WO1982002095A1

WO1982002095A1 - Linearization circuit

Info

Publication number: WO1982002095A1
Application number: PCT/US1981/001619
Authority: WO
Inventors: Inc Ametek; Joseph J Durkin
Original assignee: Inc Ametek
Priority date: 1980-12-08
Filing date: 1981-12-04
Publication date: 1982-06-24
Also published as: EP0065572A1; JPS57501891A

Abstract

A linearization circuit (100) receives a non-linear input signal and provides a compensating signal. The amount of non-linear compensation is predetermined to produce a linear output signal. In one particular application, the linearization circuit (100) is used with a pressure transducer (102) to compensate for non-linear output produced by such transducers.

Description

LINEARIZATION CIRCUIT

DESCRIPTION OF THE INVENTION

This invention relates in general to a linearization circuit, and more particularly, to a voltage ramp circuit adapted for linearizing a non-linear voltage ramp input having a positive or negative nonlinearity.

In instrumentation and control systems, it is desirable to provide an electrical output signal, i.e., a voltage ramp, the magnitude of which is proportional to the value of some physical parameter being measured, such as air pressure, temperature, speed, weight, and the like. The electrical output signal can be used to operate a recording device such that a record can be made of the magnitude and variations in the measured parameter; and, as a control signal indicative of the magnitude of the measured parameter for use within a process control system to effect desired changes in the process . The electrical output signal can also be used to directly control measuring instrumentation such that the magnitude of the measured parameter can be readily determined at any given time and variations thereof noted. To this end, various types of instrumentation and control systems have been developed which have included conventional types of transducer elements such as piezoelectric strain elements, linear variable differential transformers, and the like, for measuring the physical parameter. The transducer element typically has a transducer parameter, such as resistance, which changes in magnitude relative to a reference value responsive to variations in the physical parameter being measured. These changes in the transducer parameter can be measured and used to provide the electrical output signal. Generally, in instrumentation and control systems, it is preferred to have these electrical output signals linearally proportional to the physical parameter being measured to ensure accuracy. For example, if a transducer element provides a one-volt electrical output signal for one-unit of the parameter being measured, it should provide a one-half volt electrical output signal for one-half unit of the measured parameter. However, the prior art transducer elements generally produce an electrical output signal that is a non-linear function of the parameter being measured. This non-linearity results in the magnitude of the measured parameter being inaccurate, which has heretofore caused problems in the use of conventional transducer elements in a precision instrumentation and control system.

A variety of approaches have been used in transducer systems to solve the problem of compensating for the non-linear voltage output obtained from these transducer elements. For example: in U.S. Patent No. 4,192,005 a complex memory storage digital processing system has been used for compensating a bridge type piezoresistive pressure transducer; in U.S. Patent No. 3,350,927 a servo-controlled potentiometer has been used to provide a compensating circuit which exhibits a parabolic output to compensate for the normally parabolic response of a pressure transducer; in U.S.

Patent No. 3,358,501 one or more additional semiconductor strain elements arranged in a T-network have been used to compensate for the non-linear operation of a strain transducer; and, in U.S. Patent No. 3,283,923 a pair of varactor diodes have been used to compensate for movement of the core in a linear variable differential transformer. In addition, a diode connected across the output has been used as has the well-known root extractor circuit utilizing an operational amplifier and transistor between the input and output. These prior art approaches used in constructing a transducer system provide a linear voltage ramp output, however, they are relatively complex, requiring numerous active components and increased cost of manufacture. There is thus an unsolved need for providing a linearization circuit to compensate for the non-linearity of the electrical output voltages, and in particular for use with various types of transducer elements to provide precision measuring instrumentation and control systems which are both simple and cost effective.

It is broadly an object of the present invention to provide a circuit to compensate for the non- linear voltage input signals. More specifically, it is an object of this invention to provide such a circuit for use with a transducer element which overcomes or avoids one or more of the foregoing disadvantages resuiting from the use of these prior art approaches. It is a further object of the present invention to provide a linearization circuit which produces a predictable non-linear voltage ramp with the amount of non-linearity being predetermined to produce an electrical output signal linearally proportional to the variable parameter being measured by the transducer element.

Still further, it is an object of the present invention to provide a linearization circuit which provides a nearly perfect linear voltage ramp output in response to a non-linear voltage ramp input.

