Video of Design and build of precision Kelvin banana jacks and header pins

[robrenz]

I needed a way to make consistent, low error, zero bulk resistance connections in designing and calibrating my decade resistance box design. Both 4mm banana and .025" header pins and sockets needed to be probed using 4 wire kelvin offset compensated resistance measurements. It took me a while to figure out how to design these to have zero bulk resistance and this is the design I came up with. (yeah, I know it probably already exists) This video covers the design, construction and testing of these connectors for Micro-Ohm measurements.

Video 


Dave Cad drawing of a semi Kelvin banana jack on top half of section view. The bottom half of the section view shows what I call a full Kelvin zero bulk resistance banana jack which is built and tested in the video.


Zero bulk resistance hermaphrodite Kelvin banana jacks built in the video.


Zero bulk resistance Kelvin male and female .025" round header pin and socket built in the video.

[HLA-27b]

Awesome engineering and craftsmanship!

This would probably be of interest to some R&D institutions that do low ohms work.  So it may be useful to let it be known that you have a few sets at hand should they need some. Mush cheaper and faster than having their own workshop produce something similar.

On the other hand, if the OEM probes that come with their gear is insufficient in some way, then they probably need some custom jig or fixture. I am sure they'd love to give the work to somebody who knows his stuff.

[saturation]

Magnificent work, robrenz, both electrical and esthetic.  You could sell these.  You coul poll and see how many are interested.

[robrenz]

Thanks guys,  with the current interest level in this I don't think a poll is necessary. Makes me wonder if it is worth the huge time and effort to document these things.

[ElektroQuark]

Very good work!
I think there are more people interested than those who have post here. There are a lot of lurkers.

[grenert]

I think there are few here who have any where near the skills to attempt to re-create your devices, and also few who have a true (read: occupational) need to purchase what are likely to be expensive (rightly so) accessories.  Hobbyists and beginners, who seem to be the biggest part of this forum, usually don't spill big $$$.  I do think that you would find more commercial interest in a different forum with a more deep-pocketed, experienced readership (some kind of metrology forum?).  Or maybe from the insane readership of the volt-nuts list.

What I enjoy about your work is the approach to the problem and amazing skill and creativity which you put into your work.  It reminds me of the guy who DIYed vacuum tubes.  I don't think he sells them, and I certainly don't think many went out and made some.  But the video sure is cool.

So, while I am doubtful that this forum will provide you a lot of financial benefit, you can at least know that you've taught and inspired us a lot

[robrenz]

Thanks grenert,  My main purpose is to inspire people that have limited mechanical background with seeing how the mechanical aspects of an idea can be accomplished. That is why I try to capture most all the steps that would not be familiar to someone with a mainly electronic background. (also because 99.9% of my knowledge is mechanical vs electronic) I am not trying to sell these or the kelvin clips but I will be selling the decade box shown.  There will be many extensive videos on the design and build of the decade box that will have a lot of hopefully educational content.

I am surprised there is no discussion on my statements about zero bulk resistance and contact resistance.  I thought the metrology fans would have some input.  What I state in the video (maybe too authoritatively) are just my opinions from my learning experience, not necessarily technically correct.

[SeanB]

I do like the attention to detail shown. Very interesting and probably you could make a niche market out of selling this to metrology labs, as you are more likely to get appreciated as a supplier of quality test leads that are very accurate and good quality. I would suspect the manufacturers of high quality equipment would be interested in this style of connector, and would most likely license them if they were going to make them.

[HLA-27b]

I am surprised there is no discussion on my statements about zero bulk resistance and contact resistance.  I thought the metrology fans would have some input.  What I state in the video (maybe too authoritatively) are just my opinions from my learning experience, not necessarily technically correct.

One thing which I was wondering about is the Hertzian contact stress on the springed contacts. Do you think that variation of contact stress might have caused the slight variations in resistance? In other words, at the very light contact pressures you are dealing with, a slight variation in contact pressure might cause a huge variation in contact area thus changing the resistance. This is pure speculation on my side of course.

