How much do you trust your DMM Amp readings?

Flashlight Modding and DIY Parts
1

Some more testing and rambling.

Background:

There are some people that don't trust the accuracy of the Opple Light Master III (and I agree it's not the end all, be all) but their argument was that it's not consistent, or is too much off from a $1500 spectrometer. ($30 vs $1500)

And I've been wondering about my Amp readings, because they didn't seem to be consistent. And sometimes readers of my reviews reached out, asking about my Amp measurements. (Not a big fan of showing Amp measurements either, but I sometimes add them).

So I went ahead and did some testing, and oh boy.. I don't know which device I should trust now.

I used an old Fluke 77III (not calibrated) with Fluke probes, a cheap DMM, and a Fluke 325 clamp meter (and its probes).

I'll tell you, the numbers are probably less trustable than the numbers on the Opple Light Master..

I really wish these inconsistancies didn't exist, and I thought having a Fluke (albeit an older one) would be mort trustworthy.

Test: Convoy S2+ with 5 modes (8*7135)

I used:

  • Fluke 325 clamp meter set to DC amp testing
  • Fluke 77 III dmm original probes, set to 10A setting
  • Fluke 77 III dmm original probes, set to 300mA setting (for lowest 3 modes)
  • Cheap DMM Xiole XL830L set to 200mA (for lowest 3 modes)
  • Cheap DMM Xiole XL830L set to 10A
  • Fluke 77 III with probes of Fluke 325, set to 10A setting
  • Fluke 77 III with probes of Fluke 325, set to 300mA setting (for lowest 3 modes)

Here are the results...

Mode: Fluke 325 clamp (DC setting) Fluke 77III 10A setting Cheap DMM (10A setting) 77II +325 probes 10A Fluke 77III 300mA setting Cheap DMM (200mA setting) 77II +325probes 300mA
1 0.05A 0.01A 0.01A 0.01A 7.99mA 8.2mA 7.38mA
2 0.14A 0.10A 0.09A 0.10A 79.5mA 94.0mA 81mA
3 0.35A 0.21A 0.24A 0.22A 151.6mA 170.7mA 119mA
4 1.03A 0.70A 0.73 0.65A
5 3.04A 2.75A 2.82A 2.82A

Which would you trust most?

How many DMM's do you have? And why do you trust its readings?

Questions and observations:
Maybe I should clean my DMM contacts (where wires connect to the DMM)? What's the best way to do this?

Calibrating is out of the question at this point, because that's more expensive than just buying a second hand Fluke (in some cases).

It also depends on how strong you push the probes + battery into the flashlight. So there is another inconsistency.

About the clamp meter:

The Clamp meter is slow in reading.. meaning it will first show 0.01, then slowly moves to 0.02, all the way to 0.05.

Also, whenever I turn it on, I have to set it to 0, because it usually shows some current numbers, even though nothing is connected yet.

I bought it to measure high (read 10+) Amps because I don't trust the DMM for that.

But how much can you trust a clamp meter with these smaller currents?

About the Convoy S2+

From what I understood in the past 5-8 years or so, is that each AMC7135 is actually providing 380mAh, instead of 350mA!?!?!?

So, would the clamp meter be more correct, to measure 3.04A instead of the 2.8 from the DMM?

4th mode is somehow unstable with all devices. When I try to measure it, it goes either up or down the moment I connect it. It's much worse that in the other modes. Any explanation?

2

Do you use the same cables ?

Regards Xandre

3

Hi Xandre

Fluke 77III: I used the same cables for the 10A and 300mA readings

Then also tested on the 77 III with the cables of the 325 clampmeter, both on 300mA and 10A setting

For the cheap DMM, I just used the included cables

4

If you measure current, especially on a flashlight with internal nonlinearity, you must get different values if measured the current directly, because the internal resistance of the amp-meter limits the current draw and so is reducing the reading, while an inductive clamp measures rather exactly (depending on quality), because its influence on the circuit is extremely low.

5

For a real test it is necessary
to use same cables otherwise no comparison.

Regards Xandre

6

Good stuff.

So, basically, if somebody only has 1 DMM, he/she won't know if its readings are accurate?

7

I think you miss the point of my test.

How much can you trust your DMM measurements?