Another object of the present invention is to provide a linearization circuit which linearizes a voltage input having a negative bow, positive bow or S-shaped bow non-linearity. In accordance with the present invention, there is provided a linearization circuit for producing a predictable non-linear voltage ramp in response to a non-linear input voltage. The circuit includes a divider network and means operatively coupled to at least one element of the divider network selected to provide a non-linear voltage having a predetermined amount of non-linearity for producing a linear output voltage in response to the non-linear input voltage. In one embodiment of the invention, the dividing network is a resistance divider. Coupled across one of the resistors in the divider network is a series connected resistor and diode. The resistance value of the resistor is chosen so that voltage across the diode compensates for the non-linear input voltage thereby producing a linear output voltage.

The above brief description as well as further objects, features, and advantages of the present invention will be more fully understood by reference to the following detailed description of a presently preferred but nonetheless illustrative linearization circuit in accordance with the present invention when taken in conjunction with the accompanying drawings, wherein:

FIG. 1 is a graph showing an ideal linear voltage ramp and non-linear voltage ramps having positive bow, and S-shaped bow non-linearity for which compensation is provided by the linearization circuit of the present invention;

FIGS. 2a and 3 through 6 are circuit diagrams of various embodiments of the linearization circuit according to the present invention which provide compensation for the non-linear voltage ramps shown in FIG. 1;

FIG. 2b is a Thevenin equivalent circuit diagram of the circuit shown in FIG. 2a; and

FIG. 7 is a combined block and circuit dia gram of a transducer system incorporating the linearization circuit of the present invention.

The linearization circuit of the present invention is adapted to generate a predictable non-linear voltage having a predetermined amount of non-linearity to compensate for a non-linear input voltage and thereby providing a linear output voltage. Fig. 1 is a graphical representation showing possible voltages which may be produced, for example, by a pressure transducer measuring applied pressure. The ideal voltage would be linear. However, in practice, the voltage would have a positive bow, a negative bow or be S-shaped. The linearization circuit generates a predictable non-linear input voltage and produces an output which approaches the desired ideal linear voltage.

As shown in each of the embodiments of FIGS. 2a and 3 through 6, the linearization circuit of the present invention includes a resistor voltage divider having a first resistor RA and a second resistor RB. One or both of resistors RA, RB is shunted by the series combination of resistor RS and diode CR1 which produces a non-linear voltage ramp having a predetermined amount of non-linearity. The polarity of the voltage determines the polarity of insertion of the diode CR1. if a positive voltage having a negative bow is to be linearized, resistor RB is shunted by the series combination of resistor RS and diode CR1 as shown in FIG. 2a. If a positive voltage having a positive bow is to be linearized, resistor RA is shunted by the series combination of resistor RS and diode CR1 as shown in FIG. 3. For a negative voltage, negative and positive bow linearization are provided respectively by the circuits of FIGS. 4 and 5. If either a positive or negative voltage ramp having an S-shaped bow is to be linearized, both resistor RA and resistor RB are shunted by a series combination of resistors RS1, RS2 and diodes CR1, CR2 as shown in FIG. 6. The theory of operation of the linearization circuit in accordance with the present invention can be explained by reference to FIG. 2a. The input voltage V_in is typically provided from a transducer element and is a non-linear voltage proportional to the measured parameter normalized over a desired full scale range, for example, 0 to 2.0 volts. An increasing input voltage from the transducer element produces a current through resistor RS, which current, I_D also flows through diode CR1. The diode current I_D is directly proportional to the diode voltage V_D which can be operatively controlled to be non-linear when by maintaining the diode voltage between the turn-on voltage and the saturation voltage of the diode. For example, where a conventional silicon diode is used, this non-linear operating range of the diode will be generally about from 0.2 volts to 0.7 volts. By selecting the proper value for resistor RS in the series/parallel combination of RA, RB and RS, the diode voltage V_D can be determined which generates a non-linear voltage, equal in magnitude and opposite in polarity to the applied input voltage, thus producing an linear voltage output, V_out. The voltage drop across resistor RS and diode

