the battery

216

through sense resistor

258

, thereby creating a negative voltage potential with respect to ground at vin. With vin as a negative input voltage, the output va of the left-hand rectifier circuit is va=0. Accordingly, one input to the summing circuit has a value of zero. However, vin is also applied as an input to the summing circuit. The gain for this input is set up by the resistors

236

and

238

and is equal to −5 using the well-know equation for inverting amplifiers −Rf/Ri. Since vin is negative and the gain of the circuit is also negative, the output v0=−5−vin and will be positive for this condition.

When emergency unit equipment utilizing the circuitry of

FIG. 5

is powered by alternating current and is charging the battery

216

, charge current flows through the battery

216

and the sense resistor

258

to ground, thus creating a positive voltage at vin. In this condition, the output of the left-hand rectifier portion of the circuit va=−vin. The voltage va appears as one input to the summing circuit, and the gain for that input is −15. As before, vin also appears directly as an input to the summing circuit. The net output is then v0=−5vin−15va=−5vin−15(−vin)=10vin. The output of the circuit is therefore positive with a gain of 10 which provides more amplification of the smaller input voltages which will occur once the battery

216

is charged and the current flow is reduced to a trickle charge.

The circuitry thus described provides a number of benefits for accomplishment of the self-diagnostic function. In particular, the microprocessor

18

has the ability to calibrate itself to the output of the circuitry during the manufacturing process, thereby avoiding the use of high tolerance components or trim pots while still maintaining an accurate measurement of current magnitude. Further, the transformation of both negative and positive input voltages to positive output voltages meets the requirement of the microprocessor Analog to Digital (A/D) input requirements for a positive input voltage less than a 3.5V dc. The present circuit also provides the ability to provide a higher gain for smaller voltage drops across the sense resistor

258

when the unit is charging to allow the circuitry to more accurately measure charge current. The ability of the present circuit to provide lower gain when operating in the emergency mode and when the voltage drop across the sense resistor

258

is therefore greater, allows the circuitry to measure the larger currents which flow when the lamping

26

is on without exceeding maximum input voltage on the A/D input of the microprocessor

18

. While specific gain values are listed herein, these gain values can change with wattage.

The algorithms necessary for production of software for programming of the self-diagnostic system seen in

FIGS. 3 and 5

are hereinafter provided in tabular form for purposes of simplicity. The language of the algorithm presentations is not numbered to conform to particular hardware of the circuits of

FIGS. 3 and 5

but are described in language which allow recognition of hardware components where necessary. The algorithm presentations begin with Table I which describes “Variables”. The algorithm presentation beginning with Table I relates essentially to exit signage using NiCd batteries. The algorithm presentation beginning with Table II relates essentially to emergency unit equipment using lead-acid batteries.

TABLE I

VARIABLES

Range of

Variable

Type

Description

values

size

units

default

Version

Config

Version string

X.XX

4 char

1.00

Date_manufacture

Config

A string containing the manufacturing date

971101

6 char

Product_name

Config

A string for human recognition only. Probably the assembly

XXXXXXXXX

10

product

number

char

Product_number

Config

A unique number, identifying each product, incremented each

24

0

time a product is tested and configured on the ATE

Product_type

Config

Defines product type.

1 = EXIT with

8

product

This variable also determines pin usage on the micro

inverter

2 = unit

3 = inverter

Features_enabled

Config

Defines enabled features

see separate

8

product

Combined with SD2 switch settings as a mask to determine

table

exact product functionality

Batt_type

Config

We expect to use NiCad and Lead acid battery chemistries.

1 = Nicad

8

product

Nicad requires 2-level charging (full or trickle) controlled by

2 = Lead Acid

PA3. No need for temperature compensation. Actual charge

current and voltage levels are determined by external hardware.

Lead acid has full, and trickle voltages and has calibrated temp.

compensation of charge voltage. The charge voltage (V_full)

should reduce by 2.1 mV/° C. Voltage level controlled by micro

by PWM integration. Current limited by external electronics.

V_LVD

Config

Low voltage disconnect voltage

16

millivolts

product

This is the voltage at which the battery must be disconnected

the load to prevent “deep discharge” damage. The load must

always be disconnected at this voltage, regardless of other

requirements.

V_full

Config

Full charge voltage

16

millivolts

product

This is the max. battery voltage allowed for full current

charging. If the product is lead-acid, this is the voltage that the

output is regulated to in full charge mode. External current

limiting may reduce the achievable voltage below this level.

V_OK

Config

Battery OK voltage (upper limit)

16

millivolts

product

This is the max. allowable battery voltage. If a voltage is

measured above this, then the battery is assumed to have been

disconnected or gone open circuit or otherwise damaged.

Indicate BATTERY_FAIL.

V_low_batt

Config

Battery low voltage (lower limit)

16

millivolts

product

This is the min. allowable battery voltage. If a voltage is

measured lower than this, then a cell is assumed to have gone

short circuit, or is otherwise damaged. Indicate

BATTERY_FAIL.

V_trickle

Config

Lead-acid only. Trickle charge voltage (charge voltage is

16

millivolts

product

regulated to this level)

I_trickle

Config

Lead-acid only. Trickle charge current (current threshold used

16

mA

product

to decide that the battery is fully charged).

V_brownout

Config

The mains voltage level below which we go into emergency

16

Raw A/D

cal

mode

reading

V_charge

Config

The mains voltage level above which we go back into charger

16

Raw A/D

cal

mode

reading

I_load_min

Config

lamp load must be greater than this otherwise we indicate

16

mA

product

LAMP_FAIL

batt_cap

Config

Defines the total battery capacity

16

mA hours

product

Frequency_test_short

Config

Defines how often the short test is run

16

hours

720

Seconds_test_short

Config

time for short self-test

8

secs

30

Seconds_test_man

Config

time for manual test (button / remote input)

8

secs

30

Frequency_test_long

Config

Defines how often the long test is run

16

hours

4320

Mins_test_long

Config

factory configured time for long self-test

8

mins

30

Hours_reschedule

Config

The time to offset a test if the battery is not fully charged

8

hours

24

Hours_delay

Config

The time to offset a test if the user presses the test button (or

8

hours

8

remote) during a (successful) scheduled test

Batt_cap_short

Config

Defines the min. batt % for the short test or manual test to be

8

%

70

run. If < this %, a short test is delayed by hours_reschedule,

manual test is aborted.

