Author Topic: 4 wire & 2 wire resistance tests, various cheap Kelvin clips, various meters  (Read 2435 times)

0 Members and 1 Guest are viewing this topic.

Offline Electro Fan

  • Super Contributor
  • ***
  • Posts: 2527
After reading lots of posts about the DE-5000 and it's 4 wire alligator leads (vs. modified longer 4 wire leads which I haven't got around to making yet) vs various Kelvin clips vs 2 wire vs other LCR and some ESR meters, below are some tests of a 1 ohm Dale resistor.

Preliminary conclusion: 
While 4 wire is (potentially/probably) superior to 2 wire (although with null you can make a case for 2 wire), my results don’t seem to be highly repeatable.  Might be temperature changes, air movement, and very possibly how exactly the clips are connected, and maybe even how well the banana plugs are seated in the meter.  Might also be that I have lousy Kelvin leads.
Net, net:  nothing definitive, it's hard to get repeatable tests results; I have one more hopefully better set of Kelvin leads on the way.

Measurements in ohms:

1.021 DE-5500 with 4 wire standard alligator clips at 1kHz, 1.02 at 100Hz
1.022 DE-5500 with 4 wire standard alligator clips at 1kHz, 1.02 at 100Hz 2nd test

33465A setup - let connections for each test settle for 10 seconds, then started 3 min test and took average measurement.

Kelvin leads used with 33465A
Set #1 Black leads into one handle on each clip - most confidence, but it's mostly a hunch so far, nothing definitive
Set #2 Grey leads into one handle on each clip - least (not much) confidence
Set #3 Red leads into both handles, Black leads into both handles - confidence TBD

1.02010 33465A 4 wire Kelvin clips Set #1 (Black leads into one handle on each clip) 1st test (this test result is suspect due to next 3 tests getting very different results - but maybe it should be the best result since it is close to the DE-5000 results, in which case all 3 sets of cheap Kelvin leads are suspect)
1.01473 33465A 4 wire Kelvin clips Set #1 2nd test
1.01586 33465A 4 wire Kelvin clips Set #1 3rd test
1.01501 33465A 4 wire Kelvin clips Set #1 4th test

1.01697 33465A 4 wire Kelvin clips Set #2 (Gray leads into one handle on each clip) 1st test
1.01215 33465A 4 wire Kelvin clips Set #2 2nd test

1.01474 33465A 4 wire Kelvin clips Set #3 (Red leads into both handles, Black leads into both handles) 1st test
1.02210 33465A 4 wire Kelvin clips Set #3 2nd test

1.01579 33465A 4 wire (4 individual leads) 1st test
1.01048 33465A 4 wire (4 individual leads) 2nd test

1.06371 33465A 2 wire

The following handheld meter tests settled much faster than the 33465A.

Fluke 179 1.1
Aneng 8008 1.10
Radio Shack 22-812 1.0-1.1 (flutters somewhat)

 

Offline bob91343

  • Super Contributor
  • ***
  • Posts: 1138
  • Country: us
As has been variously said, dissimilar metal interfaces and the details of the connections are important for such precise measurements.  Measurements using dc are susceptible.  Measurements using ac may cancel the thermal sources but are subject to skin effect, depending on frequency.
 

Offline srb1954

  • Regular Contributor
  • *
  • Posts: 192
  • Country: nz
  • Retired Electronics Design Engineer
As has been variously said, dissimilar metal interfaces and the details of the connections are important for such precise measurements.  Measurements using dc are susceptible.  Measurements using ac may cancel the thermal sources but are subject to skin effect, depending on frequency.
Kelvin measurements using AC are also susceptible to errors due to mutal inductance and capacitive coupling between the sense and drive leads.
 

Offline WattsThat

  • Frequent Contributor
  • **
  • Posts: 306
  • Country: us
Lead sets 1 & 2 are a complete :palm: and useless for 4 wire measurements since they aren’t 4 wire devices.

Set 3 is correct and the connectors need to be oriented vertically in the meter.

With the same lead set and a warmed up meter: Connect, stabilize, measure, disconnect. Repeat.

Now what do the results tell you?
 

Offline Electro Fan

  • Super Contributor
  • ***
  • Posts: 2527
Lead sets 1 & 2 are a complete :palm: and useless for 4 wire measurements since they aren’t 4 wire devices.

The routing of the four wires is maybe not as important as where they are connected to the clip jaws, no?

