On Amplitude Comparison
Abstract Generated abstract
This article presents a circuit for comparing measured voltage amplitudes with a stabilized reference using bipolar rectangular pulses, electromechanical commutation, diode limiting, and an oscilloscope as a null indicator. It describes the formation of the reference voltage, the use of low capacitance GaAs mesa diodes as limiting elements, and experimental determination of device sensitivity, including comparison with a theoretical sensitivity curve. The analysis identifies an optimal operating current of about 40 to 50 microamperes and evaluates errors arising from resistance accuracy, load current, and visual reading of oscilloscope traces. The device is reported to compare unipolar and bipolar rectangular pulses of about 9 V over durations from 0.1 s to 20 microseconds with an error no worse than 0.01 percent and an input resistance of 220 kilohms.
Full Text
On the Issue of Amplitude Comparison
G. I. Zaidman
(Presented by the scientific seminar of the Department of Radio Engineering)
This article describes the circuit of a device for comparison by amplitude values of the measured voltage, explains its operation, and analyzes measurement errors.
The electrical circuit of the comparison device is shown in Fig. 1. As can be seen from Fig. 1, to form bipolar rectangular voltage pulses, the voltage of galvanic batteries is used, stabilized by two electronic stabilizers:
- a stabilizer with feedback;
- a parametric stabilizer.
The midpoint of the connection of the Zener diodes is grounded. Periodic closing of contacts a, b of an electromechanical commutator based on a self-oscillating multivibrator and an RP-4 type relay makes it possible to form bipolar pulses with an amplitude of 9 V and a repetition frequency of 50 Hz, used as a reference voltage. With the aid of a second electromechanical commutator, the reference voltage pulses through contact v and the measured voltage through contact g are periodically applied to a series bilateral diode limiter, whose diodes are biased to a preset level. The reference-voltage and measured-voltage pulses, limited from above to a level of 15–20 mV, are fed to a C1-19 type oscilloscope, which is used as a null indicator.
The frequency of the second electromechanical commutator is varied over wide limits; this provides convenience in selecting the corresponding sweep duration and makes it possible to measure infralow frequencies of the investigated voltage down to direct current.
As limiting elements we used GaAs mesa diodes with a junction capacitance near the zero point of the current-voltage characteristic equal to 3–5 pF. Figure 2 shows the forward branch of the current-voltage characteristic of such a diode. On its basis the dependence of the dynamic resistance on the magnitude of the forward current was constructed, which is presented in Fig. 3. This dependence is subsequently used to construct the theoretical sensitivity curve.
The sensitivity of the comparison device was determined according to the block diagram shown in Fig. 4:
1 — stabilized DC voltage source,
2 — DN-1 voltage divider from a PPTV-1 potentiometer,
Fig. 1.
Fig. 2.
Fig. 3.
Fig. 4.
3 — P-309 potentiometer,
4 — comparison device,
5 — voltage divider for the P-309 potentiometer,
6 — S1-19 oscillograph.
A constant voltage from the stabilized source VSP-30 was applied to a parametric stabilizer composed of a resistance \(R\) and a thermostated stabilitron D-818E. The absolute value of the voltage on the stabilitron was simultaneously measured by means of a P-309 potentiometer (input \(X_1\)), using a DN-1 voltage divider and the comparison device. The reference voltage of the comparison device was also monitored by the P-309 potentiometer (input \(X_2\)), using the voltage divider from the P-309 potentiometer. For this purpose, one of the ballast resistances \(R_{\mathrm{b}}^{*}\) or \(R_{\mathrm{b}}^{**}\) was first disconnected, and in its place the voltage divider to the P-309 potentiometer was connected. Such switching made it possible to measure the voltages of the reference and measured stabilitrons under conditions close in load current. Figure 5 shows the dependence of the sensitivity of the comparison device on the magnitude of the input current for positive and negative pulses of the reference voltage, obtained experimentally (curve 1) and constructed theoretically (curve 2).
Fig. 5.
It is seen from Fig. 5 that it is desirable to choose the value of the operating current within the range \(40\text{–}50\ \mu\mathrm{A}\).
In determining the absolute value of the reference voltage, two errors arise which must be taken into account. The first error is due to the difference in the accuracy classes of the ballast resistance \(R_{\mathrm{b}}\) of the comparison device and the divider \(R_g\) of the P-309 potentiometer. It is written as follows:
\[ \gamma = \frac{R_{\mathrm{dyn}}}{R_{\mathrm{d}}}\cdot K \cdot 100\%, \tag{1} \]
where \(R_{\mathrm{dyn}}\) is the dynamic resistance of the D-818E stabilitron at a current of \(7\ \mathrm{mA}\),
\(R_g\) is the input resistance of the divider of the P-309 potentiometer when a voltage of about \(9\ \mathrm{V}\) is applied to its input,
\(K\) is a coefficient taking into account the scatter in the accuracy class of the resistance \(R_{\mathrm{b}}\) relative to the resistance \(R_g\) of the P-309 potentiometer.
The second error in determining the absolute value of the voltage of the reference source is associated with the magnitude of the load current
\[ \beta = R_{\mathrm{dyn}}\cdot \frac{U_{R_{\mathrm{n}}}(\pm a)}{R_{\mathrm{n}}(\pm b)} \cdot \frac{1}{U_{\mathrm{st}}}\cdot 100\%, \tag{2} \]
where \(U'_{\mathrm{st}}\) is the voltage of the Zener diode,
\(R_{\mathrm{n}}\) is the value of the load resistance,
\(U'_{R_{\mathrm{n}}}\) is the voltage drop across the load resistance.
In parentheses is indicated the accuracy class of the instrument by means of which the voltage across the load resistance is determined, and the accuracy class of the resistance. The sign of the error \(\beta\) is known, and it is easy to take it into account with the accuracy determined by the instrument with which the voltage across the resistance \(R_{\mathrm{n}}\) is measured.
The next error is associated with determining the exact value of the distance between two sighting lines: the reference line and the controlled line. Let us call this error \(Z\). With a beam thickness of \(2\ \mathrm{mm}\), the maximum error will be determined by the properties of the operator’s eye and, in the worst case, may amount to \(2\ \mathrm{mm}\), which corresponds to a measurement error on the order of \(0.01\%\). When a magnifying glass is used, this error can be reduced by a factor of 3–4.
An analysis of all the errors has shown that reliable comparison of unipolar and bipolar rectangular voltage pulses with an amplitude value on the order of \(9\ \mathrm{V}\) can be carried out over a duration range from \(0.1\ \mathrm{s}\) to \(20\ \mu\mathrm{s}\) with an error no worse than \(0.01\%\); in this case the input resistance of the comparison device is \(220\ \mathrm{k}\).