Batt_cap_long

Config

Defines the min. batt % for the long test to be run - otherwise it

8

%

90

is delayed by hours_reschedule

Discharge_ratio

Config

Estimate of the ratio between the load voltage and charge

8

Exact

Product

voltage for the given product, including an efficiency factor if

ratio

needed. For exits that use inverters, the load voltage is˜lamp

voltage. Used to determine batt_cap

Sense_res

Config

A value allowing s/w to calculate the relationship between

16

milliohms

Product

measured voltage and actual charge current

Off_time

Config

A loop count value, giving the off-time for the inverter (LED

8

counts

Product

exits only). 6 μs per count. 0 = 5 μs, 1 = 11 μs, 2 = 17 μs etc

DC_scale

Config

The scale factor generated as a result of calibration

16

Raw A/D

0

V_dc = (a2d_delay * DC_scale) / 65536

reading

Time2ShortTest

Config

The number of hours until start of the next short test

16

Hours

30

Time2LongTest

Config

The number of hours until start of the next long test

16

Hours

168

OfftimeIncdelay

Config

The number of minutes between increments of the off_time

8

counts

Product

MaxOffTime

Config

The maximum value of the off-time variable for the inverter

8

counts

Product

(LED exits only).

Mins_batt_fail

Config

If a discharge lasts longer than this, the number of battery

8

Mins

60

failures in incremented in the log, but no error is shown (used

for product evaluation). Only updated if the battery was fully

charged at start of discharge

Temp_comp

Config, on

A variable (or variables) written by the ATE during calibration

??4

Cal

ATE

to define the relationship between the internal temp diode

char

reading and the offset or factor to apply to V_full. This might

be a table.

Handshake

Config, on

A location for the ATE and the micro to write their status bytes

8

0

ATE

to perform handshaking

Min_battery_runtime

Log

The minimum number of minutes that the unit has run in test or

8

Mins

255

emergency mode to LVD since day 0. Only updated if the

battery was fully charged at start of discharge

Charge_level

Log

Estimate of the current battery charge level

32

mA mins

0

I_load_learnt

Log

Current recorded during learn

16

mA

0

Count_brownouts

Log

Number of power failures

16

events

0

Count_LVD

Log

Number of times went into LVD, from a fully charged state

16

events

0

Count_user_test

Log

Number of times user has pressed test button or remote

16

events

0

Count_short_test

Log

Number of times short test has run

16

events

0

Count_long_test

Log

Number of times long test has run

16

Events

0

Count_lamp_fail

Log

16

Events

0

Count_batt_fail

Log

16

Events

0

Mins_running

Log

Total time product has been running since day 0. Can take the

24

Mins

0

modulo of this value to determine when to do a test

Mins_hicharge

Log

Total time product has been charging at full since day 0

24

Mins

0

Mins_tricklecharge

Log

Total time product has been charging at trickle since day 0

24

Mins

0

Mins_total_discharge

Log

Total time product has been in discharge since day 0

24

Mins

0

Mins_Total_LVD

Log

Total of the minutes running in discharge, when the discharge

24

Mins

0

ended in LVD, and started from a fully charged state. Used with

“Count_LVD” to determine average time to LVD over product

life.

Error_code

Log

Log of the last error. 8 bit error code is a number representing

16

error

0

the mnemonics listed in “LED states” with any additional data

code

relevant to the failure in the other 8 bits

Mins_total_test

Log

Total time product has been in test since day 0

24

Mins

0

Mins_last_test

Log

The timestamp of the last test - copy of Mins_running at the

24

Mins

0

time of the last test

Temp_ref_read

Log

Measured temp ref value during power-up

16

Raw A/D

0

reading

ROM_checksum

Log

Calculated during calibration, so ATE can compare with the

16

raw value

0

published value

F_FA_Fl_offtimeadj

Config

When Flashing, Flashing/Audible, or Fire Alarm Interface

8

counts

Product

options are enabled, inverter off time is adjusted by this value

Note:

calibration values should be stored in a fixed area at the start of EEPROM so that the product configuration can be changed later without changing the ATE calibration. E.g. reserve 1st 8 bytes for calibration values. It may be convenient to group all ATE changeable parameters into a block, e.g. calibration, date_manufacture, product_number

A 16 bit, simple byte-additive checksum will be applied to the config part of the data to ensure integrity.

SD1 Operation (Re FIG. 3)

Function

What to do

When

Special conditions

Indicator states

CHARGE

See separate table

See separate table

see separate table

see separate table

BATTERY

EMERGENCY

Turn load on (Exits - start inverter -

IF V_input < V_brownout

Ignore batt_cap value here.

EMERGENCY

PA5, Units - turn on PA5 full) - see

AND V_Batt > V_LVD

DO NOT indicate LVD or lack of

“inverter operation”

Keep going until V_LVD

charge as an error condition,

Decrement charge_level:

even if it happens after a few

Discharge_ratio * mins * ((voltage

secs. Brownout operation is

[I_load] − voltage [v_batt]) /

required for functional test on

sense_res)

ATE and mech assy.

Increment mins_total_discharge

Perform LVD actions below, if

conditions met.

LVD

Turn off load (stop inverter - PA5)

When V_Batt <= V_LVD

Monitor V_Batt

SELF_TEST or . . .

Zero charge_level

AND in test mode OR brownout

if discharge started when

BATTERY_FAIL

Log status

charge_level >= batt_cap, then

log min_battery_runtime, and

log count_LVD

MANUAL

Stop charging (turn on PA4)

When test button voltage (PB4) <

Monitor V_Batt

SELF_TEST or . . .

TEST

Turn on load (start inverter - PA5) for

1V

if V_Batt < V_LVD, end test,

BATTERY_FAIL

the duration of Seconds_test_short

OR remote control message

zero charge_level, log stats.

LAMP_FAIL

Check battery voltage

received (I

2

C)

monitor I_lamp:

INSUFFI-

Check lamp current

AND charge_level >

if I_Load < I_Load_min, OR

CIENT_CHARGE

Decrement charge_level:

(batt_cap_short * batt_cap)

(I_load_learnt * 1.1) [if

(while test switch

Discharge_ratio * mins * ((voltage

I_load_learnt > 0 and

held)

[I_load] − voltage [v_batt]) /

depending on

sense_res)

features_enabled] then end

Increment mins_total_test,

test

mins_total_discharge and

If charge_level <

mins_manual_test

batt_cap_short, then quit

Store mins_last_test

During test, the test switch or

perform ROM and EEPROM

remote becomes “kill test”

checksum

SHORT TEST

Stop charging (turn on PA4)

When Time2ShortTest = 0 and

Monitor V_Batt

SELF_TEST or

Turn on load (start inverter - PA5) for

not already in test

if V_Batt < V_LVD, end test,

BATTERY_FAIL

the duration of Seconds_test_short

AND charge_level >

zero charge_level, log stats.