For example:
https://www.testequity.com/product/Keithley-5806-Kelvin-Clip-Leads-19218-1
 

Offline MosherIV

  • Super Contributor
  • ***
  • Posts: 1526
  • Country: gb
Quote
The routing of the four wires is maybe not as important as where they are connected to the clip jaws, no?

Yes. The poster is uncertain if the first 2 are connected correctly due to the way the wires enter the kelvin clamp. You would have to inspect or test if each wire goes only to one side of the clamp.

With reputable manufacturers like Keithley, they will have connected it correctly.

As you are finding measuring at these low values is quite difficult.
 
The following users thanked this post: Electro Fan

Offline GreyWoolfe

  • Super Contributor
  • ***
  • Posts: 3447
  • Country: us
  • NW0LF
I did the mod to my DER EE 5000 from leads I bought off of evilbay.  I found on one clip that even though the wires were attached to each half of the clip, there was still a short between the 2.  A little bit of work and that fixed the issue.  I just have to remember to do the calibration when I want to use them.  When I made my 4 wire DMM leads, I bought the clips and the retractable banana plugs and I had silicone wire already and they also work quite well.  They aren't $300 test leads but good enough for me.
"Heaven has been described as the place that once you get there all the dogs you ever loved run up to greet you."
 
The following users thanked this post: Electro Fan

Offline Electro Fan

  • Super Contributor
  • ***
  • Posts: 2527
Ok, I tested each lead for continuity out to the jaw end.  They were all wired properly and consistently but I took the time to mark with tape each related banana plug and jaw and clearly designated a force side vs a sense side for full testing consistency.

In the process I observed the construction a little more closely as well as paid more attention for intermittent behavior during the continuity tests.  It is confirmed that the Kelvin clip Set #2 is the least desirable of the bunch.  I had a hunch that might be the case given the very cheap banana plugs.

I'm going to test Sets #1 and #3 again.

One thing that I'm pretty sure of is that my presence near the equipment and DUT (within 3-4 feet) will impact the temperature (by nearly 1 degree F) which will impact the resistance measurements (I think by ~.393 ohm per degree C) but it seems like that should be a fairly consistent variable across each Kelvin lead test.

I don't know how far I want to with the volt nut, ohm nut, temp nut stuff but this is enough to make me appreciate whatever is being measured at Cern a lot more than I already thought I did. Seems like even just in the milli and barely micro range not much in the test procedure is enough to throw the results off in some direction.
 

Offline Electro Fan

  • Super Contributor
  • ***
  • Posts: 2527
I can get the DE-5500 and the 34465A to agree to within 1 milli ohm using plausible testing methods and data interpretation but it really depends on how the tests are run (duration plus temp control and more).  Eventually it depends somewhat on how you want to interpret the data.  I think 1 milliohm is about as good as it's going to get in my "lab" so I'm inclined to pick my favorite Kelvin leads, rely on those, and keep moving.

I just like it when I can get the same result consistently with a piece of test equipment and when two pieces of test equipment agree with each other.   :)
 

Offline SilverSolder

  • Super Contributor
  • ***
  • Posts: 3392
  • Country: 00
As has been variously said, dissimilar metal interfaces and the details of the connections are important for such precise measurements.  Measurements using dc are susceptible.  Measurements using ac may cancel the thermal sources but are subject to skin effect, depending on frequency.
Kelvin measurements using AC are also susceptible to errors due to mutal inductance and capacitive coupling between the sense and drive leads.

If you do it at e.g. 400Hz, those effects would be almost unmeasurable, for low/medium resistance values at least?
 

Online precaud

  • Frequent Contributor
  • **
  • Posts: 555
  • Country: us
    • LinearZ
Re: 4 wire & 2 wire resistance tests, various cheap Kelvin clips, various meters
« Reply #10 on: September 12, 2020, 03:34:17 pm »
I think 1 milliohm is about as good as it's going to get in my "lab" so I'm inclined to pick my favorite Kelvin leads, rely on those, and keep moving.

That has been my experience as well; reliable 1mOhm resolution that that fits well in my workflow is "good enough", and sufficient for my needs most of the time. Although, it is nice to have an extra displayed digit to see if the connection is trending one way or another...
 

Offline srb1954

  • Regular Contributor
  • *
  • Posts: 192
  • Country: nz
  • Retired Electronics Design Engineer
Re: 4 wire & 2 wire resistance tests, various cheap Kelvin clips, various meters
« Reply #11 on: September 13, 2020, 03:31:15 am »
As has been variously said, dissimilar metal interfaces and the details of the connections are important for such precise measurements.  Measurements using dc are susceptible.  Measurements using ac may cancel the thermal sources but are subject to skin effect, depending on frequency.
Kelvin measurements using AC are also susceptible to errors due to mutal inductance and capacitive coupling between the sense and drive leads.

