May 1959 Popular Electronics
Table of Contents
People old and young enjoy waxing nostalgic about and learning some of the history
of early electronics. Popular Electronics was published from October 1954 through April 1985. All copyrights
are hereby acknowledged. See all articles from
Electronics ran a 5-part series on test equipment usage. This final
article is on the use of a vacuum tube voltmeter (VTVM) for making
accurate AC and resistance measurements. Also in this edition is
a construction article for RCA's
VTVM kit, so the two compliment each other. Author
Larry Klein discusses mainly the AC and ohmmeter functions, providing
both functional descriptions of the circuits and how to use them
for making accurate measurements. FET-input digital multimeters
(DMMs) have largely replaced VTVMs, but they can still be found
in some older electronics development labs and hobby benches.
Test Instruments Part 5
By Larry Klein
The Vacuum-Tube Voltmeter
- A.C. and Ohmmeter Ranges
A simplified d.c. voltage measuring
circuit showing the range switch and bridge circuit.
Two typical voltage-doubling rectifiers
used in VTVMs. The diodes' contact potentials buck each other
in (A) and the "A.C. Bal" pot selects the zero point. In (B)
the negative potential is bucked against B-plus voltage tapped
off the "A.C. Bal" pot.
Typical waveforms from a standard
Note relationship between r.m.s. and
P-P scales. P-P scale is 2.83 times larger than r.m.s. scale.
Circuit diagram of an ohmmeter section
of a vacuum-tube voltmeter.
Simplified input and range circuits
of VTVM ohmmeter section.
month we looked into a vacuum-tube voltmeter, examined the bridge
circuit and saw how it measured d.c. voltages. As a review, let's
look at Fig. 5, a diagram of the d.c. measurement circuit. (Figures
1 to 4 appeared in April.)
The unknown d.c. voltage connected
to the input terminals is applied across the entire range switch
voltage divider. Maximum on-scale reading is obtained by setting
the range switch at the proper voltage divider tap. The unknown
d.c. voltage is now applied to the input grid of the bridge - unbalance
of the triodes results and the meter deflects. So much for the d.c.
bridge. A.C. Voltage Measurement.
do we have to do to enable the d.c. bridge to respond to a.c.? Why
not simply rectify the unknown a.c. voltage and then apply the resultant
d.c. to the bridge input as we would any d.c. voltage? That's actually
what the standard VTVM does. Unfortunately, however, a number of
electronic bugs appear which prevent a simple diode circuit from
being used, and the circuits in actual practice usually look like
those in Fig. 6. Why the complications? Let's take a close look
at Fig. 6 (A).
On one half of the cycle, the a.c. voltage
to be measured is fed through capacitor C1 to the cathode of one
diode of the 6H6 tube, and thence to ground, The capacitor, of course,
gets charged in the process. On the positive-going part of the a.c.
cycle, no current flows through the first diode, C1 discharges and
adds its voltage to that developed across the three resistors connected
to the plate of the second, conducting diode.
If we look
carefully at the circuit, we'll recognize a type of voltage doubler.
Why a voltage doubler? Well, remember we need to get a d.c. voltage
out of the rectifier circuit which is at least as high as the a.c.
input voltage. Taking into account the voltage drop across the various
components in the circuit, obviously some technique is needed to
soup up the d.c. output ... and that's what the doubler does.
Further circuit complications arise from a phenomenon called
contact potential. It seems that vacuum tubes, including diodes,
tend to develop a small potential between the elements. If allowed
to remain, this slight voltage in the 6H6 would cause a spurious
reading on the low a.c. ranges. However, placing the a.c. balance
control between the two oppositely connected diodes, exact compensation
can be made by bucking out the opposing contact voltages.
Since the center contact of the a.c. balance potentiometer is
also the take-off point for the d.c. output, about half the d.c.
developed across the three resistors is lost by tapping off at this
point. Actually, this is of small consequence, because the d.c.
voltage across the three resistors is equal to more than the peak
of the r.m s. a.c. input voltage, so we have volts enough to spare
to provide an r.m.s. reading. R.M.S. and P-P.
The key words in that last sentence were "r.m.s, reading," which
brings us to Fig. 6(B). Slightly more complicated than the rectifier
discussed above, this circuit also makes use of a doubler circuit.
Because of the low breakdown voltage of the 6AL5 tube, a
voltage divider (in addition to the one in the grid of the bridge
tube) is needed to prevent the tube from "arcing out" at the higher
peak voltages. As shown, the a.c. input voltage divider is part
of the range switch and is, therefore, mechanically coupled to the
Perhaps you're wondering why the extra resistors
at the a.c. input don't cause a large difference in scale calibration
between the a.c. and d.c. ranges. The VTVM takes care of that by
switching the last three bridge voltage divider resistors out of
the grid circuit when set up for an a.c. reading.
the job of the second diode in Fig. 6(A) is mainly to cancel out
the contact potential of the first diode, the second diode of Fig.
