DIY Low EMF cable and connectors

[quarks]

Hello robrenz,

thanks a lot for your offer.
But I am not searching anymore  and also would feel bad if, I would even ask you to spend time for this.

I guess we are a little bit carried away with the DIY gold plating question from wiss.
Because I only answered his questions with what I found out during my research and tests.

Right now I already have more parts than I will probably ever need in my lifetime.
Maybe I should even think about selling or trading some of the parts I bought and build for this research.

Bye
quarks

[robrenz]

No problem, just trying to help out on the never ending quest 

[bingo600]

So far (besides ebay) I have not found DeoxIT in Germany.
And with equivalent products from "Kontakt Chemie" I am not happy with my handmade lugs (see the one in the middle).

quarks i got my deoxit from Holland , shipping shouldn't be bad to .de

http://www.schreeven.nl/3313-deoxit
http://www.schreeven.nl/3315-deoxit

/Bingo

[eas]

Why do you suspect the existing 4mm banana sockets on your DMM (K2015?)?

Also, think about the thermal characteristics of the long, thin pin used to connect the lead from the PCB to the back of the stock banana jacks.

[macboy]

There is no reason to suspect that the Keithley input jacks and the leads from them to the board are anything but great. Keithley are among the masters of low level measurements. The input jacks are likely made of tellurium-copper. This has two very important properties: it has nearly identical thermo-electric properties to pure copper which result in low thermal EMF connections (to copper and like materials), and it is much harder than pure copper, making it more durable and easier to machine/manufacture. They are likely gold plated, which adds virtually no thermal EMF effect, since the gold layer will be at a constant temperature throughout. Then any thermal EMF generated from plug to gold plate is cancelled by the opposite thermal EMF from gold plate to CuTe jack bulk metal. The gold plating is there to minimize oxidation, which would cause much worse thermal EMFs than the gold does, so it is a compromise of sorts. The same goes for the pin connection at the back of the jack. That lead is probably terminated in a gold plated CuTe connector. Don't forget too that these are inside the controlled environment of the meter, and after it has been properly warmed up and stabilized, no additional thermal EMFs will creep in (those that so exist are constant and will be nulled out during calibration). Remember that the value of low thermal EMF connections is to minimize EMFs due to temperature gradients that are beyond your control. Sticking your finger onto an internal connection of a meter running without its cover doesn't count 

[Huluvu]

I don't get the Idea or reason behind why Tellurium should be good for Low EMF parts.
Tellurium itself has a much worse thermal EMF than Gold or Copper 
The only technical reason I found was the faster machine / manufacturing cause.

Any other explanation for the Tellurium Hype?

[timb]

I don't get the Idea or reason behind why Tellurium should be good for Low EMF parts.
Tellurium itself has a much worse thermal EMF than Gold or Copper 
The only technical reason I found was the faster machine / manufacturing cause.

Any other explanation for the Tellurium Hype?

From what I've seen, it's not. Pure copper is best, but it oxidizes quickly.

Pure Copper + DeoxIT would be the best solution. It just takes a lot of maintenance.


Sent from my Smartphone

[eas]

Why do you suspect the existing 4mm banana sockets on your DMM (K2015?)?

Also, think about the thermal characteristics of the long, thin pin used to connect the lead from the PCB to the back of the stock banana jacks.

yea i did

there was a instance while cover was open, i have my finger on the pin with connector. the DMM is running and python logging. the reading ran off about 8ppm. i thought that was interesting. there is also the point where wire solders to the PCB, there should be some EMF there, and there is no cadmium solders.

so instead of gold, it should practically all be copper + some de-oxit?

How important is it to minimize thermal EMF when the case is open and you are touching things? I'd think it isn't very important at all. With the case on and the device given its proper warmup period, I'd guess that both sides of the PCB joints are effectively the same temperature, as are both sides of the switch contact points, and both sides of the internal connection to the tip of the panel jacks. Moreover, they are all paired, and the pairs are probably at similar temperatures. All of which adds up to minimal thermal EMF in actual operation, and what does exist, is mostly cancelled out.