If you do it at e.g. 400Hz, those effects would be almost unmeasurable, for low/medium resistance values at least?

The manual for the HP 4350A capacitance meter recommends twisting the sense leads to minimise coupling when measuring low-impedance components and that meter only uses a 120Hz  test signal.
[attachimg=1]
 
The following users thanked this post: Electro Fan, SilverSolder

Offline Electro Fan

  • Super Contributor
  • ***
  • Posts: 2527
Re: 4 wire & 2 wire resistance tests, various cheap Kelvin clips, various meters
« Reply #12 on: September 13, 2020, 03:51:33 am »
As has been variously said, dissimilar metal interfaces and the details of the connections are important for such precise measurements.  Measurements using dc are susceptible.  Measurements using ac may cancel the thermal sources but are subject to skin effect, depending on frequency.
Kelvin measurements using AC are also susceptible to errors due to mutal inductance and capacitive coupling between the sense and drive leads.

If you do it at e.g. 400Hz, those effects would be almost unmeasurable, for low/medium resistance values at least?

The manual for the HP 4350A capacitance meter recommends twisting the sense leads to minimise coupling when measuring low-impedance components and that meter only uses a 120Hz  test signal.
[attachimg=1]

That’s good info on twisting the sense leads.  Thanks
I haven’t been able to find documentation on the frequency used by the 34465A. 
 

Offline Shock

  • Super Contributor
  • ***
  • Posts: 3388
  • Country: au
Re: 4 wire & 2 wire resistance tests, various cheap Kelvin clips, various meters
« Reply #13 on: September 13, 2020, 05:25:24 am »
While 4 wire is (potentially/probably) superior to 2 wire (although with null you can make a case for 2 wire), my results don’t seem to be highly repeatable.  Might be temperature changes, air movement, and very possibly how exactly the clips are connected, and maybe even how well the banana plugs are seated in the meter.  Might also be that I have lousy Kelvin leads.

There is no question that Kelvin sensing measurements are superior for precision measurements. However as always the devil is in the detail, as you can probably deduce. The TL-21 interface has a 5th wire the guard connection as well.

I'm not familiar with the idiosyncrasy side of the DE5000s calibration and nulling but it's possible to use the meter to troubleshoot and improve on your measurement accuracy by solving deficiencies in your test leads and technique.

If you have not seen it check out Robrenz's video on the DE5000s repeatability. The other instrument he is using in the video is the LOM 510A which measures down to single digit micro ohms. If you find that interesting I recommend watching his other videos on the LOM 510A and making his various Kelvin leads. I'm not saying go to that much effort, more to illustratate how much they influence results.

« Last Edit: September 13, 2020, 05:28:16 am by Shock »
Soldering/Rework: Pace ADS200, Pace MBT350
Multimeters: Fluke 87V, 117, 27/FM       >>> WANTED STUFF <<<
Oszilloskopen: Lecroy 9314, Phillips PM3065, Tektronix 2215a, 314
 
The following users thanked this post: Electro Fan, wd5jfr, Simon_RL

Offline BravoV

  • Super Contributor
  • ***
  • Posts: 7107
  • Country: 00
  • +++ ATH1
Re: 4 wire & 2 wire resistance tests, various cheap Kelvin clips, various meters
« Reply #14 on: September 13, 2020, 05:43:01 am »
Sharing my experiences with various Kelvin clips.

The one used by the OP is my least favourite one, as their jaw biting power is not strong when measuring big DUT with big terminals like power resistor, big inductor and etc. Also the plastic part tends to get brittle overtime.

Better one is the one made by GWInstek, as its has better precision mechanic and built.



We have member reviewed it while ago -> GWInstek Kelvin clips review

For example:
https://www.testequity.com/product/Keithley-5806-Kelvin-Clip-Leads-19218-1

This model is my fav and mostly grabbed clips as it has powerful jaw that made measurement process not fiddly at big DUTs or big leaded components, especially when we moved alot and sometimes bumped on the clips or even the wires can fluctuate the reading as the teeth moved abit. Mine sourced from Franky's shop years ago, and the metal and the clips teeth are quite strong and wont chipped easily when clipped on hard metal terminals. Also it has very simple mechanism is almost fail and fool proof, and the jaw biting power can be enhanced too if needed.