6(E) has a different story to tell. Both diodes in Fig. 6(B) are
used in a complete voltage-doubler hookup which charges C2 to the
full peak voltage of the incoming waveform. Contact potential cancellation
voltage is obtained from a tap across the VTVM's B-plus supply.
The waveforms shown in Fig. 7 are taken from a standard
TV set. You can imagine the difficulties an r.m.s. calibrated a.c.
meter would have translating them to any sort of meaningful reading.
Even putting a peak-to-peak reading scale on the meter face (it
would be the r.m.s. scale x 2.83) wouldn't help much because the
reading would still only be accurate for sine-wave inputs.
However, the P-P a.c. rectifier finds no difficulty in smoothing
down these weird-looking spikey TV waveforms into an exact d.c.
equivalent and then feeding them to the bridge circuit. The exact
relationship between the P-P scales on a standard peak-reading VTVM
is shown in Fig. 8. Resistance Measurement.
One of the first things that hits your eye in the ohmmeter section
of the VTVM is the R x 1 meg. range switch position. With the last
scale division on the meter face marked 1000, this means that the
VTVM can read up to a 1000 x 1 million or a billion ohms!
The ohmmeter section of the average VTVM resembles the one shown
in Fig. 9. The string of seven resistors may differ in value somewhat
depending on the exact scales used and whether they are arranged
in series, as shown, or switched individually. But the principle
of operation remains the same, as we shall see.
we redraw the range switch and input circuit of Fig. 9 into the
form of Fig. 10. We will use only one range resistor (Rrange
and connect the resistor to be measured (Rx) to the VTVM's input
terminals. The bridge circuit remains the same and we will ignore
it for now.
The first thing to do when using a VTVM ohmmeter
is to "zero" it. Short the input leads together and adjust the Zero
Adj. control for a zero reading on the meter scale. Then, unshort
the leads of the VTVM and the needle will immediately swing to the
right-hand side of the meter face. Now adjust the meter to ∞
Let's see what the preceding adjustments
have accomplished in terms of the internal electronics of the VTVM.
Zero-adjusting the meter with the leads shorted has shorted
out the battery through resistor Rrange
to ground and
removed the voltage from the grid of the bridge tube. Unshorting
the test leads restores the battery voltage to the grid and the
meter swings full scale. The Ohms Adj. knob, which is in the same
spot as the A.C. and D.C. Cal. controls in the other circuits, adjusts
the sensitivity of the meter so that the applied battery voltage
swings the meter needle exactly to the infinite ohms scale marking
on the meter face.
a 100-ohm resistor (Rx) is connected across the input leads and
is also set at 100 ohms. The voltage present at
the grid of the bridge tube will be exactly halved, and the meter
will read half scale. Now if you look at the top scale of the meter
face shown in Fig. 8, you'll see that the center of the scale indicates
If Rx were a 30-ohm resistor, for example, the
shunting effect across Rrange
would be increased and
even less voltage would reach the bridge tube. A higher value resistor
as Rx and a higher meter reading results. The only trick involved,
and the reason why it's so difficult for some home constructors
to build their own ohmmeters, is the scale calibration. As can be
seen in Fig. 8, the scale divisions are widely spaced at the right
side of the meter face and narrow down towards the left. A little
thought as to how parallel resistors divide current will tell you
why that is so. The Function Switch.
talking about the VTVM, we've left out practically any reference
to the function switch. Since these switches are so difficult to
show schematically in an understandable way without a prolonged
discussion of each switch position and what it accomplishes, we
thought it best to save them till last.
The function switch
is usually specially made for each manufacturer's VTVM and, if analyzed,
generally works out to be a five-pole, five-position unit. Some
of its jobs include switching the input jacks to the proper circuit,
connecting in the correct calibration control for each function,
reversing the meter movement connections for plus and minus d.c.
and, in some cases, even turning the VTVM on and off.
you're curious, a complete schematic of the RCA "VoltOhmyst
VTVM kit is shown on page 79 of this issue and should answer any
questions you may have about the specific connections of the function
Next month we will put the VTVM "to work in an area
in which it's practically indispensable - repairing a hi-fi amplifier.
The basic Williamson amplifier should be a good subject, and we
will learn how to troubleshoot one and what sort of measurements
the VTVM will turn up in working and non-working models.