Can a DMM be trusted?

[Kalee20]

The "True RMS" ones are believable for lower frequencies, the question is what do you want to measure?  Non sinusoidal filament voltage/current at AC then RMS is what you want, sine wave AC or DC any meter will do.

Hmm... You'd think so, but my Fluke 177, 'true RMS' isn't any good!

It filters out any DC and does a true RMS on the AC content only. That means if there's any DC component, such as a diode dropper, you won't get a true RMS on the whole - and your valves will overheat.

It's easy to see if this is the case. Just select AC, and measure a DC voltage or current, True 'true RMS' will give the DC value; AC-filtered 'true RMS' will show zero (after a brief reading).

[Mike Watterson]

Some DVMs have a rapid response LCD "needle" or bargraph. A friend has one but I've never needed the feature, I have a scope.
A decent damped analogue movement will show the average.
I'd want the scope, not a fast response analogue or digital meter to see what's happening if the DMM wasn't sufficient. I've ONLY found the AVO better for over 500V and resistance of very large inductances.
My DMM transistor tester and my universal R-L-C and Active component testers test Germanium transistors better than the AVO. The cheap DMMs with transistor testers may get confused by Germanium transistors.

I encountered plenty of rubbish Analogue meters since 1965 and the cheap DMMs are in the same league, though better than the 1960s & 1970s cheap analogue meter.

It's always important to have good quality tools and understand the limitations and how they work. So you can't "trust" any kind of meter without understanding all the issues.

I have a cheap dedicated 9V & 1.5V battery meter that appropriately loads the battery or cell. Otherwise I put a suitable resistor load on the battery or power supply (a car headlamp bulb will do as load for big 5V SMPSU) and use a 20M Ohm input DMM. It also has an optically isolated RS232 out which I've used to plot filament current vs supply volts (fixed 0.1V steps with 2nd meter and just readings of current captured).

[DrStrangelove]

The last time I used an AVO 8 I used two: both set to 3kV to measure the output of a high voltage supply.

The 6.5 digit dvm I bought 2nd hand lasted all of a week before bringing up "FAULT" so that was £100 well spent.

[PerdioPal]

I have three of the last analogue meters Maplin were selling (£9.99 each), very useful with battery on load testing and centre zero volts ranges. That makes quickly checking volts on a board etc. very easy implemented by injecting a weeny (half FSD (full scale deflection)) current into the movement. +-5 and 25V. So simple and useful.

[The Philpott]

'Do you trust your Digital Multimeter?'

What i do not trust a DMM to do is to -necessarily- let me know when it's poorly.
All of my analogue multimeters are ACWEECo/AVO which means that i have been spoilt to a certain extent, however i have found that they are more likely to hand in a subtle but legible sicknote rather than soldier on and generate misleading twaddle. Familiarity with Avometers may of course mean that i find it easier to read their sicknotes; i accept that.

The amusing false positive that you'll all have experienced with a DMM is when isolating one circuit in your house; you go to retest the circuit for safety's sake and still see a significant voltage, albeit a phantom voltage generated by capacitive coupling with the remainder of the wiring loom.

Dave

[BusyBee]

I see that the higher the input impedance / resistance of the meter the more likely one is to be misled by phantom voltage pickup. It would not be helped by a better high frequency response either. The closer you get to electrometer performance the greater the issues their traits become noticeable. The good thing about an Avo is that, with their loading on the circuit, phantom pickup is almost non existant. Maybe some have it, I don't know, but there is maybe a need for having a switched input load / impedance on a DMM to compensate in some circumstances. A setting with the same load as an Avo 8, for instance, might make comparing with some circuit descriptions a lot easier.

Tracy

[ppppenguin]

All measuring instruments tell lies. Part of good engineering is understanding the lies and working around them.

A DMM will show phantom AC voltages, as will a neon screwdriver and a non-contact voltage detector. At least they are erring on the side of safety. A moving coil meter, with relatively low resistance**, will show low voltage readings in high impedance circuits. The capacitance of a scope probe will give a false impression of high frequencies. And so on.

**Here's an exception. An AVO 8 on the 1000VDC range will present 20M to the circuit. Most DMMs will be 10M.

[joebog1]

Well, at least somebody is reading this post.
I have never bothered to check my AVO 8 input impedance.
It was good enough 50 years ago, its good enough for me.
YES, in high impedance circuits low voltage, high impedance circuits, present something of a problem, especially in interpretation.
I thought that was what was taught AFTER you left engineering classes, and started working in the real world.
Its called experience.

Joe

[Mike Watterson]

The good thing about an Avo is that, with their loading on the circuit, phantom pickup is almost non existant. Maybe some have it, I don't know, but there is maybe a need for having a switched input load / impedance on a DMM to compensate in some circumstances.  A setting with the same load as an Avo 8, for instance, might make comparing with some circuit descriptions a lot easier.

I have a VVM (analogue) with a huge input impedance. Also on AC it works to VHF, so I've occasionally used it with a 50 Ohms via T-piece. It certainly can give phantom DC readings.
I'll add a suitable parallel resistance when testing unloaded PSUs or cells or batteries. I do have a handy analogue meter for 1.5V and 9V cell that has a parallel shunt inside to load the battery. A tester that did 3V to 5V (coin cells and Li rechargable) might be handy.
I use a suitable load across the DMM for HT batteries.

[The Philpott]

I think that there are Flukes with a low impedance option- (obviously you have to remember to select it.)

Dave