Device with automatic compensation of an ac attenuator

• US 4,663,586 A
• Filed: 05/02/1985
• Issued: 05/05/1987
• Est. Priority Date: 05/02/1985
• Status: Expired due to Term

• Current Assignee: Agilent Technologies, Inc.
• Original Assignee: Hewlett-Packard Company (HP Inc.)
• Inventors: DesJardin, Lawrence A., Stever, Scott, Swerlein, Ronald L.
• Primary Examiner(s): Eisenzopf, Reinhard J.
• Assistant Examiner(s): Baker, Stephen M.
• Application Number:
  US06/729,507
• Time in Patent Office: 733 Days
• Field of Search: 324/115, 324/130, 324/102, 324/126, 324/57 R, 324/57 PS, 364/483, 364/571, 364/579, 364/580, 328/162
• US Class Current: 324/115
• CPC Class Codes: G01R 27/02 Measuring real or complex resistance, reactance, impedance, or other two-pole characteristics derived therefrom, e.g. time constant by measuring phase angle only G01R25/00
• CPC Class Codes: G01R 35/02 of auxiliary devices, e.g. of instrument transformers according to prescribed transformation ratio, phase angle, or wattage rating
• CPC Class Codes: H03H 11/24 Frequency-independent attenuators

Abstract

An ac measuring instrument automatically compensates an input attenuator to achieve precision frequency response. The output of the input attenuator is applied to both the conventional measurement portion of the instrument and to an amplifier of variable gain. The output of the variable gain amplifier drives either a resistive or capacitive impedance coupled to the output of the input attenuator. The impedance is thus driven by a voltage that is a scaled replica of the output of the input attenuator. That scale factor may vary from near zero to greater than one, and the impedance acts as a virtual trimmer whose apparent value varies as the scale factor of its actual value according to the gain established for the variable gain amplifier. The proper gain therefor is determined by a servo loop that applies a test pulse to the input attenuator and subsequently samples the output thereof at least twice. An amplitude difference in the samples indicates an exponential component that arise from an uncompensated attenuator. The servo loop is responsive to the signed difference between the samples to adjust the value of the virtual trimmer until the signed difference is sufficiently near zero. Proper choice of the time interval between samples will adjust the servo loop to be most sensitive to exponential components of certain time constants rather than others. This allows the servo loop to ignore the effects of dielectric absorbtion while compensating the attenuator.

Claims

1. Voltage measure-apparatus having an attenuator automatically compensated for frequency response comprising:

a first attenuator having an input and an output referenced to a signal ground, the first attenduator itself comprising;

an input section connected between the input and the output of the first attenuator;

(ii) a grounded section connected between the output of the first attenuator and the signal ground; and

(iii) a virtual trimmer impedance connected between the output of the first attenuator and the signal output of the variable gain amplifier means recited below;

means coupled to the output of the first attenuator for producing a measurement of the value of an AC voltage applied to the input of the first attenuator;

variable gain amplification means having a signal input coupled to the outptu of the first attenuator, having a gain control input coupled to the processing means recited below and also having a signal output;

means for applying to the input of the first attenuator a step voltage having a transition in the direction of a selected polarity;

sampling means coupled to the output of the first attenuator for making a plurality of samples subsequent to the transition in and during the ensuing duration of the step voltage applied to the first attenuator and

processing means, having an input coupled to receive the samples made by the sampling means and having an output coupled to the gain control input of the variable gain amplification menas, for varying the effective value of the virtual trimmer impedance by specifying the variable gain in accordance with the values of the individual samples within the plurality thereof, thereby compensating the frequency response of the first attenuator.

2. Apparatus as in claim 1 wherein the virtual trimmer impedance comprises a capacitance, the variable gain amplification means is non-inverting and the processing mea decreases the gain of the variable gain amplification means whenever the different obtained by diminishing the value of a preceding individual sample by the value of a subsequent individual sample is of the the same polarity as the transition in the step voltage and increases the gain whenever the difference is of the opposite polarity.

3. Apparatus as in claim 1 wherein the virtual trimmer impedance comprises a capacitance, the variable gain amplification means is inverting and the processing means increases the gain of the variable gain amplification means whenever the difference obtained by diminishing the value of a preceding individual sample by the value of a subsequent individual sample is of the the same polarity as the transition in the step voltage and decreases the gain whenever the difference is of the opposite polarity.

4. Apparatus as in claim 1 wherein the virtual trimmer impedance comprises a resistance, the variable gain amplification means is non-inverting and the processing means increases the gain of the variable gain amplification means whenever the difference obtained by diminishing the value of a preceding individual sample by the value of a subsequent individual sample is of the the same polarity as the transition in the step voltage and decreases the gain whenever the difference is of the opposite polarity.

5. Apparatus as in claim 1 wherein the virtual trimmer impedance comprises a resistance, the variable gain amplification means is inverting and the processing means decreases the gain of the variable gain amplification means whenever the difference obtained by diminishing the value of a preceding individual sample by the value of a subsequent individual sample is of the the same polarity as the transition in the step voltage and increases the gain whenever the difference is of the opposite polarity.

6. Apparatus as in claim 1 wherein the processing means sustains the existing gain of the variable gain amplification means whenever the absolute value of the difference between the value of a preceding individual sample and the value of a subsequent individual sample is less than a selected value.

7. Apparatus as in claim 1 wherein the variable gain amplification means comprises a first amplifier whose output is coupled to the input of a programmable attenuator whose output is coupled to the input of a second amplifier, the signal input to the variable gain amplification means being the input to the first amplifier, the signal output of the variable gain amplification means being the output of the second amplifier and the gain control input to the variable gain amplification means being a programming input of the programmable attenuator.

