gold plated binding post lugs without the audiophoolery

I once worked in a lab where we did precision resistive measurements, down in the micro-ohm level. One very expensive bridge in particular came supplied with gold plated binding posts, and we always had trouble getting the contact resistance stable ( in the sub-micro ohm region ). Every time we re-tightened the posts, we got a different reading. Eventually we filed off the gold entirely, and there was a vast improvement in reproducibility.

Maybe it was just a bad gold plating batch, but I have always been suspicious of gold plating since then. Good for corrosion resistance, but you still have to be careful.

:)

How about this for a bad connection (one of our boards...)

The contact resistance should not matter too much for me, since these are voltage measurements at very low currents (nA). Should be sub-ohm, but that is about it.

For your resistance bridge I think they would probably use an AC or at least alternating technique, so any thermoelectric offsets would be removed.

how did you manage that? looks like there is both too little paste and reflow wasn't hot enough

-Lasse

I like these a lot. Damn strong. I've used them for wire antennas. Haven't used them in stereo applications. ;-)

Some process problem with the bare boards, simply would not solder even with a hand iron and flux. My theory was a thin (invisible) layer of solder resist remaining over the pads but don't know if that is really possible. We reported it to the Chinese supplier and were told the person responsible was now "seeking other opportunities"...

Woah, Those look great if nothing else. I like 'em!

On Mon, 20 Jan 2014 15:42:01 +0000, John Devereux Gave us:

Aside from finding and buying some, which sounds very costly if one's personal time is figured in... You could buy std miniature crimp lugs and go find a bike shop or hispanic upholsterer to plate them for you. Hell, some jewelers might do it.

Buying them is the right way, but sourcing the purchase sometimes costs more than the items themselves.

The first link I got on a google search for "fine instrumentation lugs" yielded some, but I thought they were way too expensive.

But there is that time thing again.

and the first hit...

Actually, the $39 is for a bag of 25. Not too bad, even though we know it is very thin electroplate. Look exactly like what you are after though...

On Mon, 20 Jan 2014 09:52:55 -0600, Tim Wescott Gave us:

That solder connection negates the entire purpose (and function) of the test setup.

On Mon, 20 Jan 2014 10:25:40 -0600, John Fields Gave us:

A lot of that stuff actually looks pretty good.

So, you caused me to sit and watch all four seasons of Lexx again. It was great!

Did you watch season one yet?

On Mon, 20 Jan 2014 09:05:09 -0800 (PST), snipped-for-privacy@gmail.com Gave us:

It isn't about that. The heat reference did not have to do with current flow.

On Mon, 20 Jan 2014 09:06:25 -0800 (PST), George Herold Gave us:

When two metals are sandwiched (mot alloyed) together, the bonds between the layers fall into a few categories related to the type of bonding, and efficacy of the bonds.

These nuances play into how a test instrument makes its measurements, and the chain of errors which get introduced into such test platforms, as they are typically calibrated with the correct hardware fitted to them. Using "any old coat hanger" will introduce error into the expected calibration points, thus causing skewed readings. At the very fringe of the instrument's capability, the skew can be huge. BAD figures.

So, the wire, and its plating, if any, and the contact connected to the wire and any platings it has. And finally, the terminal which the wire gets connected to and its platings all play in.

On Mon, 20 Jan 2014 14:10:33 -0800 (PST), Lasse Langwadt Christensen Gave us:

Looks like it sat too long before it got put into the reflow line. Looks like the stencil is very thin and sparse on pad match up too, OR there was a paste loss on the pull and entire areas lacked coverage.

But is looks like it could also be a lead free process issue too. They are hotter, so flux can bake out pretty quickly, which brings me back to the board sitting too long after the paste pull, and before the reflow line.

Make sure you get a kneading machine for the paste which constantly rolls it around in the jar.

Make fresh pulls, and get them to the reflow line quickly. I would not leave them out more than an hour before running them... on a bad day.

If it was a hand pick and place, it needs to go faster, The PCB looks of great (just fine) quality.

I like the PCBs folks like Seagate use. They have been doing it longer than just about anyone... by the millions.

You're clueless.

Good catch, looks like I had already visited that link in my previous searches but moved on too quickly seeing that $39. (Which was in line with the other ridiculous prices I was seeing elsewhere for the audio stuff).

Does look exactly like what I was looking for, thanks. Also the principle of looking for temperature sensing suppliers might be the way forward (thermocouples, low DC voltages).

Looked fine, but was not! You could not even hand-solder it (without scratching off whatever the invisible solder-proof surface was).

Actually it seems that may not be the case for the connectors designed for microvolt/nanovolt measurements. For example this manufacturer explicitly states direct plating with no nickel layer:

It seems that a relatively thick gold layer is used to counter the diffusing issue (which might explain some of the high cost).

From a thread I just found about all this on eevblog:

Fred Bartoli wrote:

And those are made to go onto banana plugs, not to crimp on wire.

There must be something to it; Keithley go to the extent of putting pictures in their nanovoltmeter manual, showing where to drip the "Caig deOxit" copper cleaner liquid into the panel connector.

Pretty much not. If a metallic sandwich is isothermal, which close connections usually are, there will be no thermoelectric potentials.