DIY Low EMF cable and connectors

[julian1]

I just noticed the possibility of using BNC cables for shielded measurements, versus hand-rolled twisted pair. Any suggestions or preferences for bananna to BNC connectors? These are the two that I found,  but I worry about the combination of alloys and thermocouple offsets.

  Model 1269 Pomona.
    pdf datasheet,
     https://www.pomonaelectronics.com/file/16994/download?token=cRlo0ROC
     ( https://www.digikey.com/en/products/detail/pomona-electronics/1269/603922 )

    MATERIALS: BNC (Female):
    Body: Brass, Nickel Plated, Dielectric: PTFE
    Center Contact: Beryllium Copper, Gold Plate
    Banana Plug - Body: Brass, Nickel Plated
    Spring: Beryllium Copper, Nickel Plated

  Mueller BU-00260   
  pdf datasheet,
      https://au.mouser.com/datasheet/2/279/DS-BU-00260-1622138.pdf

    Material:   Body   Nylon
    BNC  Plated brass body, gold-plated brass center
    Banana Brass body, beryllium copper nine-leaf springs,
    nickel-plated

[ermionesrl]

Hello julian1.

Electrometers have triaxial BNC connections, so the idea could work, yet current nanovoltmeters still have special low thermal EMF connectors that avoid dissimilar metals.

Normal BNC, adapters and even most coaxial cables, with dissimilar plating and alloys in the inner conductor and shield, are a nightmare on this respect (triaxial BNC are gold plated in both inner conductors, at least).

By the way, twisted pair work very well, why you don't want to use it? It's even cheaper to roll your own, even with teflon insulation.

[OscarM]

Apologies in advance if this has been covered somewhere but I have looked and failed to find it.
Is there any merit in connectors like this?

Available as FMTC-CU-M for about $5
It is meant for 'copper thermocouples' which strikes me as odd but might be suitable for this application.
Presumably Deoxit is a good idea.
Mates in chassis mounting or inline are available.
PCB mounting is also available.
I have also seen a version with round pins which is larger.

Clearly it isn't shielded/guarded but on the other hand it is small so the coupling loop for EMI is small and the pins are in close proximity and so will be close in temperature (which shouldn't matter because it is all copper)

[martinr33]

I could see that working. It ise really hard to get good copper bananas!

It would make a good connector for some kind of custom box - resistor testing, for example.

However, most of our instruments use bananas so it isn't useful for the instrument end.

[leighcorrigall]

Is there any merit in connectors like this?

I work with K-type and S/R-type thermocouples for high-temperature applications because they are standardized. These 'copper' types are unfamiliar to me.

I think there is some potential with using FMTC-CU-M as jacks. However, I would be careful about what 'pure' copper means. All metals are alloys and thermocouples work on the uniqueness of coupled alloys to make a thermal sensor.

Looking at the available thermocouples that LABFACILITY manufactures, there are no Cu/Cu types -- only Cu/Cu-alloys:
https://www.farnell.com/datasheets/2371070.pdf

This may suggest that the connectors you propose are actually pure copper (i.e., C101 or C110) and are used generically for all Cu/Cu-alloy thermocouples available (e.g., Cu/Cu-Ni and others). This is possible because typical thermocouples do not have a high tolerance (unless well characterized and used under the specified temperature range to prevent oxidation). For example, I usually assume a > ± 5 °C on K-types below 800 °C. When using these extension/generic connectors for general measurements, it is good enough.

Furthermore, I would be concerned also with the connection quality too. They might not be as reliable as one would need for metrology-grade measurements. That being said, I have never had a problem with MINI jacks when using them for temperature measurements. Buy the high-quality brand-name ones though.

More information at: https://www.thermocoupleinfo.com/

Conclusion: Give it a try, but make certain that the material is actually pure copper. 

P.S. How are you going to clean the female connector with corrosive deoxit and expect the connection to remain consistent? 

[leighcorrigall]

I have also seen a version with round pins which is larger.

Those are standard-size thermocouples:

https://www.omega.ca/en/temperature-measurement/temperature-connectors-panels-and-block-assemblies/temperature-connectors/ostw-cc/p/OSTW-CC-C-F

The TC jacks you were suggesting as a substitute for banana jacks are called MINI.