Sudden changes to the outside temperature will upset the equilibrium. Plugging leads into the jacks will also upset the equilibrium, and take a little time to equalize. If you really want to avoid the time to equilibrate when inserting room temperature leads into the case-temp jacks, it might make sense to have some extension leads that you leave plugged in.

As for wear and tear on the panel jacks exposing the underlying copper to oxidation. It could happen. Replacing the jacks shouldn't cost much, take much time, or involve much risk. I think there is a good chance that it wouldn't make a lick of difference, but just replacing them would address your concern without the time required to test the situation.  If you do replace them, I strongly suggest that you replace them with the original pin style, rather than the blade-lug style you linked to, because my guess is that the pin stretches the thermal gradient between outside temp and internal temp over a longer distance thus reducing differences between internal lead and socket temps.

Stepping back, I have a larger suggestion for your effort to improve upon the performance of you DMM.  Before you break anything else inside the DMM, consider making an external measurement pre-amp that will take low-level signals and produce a suitably low-noise +-10v output that you can feed into the DMM. Once you've got that working well enough that it matches the noise and sensitivity characteristics of the 1v and 100mV ranges on your DMM, you'll have a much better idea of what it will take to hack your DMM to actually improve its performance. You might even start with the reverse-engineered Keithley 2000 schematics and duplicate the relevant parts of its front end circuitry.

[robrenz]

......
......
The only technical reason I found was the faster machine / manufacturing cause.

Any other explanation for the Tellurium Hype?

IMO Tellurium copper is purely for machineablity. Pure copper machines like crap and the parts would cost much more because of it.  Tellurium copper is not an electrically tuned alloy, it is a free machining Copper that gets used because of its machinability period.  The rest is hype generated by tradition/lack of understanding.

[Gyro]

Teflon wire / cable is usually (not able to say always) silver plated.

[alanambrose]

Hi Quarks,

Do you have any measurements you can share with us?

A misc update:

(1) I also found some silver plated over copper / PTFE / twisted pair + screen cable which I plan to test. It was on UK eBay at £10 for 3m. I think people are buying this for audio.

(2) BTW here are jswilley prices I got in Jan '16:

SPSCTCS-4 is $7.75 per foot
STC-4 is $5.50 per foot
low thermal binding posts $25
low thermal forks and rings are $2.99

(3) I just ordered some small ENIG pcbs to fit the 3458a terminals to use as shorts / to hold standard vishay resistors. I'll let you know how these go

http://anagram.net/nuts/Low%20Emf/Images/low%20emf%203458a%20shorting%20bar.JPG

(4) For the truly OCD, I designed this little thing to shape copper wire for 3458a short:

http://www.shapeways.com/product/GAG65ZGSB/3458a-copper-short-forming-tool

I have not actually tested it out yet. Don't panic, I make exactly 0.00 on each The little lip is there to make it easy to hold in a vice if you want.

Alan

[VK5RC]

I am not in the league of you guys, I made a simple set of lower Thermal EMF banana leads using this cable.
http://www.ebay.com.au/itm/381447862950?_trksid=p2060353.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

I used these banana plugs;
http://www.ebay.com.au/itm/10-Premium-4mm-Gold-BANANA-PLUGS-Twin-Screw-Cable-Connectors-Soft-Feel-Covers-/300776381856?hash=item4607ab5da0:g:CZMAAMXQDfdRpUNy
The covers when tested with an insulation tester come out OK,
The springs on the tip are nice and firm and appear well bonded to the body (they do not rotate) , I GUESSED that a screw clamp (or two) would be closer to crimping than soldering, but I must admit that the really high thermal EMF of copper oxide has me still leaning a little toward soldering to some degree. It must be really hard to oxidise under a solder joint! I am a bit of a solder fan boy - I must come into the 20th Century.

[timb]

The high pressure of a crimped connection creates a cold weld and the metals are bonded together, so there shouldn't be any oxidation if the copper is cleaned before being crimped.

The problem with solder is that you're adding a dissimilar metal (lead) into the mix, which can create additional thermal EMF.

