Patent Application
20160216292
The present invention is directed to an improved design of a hot-wire or hot-film anemometer.
By way of background, anemometers measure the flow of fluids and take many forms. For macro measurements such as wind velocity, wind turbine and wind vanes are common. For measurements requiring a small sensor for profiling and other applications, a popular device is the hot wire or hot film anemometer. This configuration exploits the resistive temperature coefficient of materials by fabricating them into thin film or wire sensors. Additionally, these sensors may be oriented 90 degrees to each other to provide three orthogonal components of the fluid velocity vector. In practice, the sensor is heated to a temperature above ambient. As the sensing element is cooled by fluid flow, energy is carried away, lowering the temperature and causing a change in resistance. This change in resistance is quite small and therefore significant amplification is required to develop a useful signal for measurement.
There are three common methods of heating the wire and measuring its resistance: the constant voltage, the constant current, and the constant temperature method. All three of these methods use a feedback control system to maintain a parameter that is kept constant. All methods involve sensing the voltage across the heated element and use the sensed voltage for controlling the loop or making measurements. However, existing constant temperature hot wire/film anemometer designs are plagued by three particular issues, namely:
Each of these problems will be discussed in detail herein below.
Issue #1, adjustment of system damping is an elaborate and essentially unnecessary process to minimize, but not eliminate, system instabilities. The hot wire or film to be sensed is part of a probe used for measuring fluid flow at a specific location and orientation. Because the point of sensing is remote from the point of data acquisition, connections between sensors and signal processing hardware exists, facilitated with cabling.
Some form of shielded wire is used to prevent noise pickup in the interconnection between sensors and signal processing electronics. Coaxial cable and twisted pairs are used at audio frequencies in many home and commercial installations. At higher frequencies, shielded wires exhibit transmission line characteristics such as the reflection of energy if the wires are not properly fed and terminated.
The bandwidth of many hot wire/film anemometers is similar to the audio frequency range of 20 Hz to 20 kHz and transmission line effects are not considered. However, the control loops that are used to create the constant current, voltage, or temperature have bandwidths well exceeding the audio frequency range. The wires between the sensor and circuits therefore constitute a transmission line and must be treated as such. An improperly fed or terminated transmission line is capable of exhibiting time delay effects and can present a complex impedance; i.e. having a reactive component. The confluence of sensor, cable, and circuit can then satisfy the Barkhausen Criteria, pushing the circuitry into oscillation. This is the reason for the complicated process to adjust the damping, partially stabilizing the system. However, if the sensor or cable is exchanged, the frequency of oscillation shifts, and the tuning process must be repeated.
A control loop with a complex impedance present at any point is subject to instabilities. While hot wire or film anemometer systems have employed additional components and lengthy procedures to mitigate ringing of an underdamped system as well as other effects, these attempted solutions have not eliminated such problems.
Issue #2, the cumbersome process for setting the sensor drive, exists because of the need to adjust the drive based upon the cold resistance of the individual sensor. Typically, the drive resistance is set to approximately 1.8 times the cold resistance of the sensor. When the unit is energized, the bridge excitation increases to heat the sensor to raise its resistance to match the drive resistance. Existing designs require measuring the sensor cold resistance, performing the necessary mathematics, and then adjusting the drive setting resistance to the necessary value to achieve the desired drive. This is typically done either by disconnecting and reconnecting components or, by using expensive resistor switching units with numerical displays that must be set accordingly.
Issue #3, the too-frequent destruction of the sensor element from circuit transients, results from the propensity of existing hot-wire anemometer designs to produce and transmit large and fast current spikes to the connected sensor. This occurs when the sensor parameters are changed too quickly or functions are performed ‘out of sequence’.
The present invention provides an anemometer design that encompasses three new circuit topologies addressing these and other problems in the art.
Those skilled in the art will appreciate the above stated advantages and other advantages and benefits of various additional embodiments upon reading the following detailed description of the embodiments with reference to the below-listed drawing figure.
According to common practice, the various features of the drawing discussed below are not necessarily drawn to scale. Dimensions of various features and elements in the drawings may be expanded or reduced to more clearly illustrate the embodiments of the disclosure.
