Load resistor when using a DMM?

[rontech]

Not sure if this is in the right forum.

I have noticed when people are restoring radios they often list a set of voltages they have measured and compared them with the values in the relevant service manuals / trader sheets. The measured values are often higher and subsequent posts say this is due to the very high input impedance of a DMM.

Would it be worthwhile for these type of measurements to fit a ( temporary ) load resistor across the input of the DMM to bring the impedance down to Avometer levels._ Say somewhere between 100k and 1 M ohm?

[Wendymott]

Sounds about right to me..... if you know the AVO range then 20K/V is good to me.

[barrymagrec]

Depend what you are measuring - not much point if you are measuring cathode voltage across a 470 ohm resistor but possibly some use if you are on a highish impedance point.

PSU voltages shouldn't alter much - a 10meg DMM is equivalent to an AVO 8 on the 500 volt range anyway.

[snowman_al]

Remember a lot of the measurements quoted in the Trader Sheets pre-date the 20,000Ω / volt AVOs. Many are quoted as 1000Ω / volt or even 500Ω / volt.
They had/have a significant effect on the higher voltages readings. (250 volts DC on a 20kΩ meter loads the circuit with 5MΩ. A 1k meter loads the same circuit by 250kΩ.)
So depending on what or where you are measuring there may be a difference between a DVM reading (with 10MΩ / volt) and the figure in the Trader.

You need to be aware of it rather than compensate for it with a 'resistor'.
In fact you would need a range of values depending on the sensitivity of the original meter used and the range it was set to to take the measurement...

[Sideband]

Originally Posted by barrymagrec

Depend what you are measuring - not much point if you are measuring cathode voltage across a 470 ohm resistor but possibly some use if you are on a highish impedance point.

PSU voltages shouldn't alter much - a 10meg DMM is equivalent to an AVO 8 on the 500 volt range anyway.

Absolutely spot-on!

The only time you need to consider the difference between the old Avo reading and a DMM reading is in high-impedance circuits like AGC rails and (typically) the anode voltage of say an AF amplifier valve with a 220K anode load resistor.

[G6Tanuki]

I'd only bother with this if trying to measure the likes of an AGC rail [where, if anything, a DMM's high impedance is a good thing] or if looking for DC volts on a local-oscillator grid [where I suspect an old AVO or a modern DMM will add so much extra stray capacitance that the test is spurious anyway].

[rontech]

Thanks for your responses. Very informative.

[Craig Sawyers]

The other thing worth bearing in mind is measuring AC voltages, such as ripple on an HT line. A DVM usually calculates the RMS value of the waveform digitally. What you get on an AVO will be different. A moving coil meter responds to the average value of the waveform. If that waveform is sinusoidal all is well - because the AVO scales are calibrated for the RMS value of a sine wave. It is described as an average reading, RMS calibrated instrument - but only for a sine wave.

But if the wave is not a pure sine wave - such as the ripple waveform - the calibration will be off. So you will get a different ripple voltage measurement with an AVO as compared with a DVM.

Craig

[Oldcodger]

Mention of AC reminds me of an explanation I was given as to why rail signal testing requires that on track DC measurements require that the meter is fitted with a high (memory suggests 100k) shunt). I thought it strange that these meters (Fluke 187 etc) needed to be shunted as the impedance of these is high compared to the older meters that would have been used.
The explanation was to prevent AC pickup from the 25kv overhead.
As for using meters of higher impedance on circuits where something like an AVO (20KOHM/V) might be specified, surely it's just a matter of common sense to adjust the readings to allow for the impedance differing.

To my mind this is very like the old arguments about RC time constants. As engineers, we work out the time constant as R times C whereas our scientific colleagues revert to some long-winded mathematical formula(apologies to any scientist reading) and make mental adjustments.

[TonyDuell]

Originally Posted by Oldcodger

To my mind this is very like the old arguments about RC time constants. As engineers, we work out the time constant as R times C whereas our scientific colleagues revert to some long-winded mathematical formula(apologies to any scientist reading) and make mental adjustments.

I'm a scientist, at least by my qualificatons, and I can assure you I calculate a time constant as R*C. I also normally work in megohms and microfarads when doing so.

Getting back to the original question, I wonder how much of a difference the meter resistance makes in a lot of cases. Certainly for things like HT rails, heater supplies,output valve cathode voltages (which you measure across a relatively small resistor anyway) I can't see it being a problem. The voltages given in the service sheet are not exact anyway.

[Oldcodger]

Tony- my point using R*C as an example. As to how two different branches view the same problem and as long as the result falls in the ballpark then all is OK.
Same as voltages as the older stuff used resistors of varying tolerances and when stuck it was (in my training) to compare a working bit of equipment with the faulty one. Experience tells you when a voltage is outside what you'd look at as reasonable. Is the problem loading caused by a component presenting excess impedance( thus dropping the voltage by ohms law) or high voltage caused by a circuit /part of it not drawing the expected current.
However, how many of ye old school GPO engineers remember being taught to find HR contacts with a Detector no5, whereas an AVO or something of 20k/v would not see the volt drop.

[GrimJosef]

Slightly OT, but I've just temporarily wired a 47R across my DMM's probes. The four AA batteries powering another DMM, which was complaining of 'low battery volts', all measured OK - around 1.32-1.35V - without the resistor. Three of them maintained this across the 47R. The fourth dropped to 0.65V, so it's obviously gone high-resistance inside. Now I only have to throw one of them away.

Cheers,

GJ

[TonyDuell]

I suspect a 20000 ohms/volt meter like an AVO8 would have shown very similar results on those AA cells. Even a 1000 ohms/volt meter may well have done. There are certainly times when you want a voltmeter to draw significant current. Testing primary cells is one of them.

I think what this illustrates is something that I was taught many years ago. 'It is a lot easier to make measurments than to understand what you are measuring'. In other words you can stick your voltmeter probes on a device. But whether that voltage reading is meaningful is another matter.

You can't possibly own one of every multimeter type ever made. Even having the right sensitivity is not enough. If the readings on the service sheet were taken with a 20000 ohms/volt meter on the 500V range and you meter only has 100V and 1000V ranges you don't have something of the right resistance that will load the circuit in the same way. So you either have to learn how to add load resistors or how to make an allowance for the current drawn by the meter -- or preferably both.

Incidentally, if those 4 AA cells were from the same batch and fitted at the same time, I think it's unlikely that the other 3 will last much longer.

[SteveCG]

Re - testing AA and AAA cells:

I use an old AVO meter on its 6V scale that requires 6ma for FSD for just the purpose of giving the cell a bit of a load.

[John10b]

When testing the condition of small batteries I usually select 10A range on my meter, this quickly lets me know state of battery.
I agree with Tonyduell, you got to understand what the measurements mean.
John