[quarks]

when I made my comparisons I compared all my different cables and connectors with the same settings on the DMM (Fluke 8508A) and on the same 10 VDC reference (Wavetek 4808). The results between my DIY and the Fluke low EMF "reference" cables were practically identical. The averraged results were mostly only in the 0.1 µVolts, which is not really meassured, because that is the last digit, which is not always perfectly identical, even when you use the same cable again.

With my other good cables I saw differences around 1 µV. But that is still ok for 10VDC because that is only 0.1ppm.

My conclusion and ranking was:

1. use good quality cables with high quality insulation material like PTFE (avoid PVC)
2. use crimped gold plated copper spade lugs for all precision measurement (avoid nickel plating)
3. use low emf 4mm banana if you do not have the above (i.e. Pomona, MC)
4. use good 4mm banana (in my case Hirschmann and MC, both with welded on cable), if you do not have the above

And use the very same cable, when you do high precision comparison meassurements or calibration, whenever possible and wait long enough for temperature equilibrium.

att. you can see some cable madness (the pile shows most of my low emf reference cables from Fluke, Pomona and ESI)

[quarks]

Hi Quarks,

Do you have any measurements you can share with us?

A misc update:

(1) I also found some silver plated over copper / PTFE / twisted pair + screen cable which I plan to test. It was on UK eBay at £10 for 3m. I think people are buying this for audio.

(2) BTW here are jswilley prices I got in Jan '16:

SPSCTCS-4 is $7.75 per foot
STC-4 is $5.50 per foot
low thermal binding posts $25
low thermal forks and rings are $2.99

(3) I just ordered some small ENIG pcbs to fit the 3458a terminals to use as shorts / to hold standard vishay resistors. I'll let you know how these go

http://anagram.net/nuts/Low%20Emf/Images/low%20emf%203458a%20shorting%20bar.JPG

(4) For the truly OCD, I designed this little thing to shape copper wire for 3458a short:

http://www.shapeways.com/product/GAG65ZGSB/3458a-copper-short-forming-tool

I have not actually tested it out yet. Don't panic, I make exactly 0.00 on each The little lip is there to make it easy to hold in a vice if you want.

Alan

Hello Alan,

about the shorts, that is very interesting please share where to buy.
Also the "3458A Copper Short Forming Tool" looks very good.

Att. is a collection of shorts I mainly use.
As you can see my own 3458A short does not look good at all. So it would be very handy to have this jig.

bye
quarks

[quarks]

here are some links (Transmille videos) to the low emf cable subject:

Making Good Measurements - Effect of Test Lead Resistance on Measurement Error


Making Good Measurements - How to avoid thermal EMF errors

[alanambrose]

>>> I also found some silver plated over copper / PTFE / twisted pair + screen cable which I plan to test. It was on UK eBay at £10 for 3m. I think people are buying this for audio.

BTW I did test this (10V from 732a into 3458a at NPLC=2, AZERO=ON) and it is not much of improvement over cheap twisted-together banana leads.

For both setups, I'm getting 0.06-0.07 ppm rms noise - which I think is pretty good for the bananas actually. This is with soldered ends on the PTFE leads, no lugs. Neither using the shield as guard or running the 732 off battery seems to make any difference. I think I need to do some more careful measurements...and get some lugs to crimp on some new cable...

A.

[acbern]

I have seen some considerable EMF voltage from certain banana plug cables, but not on all. So indeed, depending on the cable, sometimes no relevant difference when using standard cable. But you do not want to guess that, so desoite this, for precision measurements, I dont take the risk, not worth it.

[martinr33]

I've been looking for copper banana plugs, because I am getting measurement offsets with my current cables (Nakamichi jacks on copper wire).  I'v purchased many types from eBay, including Nakamichi from, several suppliers. I run them on my sander to see if they are really copper.

Bottom line is everything I have found that is gold plated is always brass, no matter what claims are on the web site. 

I did find some on eBay with the description of "pure red copper". Although some are plated copper (and say so) the ones I picked up are copper through and through. They run about $4 each. I use them with silicon insulation 10GA wire, works well.

I also picked up some Mueller copper alligator clips. I soldered the end of the wire to the inside of the alligator clip, and then crimped to body of the plug to  the wire.