CR1 is determined by the value of resistor RA and resistor RB in the voltage divider circuit. The values of resistor RA and resistor RB are selected to cause the voltage drop across resistor RS and diode CR1 to be generally within the non-linear operating range of diode CR1. Typically, the value of RA is selected to be equal to the value of RB. If the voltage drop across resistor RS and diode CR1 is greater than the saturation voltage of diode CR1, the diode Voltage V_D will be linear and the circuit will function as a conventional linear voltage divider. If the voltage drop across resistor RS and diode CR1 is less than the turn-on voltage of diode CR1, the circuit will also generate a linear voltage. Typically, the non-linearity of the voltage generated by a transducer element is symmetrical about an ideal linear voltage. The maximum error of the non-linear voltage from the ideal voltage will therefore occur at 50% of the normalized full scale. By selecting the value of resistor RA and resistor RB such triat the voltage drop across resistor RS and diode CR1 is within the operating range of the diode, the maximum error from the ideal voltage ramp will occur at about 50% of full scale. However, if the non-linearity of the input voltage is not symmetrical about an ideal linear voltage, the maximum error will occur at other than 50% of full scale. In linearizing a non-symmetrical voltage ramp, the voltage drop across resistor RS and diode CR1 is provided partially within the linear operating range of the diode causing the maximum error to be shifted either upward or downward an appropriate amount. If the voltage drop across resistor RS and diode CR1 is less than the turn-on voltage of the diode, the maximum error will be shifted downward from 50% of full scale and if the voltage drop is greater than the diode saturation voltage, the maximum error will be shifted upward from

50% of full scale.

Referring to FIG. 2a, an example incorporating the principles of the present invention will now be described for determining the correct value of resistor

RS for a negative bow correction where V_in = 0 to + V.

The Thevenin equivalent circuit of FIG. 2a is shown in

FIG. 2b where:

RT = RAxRB RA+RB

and, the current Io through diode CR1 is

where I_O = reverse saturation current k = Boltzman's constant q = electronic change T = ambient temperature V_D = diode output voltage.

From FIG. 2b, the Thevenin equivalent input voltage V_T is

and the input voltage as a non-linear function of (V_in) is

or

The output voltage V_out from FIG. 2b is

The proper value for resistor RS can be determined by solving Equations 1 and 2 such that resistor RS in combination with diode CR1 will have a voltage dependent parallel shunting effect on resistor RB to cause the linearization circuit to generate the desired voltage having a predetermined amount of non-linearity.

The equations for the output voltage V_out as a function and the input voltage V_in for the linearization circuits of FIGS. 3-6 can be derived in a similar manner and would be known to those having ordinary skill in the art by the application of the principles of the present invention disclosed with reference to FIGS. 2a and 2b. In general, the appropriate value for RS can be determined as follows. The normalized uncompensated voltage output from the transducer element is determined at selected percentages of full scale, such as at 0%, 25%, 50%, 75% and 100%. These normalized uncompensated voltages, V_in are entered into Equation 1 in addition to values for q, k, T, and resistor RA, resistor RB, I_O of the specific diode used, and an initial estimated value for resistor RS. Equation 1 is solved for the diode voltage V_D for each of the normalized uncompensated voltage readings. The corresponding diode voltages V_D are entered into Equation 2 to solve for the output voltage V_out of the linearization circuit. For each value of V_out determined by Equation 2, the percentage error is calculated and the resulting maximum full scale error determined. Equations 1 and 2 are again solved using a different estimated value for resistor RS until the resulting maximum full scale error is determined to be a minimum value. The value for resistor RS which generates a minimum full scale error is used in the linearization circuit in combination with the selected diode CR1. These calculations can best be performed with the aid of a programmed computer. The actual program for the computer would be well worth the skill of the art and need not be discussed further.

Resistor RA and resistor RB reduce the gain of the normalized uncompensated voltage according to the ratio of their magnitude and resistor RS reduces the full scale output voltage of the linearization circuit by the compensated percentage error at full scale. The lost gain and full scale output voltage can be recovered by providing an additional amplifier coupled to the output of the linearization circuit. The linearization circuit of the present invention can generate a linearized output voltage having precision of about 1% of the uncompensated error of the non-linear voltage input.

Referring to FIG. 6, the linearization circuit which compensates for an S-shaped bow non-linearity includes resistor RS1 and resistor RS2. The resistance values for resistors RS1 and RS2 are determined in the general manner described for the circuit of FIG. 2a. In particular, the non-linear voltage input to the circuit is normalized from -1 volt to +1 volt, i.e., 2.00 volts full scale. The value for resistor RA is selected to equal the value of resistor RB causing a one-volt drop across resistor RS1 and diode CR1 and resistor RS2 and diode CR2. Diode CR1 and diode CR2 will operate in the non-linear range. The value for resistor RS1 is determined which causes the maximum error over the normalized scale of 0 volts to +1 volt to be a minimum. Likewise, the value of resistor RS2 is determined which causes the maximum error over the normalized scale of -1 volt to 0 volts to be a minimum.