8
Amazing how good the cheap DMM is at the 3 A mark. Thanks for reporting this values!
9

Yes, exactly. But I bought the Fluke because I didn't trust the cheap one :D

But now I don't know which to trust anymore :)

10

A DMM measures the current by reading the voltage drop across a current shunt, this shunt is usually not a very low value, this is indicated by the burden voltage spec, it’s not always given in the manual/datasheet, I couldn’t find it for the the Fluke 77III, the cost of the DMM doesn’t really matter in that regard, looking at some Fluke review by HKJ the two of them have a 0.57V drop at 10A, I.e 57mΩ shunt (worse than my cheaper uni-t 139c, 0.36V), to that must be added the resistance of the probe, something like 30~50mΩ for each, plus contact resistance. In total we’re looking at 100~150mΩ added to the circuit, which will affect the flashlight circuit, if the drop is enough to decrease it below the Vf of the LED (for a linear driver such as a S2+) then the reading will be false, which nearly always happens in high modes.

Conclusion a DMM is not the right tool for the job, no matter how expensive or accurate it is.
A Clamp is less accurate (2% +5 digits for the Fluke 325) but doesn’t affect the circuit, it’s a good tool for the job.

11

Well, it would be good to know the internal resistance of the meter, so one could calculate a compensation, if the DMM does not do that already.
As said, with nonlinear electronics, in-line current readings become more of an estimate, while they are quite OK on purely resistive loads.
The inductive clamp measures the distortion caused by the current in a known alternating weak magnetic field and calculates the current number (short explanation, see literature for more).
That’s why it may take a while to become accurate in DC mode and it will need nulling first on the spot, to take out environmental fields from the equation.
So, one single good Clamp for current and one good DMM for the rest would be enough.

And yeah, normal probe cables will have too high resistance for exact in-line current reading. They are not meant to be used that way, so use some lab cables with 4mm² instead for measuring current…

12

Interesting.

@thefreeman

That's not good then, and pretty sad. What would be the best solution do you think?

A desktop multimeter wouldn't have this drop?

@flashburn

What would your advice be, seeing these results?

13

A low value precision current shunt with a precision DMM. That way the voltage drop is very low. I just ordered a board to use These 0.1% Kelvin current shunts to be able to do precise efficiency measurements.
But even then, for a direct drive light, a 5mΩ shunt (+wire) will still affect the current.

14

Once a DMM is calibrated (which can mean exchanging or laser-trimming resistors) , it does not really matter which brand it is.

As I’ve written up there, probe cables are never meant for in-line current probing, at least in higher than a few mA range, because they are too thin and have a too high resistance.

For this, use 4mm² lab cables instead.

15

About the low modes, they are also all over the places because it’s a PWM driver, it’s either on or off at full current, it’s even worse now because on a lower range the current shunt has a higher value. With a constant current non PWM driver you should normally get closer readings.

16

It would be easy to make a calibration check yourself, get a battery or dc power supply and a resistor. You then will momentarily connect the resistor in a circuit as the load.

Measure the voltage of the source.
Measure the resistance of the load circuit.
Calculate the current using Ohms law.
Connect the circuit and Measure the current using the various meters/current probes, and compare to the calculated value.

Change the resistor value to check a different current level and repeat.

[edit] another approach is to purchase a dc current shunt in the range that is desired and use that to check/calibrate the meter performance.

Only when you have confidence in reading fixed known values can you use your tools to investigate an unknown.

i have an old 77 and trust it for readings, but i will do a quick and simple little check such as this as a sanity check if there is any question or puzzle in the current.

17

Hi

a fluke clamb meter and a cheapo yellow unbranded.

I trust both,cause they are showing the same :smiley:

But a clamp meter is the best choice.

Regards Xandre

18

Kennybobby, but as thefreeman said, we’re affecting what we measure in a way that depends on driver and LED. Calibration is not possible with just the DMM in focus, only the error can be reduced by limiting the resistance. I find the idea to measure the voltage drop over a small resistor intriguing. But he might not have something like an Aneng AN8008 in mind when mentioning a precise DMM :weary: .

Oh well, just bought a clamb meter.

19

hahaha.. Unheard.. sorry about letting you buy a clamp meter!
Which one did you buy?

20

a word about accuracy

from where I sit, with no experience measuring with a clamp meter… all those measurements are accurate enough, and the variation between them is not significant, to me.

I think some people are overly invested in the idea of “accuracy” when in fact there are variations and differences in the measurements, that are due to variations not only in the tool doing the measuring, but also in the specific individual light being tested.

I do not expect accuracy to mean identical, when comparing data from different sources.