LAMP_FAIL

check battery voltage

(batt_cap_short * batt_cap)

monitor I_lamp:

check lamp current

if I_Load < I_Load_min, OR

decrement charge_level:

(I_load_learnt * 1.1) [if

discharge_ratio * mins * ((voltage

I_load_learnt > 0 and

[I_load] − voltage [v_batt]) /

depending on

sense_res)

features_enabled] then end

increment mins_total_test,

test

mins_total_discharge, and

if charge_level <

mins_short_test

batt_cap_short, the test is held

store mins_last_test

off for hours_reschedule

perform ROM and EEPROM

During test, test switch <1V or

checksum

remote becomes “kill test and

hold off for hours_delay”

LONG TEST

As above (replace any occurrencesd

When Time2LongTest = 0 and

as above (replace any

SELF_TEST or . . .

of “short” with “long”)

not already in test

occurrencesd of “short” with

BATTERY_FAIL

AND charge_level >

“long”)

LAMP_FAIL

(batt_cap_long * batt_cap)

TEST DELAY

Quit test in progress and re-schedule

If in a SCHEDULED test

normal indicator

test in hours_delay

AND test button pressed

priority

OR remote control message

received (I

2

C)

CHIRP

500 ms chirp every 15 mins (like a

When any test failed, until test

all products, depends on

smoke detector)

button pressed (input <1V) OR

features_enabled

remote control message received

(I

2

C) - i.e. error cleared

(test button acts as a reset, then

reverts to being a test button)

CHECK DATA

EEPROM data integrity verified

At power up (except in calibration

ELEC-

against the byte additive checksum

phase 2)

TRONICS_FAIL

Check EPROM checksum

As a part of each test

CALIBRATE

Read internal temp ref. and dump to

At power up, dependent on the

If out of range (TBD)

ELEC-

log

value of the handshake variable

TRONICS_FAIL

Battery Charging Algorithms

These battery charging algorithms are a compromise between simplicity and good battery maintenance to enhance capacity and life.

Function

What to do

When

Special conditions

Indicator states

CHARGE

Turn off PA3

When not in TEST

if V_Batt > V_OK OR

BATTERY -

Increment charge_level =

AND V_Input > V_charge

V_Batt < V_low_Batt

FULL

mins * ((voltage [I_load] − voltage [v_batt]) /

AND V_batt <= V_full (full

then BATTERY_FAIL

CURRENT

sense_res)

charge threshold)

by default, indicate

If V_batt < V_LVD, then Charge_level = 0

AND charge_level < batt_cap

HI_CHARGE

If V_batt > V_full, then charge_level =

batt_cap

CHARGE

Turn on PA3

When not in TEST

Note that PA3 being ON

if V_Batt > V_OK OR

BATTERY -

This is not assumed to add any charge to

AND V_Input > V_charge

has the added benefit of

V_Batt < V_low_Batt

TRICKLE

the battery. Just maintains charge_level =

AND V_Batt > V_full

keeping a decent lamp

then BATTERY_FAIL

CURRENT

batt_cap

OR charge_level >= batt_cap

current flowing if the

by default, indicate OK

If V_batt < V_LVD, then charge_level = 0

battery has been removed.

Function

What to do

When

Special conditions

Indicator states

CHARGE

PWM cycle at ??Hz regulate V_batt to

When not in TEST

as NiCad

BATTERY -

V_full * temp_comp * V_tcomp_diode

AND V_Input > V_charge

FULL

Increment charge_level =

AND ((voltage [I_load] − voltage

CURRENT

mins * ((voltage [I_load] − voltage [v_batt]) /

[v_batt[) / sense_res) > I_trickle

sense_res)

AND charge_level < batt_cap

If V_batt < V_LVD, then charge_level = 0

CHARGE

PWM cycle at ??Hz regulate to V_batt to

When not in TEST

Cycle back to full current if

as NiCad

BATTERY -

V_trickle * temp_comp * V_tcomp_diode

AND V_Input > V_charge

the current rises above

TRICKLE

This is not assumed to add any charge to

AND ((voltage [I_load] − voltage

I_trickle (indicates battery

CURRENT

the battery. Just maintains charge_level =

[v_batt]) / sense_res) < I_trickle

discharging), and reset

batt_cap

OR charge_level >= batt_cap

charge_level to batt_cap

If V_batt < V_LVD, then Charge_level = 0

Note: hysteresis needed

here to prevent fast cycling

Inverter Operation

The inverter uses several variables to define its operation. The idea is to optimize the battery discharge to ensure that the UL requirement (of having 60% of initial (@1 minute) brightness at 90 minutes) is met with the smallest and therefore lowest cost battery.

The software will increase the off_time every OffTimeIncDelay minutes, until MaxOffTime has been reached, and the Offtime will remain at that value until the end of inverter operation (end of test or LVD). These values will be used whenever the inverter is running.

For F/FA/FI options, the off time operation must be adjusted (see next section).

SD2 Operation

SD2 will be configured by features_enabled stored in an EEPROM (different slave address) on the SD2 PCB. The SD2 PCB houses the buzzer too.

Each of the configuration options below act as a mask to modify the features_enabled variable read from the SD2 EEPROM:

Function

What to do

When

Special conditions

Indicator states

FLASHING (F)

Lamp flashes at 1 Hz, 50% duty cycle

EMERGENCY

While running in EMERGENCY,

EMERGENCY

generate off time variables by

subtracting F_FA_Fiofftimeadj

from off_time and MaxOffTime

FLASHING/

Lamp flashes at 1 Hz, 50% duty cycle;

EMERGENCY

See FLASHING

EMERGENCY

AUDIBLE (FA)

Audible (PA2) is turned on and off at

1 Hz (on 0.5 seconds, off 0.5 seconds)

FIRE ALARM

Lamp flashes at 1 Hz, 50% duty cycle

The test input (PB4) is

See FLASHING

See separate table

INTERFACE (FI)

detected in the range 2.4 to

2.8 V for >300 ms

Flashing in EMERGENCY can be accomplished by shutting off the boost converter (PA5) to turn off the lamp.

Flashing while charging can be accomplished by clamping the bridge (turning on PA4) to turn off the lamp.

SD “Features_enabled” Variable

SD1 Features_enabled

The features_enabled variable for SD1 is defined as follows. All values are binary.

Value

Name

What is does

00000001

flashing

Flashes when in brownout

00000010

load_learn

Enables load learning function

00000100

remote

Enables remote control features

00001000

time_delay

Enables time delay of 15 min.

00010000

audio_enabled

Enables chirp

00100000

nominal_v_test

Enables check for battery voltage

not falling below nominal during

first 30 seconds of test

01000000

undefined

10000000

SD_enabled

Enables all SD functions. If this bit

is not set, then the LED only

indicates charging mode (normal/

hi-charge), and the test switch

just activates load from the battery

(via inverter - or relay if it's a

unit) while the test switch is pressed.

SD2 Features_enabled (EEPROM on SD2 PCB)

The features_enabled variable for SD2 is defined as follows. All values are binary.

Value

Name

What is does

00000001

flashing

Flashes when in brownout

00000010

load_learn

Enables load learning function

00000100

remote

Enables remote control features

00001000

time_delay

Enables time delay of 15 min.