« Last Edit: September 13, 2020, 05:46:35 am by BravoV »
 
The following users thanked this post: Electro Fan

Online Wytnucls

  • Super Contributor
  • ***
  • Posts: 2962
  • Country: be
Re: 4 wire & 2 wire resistance tests, various cheap Kelvin clips, various meters
« Reply #15 on: September 13, 2020, 07:26:55 am »
3 sources of error with Kelvin measurements:

Ohmic contact: Use gold plated probe tips
Thermoelectric EMF offset: Reverse current flow and take the mean of measurements
Heating (I^2R): Lower current source (Range) and reduce measurement time (Trigger).

1 ohm resistor 4W measurement accuracy (1 year):
Keithley 2000: 4mohm
Rigol DM3058E: 6mohm

Both use 1mA test current on their lowest range



« Last Edit: September 13, 2020, 09:55:43 am by Wytnucls »
 
The following users thanked this post: Electro Fan

Offline Electro Fan

  • Super Contributor
  • ***
  • Posts: 2527
Re: 4 wire & 2 wire resistance tests, various cheap Kelvin clips, various meters
« Reply #16 on: September 13, 2020, 02:00:33 pm »
Yep, I’ve watched Robrenz's Kelvin videos several times.  He is in the low resistance measuring HOF (Hall of Fame), and the precision machining HOF, and the beautiful restoration HOF.
 

Offline SilverSolder

  • Super Contributor
  • ***
  • Posts: 3392
  • Country: 00
Re: 4 wire & 2 wire resistance tests, various cheap Kelvin clips, various meters
« Reply #17 on: September 13, 2020, 02:17:03 pm »

I don't understand why nobody makes those super cool Kelvin hooks as a commercial product...   Such an elegant concept!
 

Offline Electro Fan

  • Super Contributor
  • ***
  • Posts: 2527
Re: 4 wire & 2 wire resistance tests, various cheap Kelvin clips, various meters
« Reply #18 on: September 13, 2020, 03:13:52 pm »

I don't understand why nobody makes those super cool Kelvin hooks as a commercial product...   Such an elegant concept!

+1
Seems like Pomona would be a natural to sell these; maybe they should hire Robrenz to over see the design, manufacturing, and QA testing, and the marketing  :)

Not sure how big the market is 🤔 but if someone is paying $300 or more for a set of Kelvin leads there must be a fair amount of perceived value.
 

Offline Electro Fan

  • Super Contributor
  • ***
  • Posts: 2527
Re: 4 wire & 2 wire resistance tests, various cheap Kelvin clips, various meters
« Reply #19 on: September 13, 2020, 06:55:23 pm »
Big Disclaimer/Request:  no laughing....  :-DD

No doubt low value resistance measurements have a lot of potential variables including finding the right Kelvin clip test leads, and much more (temp, materials, settling, etc, etc).   But.... back to basics:

Given two 4 wire measurements, one showing .0092 ohms and one showing .0097 ohms, is the difference within the specs for the 34465A?  (I'm having a difficult time interpreting the specs.  Thx)

https://www.keysight.com/us/en/assets/7018-03846/data-sheets/5991-1983.pdf

Specifications 34465A
34461A accuracy specifications (pdf says it's for 34461A but I think it's a typo and is for 34465A as well as the 34461A and 34470A):
± (% of reading + % of range)

Relevant Spec?
100 Ω 0.0030 + 0.0030
100 Ω is the range?

So, if we start with .0092 ohms
then we multiply .0092 x .0030% and we get 0.000000276 ohms (not much?)
then we multiply .0030% x 100 ohms (the range) and we get .003 ohms (this is the big one?)
then we add it all up:

.000000276 + .0039476 = .003947876

This would indicate that for .0092 ohms the error could be +/- .003947876 ohms

If that is correct (and I'm betting it's not) that would indicate there is no chance of accurately distinguishing between .0092 and .0097 ohms.

Ok, feel free to issue my F grade for interpreting the specs.   :palm:

Thanks
 

Offline Electro Fan

  • Super Contributor
  • ***
  • Posts: 2527
Re: 4 wire & 2 wire resistance tests, various cheap Kelvin clips, various meters
« Reply #20 on: September 13, 2020, 11:32:40 pm »
Still hoping that someone(s) will comment on my attempt to interpret the 34465A spec above but in the meantime I might have made some progress.