The present invention eliminates the instabilities for any configuration of sensor and transmission line by considering the physics of transmission lines and insuring proper termination to prevent reflections and complex impedances. Referring to the block diagram of
In the calibrate mode of
The switching transistor Q3 shown at 45 and 145, the slewing network shown at 50 and 150, and the switching element shown at 60 and 160 as depicted in
The sensor element is connected to the control system by use of a controlled characteristic impedance transmission line, which could be a twisted pair, shielded or unshielded, or a coaxial cable. This system can be adapted to any type of controlled impedance cable including Twinax, Triax, double shielded, etc. These cables are common and used for high speed data transmission and radio communication—applications that involve high frequencies and are subject to distortion from reflections. However, as shown in
The resistance of the hot wire or film sensor 30, 130 will not be close to the characteristic impedance of practical transmission lines and therefore will not present a matched source impedance. The addition of resistance is unattractive because of the resulting signal loss and, because sensor impedance is dynamic and therefore not amenable to the selection of a specific component. This system accepts that the source (i.e. the hot wire or film sensor 30, 130) is not impedance matched to the transmission line 20, 120 and an alternate method of eliminating the effects of reflections is used.
The operate mode shown in the block diagram of
Connected across the Wheatstone bridge 175 is termination resistor R9 which is equal to the characteristic impedance of the transmission line 20, 120. Because the transmission line 20, 120 is terminated with its characteristic impedance at the amplifier 195, reflections from any source can make only one pass through the transmission line 20, 120 before being completely absorbed by the termination resistor R9. This termination resistor R9 does not cause significant signal reduction because of the high loop gain. The transmission line 20, 120 is connected to the summing junction N1 of the feedback loop, which is known as a virtual ground, i.e. both inputs of the differential amplifier 195 are practically at the same potential. At higher frequencies (i.e. those above the bandwidth of the control loop) this node ceases to function as virtual ground as the control loop runs out of gain. With the diminution of the virtual ground, the termination resistor R6 presents a limit to the amplifier impedance and thereby inhibits the circuit from going into oscillation. Additionally, as shown in the Operate Mode of
In
The Whetstone Bridge 175 includes first, second, third and fourth nodes N1, N2, N3, N4, respectively. The ends of first and second bridge resistors R3, R4 meet at the first node N1. The termination resistor R9 is connected between nodes N2, N3 and therefore is connected across first and second bridge resistors R3, R4. In the operate subcircuit, the drive potentiometer 140 is connected between the third node N3 and the fourth (ground) node N4. At third node N3, the drive potentiometer presents a drive resistance signal POT. Meanwhile, the sensor 30, 130 output is connected to node N2 via the transmission line 20, 120, where is presents the sensor output signal SNSR. Excitation for the Whetstone Bridge 175 is provided by the feedback (power) amplifier 185 which is implemented as a complementary pair of power transistors Q1, Q2, the latter of whose base is controlled by a feedback signal 187.
Adjusting the drive potentiometer 140 to the typical value of 1.8× the cold resistance of the sensor 30, 130 produces a differential voltage at the bridge output across termination resistor R9. This voltage is presented to the differential inputs of instrumentation amplifier U1 (which serves as the differential amplifier 195) via first amplifier input circuit 177. In the shown embodiment, the instrumentation amplifier U1 is implemented as an INA103KP. The differential voltage is amplified by the instrumentation amplifier U1 to produce first amplified output signal 197. In the shown embodiment, the first amplified output signal 197 is input to a second amplifier U2B, where it is further amplified. Second amplifier U2B may constitute one of the four op amps of a LM49740 quad amplifier. The output of second amplifier U2B is a second amplified output signal 199. Second amplified outlet signal 199 is used as the feedback signal 187 applied to the base of transistor Q2 of the feedback amplifier 185.
As long as the resistance of the drive potentiometer 140 is higher than the resistance of the sensor 30, 130, the feedback signal 187 will become more positive. This has the effect of increasing the drive to the bridge 175, raising the temperature of the sensor 30, 130. As the temperature of the sensor 30, 130 increases, its resistance will increase until it matches that of the drive potentiometer 140. In this manner, the feedback loop will work to maintain the sensor 30, 130 at a constant temperature. Air moving past the sensor 30, 130, in carrying away heat, is met with a change in potential at the output of amplifier U2B which closely tracks the variations of the air movement. This is the basis of anemometer operation.
