Using a modern high impedance meter does not load the circuitry in a set like the meters of yesteryear did. Often you will see on a schematic that the voltage readings were taken with a 1000 ohms per volt or a 20,000 ohms per volt meter.<P>Todays meters have imputs in the megohm regions and the old standard for vacuum tube volt meters (VTVM) was 11 megohms, or 10 megohms on AC.<P>This can cause readings to indicate high when going with a schematic that calls out using the older meters. It does not mean that your resistors or capacitors are bad, but simply bad science involved in the testing apparatus.<P>I mentioned this on a previous forum, but thought it would be nice here to let everybody know how I get around that problem.<P>Since a 1000 ohms per volt meter in the 250 volt position has an impedance of 250K ohms and in the 500 volt position a 500K impedance (math is simple guys and gals), 1000 times whatever voltage range the meter is set for. I simply put a "loading" resistor in shunt with the meter leads. The additional loading presented by a modern meter is insignificient, so I disregard it. Since most values of DC voltages we look at in sets is in the 0 to 250 volt range, I imagine the meter being set to the 250 volt range. This calls for a 250K ohm resistor shunting the meter. If you are measuring higher voltages that would call for the 500 volt range, simply substitute a 500K ohm resistor.<P>Please note that the value is not that critical. You just want something that will get close to the loading the original meter would have. If you want to separate the fly specs from the pepper, you could figure out, using Ohm's law the exact value, but that is uncalled for in a world of minimum +/- tolerances of 20% or worse we see in vintage sets.<P>My meter has banana plugs where the leads plug in and I simply put the resistor there. A one watt resistor would work in all cases and fits the twin banana plug just fine.<P>Hope this suggestion helps with some of the "too high" voltage readings we often see on the forum threads.<BR>Curt<P>------------------<BR>Curt, N7AH<BR>(Connoisseur of the cold 807)<BR>QCWA# 25085 AMI# 242<BR>CW forever

Thanks Curt, that's very helpful!<P>------------------<BR>Richard <BR><A HREF="http://www.dressner.com/radio.htm" TARGET=_blank>www.dressner.com/radio.htm</A>

Good info, Curt. <BR>Thanks!<BR>Tom<P>------------------<BR>

There's another thing these digital meters do that drives me absolutely nuts. The last time the battery got weak in my Fluke, I was troubleshooting in a solid state circuit and it would read voltages that seemed entirely nominal for the circuit, BUT WERE IN FACT NOT THERE! I mean, the votage was in fact zero! Wasted over an hour til I checked with a second meter and figured out what was happening. It's like with computerized equipment, I guess. This thing gets asked for a voltage, its battery is weak so it can't read one, so it's just makes something up?!?<P>------------------<BR>OZ

What I hate about digital meters is how unstable and erratic they can get when they are exposed to RF fields. I wanted to check how accurate the 40+ year old filament voltmeter was on the Gates BC-1T broadcast transmitter and the only way I could get my digital meter to give any sort of a proper reading was to momentarily kill the carrier.<BR>Curt<P>------------------<BR>Curt, N7AH<BR>(Connoisseur of the cold 807)<BR>QCWA# 25085 AMI# 242<BR>CW forever

I have run into this same scenario before using VTVM and/or FET-VOM's. I have a B&K Transistor Radio Analyst that incorporates a meter with 5000 ohm/volt sensitivity and it will usually read voltages close to service specs listing a 1000 ohm/volt on schematic. Since it has a 500V scale, it works perfect for tube radios too. I guess we should not have tossed/gifted/trashed those cheap 1k ohm/volt meters when we "stepped up" to better stuff. I know I got rid of my Rat Shack Archer Kit one many moons ago because I wanted to improve my bench.(1) VTVM, (2) FET's, (1) DMM is what I use plus the B&K one when I just gotta know if it matches spec. I only use FET or VTVM around critical stages such as Oscillator.<P>Gary<P>------------------<BR>You did WHAT???

