Anybody with siglent multimeter ?

[tonycstech]

I have  SDM3045X desktop multimeter from siglent and i am having misunderstanding about its reading of resistance.

I have a circuit that has 900 ohms.
In automatic mode, it will set its self to 600 ohm range and measure 500 ohms for what ever reason.
If i manually set range to 6K, it will still give me 500 ohms reading.
If i set it to 60K range, it will give me correct 900 ohms.

Every multimeter i have, digital or analog, all read 900 ohms manual or auto.
This one reads 500 unless i set range at 60K or higher.

What am i measuring ?
I dont want to get into this rabbit hole because it shouldn't matter when every other multimeter measures this correctly.
But in case you are one of those, i am measuring resistance of a circuit containing buck converter.

[bdunham7]

Your circuit is interacting with the current source of the meter and doing....something?  Is there an inductor in the part of the circuit you are measuring?  Or a junction (diode or other) that could be conducting?

[tonycstech]

There is a coil.
I am measuring video card main power phases.

I contacted tech support they are not very helpful and i am trying to get an RMA but it takes almost a week to get a reply and does not seem to go anywhere for weeks now.

[bdunham7]

I contacted tech support they are not very helpful and i am trying to get an RMA but it takes almost a week to get a reply and does not seem to go anywhere for weeks now.

I doubt that the meter is 'broken', so an RMA will likely just result in more confusion.  And even if it is defective, getting them to do anything about it will require an explicit explanation of what is wrong.  What you refer to as going down the rabbit hole is just understanding what DMMs do when they measure resistance.  The fact that this one appears to give you what you think is an incorrect result while others are correct may be a design quirk or defect--or it may simply be the result of different meters using different test stimuli.  I have a sneaking suspicion from other discussions that your meter has a combination of both due to the way ranges are set up.

What you need to do first is put a high-impedance voltmeter in parallel with the DMM that is testing resistance.  If you don't have any high-impedance voltmeters, a 10M input (most handhelds) will be fine at this low of a resistance.  Try measuring the voltage across the test resistor during the resistance test with the Siglent and then at least two of the meters that give you correct results.  You can also confirm the test current by connecting the Siglent to another DMM set to measure (small) current and then measuring the resulting current when you manually select each resistance range on the Siglent.  Post your results and we'll go from there.

[Grandchuck]

Do you have ordinary resistors available?  If so, check them with your meters.

[tonycstech]

measuring plain resistor does show expected result.
Measuring 1K resistor, auto range sets to 6K
Connecting handheld while to the leads while its measuring 1K resistor shows around 1v
So i am guessing 1v=1Kohms
With 2K resistor, it shows 2v

[bdunham7]

measuring plain resistor does show expected result.
Measuring 1K resistor, auto range sets to 6K
Connecting handheld while to the leads while its measuring 1K resistor shows around 1v
So i am guessing 1v=1Kohms
With 2K resistor, it shows 2v

Now try measuring your original problem circuit the same way with the 6K range manually selected.  Then do the same using one of the meters that reads the 900 ohms 'correctly'.

And can you try a 3K or so resistor in in the 6K range as well?  According to the datasheet, the test current for the 6K range should be 0.1mA, but your measurement shows it to be 1.0mA.  That may change above 2K.

[tonycstech]

6k manually selected shows same thing.

I dont have that exact circuit but i found a similar one that shows me 820 ohms instead of 935.

If i change the range to 60K, reading is 935 as expected.

Voltage 0.093v when range is set to 60K and resistance is 935 ohms
Voltage is  0.82v when measuring in 6K and resistance is 820 Ohms

[bdunham7]

Voltage 0.093v when range is set to 60K and resistance is 935 ohms
Voltage is  0.82v when measuring in 6K and resistance is 820 Ohms

One explanation for those numbers is that your circuit is not a resistor and does not obey ohms law under those test conditions.  The first reading is consistent with a test current of 0.1mA and the second with 1.0mA, but with the circuit having a lower apparent resistance with the higher voltage.  A semiconductor component or junction that is starting to conduct might be the cause for that. 

Try the rest of what I suggested to confirm this.  Your other meters may have lower test currents and thus lower indicated voltages.

[tonycstech]

What do you want me to do again ?
Auto range selects 6K range and i already gave you the values for that.
It does not have 1K range, only 600,6k,60k,600k, etc

Yes, there are a ton of semi conductors in the circuit. The reason why i bought this meter is because i needed some low resistance measurements that my handhelds cannot detect.
Didnt expect to get wild results everywhere else. Even the low resistance is wacky. Its either 0 ohms or something wild allover the place.