A typical application for the linearization circuit to linearize the output from a transducer is shown in FIG. 7. The transducer system 100 includes a transducer element 102 of the type described above for providing a voltage output proportional to the parameter being measured. The output of the transducer element 102 is connected to the input of first amplifier 104. The output from amplifier 104 is coupled to the input of a second amplifier 106 through a linearization circuit according to the present invention. If the non-linear voltage ramp produced by transducer element 102 has a negative bow non-linearity, resistor RA is shunted by the series combination of resistor RS and diode CR1 and resistor RS' and diode CR1' are eliminated. Conversely, if transducer element 102 produces a positive bow non-linearity, resistor RB is shunted by the series combination of resistor RS' and diode CR1' and resistor RS and diode CR1 are eliminated. The diodes CR1 and CR1' are provided as the PN injunction of a standard semiconductor transistor. The output from the second amplifier 106 is connected to a gauge indicator 108 which indicates the measured parameter.

In operation, in response to the variable parameter being measured, transducer element 102 provides a non-linear voltage output which is normalized from 0 to 2 volts full scale by amplifier 104. The normalized uncompensated output voltage from amplifier 104 is linearized by the linearization circuit to provide a linear voltage output in the manner described above. The value of resistor RS or resistor RS' is determined in the manner described in the above illustrative examples. The linearized voltage output is applied to amplifier 106. The precise magnitude of the measured parameter is recorded by guage indicator 108 which may be included in an instrumentation or control system. The invention herein has been described with reference to particular embodiments, and it is to be understood that these embodiments are merely illustrative of the principles and applications of this invention. Thus, it is to be understood that numerous modifications may be devised without departing from the spirit and scope of this invention.

Claims

WHAT IS CLAIMED IS:

1. A circuit for producing a linear output voltage in response to a non-linear input voltage comprising, a divider network coupled to receive said input voltage, and compensating means operatively coupled to at least one element of said divider network selected to provide a non-linear voltage having a predetermined characteristic to compensate for the nonlinear input voltage such that the air circuit produces a substantial linear output voltage.

2. The circuit of Claim 1 wherein said compensating means provides a positive shaped bow having a predetermined amount of non-linearity.

3. The circuit of Claim 1 wherein said compensating means provides a negative shaped bow having a predetermined amount of non-linearity.

4. The circuit of Claim 1 wherein said compensating means provides an S-shaped bow having a predetermined amount of non-linearity.

5. A circuit for producing a linear output voltage in response to a non-linear input voltage comprising, a divider network coupled to receive an input voltage and a series connected resistor and diode operatively coupled to said divider network for providing a voltage having a predetermined non-linear characteristic to compensate for the non-linearity of said input voltage such that the output of said circuit is a substantially linear voltage.

6. A circuit for producing a linear voltage output in response to a non-linear voltage input comprising, first and second resistive elements constructed and arranged in a divider network and coupled to receive said non-linear voltage input, and a compensating network operatively coupled across at least one of said first and second resistive elements for providing a voltage having a predetermined amount of non-linearity to compensate for the non-linearity of said voltage input such that said circuit produces a substantial linear voltage output.

7. The circuit of Claim 6 wherein said compensating network includes a series connected diode and resistor.

8. The circuit as set forth in Claim 7 wherein said compensating network is connected in parallel across either of said first and second resistive elements.

9. The circuit of Claim 7 further including a second compensating network wherein the first and second compensating networks are connected in parallel respectively with said first and second resistive elements.

10. A transducer system comprising, means for providing a non-linear voltage input signal representative of a variable parameter being measured by said transducer system, means coupled to receive said voltage input signal and for linearizing said voltage input signal including a divider network, and a compensating network operatively coupled to said divider network for providing a voltage having a predetermined amount of non-linearity to compensate for the non- linearity of said voltage input signal such that the output voltage is a linear function related to said variable parameter, and means coupled to said linearization means and responsive to said output voltage for indicating the magnitude of said variable parameter measured by said transducer system.

11. The transducer system of Claim 10 wherein said divider network includes a pair of resistor elements and said compensating means includes a series combination of a resistor element and a diode element connected in parallel with at least one of said resistor elements.