In the first photo is a measurement showing an average of .00046 ohms - this is just the two Kelvin clip/leads hooked to each other.  If I am interpreting this correctly it is showing that the the combined 4 wires (two of which are sense) and two clips have a total resistance of about half a milliohm (~460 microohms).

This I think this helps explain (maybe partially) why I have not been able to get the consistent 9.2 - 9.3 ohm measurement  that the theory says I should be getting for my reference 45cm 18 gauge copper wire.  With the better (best of the bunch) set of Kelvin clips I was getting about 9.7 - 9.8 milliohms and I think that part of the reason it was too high was that I wasn't accounting for the roughly 0.46 milliohms in the Kelvin clips/leads.

After applying the Offset Compensation with the 34465A the Kelvin clip leads shows -0.00004 which I think is minus? 40 micro ohms - which is still not zero but I can believe we are reaching the threshold of what the meter can do, which I think is probably significantly better than my interpretation of the spec above.  Or maybe my spec interpretation was correct and the meter just exceeds the specs?

Still lots of unknowns to be addressed and learnings to be made.

Looking forward to constructive critiques.... Thx!

 

Online 2N3055

  • Super Contributor
  • ***
  • Posts: 3001
  • Country: hr
Re: 4 wire & 2 wire resistance tests, various cheap Kelvin clips, various meters
« Reply #21 on: September 14, 2020, 07:04:47 am »
Big Disclaimer/Request:  no laughing....  :-DD

No doubt low value resistance measurements have a lot of potential variables including finding the right Kelvin clip test leads, and much more (temp, materials, settling, etc, etc).   But.... back to basics:

Given two 4 wire measurements, one showing .0092 ohms and one showing .0097 ohms, is the difference within the specs for the 34465A?  (I'm having a difficult time interpreting the specs.  Thx)

https://www.keysight.com/us/en/assets/7018-03846/data-sheets/5991-1983.pdf

Specifications 34465A
34461A accuracy specifications (pdf says it's for 34461A but I think it's a typo and is for 34465A as well as the 34461A and 34470A):
± (% of reading + % of range)

Relevant Spec?
100 Ω 0.0030 + 0.0030
100 Ω is the range?

So, if we start with .0092 ohms
then we multiply .0092 x .0030% and we get 0.000000276 ohms (not much?)
then we multiply .0030% x 100 ohms (the range) and we get .003 ohms (this is the big one?)
then we add it all up:

.000000276 + .0039476 = .003947876

This would indicate that for .0092 ohms the error could be +/- .003947876 ohms

If that is correct (and I'm betting it's not) that would indicate there is no chance of accurately distinguishing between .0092 and .0097 ohms.

Ok, feel free to issue my F grade for interpreting the specs.   :palm:

Thanks

The math seems correct, but I would use 1 year specs for it, which are 100 Ω : 0.0060% of value + 0.0040% of range ( unless your instrument was calibrated in last 24 hours..):

(0,0092*60e-6)+(100*40e-6) =0,000000552+ 0,004= 0,004000552 which gives 0,0092 +/- 0,004000552 Ohm as your interval..

 
The following users thanked this post: Electro Fan

Offline SilverSolder

  • Super Contributor
  • ***
  • Posts: 3392
  • Country: 00
Re: 4 wire & 2 wire resistance tests, various cheap Kelvin clips, various meters
« Reply #22 on: September 14, 2020, 04:17:42 pm »

I would expect the repeatability to be much better than those specs, if the measurements are made within a few minutes of each other?

 

Offline tooki

  • Super Contributor
  • ***
  • Posts: 5640
  • Country: ch
Re: 4 wire & 2 wire resistance tests, various cheap Kelvin clips, various meters
« Reply #23 on: September 14, 2020, 04:42:22 pm »
Lead sets 1 & 2 are a complete :palm: and useless for 4 wire measurements since they aren’t 4 wire devices.
Yes, they are, and it's even visible in the photos, where you can see the second wire folding back near the vertex of the clips. It's very, very common to wire them that way.
 

Online 2N3055

  • Super Contributor
  • ***
  • Posts: 3001
  • Country: hr
Re: 4 wire & 2 wire resistance tests, various cheap Kelvin clips, various meters
« Reply #24 on: September 14, 2020, 05:00:28 pm »

I would expect the repeatability to be much better than those specs, if the measurements are made within a few minutes of each other?
That is absolute measurement accuracy. Short term it should be well qirhin 24h specs for relative measurements.
 
The following users thanked this post: SilverSolder


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf