Transferrability of diode mode measurement values between DMMs

[Rooster Cogburn]

I've lately been trying to incooperate diode mode measurements into my repair/fault-finding routine. All the cool kids on YouTube do it, so who am I to deviate.

Something I don't understand is how these measurements could possibly transfer between different DMMs. I've seen people make tables of these values and/or share them in forums etc., so that would seem rather pointless if they are only valid for the author's particular DMM.

From what I understand, the diode mode uses a constant current source to power the DUT and then measures the voltage drop. There is no standard for the current of said constant current source. I can see how that would work for an actual diode as the resistance of a diode should drop low once its forward voltage is exceeded. But in the case of repair work / troubleshooting we're measuring all kinds of points on the PCB, not just ones with diodes directly connected to ground.

Where I can't see this working is simply a resistor. If your CC source is twice as high you should observe twice the voltage drop across the resistor, right? I gave it a try with two DMMs. I know they use a different valued CC source because one of them lights up LEDs in diode mode significantly brighter. They give similar readings for diodes, but if I simply measure a resistor in diode mode they are wildly different, as I expected.

I can't see how values for many test points on a to be troubleshooted PCB could possibly be transferable between DMMs since they likely include all kinds of components in the signal path, including resistors. Additionally, since most people, for some reason, place the positive lead on ground the test points that actually have a diode in the path would all be just OL.

I don't get it.

[Doctorandus_P]

In general it's not much use to try to measure the voltage drop accurately with a DMM.
It also does not make much sense to compare the measurements of a diode with different DMM's.

Diode voltage is very non linear (quite exponential actually) with current, and it's also quite temperature dependent (approx -2mV per degree centigrade). You will probably see the voltage drop if you heat the diode with your finger.

Diode measurement with a DMM is mostly intended as a quick Go / No Go test. Small signal diodes are usually around 600mV, Scottky diodes between 200mV to 400mV, while high power diodes can be 800mV (even over a volt with higher currents)

A part of learning electronics is to know hen accuracy is important, and to also realize that accuracy is often not needed at all, Always have in mind with what sort of tolerance you want to do your measurements or design your circuits.

[Rooster Cogburn]

Thanks, but you gave me a very general answer to a much more specific question. I'm not particularly concerned about accuracy or even diodes in general. I'm very specifically interested in using diode mode for electronics troubleshooting and repair and trying to understand why certain practices that seem strange to me are so common in the repair community. Why do people generally reverse the leads/polarity from what would be the normal usage and how do people deal with the fact that most circuits will cause them to do resistance measurements with diode mode which should be completely meter dependent.

[bdunham7]

Why do people generally reverse the leads/polarity from what would be the normal usage and how do people deal with the fact that most circuits will cause them to do resistance measurements with diode mode which should be completely meter dependent.

I'm assuming that they've found that this practice gives them a signature of sorts which can be compared to other signatures from testing identical boards.  You're right, this will not work as well if you don't have a standardized test current.  1mA is very common, so perhaps it works enough of the time to have some value?  Or the results are so stark that minor variations don't matter--perhaps because of the reversed polarity.  I don't know for sure, it sounds a bit like witchcraft to me, but there's probably some validity in there somewhere.  Can you link specific examples that we could discuss?

[magic]

People do all sorts of silly things.

The diode mode is for measuring diodes. Or other PN junctions inside transistors or integrated circuits. If you measure between circuit points which are connected by a diode, the result will be reasonably repeatable - within 0.1V or better. Maybe in most cases that you have observed this was what they were doing and then it makes sense.

If you measure something else, resistors in particular, they will vary quite a bit. Not only the test current varies between DMMs, but sometimes the cheapest ones aren't even very accurate at all.

[Rooster Cogburn]

I'm assuming that they've found that this practice gives them a signature of sorts which can be compared to other signatures from testing identical boards.  You're right, this will not work as well if you don't have a standardized test current.  1mA is very common, so perhaps it works enough of the time to have some value?  Or the results are so stark that minor variations don't matter--perhaps because of the reversed polarity.  I don't know for sure, it sounds a bit like witchcraft to me, but there's probably some validity in there somewhere.  Can you link specific examples that we could discuss?

Yeah, I see this being useful if you got a working and a faulty board and just want to compare readings so you know which part of the circuit is different / defective.

Those two videos show what I'm thinking about:




I just found it strange that people in forums share their diode mode readings for all kinds of test points.

People do all sorts of silly things.

The diode mode is for measuring diodes. Or other PN junctions inside transistors or integrated circuits. If you measure between circuit points which are connected by a diode, the result will be reasonably repeatable - within 0.1V or better. Maybe in most cases that you have observed this was what they were doing and then it makes sense.

If you measure something else, resistors in particular, they will vary quite a bit. Not only the test current varies between DMMs, but sometimes the cheapest ones aren't even very accurate at all.

Ok, thanks for confirming, that's what I thought. It would make sense to measure an LED to ground, but if you also include the current limiting resistor the results would not be repeatable.