8. Apparatus as in claim 1 wherein the variable gain amplification means comprises a first amplifier having an output coupled to a first input of an analog multiplier, a digital to analog converter having a signal input operated at a reference voltage and a signal output coupled to a second input of the analog multiplier, the signal input to the variable gain amplification means being the input to the first amplifier, the signal output of the variable gain amplification means being an output of the analog multiplier, and the gain control input to the variable gain amplification means being a programming input of the digital to analog converter.

9. Apparatus as in claim 1 wherein a preceding individual sample and a subsequent individual sample within the plurality thereof each occur after the transition in the step voltage by respective amounts of time selected to maximize the degree to which the difference between the values of those samples is indicative of mis-compensation of the first attenuator.

10. Apparatus as in claim 1 wherein a preceding individual sample and a subsequent individual sample within the plurality thereof each occur after the transition in the step voltage by respective amounts of time selected to maximize the degree to which the difference between the values of those samples represents an exponential voltage waveform having a time constant equal to that of the first attenuator.

11. A method of compensating the frequency response of an attenuator comprising the steps of:

first,

a. applying an abrupt step in voltage to the input of an attenuator to produce an attenuated voltage step;

b. sampling, subsequent to step a, the attenuated voltage step a plurality of times;

c. determining for the plurality of samples a difference therebetween;

d. amplifying the attenuated step by a gain selected in accordance with the determined difference;

e. coupling the amplified attenuated voltage step through an impedance to the output of the attenuator;

f. if the different of step c exceeds a selected limit then changing the selected gain used in step d to produce a smaller difference in step c; and

g. repeating steps a-f until the difference of step c no longer exceeds the selected limit in step f;

then,

h. applying a work voltage to be measured to the input of the attenuator;

i. amplifying the attenuated work voltage by the changed selected gain determined in step f; and

j. coupling the amplified attenuated work voltage through the impedance to the output of the attenuator.

12. A method as in claim 11 wherein:

the impedance coupled in steps e and j is a capacitance;

the amplifying steps d and i are non-inverting;

the determining step c further determines whether a preceeding sample of the plurality in step b diminished by a subsequent sample of that plurality produces a difference of the same polarity as the abrupt step in voltage of step a, or if that difference is of the opposite polarity; and

the changing step f decreases the gain used in steps d and i whenever the determining step c produces a difference of the same polarity and increases the gain used in steps d and i whenever the determining step c produces a difference of the opposite polarity.

13. A method as in claim 11 wherein:

the impedance coupled in steps e and j is a capacitance;

the amplifying steps d and i are inverting;

the determining step c further determines whether a preceeding sample of the plurality in step b diminished by a subsequent sample of that plurality produces a difference of the same polarity as the aburpt step in voltage of step a, or if that difference is of the opposite polarity; and

the changing step f increases the gain used in steps d and i whenever the determining step c produces a difference of the same polarity and decreases the gain used in steps d and i whenever the determining step c produces a difference of the opposite polarity

14. a method as in claim 11 wherein:

the impedance coupled in steps e and j is a resistance;

the amplifying steps d and i are non-inverting;

the determining step c further determines whether a preceeding sample of the plurality in step b diminished by a subsequent sample of the plurality produces a difference of the same polarity as the abrupt step in voltage of step a, or if that difference is of the opposite polarity; and

the changing step f increases the gain used in steps d and i whenever the determining step c produces a difference of the same polarity and decreaes the gain used in steps d and i whenever the determining step c produces a difference of the opposite polarity

15. A method as in claim 11 wherein

the impedance coupled in steps e and j is a resistance;

the amplifying steps d and i are inverting;

the determining step c further determines whether a preceeding sample of the plurality of step b diminished by a subsequent sample of that plurality produces a difference of the same polarity as the abrupt step in voltage of step a, or if that difference is of the opposite polarity; and

the changing step f decreases the gain used in steps d and i whenever the determining step c produces a difference of the same polarity and increases the gain used in steps d and i whenever the determining step c produces a difference of the opposite polarity.

16. A methodas in claim 11 wherein the sampling step b samples the attenuated voltage step at a plurality of times selected to maximize the degree to which the difference of the determining step c is indicative of miscompensation of the attenuator.

17. A method as in claim 11 wherein the sampling step b samples the attenuated voltage step at a plurality of times selected to maximize the degree to which the difference of the determining step c represents an exponential voltage waveform having a time constant equal to that of the attenuator.

18. A method of compensating the frequency response of an attenuator comprising the steps of:

creating a scaled replica of the voltage appearing at the output of the attenuator;

driving the output of the attenuator with the scaled replica through a virtual trimmer impedance; and

varying the effective value of the virtual trimmer impedance be changing the scale of the replica.

19. A method of compensating he frequency response of an attenuator comprising the steps of:

creating a scaled replica of the AC components above a selected frequency in the voltage appearing at the output of the attenuator;

driving the output of the attenuator with the scaled replica through a virtual trimmer capacitance; and

varying the effective value of the virtual trimmer capacitance by changing the scale of the replica.

20. A method of compensating the frequency response of an attenuator comprising the steps of:

creating a scaled replica of the DC and AC components below a selected requency in the voltage appearing at the output of the attenuator;

driving the output of the attenuator with the scaled replica through a virtual trimmer resistance; and

varying the effective value of the virtual trimmer resistance by changing the scale of the replica.