So we'll see how these do.

[ManateeMafia]

Do a search for Furez TSTWP30NP. I bought some and used them for a Keithley 181 cable. Yes, these do appear to fall into the audiophoolery category but they seem to be real. Real expensive if you include shipping. Also they are not shielded.

It is prone to scratching so a clean binding post will save on the wear and tear.

[Robert763]

Raychem make heatshrink transition boots e.g. Raychem 462a011-25-0 Check ebay.
The gold plated amp lugs may be standard tin plated ones that have had the tin plating stripped and then been gold plated. See http://www.finishing.com/319/63.shtml
 I've used Van Damme professional audio cables for test leads with good results. Some national labs just use 2.5mm solid copper mains installation cables without lugs. They rely on air spacing for insulation and just scrap them when damaged or too short from repeated stripping.

[martinr33]

Well - it is confusing. When you are looking at microvolts, everything make a difference. I am not saying that the laws of physics are suspended, but that we do need to think very carefully about where the various EMFs come from

Thermal EMFs come from two dissimilar metals in contact. In the case of plating, assuming that the copper is the same as the copper in the wire (it isn't, but that is OK for now), you end up with a copper to brass junction.

Now, the copper test lead wire creates a canceling brass to copper  junction, and as long as the jack is at the same temperature all through, there's no voltage because the two junctions cancel out. But if the tip is warmer than the body, there is a net voltage  - and that matters at the microvolt level.

I am still messing with this. My new leads are definitely better, but I still have offsets. I might have copper oxide in my meter connections. The jacks are really, really red - my wire is more yellow.

And I would like to have gold plating, because I know I will be cleaning my copper. 

Sorry I'm feeling a little confused about the plating comment.
I understand it is bad if something isn't as described by the seller.
I understand that the bulk of the material as well as the plating will matter for mechanical qualities, durability, wear life, corrosion, et. al.
I understand that plating may or may not be absorbed into or affected by a soldering process for better or worse.
I understand the skin depth and that at some frequencies some amount of the DC/AC current will flow in the interior (sub-plating) layers of a conductor.
But with respect to thermal/contact EMFs I think I am forgetting something having to do with multple-junctions or something or just not understanding something about when/how a plated substrate will perform differently near DC in terms of thermal EMFs than if the whole conductor was made of the plating material only.  If it is copper plated brass or nickel or steel or something but the copper plating is not removed/damaged by the soldering/termination, and the copper plating fully surrounds the inner conductor (ignoring wear issues), in what way is the copper plating to substrate junction going to affect the voltage seen along different points on the copper exterior of the connector (ignoring corrosion issues)?
Where are the thermal gradients and where are the potential differences developed in the hypothetical scenario?
I can sort of see it but can't quite visualize it yet because of the distributed and exterior nature of the plating vs core and the geometric complexity.
I guess if you assume you can have arbitrary thermal gradients along and into the material and finite conductivities then you could have thermally driven potentials and currents variously distributed in/on the object?

[acbern]

What meter are you using and what offset do you see?

[martinr33]

I won't be back to my meter for a week or two. It is a K2001. I'll measure and post.

[VintageNut]

I have something to add to this discussion. I am in the process of creating a good 100:1 and 10:1 Hamon divider. As part of the parts selection and assembly process, I have to measure single-ppm differential null voltages.

Today I watched the a null reading go from 10s of ppm to single ppm over a minute. I thought there was a problem with the source. The problem was cheap banana/alligator cables connected to my divider parts on the bench.

I disconnected the voltage sources and divider and just shorted the null meter alligators together. There were were over 10 microvots off offset that slowly drifted to zero after more than one minute. I grabbed one of the shorted alligator clips and held it a few seconds. The offset measurement climbed back up to over 10 microvolts.

My point is that is not just a dissimilar metal problem. It also a temperature gradient problem that you have to be looking for.

For me, the cure will be to make better alligator cables and to place enough insulation where the alligator is touched to mitigate transfer of body heat to the clip.

Touching cables and plugs in the signal chain can create a temporary offset that you have to be aware of. If you do create an offset, you have to wait for the signal chain to come to thermal equilibrium.