EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: abdulbadii on September 10, 2021, 11:48:29 pm
-
What would be the most accurate low, ~ 50 mili - 15, ohm meter possible either as finished product or as components to be assembled on PCB (shown with its circuit diagram)?
-
No matter what you desire, it will all come down to how much money do intend to spend?
Then, perhaps after answering twenty to thirty additional questions, you might start converging on an answer to your question.
-
An ohmmeter is a voltmeter along with a precision current source. Voltmeters are a known quantity, we all sort of know how low and precise you can reasonably go without too much noise, instability or cost. See TiN's nanovoltmeter thread for a good idea what the ultimate limitations are. The primary issue with measuring low resistances is how much current can you reasonably put through it? So the answer is....it depends.
-
When you say 'best' what do you mean? The best possible, or within some limitations? At NIST they probably have nearly the best means of measuring resistance. Commercial instruments such as the HP 3458A and its successors are also very good.
For accurate measurement of such low resistance, a 4 wire measurement and controlled environment are mandatory. Very few resistance alloys have zero temperature coefficient of resistance, so the temperature must be rigidly controlled for consistent results.
Contact resistance is a problem, as is the exact location of the test connections. Thermal voltages are generated, as any two metals (unless identical in composition) form a thermocouple that interferes with measurement.
-
For this, you need to designate a budget, and accuracy requirement. I do own a few milliohm meters. Even some bench meters do a good job.
What I own.
LOM-510 x2, 1 works, 1 needs work.
HP 34420A x2 1working, 1 has problems
Ballantine 3205b works.
-
V/I=R
1mV/100mA=10mOhm
So 50mOhm should be trivial to reach even with the most basic test equipment.
-
I don't think so. Read my post above about the specific issues. Plus, now you need to measure the current and voltage to the desired accuracy.
Of course, if 10% is close enough, it gets simpler.
-
As mentioned "most accurate" depends on other factors like budget other limitations, as well as the ability to calibrate it.
For a simple milliohm measuring device you could start with the Shorty design discussed elsewhere on this forum and at https://hackaday.io/project/3635-shorty-short-circuit-finder
-
Yeah, I am not trying to say toss your nice milliohm meter or that sub-1mOhm won't require some attentiveness. I'm just presenting the idea that maybe the OP doesn't need to buy something to accomplish what they are looking for.
Quick note, the DE-5000 is pretty good, typically within 1-2mOhm.
Anyway, I ran a few tests with some 1% 0.1Ohm resistors:
1 resistor: 101mOhms typical (99-102 via various equipment/methods)
2 resistors: 51mOhms (calculated value is 50.5mOhms)
4 resistors: 26mOhms (calculated value is 25.25mOhms)
This is with a $50 power supply and a $50 DMM (gave away my cheaper DMMs a while ago).
I did set up my 6.5 digit bench DMMs where I measured the voltage and current to a couple more decimal places and the above results were basically spot on, at least to the resolution displayed by the lower-end equipment. I also experimented with 10mA instead of 100mA and had excellent results. Of course nothing is preventing you from using 50mA or 500mA or etc...
-
I did set up my 6.5 digit bench DMMs where I measured the voltage and current to a couple more decimal places and the above results were basically spot on, at least to the resolution displayed by the lower-end equipment. I also experimented with 10mA instead of 100mA and had excellent results. Of course nothing is preventing you from using 50mA or 500mA or etc...
You're going to need to pay special attention to the heating effect of pushing that much current through the resistor you're testing though.
-
I as well as a few others have LOM510A microohm meters, picked up for a few hundred secondhand, they are over $2000 new.
If you want a cheap milliohm resolution something like a DER EE DE5000 is good enough for most people as well as being a fully featured LCR meter. Check this Robrenz video (a forum member) showing them. The DE5000 looks fairly repeatable milliohms in ESR mode, in DCR mode 0.01 ohms resolution (so one digit less). It's around $100 brand new.
https://www.youtube.com/watch?v=D111PZqtAis (https://www.youtube.com/watch?v=D111PZqtAis)
-
the DE5000 is a nice piece of measuring device.
for measuring short circuits, I prefer my modified shorty where I added a display.
the original shorty doesn't measure anything, it's only an audio frequency changing with low ohm.
you can't tell if it's 10 or 20mΩ with it, difficult to hear the difference.
with the display, I can detect 0.2mΩ difference, but it is not a precision instrument so the repetability of the measure is not very precise.
http://kripton2035.free.fr/Projects/shorty-display.html (http://kripton2035.free.fr/Projects/shorty-display.html)
[attachimg=1]
(https://kripton2035.free.fr/Resources/img0697.jpeg)
(http://kripton2035.free.fr/Resources/img0697.jpeg)
-
I did set up my 6.5 digit bench DMMs where I measured the voltage and current to a couple more decimal places and the above results were basically spot on, at least to the resolution displayed by the lower-end equipment. I also experimented with 10mA instead of 100mA and had excellent results. Of course nothing is preventing you from using 50mA or 500mA or etc...