I already packaged the unit for RMA, hopefully they can figure out whats causing this issue and maybe introduce a firmware update with some kind of a mode switch that apparently handheld meters dont need 

[bdunham7]

What do you want me to do again ?

With a parallel meter in the same manner with that same circuit, measure the test voltages of one of your handheld DMMs that you think is giving you correct readings.  It would help to know which model handhelds you have.  If the handheld uses a lower current and voltage, that is the explanation for what you are seeing.

I already packaged the unit for RMA, hopefully they can figure out whats causing this issue and maybe introduce a firmware update with some kind of a mode switch that apparently handheld meters dont need 

Unfortunately I think your frustration will continue when they tell you there's nothing wrong with the meter.  In-circuit testing like that is very tricky and low resistance measurement can be even more so.

[tonycstech]

Yes, handheld meters i used used maybe 150mv or even less to measure same circuit, while this one used almost 1v.
I used etekcity  msr-a1000

If 6K shows wrong value and 60K shows correct value, pretty sure something is wrong with the meter.
Maybe they need to make an option to use lower voltage measurement. 1v to 1k ohms 1:1 ratio was easiest to implement i guess.

[bdunham7]

If 6K shows wrong value and 60K shows correct value, pretty sure something is wrong with the meter.

Not really.  It just shows that you aren't measuring an actual resistor. 

Maybe they need to make an option to use lower voltage measurement. 1v to 1k ohms 1:1 ratio was easiest to implement i guess.

Bench meters often use higher test currents because that results in more precision--for true resistors anyway.  Some very high end models actually do have selectable test power levels.  Otherwise, the solution is often to use a higher range, although the specific ranges on your meter probably make that a bit awkward.

[tonycstech]

That sucks. I guess instead of investing in a desktop meter, i'd be better off buying a milliohm meter instead.

[bdunham7]

That sucks. I guess instead of investing in a desktop meter, i'd be better off buying a milliohm meter instead.

Maybe.  Milliohm meters typically use even higher test currents and sometimes even AC.  What are you typically trying to do?  Very low resistance or several hundred ohms like in this example?

[tonycstech]

Multipurpose.
I need to measure very low resistance as well as up to 1 megaohm.
This junk doesn't seem to be able to do it, while real junk can. 🤔
Makes me question my purchase choices.

I need to be able to see resistance below 1 ohm fairly accurately, not super perfect but 10% error is ok. My handheld can't see anything lower then 1 ohm it seems, unless pcb is ice cold. Room temperature reading doesn't work.

I already shipped siglent for a factory evaluation. If they don't find anything wrong with it, I'll put it up for sale just to get my money back. I got it brand new very cheap and it still has only 20 30 startups.

[bdunham7]

This junk doesn't seem to be able to do it, while real junk can. 🤔
Makes me question my purchase choices.

That happens sometimes.  Your evaluation of 'good' and 'junk' is based on how well the tool's characteristics match your needs.   Expensive instruments aren't magic.

I need to be able to see resistance below 1 ohm fairly accurately, not super perfect but 10% error is ok. My handheld can't see anything lower then 1 ohm it seems, unless pcb is ice cold. Room temperature reading doesn't work.

To do low-ohms well, you need a 4-wire setup, or something close to it--and then you still have to worry about offsets due to thermal voltages.  The SDM3045X nominally has 4W ohms, but since the lowest range is 600R, I'm not sure how great it will be.  But measuring down to 100 milliohms with a 10% error might be doable with careful lead setup. 

[bateau020]

Voltage 0.093v when range is set to 60K and resistance is 935 ohms
Voltage is  0.82v when measuring in 6K and resistance is 820 Ohms

You were trying to measure in-circuit. When doing so, it is almost always needed to keep the test voltage below 0.5V, and better to keep it below 0.2V, as many types of semiconductors will start to conduct above those voltages.
You went way above 0.2V in the last measurement. So it is likely that you saw some semiconductors with your measurement, and also likely that the meter is technically OK. It is just that that specific range is not fit for this particular use case.
The meter uses a rather high current at that range, which is bad for in-circuit testing, but good for out of circuit milli-ohm testing. You must be aware of the specifications of your tools and adapt the settings for how and what you test.

As explained above, some meters have a "low power" mode for resistance measurement, which might help you for in-circuit testing.
I looks however that the SDM 3045 does not have such a mode (the SDM 3055 doesn't, just extrapolating here). But even if you have that "low power" mode, in-circuit testing is very tricky, as coils and caps can easily influence the results as well. Just imagine a low test current, and you are testing over a big cap. That will suck up all the current, and again you will end up with a value that is too low. Not the meter's fault, just a side effect of how you measure.

[tonycstech]

Correct. There is a ton of semi conductors.
Your explanation makes perfect sense and i did not know any better.