OTOH it would be interesting to see for example the contact resistance of a ball bearing and a surface, varying with contact pressure.

[robrenz]

IMO the actual contact stress is not directly a factor, but is a corollary to the normal contact force which does directly influence the contact resistance.  Higher contact force does give lower contact resistance to a point. This is directly related to the study of tribology. Higher force causes contacting asperities of each surface to deform allowing more to come into contact and also deform/weld. Contact resistance will continue to decrease as peaks and valleys are smashed into each other and oxide layers are punctured. Increased force will start to elastically deform the bulk material of the contacting surfaces which will usually increase (depending on the contact geometry) the contact patch, also decreasing the resistance.  Mutiple contact fingers have thier contact resistances in parallel so for the same contact geometry and force per finger the total contact resistance will be inversely proportional to the number of fingers = more is better.

Intro to contact resistance http://www.google.com/url?sa=t&rct=j&q=&esrc=s&frm=1&source=web&cd=1&cad=rja&ved=0CB8QFjAA&url=http%3A%2F%2Fwww.swtest.org%2Fswtw_library%2F1998proc%2FPDF%2FS01_kister.PDF&ei=nelhUKXaGdGH0QG6iYCgAg&usg=AFQjCNGI4e9YsFIGNMKCxfdzjZihV5vyNg&sig2=v6bzxxmv_ROrl3AXm3YVgw

[HLA-27b]

My terminology is sloppy - fact.
I guess what I was trying to say is that spring forces seem too light for repeatable contact. Anyway, here is what was going though my mind as a way to improve repeatability of the contacts. Is this an actual  problem that needs solving? - I have no idea. Maybe I am just interfering (Sometimes I do that inadvertently).

[robrenz]

Wasn't trying to correct, and I never think of discussion or constructive criticism as interference. I was just making sure we were on the same page. 

The force to insert these pins is fairly high already.  These sockets are designed to handle .015 diameter to .025 diameter pins. Once you insert a .025 pin they will not hold a .015 pin anymore because the compliance (springback) value is not high enough. These values are in the Mil-Max catalog.  So that means the contact fingers actually yield in the process of accepting a .025 pin (from virgin condition). If you put a virgin pin into a virgin socket and then inspect the pin under a microscope you will see that the gold plated brass pin has four tracks where the pin has been mildly deformed by the contact fingers of the  socket. That is why I pre size the sockets with the hardened polished gage pin first in an attempt to burnish the gold contact zone instead of possibly smearing it and to pre yield the contact fingers for the .025 diameter. In speaking with Mill-Max engineering this technique and having the pin concentrically guided into the socket are key to having these last in high repetition applications (up to 50000 insertions). 

The repeatability is pretty good on these pins and sockets.  Its easy to forget we are taking about a 20 or 30 micro Ohms variation here.  The o-ring idea also crossed my mind while i was working on these but these decade boxes are already going to be fairly expensive so I don't want to add any customization that would make that worse.

Also easy to forget how small this is. That o-ring gap is only 0.008" (0.2mm) and the magenta colored sense leads are only 0.002" (0.05mm) thick.

[robrenz]

I sent information on this design to IET Labs engineering dept. to see if there is any interest.  I wonder if I will get any feedback.

[felixd]

Good luck with your design !

[quarks]

Great work!!!
Thanks a lot.

[cwalex]

amazing work. the precision machining you do is just awesome! don't be shy to make more videos even if they aren't electronics related.

[robrenz]

Thanks guys for the encouragement.  There are 2 others on here from another thread
https://www.eevblog.com/forum/projects-designs-and-technical-stuff/video-of-making-the-ultimate-kelvin-connection-(major-hacking-of-a-pomona-clip)/msg132596/#msg132596
Part 2 on the same thread https://www.eevblog.com/forum/projects-designs-and-technical-stuff/video-of-making-the-ultimate-kelvin-connection-(major-hacking-of-a-pomona-clip)/msg137130/#msg137130
more to come