You're going to need to pay special attention to the heating effect of pushing that much current through the resistor you're testing though.
I'm not seeing that in the math, what are you seeing?
-
I have two that do the job very well
Cambridge Technology, Model 510 Micro-Ohmmeter
Keysight 34420A
-
I have a Hioki RM3545 resistance meter, 0.006% basic accuracy, 0.01 μΩ max resolution, 1A max testing current. I think it's designed for rapid testing on the production line, for laboratory use, Keysight 34420A is a better choice.
https://www.hioki.com/global/products/resistance-meters/resistance/id_5876 (https://www.hioki.com/global/products/resistance-meters/resistance/id_5876)
https://youtu.be/TPB8mKlJwUc (https://youtu.be/TPB8mKlJwUc)
-
I did set up my 6.5 digit bench DMMs where I measured the voltage and current to a couple more decimal places and the above results were basically spot on, at least to the resolution displayed by the lower-end equipment. I also experimented with 10mA instead of 100mA and had excellent results. Of course nothing is preventing you from using 50mA or 500mA or etc...
You're going to need to pay special attention to the heating effect of pushing that much current through the resistor you're testing though.
I'm not seeing that in the math, what are you seeing?
Depends on the ppm/°c of the resistor I guess, but remember if you start heating the DUT you're also going to increase the thermal gradient between it and the probe tips.
Basically what I am saying is, it's preferable to use as low a current as possible, to reduce these factors. You have to juggle that with the lower voltage and therefore lower resolution you'll get with a lower test current.
-
Power dissipation in the resistor will depend on its value, package, etc. You need to do the math yourself to see if the temperature error is within allowable margin (+headroom for gradient effects)
Why are you using a weird font?
-
Yes, just pointing out that there is no one-size-fits-all solution, as much as anything.
It's the same font as yours, just different pt/color/b. Force of habit mostly.
-
In general, for good resolution at low resistance levels with a four-terminal measurement, low-frequency AC is preferred to DC, since you can avoid the DC offset of the voltmeter section using an AC-coupled amplifier and synchronous detection (as in a lock-in amplifier).
-
If you search for short circuits between vsupply and ground, using an AC ohmmeter, all the capacitors will appear as a short.( or a very low ohm value)
-
It’s a quantitative question. At 32 Hz (200 rad/s), 100 uF has a reactance of 50 ohms, but the phase-sensitive detector I mentioned will note the reactive phase angle. Far from a “short”.
-
If you search for short circuits between vsupply and ground, using an AC ohmmeter, all the capacitors will appear as a short.( or a very low ohm value)
Using a cheap VNA to look at the impedance of power distribution networks.
https://www.youtube.com/watch?v=Y8ouApeex78 (https://www.youtube.com/watch?v=Y8ouApeex78)
-
Yes, just pointing out that there is no one-size-fits-all solution, as much as anything.
It's the same font as yours, just different pt/color/b. Force of habit mostly.
Yep, I did the math before setting up my test, 0.001W with a 1W resistor.
I copied the test current for the resistance range from this thread since Vgkid mentioned it:
https://www.eevblog.com/forum/testgear/ballantine-3205b-microhm-meter/ (https://www.eevblog.com/forum/testgear/ballantine-3205b-microhm-meter/)
Using two 6.5 digit DMMs to take simultaneous readings, a large battery as a less noisy current source and some Kelvin clips I got some pretty amazing results.
-
Yes, just pointing out that there is no one-size-fits-all solution, as much as anything.
It's the same font as yours, just different pt/color/b. Force of habit mostly.
Why do you do that? All three of those attributes reduce the readability of the text a little bit. All three simultaneously make your posts noticeably more difficult to read. It adds nothing, since highlighting everything means you're highlighting nothing, but definitely takes away. It baffles me why anyone would go to the effort (on every single post!) to make their contributions harder to read. :-//