BreakPoint
4-wire (Kelvin) resistance measurement
Cross-posted on LAVA here.
I just witnessed a 4-wire resistance measurement across the barrel of a metalized connector made with a standard set of alligator clip-like probes.
One operator makes this particular measurement with the clips only slightly opened, and the resistance exceeds the test limits. This triggered the Manufacturing Engineer to get involved and his habit is to make this measurement with the clips almost fully opened. The resistance, when doing it this way, is about 1/3 of the first operators method.
I tried to imagine why this might occur, but I'm stumped. I thought about the length of the current paths through the material and how the corresponding change in distance would increase the resistance when the probes are closer together, but this doesn't explain the 3x factor. This would also imply that turning one clip 180 degrees should result in a change in the opposite direction, but this isn't observed. I'll keep guessing.
Have you seen this? Can you explain it?
@Intaris wrote:
I don't think the resistance is measured between the two prongs of the clip but between the two clips. As such the pathways the current flows will be affected by the position of contact but the length of the circuit not.
That's right.
@johnsold wrote:
Jim,
That certainly is not a 4-wire measurement of the test object. It may use the 4-wire method of the instrument, but it is only compensating for the drop in the first ~15 cm of the lead wires. I would run 4 wires all the way to the clips. [...]
That's perhaps not the best picture. In fact, the four wires go all the way down to the clips and each half of the clips are electrically isolated from each other.
@Darin.K wrote:
I could throw around terms like Conformal Mapping and Poisson's equation, but let me just say that you are effectively measuring a sheet resistance which is strongly dependent on the location of the contacts. Look into the theory behind wafer probes or the van der Pauw method for more details. The upshot is that this type of test in a QC setting probably requires some simple fixturing.
Yeah! I get to push it back onto the Design Engineers. 😄 And to think that my oldest son, just a few weeks ago, told me that his least favorite subject was Physics. 😞
Darin.K's answer made me think of http://xkcd.com/356/ which, naturally, reminded me of http://xkcd.com/730/ (which more closely represents my situation).
There is heatshrink in two locations on each wire. Looking at the picture it looks like the sense lead is terminated at the first heatshrink on each lead. The sense lead really needs to go all the way to the clips. And not at the handle either but right down to where it touches.
@jcarmody wrote:
[...]
That's perhaps not the best picture. In fact, the four wires go all the way down to the clips and each half of the clips are electrically isolated from each other.
[...]
There are actually four, independant wires; each begins with a banana plug and end in a half-alligator clip. The only point of electrical contact is at the clips, when they're closed; opening the alligator's mouth opens the circuit.