The termination resistor R9, by virtue of being connected across the input terminals of instrumentation amplifier U1, is capable of absorbing reflections that develop on the transmission line 20, 120 connecting the sensor to the instrument. Such reflections develop due to impedance mismatch between the sensor 20, 120, transmission line 30, 130, and the instrument amplifier U1 under the conditions of a high-bandwidth feedback loop. In contrast to prior art designs which attempt to manage the instability, the present design removes the source of the instability. However, it should be noted that instrumentation amplifier, U1, cannot respond unless there is some excitation on the bridge 175 to produce a voltage differential in the first instance.
The first amplified output 197 of instrumentation amplifier U1 is just one input to the second amplifier U2B. The other input to the second amplifier U2B is provided by a reference voltage from at fifth node N5, which belongs to a voltage divider 275. The reference voltage at fifth node N5 is amplified by U2B to provide the minimal excitation needed to set the process in motion.
The second amplified output 199 from the second amplifier U2B constitutes a varying amplified anemometer signal which is subjected to additional conditioning and amplification before it matures into the anemometer output signal 200.
The second amplified output 199 from the second amplifier U2B is applied as a first input to a third, inverting amplifier U2A circuit 235. A DC offset is applied to the second input of the third amplifier U2A. More particularly, at the third amplifier U2A, the DC offset is added to the second amplified output 199. As shown in
The output of the third, inverting amplifier U2A may be subjected to additional adjustable gain, as needed. The additional gain may be provided by adjustable gain amplifier 265 which in the shown embodiment comprises fourth and fifth amplifiers U2C and U2D, also belonging to the LME49740 quad op amp. Gain may be adjusted by a gain control circuit 255 comprising a potentiometer connected via connector J1. The adjustable gain amplifier 265 outputs the anemometer output signal 200 which may then be used for monitoring, display, control, and other uses.
Slewing network 150 is responsive to signal CAL, which is active when the switch 51 is moved from the operate mode of
Slewing network 150 includes a voltage divider comprising resistors R21, R22 and a capacitor C13. When S1 is moved to the calibrate position, the supply voltage VCC is applied to the gate of Q3 in
If switch S1 is in the operate position seen in
In the calibrate mode, the sensor 30 and the drive potentiometer are each connected to precise, stable current sources which in the shown embodiment are implemented by a REF200 dual current source 80. Dual current source 80 has low-side constant current terminals 82, 84 which are connected to the sensor 30 and the drive potentiometer 40, respectively. Also, sensor 30 is connected via line 82a to the primary voltage input of sensor (first) gain amplifier U4A 92 while the drive potentiometer 40 is connected via line 84a to the primary voltage input of the potentiometer (second) gain amplifier U4B 90.
In the embodiment shown, the sensor and potentiometer gain amplifiers U4A, U4B are implemented using an AD8622 dual op amp, with clamping capacitors C18, C16 helping stabilize the supply voltage VCC 96. Anti-biasing resistors R31, R34 are connected to ground at the primary voltage inputs of U4A, U4B.
Potentiometer gain amplifier U4B 90, which outputs an amplified potentiometer voltage 91P, is provided with a second feedback network 90a. Second feedback network 90a includes capacitor C15 connected in parallel with permanent feedback resistor R27. The inverting input of U4B is also connected to ground via resistor R29. Potentiometer gain amplifier U4B 90 and its second feedback network 90a are configured to amplify the voltage developed across the drive potentiometer 140 by a factor of about 100×. The sensor gain amplifier U4A 92, which outputs an amplified sensor voltage 91S, is provided with a first feedback network 92a. First feedback network 92a includes capacitor C17 connected in parallel with permanent feedback resistor R33. The inverting input of U4A is connected to ground via resistor R32. However, the first feedback network 92a also includes a selectively actuated auxiliary feedback resistor R159 controlled by selector switch S6. Senor gain amplifier U4A 92 and its second feedback network 92a are thus configured to amplify the voltage developed across the sensor 30 by a factor that depends on the position of selector switch S6, and output the amplified sensor voltage 91S.
When selector switch S6 is open, as seen in
On the other hand, when selector switch S6 is closed, the auxiliary feedback resistor R159 is included in the first feedback network, in parallel with permanent feedback resistor R33, thus dropping the effective resistance of the first feedback network 92a. With the sensor-type switch S6 closed, amplifier U4A 92 and its first feedback network 92a are configured to amplify the voltage developed across the sensor 30 by a factor of about 120×, and so the unit is configured to employ a hot film anemometer sensor.
Thus, if selector switch S6 is open (i.e., a hot wire anemometer is employed as the sensor 30, in order to get the voltage outputs of gain amplifiers U4A and U4B to match, the resistance of the drive potentiometer 140 must equal about 1.8× that of the cold sensor 30. On the other hand, if the selector switch S6 is closed (i.e., a hot film anemometer is employed as the sensor 30), in order to get the voltage outputs of gain amplifiers U4A and U4B to match, the resistance of the drive potentiometer 140 must equal about 1.2× that of the cold sensor 30.
To know when a specific ratio of the sensor and potentiometer resistances has been achieved, the amplified potentiometer voltage 91P and the amplified sensor voltage 91S (i.e., the outputs of gain amplifiers U4A and U4B) are applied to an equal-voltage detector circuit 94.
The amplified voltages 91P, 91S are first subject to a differential resistor network comprising sensor line resistors R25, R30 and differential resistor R28 connected across the line resistors R25, R39. This network simply reduces the sensitivity of the equal voltage detector, making adjustment easier. Differences in the amplified voltages 91P, 91S are reflected by the voltage across resistor R28, thereby producing comparator drive voltage input 93P and comparator sensor voltage input 93S.
The equal-voltage detector circuit 94 utilizes dual comparators U5A, U5B with open collector outputs connected to sixth node N6, located between resistor R26 and LED1 98. In the embodiment of
The present design replaces the entire measurement/calculation task with the simple push of a button and adjustment of the drive resistance. The operator pushes the ‘Cal/Operate’ switch and adjusts the drive resistance until the ‘Cal’ LED illuminates—after which the operator will push the Cal/Operate button again. Referencing
The present design incorporates power slewing circuitry that mitigates most occurrences of rapid spiking. Referring to
Power spikes are thus ignored—sidestepping sensor shock and thereby reducing sensor mortality. The specific components and their interconnections do not themselves constitute the advancement, but rather the implementation of strategic time constants that function to suppress transient events to the sensor without compromising the acquisition of data.
The foregoing description of the disclosure illustrates and describes various embodiments of the present invention. As various changes could be made in the above construction without departing from the scope of the disclosure, it is intended that all matter contained in the above description or shown in the accompanying drawings shall be interpreted as illustrative and not in a limiting sense. Furthermore, the scope of the present disclosure covers various modifications, combinations, alterations, etc., of the above-described embodiments of the present invention that are within the scope of the claims.
Additionally, while the disclosure shows and describes only selected embodiments of the present invention, it will be understood that the present invention further is capable of use in various other combinations and environments, and is capable of various changes or modifications within the scope of the inventive concept as expressed herein, commensurate with the above teachings, and/or within the skill or knowledge of the relevant art. Furthermore, certain features and characteristics of each embodiment may be selectively interchanged and applied to other illustrated and non-illustrated embodiments of the disclosure.
The present Patent Application is a formalization of previously filed, co-pending U.S. Provisional Patent Application Ser. No. 62/106,981, filed Jan. 23, 2015 by the inventors named in the present Application. This Patent Application claims the benefit of the filing date of this cited Provisional Patent Application according to the statutes and rules governing provisional patent applications, particularly 35 U.S.C. §119(e), and 37 C.F.R. §§1.78(a)(3) and 1.78(a)(4). The specification and drawings of the Provisional Patent Application referenced above are specifically incorporated herein by reference as if set forth in their entirety.
| Number | Date | Country | |
|---|---|---|---|
| 62106981 | Jan 2015 | US |