00010000

audio_enabled

Enables audible (pulses during brownout

only if flashing enabled, chirp enabled)

00100000

nominal_v_test

Enables check for battery voltage

not falling below nominal during

first 30 seconds of test

01000000

fire alarm

Enables flashing when fire alarm

interface

input is activated if flashing

enabled; enables pulsed audible when

fire alarm input is activated

if audible_enabled is enabled

10000000

SD_enabled

Enables all SD functions. If this bit

is not set, then the LED only

indicates charging mode (normal/

hi-charge), and the test switch

just activates load from the battery

(via inverter - or relay if it's a

unit) while the test switch is pressed.

Other variables undefined as yet.

Test Strategy

At ATE test, we need to do the following:

Test the hardware-passive component values and tolerances, and connectivity

Calibrate the hardware

Configure the EEPROM with the required variables

This spec. deals with the software requirements to assist in ATE test

Test Sequence

The ATE sequence will be:

1. Check passive components

2. Hold micro in RESET and power PCB at 83.5% of nominal supply voltage(100V)—the minimum battery charge voltage. The D.C. power supply simulating the nicads is set to the maximum charge voltage V_full—as specified for the particular product being tested.

3. Check the +5V supply voltage to the micro.

4. ATE programs the fixed EEPROM parameters, according to assembly number (by operator). This is a fixed set of data (per product), according to scruct EEPROM_CONFIG

5. ATE programs handshake byte (AA hex) to indicate to micro that it is in phase 1 of ATE test mode.

6. ATE releases RESET and begins monitoring the test switch (test point T

11

, U

2

pin

4

) to go low. If the test switch does not go low within TBD milliseconds, the unit under test fails. (tests basic microprocessor operation)

Note: The test switch has an pullup resistor so its value is high during reset, using this pin provides an unambiguous signal to the ATE that calibration mode has been recognized and the current calibration phase has been completed. During normal operation the test switch is an input and will never be driven low by the micro.

During the first phase of test/calibration, while the ATE is waiting for the test switch input to go low, the following sequence of events is occurring:

7. The EEPROM version string is created from the values stored in ROM, allowing the ATE to check ROM version, and determine the expected checksum.

8. If the EEPROM passes the checksum tests and the handshake byte is AA hex, the micro will begin the calibration sequence. If the checksum test fails or handshake value is anything other than AA or BB or 55 Hex , the micro will indicate ELECTRONICS_FAIL. and will not ever set the test switch low.

9. The micro will calculate the value of its internal ROM checksum, and write it to the ROM_checksum variable in EEPROM.

10. The micro writes temp_ref_read value read from internal temp diode to EEPROM.

11. The micro measures V_charge and stores the A/D reading in EEPROM. The value of V_full (=V_charge at this time) already programmed in the EEPROM is used to calculate and store DC_scale.

12. The micro updates the EEPROM checksum, and drives the test switch input low to indicate it has completed this phase of the calibration sequence.

The ATE has now detected the switch input as low, so phase 1 of the procedure is complete. The ATE now has to put the micro into RESET and reads the EEPROM, taking actions based on their contents.

1. When the ATE detects a low on the test switch it applies RESET to the micro, changes the A.C. input voltage to V_brownout.

2. The ATE reads the value of the ROM_checksum variable, and compares it with the published ROM checksum for that software version. A mismatch indicates a test failure.

3. The ATE reads temp_ref_read value from EEPROM and calculates temp_comp based on it's own readings of temperature.

4. The ATE writes the temp_comp calibration parameter(s) to EEPROM, having calculated the necessary relationship between temp_ref_read, and room temperature. (TBD)

5. The ATE writes a value of BB hex to the EEPROM handshake location to indicate to the micro that it is to perform phase 2 of the cal. process.

6. After the new voltage settings have stabilized the ATE releases RESET and once again waits for the test switch to go low. If the test switch does not go low within TBD milliseconds, the unit under test fails.

The ATE now allows the micro to run, and perform the second phase of calibration.

7. The micro measures and stores V_brownout in the EEPROM.

8. If the calibration values within acceptable limits (TBD), it writes the acknowledge value 55 Hex into the handshake byte in the EEPROM.

9. The micro drives the test switch low, indicating completion of the final calibration phase.

10. The micro recalculates the EEPROM checksum, allowing for the added new variables.

11. When the ATE detects a low on the test switch, it applies RESET to the micro and reads the handshake byte from the EEPROM. The handshake byte value is not 55 Hex then the unit under test fails.

12. The ATE then releases reset and performs functional test as desired.

Final Test

The software will support a final test mode, designed to allow any factory induced failures to be cleared out, and clear down the false charge_level variable necessary to get the product through functional test in final assembly. This function may also be used to clear down demo units or returns from the field. The product must be reconfigured via ATE if further tests need to be run after this procedure.

Shutdown Mode

If the test input (PB4) is detected in the range 1.4 to 1.8V for >300 ms AND it is in emergency mode (the inverter running in the case of exits), then the unit goes into shutdown mode. The process is:

Set the following variables:

Variable

Set to

Min_battery_runtime

255

Charge_level

0

I_load_learnt

0

Count_brownouts

0

Count_LVD

0

Count_user_test

0

Count_short_test

0

Count_long_test

0

Count_lamp_fail

0

Count_batt_fail

0

Mins_running

0

Mins_hicharge

0

Mins_total_discharge

0

Mins_total_LVD

0

Error_code

0

Mins_total_test

0

Mins_last_test

0

Following the resetting of variables, the software will:

Configure the switch input into an output, and drive it high for 1 second to drive the LEDs

Switch off the inverter and halt the processor.

LED Exit Functions

product_type=1

Load Switching Differences

In LED exits, the load is a switching output, with an off time (low) >=11 μs controlled by the variable off_time and an on-time (high) controlled by the time it takes for an interrupt to occur on the external INT pin.

The software maintains a watchdog on this process to ensure that an interrupt has occurred within 500 ms, as a failure of the current-sense comparator will destroy the inverter transistor. This could be accomplished by timestamping each interrupt. The safe level is LOW.

SD1

Signature Exit Signs

The following are the product-specific values for EEPROM configuration.

Values are in decimal unless otherwise indicated.

Variable

Type

Value

Product_name

Config

Signature

Product_type

Config

1

Features_enabled

Config

00000000

(binary)

Batt_type

Config

1

V_LVD

Config

2100

V_full

Config

2950

V_OK

Config

3200

V_low_batt

Config - EEPROM

1200

(1 cell)

I_load_min

Config - EEPROM

20

Batt_cap (mAH)

Config - EEPROM

1000

Discharge_ratio

Config - EEPROM

15

Sense_res

Config - EEPROM

10000

off_time

Config - EEPROM

2

MaxOffTime

Config - EEPROM

2

OffTimeIncDelay

Config

0

F_FA_Fiofftimeadj

Config

0

SD2

Signature

The following are the product-specific values for the SD2 EEPROM configuration. Values are in decimal unless otherwise indicated.

Variable

Type

Value

Features_enabled

Config

00000000 (binary)

LED States

The software will generate LED states as defined in the table below

All error indications are cleared by pressing the test button (or remote input). Error conditions will be held through power fail, and continue to be indicated at next power up. More severe errors (as defined by the priority level below) will be shown in preference to less severe ones.

The priority number in the table below indicates which error should be shown in preference to others. Priority 1 is highest.

Prior-

Indicator

ity

state

Type

Mnemonic

Means

1

RED

Triple

ELEC-

General bad electronics

flashing

pulse:

TRONICS_

indicator - e.g. couldn't

125 ms on,

FAIL

go into emergency mode

125 ms off,

checksum failure, or

125 ms on,

any faults not

125 ms off,

specifically covered

125 ms on,

elsewhere

500 ms off

2

RED

1:1, 1 Hz

BATTERY

—

Battery has failed in any

flashing

FAIL

of it's detected fail

modes (e.g. discon-

nected, didn't last

the test duration

before LVD etc)

3

RED

Double

LAMP

—

Lamp failed in any of it's

flashing

pulse:

FAIL

detected fail modes (e.g.

(as above)

disconnected, out of

tolerance etc)

4

GREEN

1:1, 1 Hz

SELF

—

Product is doing a self

flashing

TEST

test (of any type)

5

OFF

EMER-

Product is in emergency

GENCY

mode

5

RED/

single

INSUFFI-

There is insufficient

GREEN

pulse

CIENT

—

energy in the battery

flashing -

1:1, 1 Hz

CHARGE

to do the requested test

only when

test

button

pressed

or external

command

6

RED

ON

HI

—

Hi-charge mode (i.e. it's

CHARGE

charging at anything

other than trickle)

7

GREEN

ON

OK (normal)

Product OK (normal

indication of charge)

TABLE II

Variables

Range of

Variable

Type

Description

values

size

units

default

Version

Config

Version string

X.XX

4 char

1.00

Date_manufacture

Config

A string containing the manufacturing date

971101

6 char

Product_name

Config

A string for human recognition only. Probably the assembly

XXXXXXXXX

10

product

number

char

Product_type

Config

Defines product type.

1 = EXIT with

8

product

This variable also determines pin usage on the micro

inverter

2 = unit-VRLA

3 = unit-NiCd

4 = inverter

Features_enabled

Config

Defines enabled features

see separate

8

product

Combined with SD2 switch settings as a mask to determine

table

exact product functionality

V_LVD

Config

Low voltage disconnect voltage

16

millivolts

product

This is the voltage at which the lamp load must be disconnected

from the battery to prevent “deep discharge” damage. The load

must always be disconnected at this voltage, regardless of other

requirements.

V_full

Config

Full charge voltage @ 20 degrees C.

16

millivolts

product

This is the max. battery voltage allowed for full current

charging. If the product is lead-acid, this is the voltage

that the output is regulated to in full-charge mode (i.e.

this is the voltage setpoint to turn charger off). External

current limiting may reduce the achievable voltage below

this level.

V_OK

Config

Charger OK voltage (upper limit)

16

millivolts

product

This is the max. allowable charger voltage. If a voltage is

measured above this, then the charger is assumed to have

been damaged. Indicate ELECTRONICS_FAIL.

V_low_batt

Config

Battery low voltage (lower limit)

16

millivolts

product

This is the min. allowable battery voltage. If a voltage is

measured lower than this, then a cell is assumed to have gone

short circuit, or is otherwise damaged. Indicate

BATTERY_FAIL.

V_trickle

Config

Lead-acid only. Trickle charge voltage (i.e. - this is the voltage

16

millivolts

product

level that you turn charger on at) @ 20 degrees C.

I_trickle

Config

Lead-acid only. Trickle charge current (current threshold used

16

mA

product

in conjunction with the charge gauge to decide when the battery

is fully charged).

V_brownout

Config

The mains voltage level below which we go into emergency

16

Raw A/D

Cal

mode

reading

V_charge

Config

The mains voltage level above which we go back into charger

16

Raw A/D

Cal

mode

reading

sense_res

Config

The value of the sense resistor in milliohms

8

Ohms

Product

batt_cap

Config

Defines the one half of the total battery capacity (used ½ of total

16

mA hours

Product

capacity to conserve space in the EEPROM).

Frequency_test_short

Config

Defines how often the short test is run

16

hours

720

Mins_test_short

Config

Factory configured time for short self-test

8

mins

5

Seconds_test_man

Config

time for manual test (button / remote input)

8

secs

30

Frequency_test_long

Config

Defines how often the long test is run

16

hours

4320

Mins_test_long

Config

Factory configured time for long self-test

8

mins

30

Hours_reschedule

Config

The time to offset a test if the battery is not fully charged

8

hours

24

Hours_delay

Config

The time to offset a test if the user presses the test button (or

8

hours

8

remote) during a (successful) scheduled test

Batt_cap_short

Config

Defines the min. batt % for the short test or manual test to be

8

%

70

run. If < this %, a short test is delayed by hours_reschedule,

manual test is aborted.

Batt_cap_long

Config

Defines the min. batt % for the long test to be run - otherwise it

8

%

90

is delayed by hours_reschedule

Lamp_on_time

Config

Defines the amount of time to leave the lamps on after power is

8

mins

0

restored following a brownout. Used to provide time-delay

feature for unit equipment.

DC_scale

Config

The scale factor generated as a result of calibration is used to

16

Raw A/D

0

calculate the expected diode voltage when reading the

reading

temperature diode. V_TempDiode = (a2d_delay * DC_scale) /

65536.

Time2ShortTest

Nonvol

The number of hours until start of the next short test

16

Hours

30

Time2LongTest

Nonvol

The number of hours until start of the next long test

16

Hours

168

Mins_batt_fail

Config

If a discharge starts with a fully charged battery and then drops

8

Mins

60

below V_lvd before this amount of time has elapsed then we

count a failure.}

Handshake

Config, on

A location for the ATE and the micro to write their status byte to

8

0

ATE

perform handshaking during calibration

VdeltaC

Config

Number of millivolts/degree C. * 5 to reduce V_full and

8

Milli-

3.8

V_trickle if the temperature is above 20 degrees C.

volts * 5

mV /

Also the number used to increase V_full and V_trickle if

cell

the temperature is below 20 degrees C.

TempVZero

Config

Nominal temperature diode voltage in millivolts at 0 degrees C.

16

Millivolts

747

TempDeltaV

Config

Nominal temperature diode slope in millivolts / degree C. * 10

8

Milli-

16

volts * 10

AC_adjust

Config

Value to be added to V_brownout when charging to compensate

8

Raw

Product

for transformer I2R losses.

value

BatV_scale

Config

Raw A/D reading of pin 2 taken by A/D during calibration.

16

Raw A/D

Cal

It will be used to calculate a calibration factor for measuring

reading

battery voltage. V_Batt = (a2d_delay * BatV_scale) / 65536

I_charge_Factor

Config

Scaling factor to calibrate current measurements while charging

16

Raw A/D

Cal

reading

I_discharge_factor

Config

Scaling factor to calibrate current measurement while

16

Raw A/D

Cal

discharging

reading

Min_battery_runtime

Log

The minimum number of minutes that the unit has run in test or

8

Mins

255

emergency mode to LVD since day 0. Only updated if the

battery was fully charged at start of discharge

Charge_level

Nonvol

Estimate of the current battery charge level

32

mA mins

0

I_load_min

Nonvol

Minimum discharge current calculated during the load learn

16

mA

0

cycle. Current is calculated by multiplying the total measured

current by the min_load_percent variable. If the measured

discharge current drops below this value, indicate LAMP_FAIL

Min_load_percent

Config

Minimum percentage of learned lamp current before

8

%

80

LAMP_FAIL is declared.

Count_brownouts

Log

Number of power failures

16

events

0

Count_LVD

Log

Number of times went into LVD, from a fully charged state

16

events

0

Count_user_test

Log

Number of times user has pressed test button or remote

16

events

0

Count_short_test

Log

Number of times short test has run

16

events

0

Count_long_test

Log

Number of times long test has run

16

Events

0

Count_lamp_fail

Log

Number of times a lamp has failed

16

Events

0

Count_batt_fail

Log

Number of times a battery has failed

16

Events

0

Mins_running

Log

Total time product has been running since day 0.

24

Mins

0

Mins_hicharge

Log

Total time product has been charging at full since day 0

24

Mins

0

Mins_tricklecharge

Log

Total time product has been charging at trickle since day 0

24

Mins

0

Mins_total_discharge

Log

Total time product has been in discharge since day 0

24

Mins

0

Mins_Total_LVD

Log

Total of the minutes running in discharge, when the discharge

24

Mins

0

ended in LVD, and started from a fully charged state. Used with

“Count_LVD” to determine average time to LVD over product

life.

Error_code

Log

Log of the last error. 8 bit error code is a number representing

8

error

0

the mnemonics listed in “LED states” with any additional data

code

relevant to the failure in the other 8 bits

0 No error (ever)

1 V_LVD detected during test

2 V_LVD detected during brownout

3 V_LVD detected during normal operation

4 Battery V_OK check failed

5 Battery V_low_batt check failed

6 Lamp current check failed

7 EEPROM checksum incorrect

8 Not used

9 Invalid handshake byte in EEPROM

10 Micro Code checksum incorrect

11 Battery below nominal during first 30 secs of test)

ErrFlags

Nonvol

Bit map indicating current failures which have not be cleared by

8

Bit map

0

the user:

00000001 = ElectronicsFail

00000002 = BatteryFail

00000004 = LampFail

Mins_total_test

Log

Total time product has been in test since day 0

24

Mins

0

Mins_last_test

Log

The timestamp of the last test - copy of Mins_running at the

24

Mins

0

time of the last test

Temp_ref_read

Log

Set to the current temperature by the ATE during calibration.

16

Deg C.

Cal

During normal operation it is the measured temperature at the

time the EEPROM was last updated

ROM_checksum

Log

Calculated during calibration, so ATE can compare with the

16

raw value

0

published value

Ibatt

Log

Calculated value of measured current

16

MA

0

Note:

calibration values should be stored in a fixed area at the start of EEPROM so that the product configuration can be changed later without changing the ATE calibration. E.g. reserve 1st 8 bytes for calibration values. It may be convenient to group all ATE changeable parameters into a block, e.g. calibration, date_manufacture, product_number

4 byte in EEPROM available

Log Bytes = 40, Config Bytes = 53, Non_Vol = 11, Other Bytes = 20

A 16 bit, simple byte-additive checksum will be applied to the config part of the data to ensure integrity.

SDI Operation (Re FIG. 5)

Function

What to do

When

Special conditions

Indicator states

CHARGE

See separate table

See separate table

see separate table

see separate table

BATTERY

EMERGENCY

Turn load on (Units - turn on PA5 full)

IF V_Input < V_brownout

Ignore batt_cap value here.

EMERGENCY

Decrement charge_level: mins *

AND V_Batt > V_LVD

DO NOT indicate LVD or lack of

I_discharge

Keep going until V_LVD

charge as an error condition,

Increment mins_total_discharge

even if it happens after a few

secs. Brownout operation is

required for functional test on

ATE and mech assy.

Perform LVD actions below, if

conditions met.

LVD

Turn off load

When V_Batt <= V_LVD

Monitor V_Batt

EMERGENCY

Zero charge_level

AND in test mode OR brownout

If discharge started when

Log status

charge_level >= full, then log

min_battery_runtime, and log

count_LVD

TIME

Stop charging (PA4 on “high”)

When AC power is restored after

Time delay operation of DC

STEADY GREEN

DELAY

Turn on load (PA5 on “high”) for the

a brownout.

lamps, set by lamp_on_time

duration of Seconds_test_man

variable can be canceled at any

Check battery voltage

time during the time delay by

Check lamp current

pressing the test switch.

Decrement charge_level: mins *

I_discharge

MANUAL

Stop charging (PA4 on “high”)

When test button voltage

Monitor V_Batt

SELF_TEST or . . .

TEST

Turn on load (PA5 on “high”) for the

(PB4) < 1 V

if V_Batt < V_LVD, end test,

BATTERY_FAIL

duration of Seconds_test_man

OR remote control message

zero charge_level, log stats.

LAMP_FAIL

Check battery voltage

received (I

2

C)

monitor I_lamp:

INSUFFI-

Check lamp current

AND charge_level >

if I_Load < I_Load_min, and

CIENT_CHARGE

Decrement charge_level: mins *

(batt_cap_short * batt_cap)

depending on

(while test

I_discharge

features_enabled] then end

switch held)

Increment mins_total_test,

test

mins_total_discharge and

If charge_level <

mins_manual_test

batt_cap_short, then quit

Store mins_last_test

During test, the test switch or

perform ROM and EEPROM

remote becomes “kill test”

checksum

SHORT TEST

Stop charging (PA4 on “high”)

When Time2ShortTest = 0 and

Monitor V_Batt

SELF_TEST or . . .

Turn on load (PA5 on “high”) for the

not already in test

if V_Batt < V_LVD, end test,

BATTERY_FAIL

duration of Seconds_test_short

AND charge_level >

zero charge_level, log stats.

LAMP_FAIL

check battery voltage

(batt_cap_short * batt_cap)

monitor I_lamp:

check lamp current

if I_Load < I_Load_min, and

Decrement charge_level: mins *

depending on

I_discharge

features_enabled] then end

increment mins_total_test,

test

mins_total_discharge, and

If charge_level <

mins_short_test

batt_cap_short, the test is held

store mins_last_test

off for hours_reschedule

perform ROM and EEPROM

During test, test switch <1 V or

checksum

remote becomes “kill test and

hold off for hours_delay”

LONG TEST

As above (replace any occurrences of

When Time2LongTest = 0 and

as above (replace any

SELF_TEST or . . .

“short” with “long”)

not already in test

occurrences of “short” with

BATTERY_FAIL

AND charge_level >

“long”)

LAMP_FAIL

(batt_cap_long * batt_cap)

TEST DELAY

Quit test in progress and re-schedule

If in a SCHEDULED test

normal indicator

test in hours_delay

AND test button pressed

priority

OR remote control message

received (I

2

C)

CHIRP

500 ms chirp every 15 mins (like a

When any test failed, until test

all products, depends on

smoke detector)

button pressed (input <1 V) OR

features_enabled

remote control message received

(I

2

C) - i.e. error cleared

(test button acts as a reset, then

reverts to being a test button)

CHECK DATA

EEPROM data integrity verified

At power up (except in

ELEC-

against the byte additive checksum

calibration phase 2)

TRONICS_FAIL

Check EPROM checksum

As a part of each test

CALIBRATE

Read internal temp ref. and dump to

At power up, dependent on the

If out of range (TBD)

ELEC-

log

value of the handshake variable

TRONICS_FAIL

Battery Charging Algorithms

These battery charging algorithms are a compromise between simplicity and good battery maintenance to enhance capacity and life.

Lead Acid (This description assumes starting with a fully discharged battery)

Function

What to do

When

Special conditions

Indicator states

CHARGE

Regulate V_batt to V_full ± VDeltaC

When not in TEST

Regulate charge voltage at

as NiCad

BATTERY -

Increment charge_level =

AND V_Input > V_charge

V_full until Charge_level =

FULL

mins * I_batt

Wait for Charge_level = full

full.

CURRENT

If V_batt < V_LVD, then charge_level = 0

before going to trickle charge OR

after 168 hrs of continuous high

charge:

CHARGE

Regulate V_batt to V_trickle ± VDeltaC

When not in TEST

Cycle back to full charge

as NiCad

BATTERY -

This is not assumed to add any charge to

AND V_Input > V_charge

voltage levels if the current

TRICKLE

the battery. Just maintains charge_level =

AND V_batt <= V_trickle

rises above I_trickle

CURRENT

batt_cap

AND I_charge < I_trickle

(indicates battery

If V_batt < V_LVD, then Charge_level = 0

discharging), and reduce

charge_level by 10%

SD2 Operation

The hardware of SD2 can include a simple I

2

C port or another microdevice that decodes radio data received, reads a set of config switches and the Fire Alarm input. The SD2 PCB houses the buzzer too. Either solution is assumed to involved the main micro reading the slave in a master-read mode, as if it were reading an EEPROM at a different address.

Each of the configuration options below act as a mask to modify the features_enabled variable read from EEPROM. If no SD2 board is detected at power-up, the mask is set to FF Hex so that any features programmed on the ATE are automatically enabled.

The following functions are so far defined for SD2 hardware and software:

Configuration of “long” test duration, by DIP switches or other method. An appropriate I

2

C message is sent from the daughter board

Enable/disable buzzer by DIP switches or other method An appropriate I

2

C message is sent from the daughter board

An additional feature is the reception of serial radio data, passing it to the main processor. This may or may not be decoded by the SD2 board, depending on feasibility. If decoded by the SD2 hardware, then it will pass a simple I

2

C message to the main processor.

SD “features_enabled” Variable

The features_enabled variable is defined as follows. All values are binary

Value

Name

What is does

00000001

Unused

00000010

Unused

00000100

Remote

Enables remote control features

00001000

Unused

00010000

audio_enabled

Enables chirp

00100000

Nominal_v_test

Enables the UL Nominal voltage

test during the first

30 seconds of a test

01000000

Power_on_delay

**Test only? ? ? Do we use this

for the 5 second delay in calibration?

10000000

SD_enabled

Enables all SD functions. If this bit

is not set, then the LED only indicates

charging mode (normal/hi-charge), and the

test switch just activates load

from the battery (via inverter - or

relay if it's a unit) while

the test switch is pressed.

Other variables undefined as yet.

TEST Strategy

At ATE test, we need to do the following:

Test the hardware-passive component values and tolerances, and connectivity

Calibrate the hardware

Configure the EEPROM with the required variables

This spec. deals with the software requirements to assist in ATE test

Test Sequence

The ATE sequence will be:

1. Check passive components

2. Hold micro in RESET and power PCB at 83.5% of nominal supply voltage(100V)—the minimum battery charge voltage. The D.C. power supply simulating the battery is set to the maximum charge voltage V_full—as specified for the particular product being tested. A separate DC power supply draws 4.5 amps of current through the current sense resistor to create a negative voltage drop simulating a charge current.

3. Check the +5V supply voltage to the micro.

4. ATE programs the fixed EEPROM parameters, according to assembly number (by operator). This is a fixed set of data (per product), according to struct EEPROM_CONFIG

5. ATE programs temp_ref_read EEPROM variable with the current ambient temperature in degrees C. Micro will use this value to calibrate the temperature reading.

6. ATE programs handshake byte (AA hex) to indicate to micro that it is in phase 1 of ATE test mode.

7. ATE releases RESET and begins monitoring the test switch (test point T

12

, U

6

pin

4

) to go low. If the test switch does not go low within 3 seconds, the unit under test fails. (tests basic microprocessor operation)

Note: The test switch has an pullup resistor so its value is high during reset, using this pin provides an unambiguous signal to the ATE that calibration mode has been recognized and the current calibration phase has been completed. During normal operation the test switch is an input and will never be driven low by the micro.

During the first phase of test/calibration, while the ATE is waiting for the test switch input to go low, the following sequence of events is occurring:

8. If the EEPROM passes the checksum tests and the handshake byte is AA hex, the micro will begin the calibration sequence. If the checksum test fails or handshake value is anything other than AA or BB or 55 Hex, the micro will indicate ELECTRONICS_FAIL, and will not ever set the test switch low.

9. The EEPROM version string is created from the values stored in ROM, allowing the ATE to check ROM version, and determine the expected checksum. The micro will calculate the value of its internal ROM checksum, and write it to the ROM_checksum variable in EEPROM.

10. The micro measures V_charge and stores the A/D reading in EEPROM. The value of V_full (=V_charge at this time) already programmed in the EEPROM is used to calculate and store DC_scale.

11. The micro measures voltage I_load and calculates and stores I_discharge_factor.

12. The micro measures the voltage on U

6

pin

2

and uses this value to calculate and store a calibration factor for battery measurement BatV_scale.

13. The micro updates the EEPROM checksum, and drives the test switch input low to indicate it has completed this phase of the calibration sequence.

The ATE has now detected the switch input as low, so phase 1 of the procedure is complete. The ATE now has to put the micro into RESET and reads the EEPROM, taking actions based on their contents.

1. When the ATE detects a low on the test switch it applies RESET to the micro, changes the A.C. input voltage to V_brownout.

2. The ATE reads the value of the ROM_checksum variable, and compares it with the published ROM checksum for that software version. A mismatch indicates a test failure.

3. The ATE writes a value of BB hex to the EEPROM handshake location to indicate to the micro that it is to perform phase 2 of the cal. process.

4. The ATE draws 1 amp of current through the current sense resistor to create a positive voltage drop.

5. After the new voltage settings have stabilized the ATE releases RESET and once again waits for the test switch to go low. If the test switch does not go low within TBD milliseconds, the unit under test fails.

The ATE now allows the micro to run, and perform the second phase of calibration.

6. The micro measures and stores V_brownout in the EEPROM.

7. The micro measures voltage I_load and calculates and stores I_charge_factor.

8. If the calibration values within acceptable limits (checked against values in the ATE program), it writes the acknowledge value 55 Hex into the handshake byte in the EEPROM.

9. The micro recalculates the EEPROM checksum, allowing for the added new variables.

10. The micro drives the test switch low, indicating completion of the final calibration phase.

11. When the ATE detects a low on the test switch, it applies RESET to the micro and reads the handshake byte from the EEPROM. The handshake byte value is not 55 Hex then the unit under test fails.

12. The ATE then releases reset and performs functional test as desired.

Final Test

The software will support a final test mode, designed to allow any factory induced failures to be cleared out, and clear down the false charge_level variable necessary to get the product through functional test in final assembly. This function may also be used to clear down demo units or returns from the field. The product must be reconfigured via ATE if further tests need to be run after this procedure.

Shutdown Mode

If the test input (PB4) is detected in the range 1.4 to 1.8V for >300 ms AND it is in emergency mode (the inverter running in the case of exits), then the unit goes into shutdown mode. The process is:

Set the following variables:

Variable

Set to

Min_battery_runtime

255

Charge_level

100%

I_load_min

0

Count_brownouts

0

Count_LVD

0

Count_user_test

0

Count_short_test

0

Count_long_test

0

Count_lamp_fail

0

Count_batt_fail

0

Mins_running

0

Mins_hicharge

0

Mins_total_discharge

0

Mins_total_LVD

0

Error_code

0

Mins_total_test

0

Mins_last_test

0

Following the resetting of variables, the software will:

Configure the switch input into an output, and drive it high for 1 second to drive the LEDs

Halt the processor.

Unit Functions

product_type—emergency unit equipment

Load Switching

Lamp drive is constant, high=on

Charger

PA4 is used to control a transistor tied to the adjust pin of a voltage regulator. Making PA4 high turns on the transistor which pulls the adjust pin low turning off the charge current to the battery. The PA4 pin is also used to turn the charger off during manual and self test modes.

SD1

Time delay: Keeps lamps on for lamp_on_time mins. after power is restored.

Load learn: This feature enables the unit to determine the total amount of lamp current drawn by the unit. The unit takes the result of the load learn cycle and multiplies it by the min_load_percent variable storing the result in the variable I_load_min. The load learn cycle is initiated automatically during the first scheduled self test. The load learn cycle can also be initiated manually by pressing and holding the test switch for 15 seconds. This starts a test cycle that is terminated after 15 seconds, to give the user feedback that the load learn cycle is complete.

SD1

6V, 16W

The following are the product-specific values for EEPROM configuration.

Values are in decimal unless otherwise indicated.

These values will vary according to unit DC voltage and wattage levels.

Variable

Type

Value

Product_name

Config

6 V, 16 W

Product_type

Config

2

Features_enabled

Config

11011111

(binary)

Batt_type

Config

2

V_LVD

Config

5250

V_full

Config

7200

V_trickle

Config

6900

I_trickle

Config

50

V_OK

Config

8000

V_low_batt

Config

4000

(2 cell)

I_load_min

Config

0

Batt_cap (mAH)

Config

6500

VdeltaC

Config

114

TempVZero

Config

747

TempDeltaV

Config

16

12V, 50W

The following are the product-specific values for EEPROM configuration.

Values are in decimal unless otherwise indicated.

Variable

Type

Value

Product_name

Config

12 V, 50 W

Product_type

Config

2

Features_enabled

Config

11011111

(binary)

Batt_type

Config

2

V_LVD

Config

10500

V_full

Config

14400

V_trickle

Config

13800

I_trickle

Config

50

V_OK

Config

15500

V_low_batt

Config

9000

(4.5 cell)

I_load_min

Config

0

Batt_cap (mAH)

Config

10000

VDeltaC

Config

228

TempVZero

Config

747

TempDeltaV

Config

16

SD2

Voltmeter and Ammeter functions via I

2

C—report measured charge current and voltage to bargraph LED meters.

LED States

The software will generate LED states as defined in the table below.

All error indications are cleared by pressing the test button (or remote input). Error conditions will be held through power fail, and continue to be indicated at next power up. More severe errors (as defined by the priority level below) will be shown in preference to less severe ones.

The priority number in the table below indicates which error should be shown in preference to others. Priority 1 is highest.

While the invention has been described expressly relative to the particular emergency lighting system shown in the drawings and detailed in the present specification, it is to be understood that the invention can be configured other than expressly illustrated without departing from the scope of the invention, the scope of the invention being dictated by the recitations of the appended claims.

Number	Name	Date	Kind
4167680	Gross	Sep 1979	A
4544910	Hoberman	Oct 1985	A
4799039	Balcom et al.	Jan 1989	A
4945280	Beghelli	Jul 1990	A
5154504	Helal et al.	Oct 1992	A
5574423	Vosika et al.	Nov 1996	A
5646502	Johnson	Jul 1997	A
5666029	McDonald	Sep 1997	A
5734230	Edwards et al.	Mar 1998	A
5739639	Johnson	Apr 1998	A
5811938	Rodriguez	Sep 1998	A
5811975	Bernardo	Sep 1998	A
6069801	Hodge, Jr. et al.	May 2000	A
6072283	Hedrei et al.	Jun 2000	A
6111368	Luchaco	Aug 2000	A
6255804	Herniter et al.	Jul 2001	B1
6259215	Roman	Jul 2001	B1

Self-diagnostic circuitry for emergency lighting fixtures

Information

Patent Number

Date Filed

Date Issued

Inventors

Original Assignees

Examiners

Agents

CPC

US Classifications

Field of Search

US

International Classifications

Abstract

Description

Claims

US Referenced Citations (17)