Great idea Curt. One other item to remember is line voltage. Many sets were tested at 110 or 115 volts. These days it can be 125 volts or higher in some places. Figure 10 percent increase input and your B+ shoots up a bit!<P>------------------<BR><BR>

Thanks Curt ... I'm going to build one into a meter probe to use for this purpose.<P>------------------<BR>

OOP's ... cant put it in a probe ... have to shunt the meter like you said ... silly me... will do the double banana plug deal like you suggested.<P>------------------<BR>

Just one thing to add. Differences in voltage reading is noticed when taken through high resistance values. <P>When reading a tube plate voltage through 100K resistor there will be a difference between meters.<P>Voltages taken right from the power supply will read the same regardless of meter loading. <P>Adding a shunt as Curt suggests will make all voltages match the schematic. <P>------------------<BR>Norm

Just bringing this thread to the top again.
Curt

Great info Curt.

I have a couple of old tube meters I plan to put back in service, but I keep my digital handy most times.

I knew it's readings would not be 100% correct for those applications, but I figured I would learn to allow some with experience. Now I don't have to worry about that.

One more item for the notebook.

The good old PSM-6A has both 1k/v and 20k/v sensitivities. They show up on eBay from time to time.

RRM

Watch out for the "hold" feature on some DVM's. It will read the last voltage measured.

Carl

As basic as it sounds, that's why I bought an antique Simpson Series II VOM from good old Genoo, although newer Simpsons can be used. My .02-Gearhead

"I simply put a "loading" resistor in shunt with the meter leads."

Does that mean one lead of the resistor connects to the red meter probe and the other resistor lead connects to the black meter probe? Thank you.

Yes, exactly. What it does is to parallel the meter's resistance with the added resistor to come up with something close to what an old time meter would have had for an input impedance.

Since this thread is more than three years old, and still being found useful by some, why can't it be made into a "sticky" post at the top of the threads in the Hints and Kinks forum?

Alan.... hey Alan, where are you?
Curt

Curt was right about this 3 years ago.
This thread helped me a great deal.
Phil

gentlemen, i think i understand what's basically being said, that newer instruments might show higher voltages than the old voltmeters showed in identical circuits, but in percentages, what are we talking about here? 10, 20% higher? would 150 or 180 volts make enough of a difference as to keep the circuit from working thus encouraging me to either do some sort of math (if i used a dvm) or just keep doing what i do anyways, which is use my simpson 8 analog baby?
i'm just trying to understand the best way to proceed in light of this information and thanks for the constant education i've received over the last 10+ years i've been on here.
tony

There are no general answers to that question.

Once common misconception is in determining the meter resistance of VOMs.
The sensitivity in ohms per volt must be multiplied by the meter range selected, not by the actual reading.

For example (using a 1000 o/v meter):
If you set it on the 100-volt range, the meter resistance is 100,000 ohms (= 100 * 1000).
If you set it on the 250-volt range, the meter resistance is 250,000 ohms (= 250 * 1000).

Those resistance values do not change as a result of the actual voltage being measured.

For modern meters (≥10 Megohms input resistance), calculating the VOM range resistance as above and putting that resistance in shunt across the meter probes/leads is accurate enough.

=====

The error depends entirely on the resistance between the B+ source and the node being measured.

The actual loading error for any particular metered point (node) is easily calculated. For example:

The schematic shows 200 volts at a node (e.g. screen grid) that has a 100,000 ohm resistor to B+, which
is 350 volts. For this reading the meter would be set on the 250-volt range (= 250,000 ohms resistance).

The 200 volts actually appears at the junction of the 100,000 ohm resistor and the parallel combination
of the 250,000 ohm meter and the load (screen grid).

The current through the system = (350v - 200v) / 100,000ohms = 150/100,000 = 1.5 mA.
The current through the meter = 200v / 250,000 ohms = 0,8 mA, which is a substation portion of the 1.5 mA.
The remaining current flows through the load (screen grid??) = 0.7 mA.

You can see that in this case the meter draws more current than the actual screen grid.

If you replace the 1000 o/v VOM with a modern 10Meg meter, you can ignore the meter resistance.
The circuit current is 0.7 mA. The drop across the 100K resistor = 70 volts, so the screen = 280 volts.

That's quite a substantial difference in readings.

In practice, the screen current will be slightly higher due to the higher voltage, so its voltage will be slightly
less than the 280 volts calculated above, but not my much.

- Leigh