An Inexpensive, Very High Impedance Digital Voltmeter

Marco S. Caceci Florida State University, Tallahassee, FL 32306

The impedance of most commercial glass electrodes is in the vicinity of 108 f2, while ion selective electrodes reach im- oedances of lo9 Q and more. A voltmeter. in order to measure the difference in potential between a sensing and areference electrode. must have a much hieher input im~edance than the electrode with the highest imgedanee (i.e.: >> 109 Q). The reason for this is seen in Figure 1 where (assuming purely re- sistive impedances)

E. = E,(1+ RJR,)

E. being the actual (unknown) electrode potential; Re, the electrode resistance: E,. the ootential read a t the voltmeter: K , the resistanceof'the'voltmeter.

The actual ootential is hieher than the measured one bv a factor equal td the ratio betGeen the impedance of the volt&e source and that of the measuring instrument.

In practice, voltmeters do not behave like resistors, as far as their impedance is concerned, but rather drain small cur- rents (inputs bias currents), which develop across the elec- trodes a potential proportional to their resistance. Typical high quality pH meters are rated with an input offset current of less than 10-l2 A. I t can he easily calculated that such a figure represents an error of up to 1% when measuring 100 mV across a tvpical membrane electrode of 109 Q resistance.

The con;truction of inexpensive digital pH meters was re- cently illustrated in detail in THIS JOIIRNAI..' This paper describes an instrument which exceeds, both in accuracy A d input impedance, the meters described in note 2 as well as mbst corimercial pH meters and potentiometers at a cost of about $150 (1981). This simple compact digital voltmeter has 0.1 mVresolution on a f 2 V scale, and an input offset current of less than 10-'4A. We applied it successfully in the course of a study involving the development of a uranium-selective electrode of the coated wire type? where the very high im- pedance of the electrodes caused inaccurate reading on com- mercial pH meters.

The instrument (Fig. 2), consists essentially of only two parts: a very high impedance hybrid operational amplifier (ICH8500/A, Intersil) used as a voltage follower, and a 4% digits LED display panel meter (RP-4500, Texmate). The amplifier is rated with a maximum input offset current of

A, while the display, of the autozeroing type, is rated with a precision of 0.015% of reading i 2 digits and is set to operate a t 1.9999 V full scale (0.1 mV resolution). A lot0 f2 resistor orotects the input. A ten-turn variable ootentiometer adjusts ihe zero offsetif the operational amplkier. A poten- tiomrter f i~ r the fine rerulation of theaain is vrwided in the display panel meter. A filter is placed on the line power supply and, in order to minimize noise, a power supply separated from the digital display feeds the operational amplifier, consisting of a *15 V source scaled down to i12 V hy two Zenerdiodes.

' Diefendwfer, A. James, "Principles of Electrmic Insbumentation," 2nd ed., W. B. Saunders Company, Philadelphia, 1979, p. 265.

Warner. Brian D., Boehme. Gerhard, and Pool. Karl H., J. CnEM. Ewc., 59, 65 (1982).

Choppin. Gregory R., Bertrand. Peggy A,, and Bungli, J. C., to be published.

Care was taken to prevent both leakage of the signal to ground and nickuo of radio freuuencv noise. An NBS connector with . - ~ef lon" irklator was used for the high impedance input, the signal path from the connector to the input resistor to the amplifier pin was all in solid copper wire in air, and all the critical surfaces were carefully degreased. The operational amplifier was mounted upside down on a cooling fan glued on

Figure 1. A real vonmeter allows a finite current to flow, which introduces an error in Hm reading

Figure 2. High Impedance digital voltmeter:

Schematics:

LF: Line Filter SW: Switch DD: Digital Display, RP4500 PS: Power Supply. 15 V. 100 mA ZD: Zener Diodes. 12 V

PR1: = PR2: Resistors, 150 0. '1, W OR: lo turn Potentlamater, 20 K 0 iR: Resistor. 10'O 0

OA: Operational Amplifier. ICH 8500 A

Volume 61 Number 10 October 1984 935

E vr SCE

I -4 -3 -2 -I

10s [uo:']

Figwe 3. Uranyl selective electrode Potential (vs. SCE) vs. Uraniurn(V1) can- centration pH 4. phthalate buffer. 0.094 M.

a PVC board, and the case was connected to the output to minimize leakage. I t was also placed in a separate shielding box in order to protect it from noise from the display meter.

Since the input of the 8500 operational amplifier is inter- nally orotected by 7finer diodes, no extraordinary precautions in handling or soldering are needed to prevent breakdown due to discharge of staticelectricity, as it is usually the case with MOS devices. The input 10'" 11 resistor is needed to prevent occasional out-of-ranpe potentials from inducing destructive current flows in the internal protection circuitr;.

The device is stable within less than a minute after turning on the power. The noise level is such as to produce an occa- sional one-digit fluctuation in the least significant figure. The impedance was tested by measuring a constant voltage (-1.5 V) with and without a 4.10" Q resistor inserted in series on the icnut. The measured sienal d r o ~ ~ e d 1.3 mV on the in- sertion of the resistor, whGh corresponds to a current of 3.10-l5 A. eouivalent to an imoedance of 5.1014 Q.

Figure 3 proves how the impedance of the voltmeter influ- ences the measurement. The uranium-selective electrode was used with this voltmeter and with a Corning 130 pH meter: both instruments have 0.1 mV resolution, and both gave re- producible and drift-free readings, but the Corning instru- ment, which has a rated input offset current of 10-12 A, gave readings consistently 5 to 10 mV lower than o w voltmeter.

This research was supported by a contract with the Office of Health and Environment, USDOE. The author wishes to thank G. R. Choppin for his help in preparation of this ar- ticle.

936 Journal of Chemical Education