[mendip_discovery]

I use them in the Calibration world as they are universal connector. Often used on the end of a simulator to plug into a display.

You can get then in male/female options etc. Even in slightly larger sizes.

[OscarM]

Is there any merit in connectors like this?

P.S. How are you going to clean the female connector with corrosive deoxit and expect the connection to remain consistent? 

I was going by pp.128 of the Keysight 34420a manual.
Here, Keysight suggests putting a drop of Deoxit in the contacts of the Lemo connector.
Those contacts are copper.
They don't say remove it afterwards. In fact they end with 'Engage and disengage several times to distribute...'
They also don't specify which Deoxit product to use which is a whopping drawback.
Perhaps they mean the 'shield' version?
If you don't like Deoxit (whichever one it turns out to be) there is always Santovac 5. Very viscous and about as inert as anything.
Sadly I cleaned my diffusion pump and haven't filled it again.
Santovac 5 won't remove oxidation but it will probably prevent it if applied to new parts which is really what we are talking about here.

To test, I can either restore my HP 419A or use my HP 3456A. Neither may be ideal but they are on hand.
Tentative test plan:
1) Measure resistance of a loop using some of said parts once lubricated. Perhaps 5 mated pairs.
2) Measure Seebeck of said loop after, I don't know, heating up one side of the connector with a hot air pencil or heating up one connector with a hot air gun and then immediately mating them.
3) Put parts on shelf
4) Repeat after a month.
Constructive criticism of test plan welcomed.

[1audio]

it should be pretty obvious if there is a contact issue. Copper to copper Seebeck is nanovolts, Copper to copper oxide is millivolts. The only issue with the test is making sure only one contact pair is heated. if both are you may have cancellation. I hope it works out.

Looking into Deoxit  https://www.hagensieker.com/wordpress/2018/06/18/deoxit-what-is-it-what-isnt-it/   it seems that there are little or no corrosives in it. Its seems more to preventing corrosion. For copper that is badly oxidized I used some Brasso and then cleaned everything with isopropyl. It seems to work quite well. I would not do this unless the connector is really brown or green and otherwise a lost cause. It worked well with my vintage Keithley 148.

[martinr33]

I have some bare copper banana sockets on a 0Ohm resistance standard.

I liberally dosed them up with Deoxit Red, screwed them loosely together and put the vinyl caps back on.

Some months later, there was a tinge of green in the Deoxit and the terminals looked fine. However, that's after a very long time and with a liquid layer. Seems that Deoxit is partly oleic acid. But I would exect that acid to very quickly denature when used as instructed.

The flip side is that Deoxit is pretty hopeless at treating corroded copper spades. For that, I use vinegar immersion and a fresh water rinse.

I was skeptical of Deoxit for a long time. However, I have found it very effective in improving contact performance in batteries and potentiometers. And HP recommends it for use with the 3458A.

Overall, you should have a small bottle around.

[leighcorrigall]

For that, I use vinegar immersion and a fresh water rinse.

Acetic acid (essentially synthesized vinegar) is very effective at removing copper oxides. This is what is commonly used in metallurgy. Wear gloves and use a fume hood well-ventilated area when using the high-molar concentrate because it will burn you and is bad for the lungs. Acetic acid stinks. To prevent tarnishing, rinse with water immediately followed by drying with a clean cloth (or compressed air when the geometry is complex). C101 will remain oxide-free for months after treatment. If the part oxidizes in a few days or weeks, you may have either contaminated it or you are working with a lower purity copper (e.g., C110 and C122).

If all you can access is grocery store vinegar, heat the acid to improve its activity. An ultrasonic cleaner will also help encourage oxide stripping.

 

[martinr33]

Here's a picture of the business end of the latest Fluke metrology cables for the 8558 multimeter.

 - the cable is 22GA PTFE shielded twisted pair. It is black, not white.
 - the ends are sleeved in silicone rubber. They are very flexible, so much so that the PTFE insulation may have been stripped back inside the sleeving. They do not behave as if they are still insulated with PTFE, as they will not hold a bend. I wonder if the insulation might have been removed and the wire untwisted to make it more flexible.

" alt="" class="bbc_img" />

[TimFox]

Apologies in advance if this has been covered somewhere but I have looked and failed to find it.
Is there any merit in connectors like this?
(Attachment Link)
Available as FMTC-CU-M for about $5
It is meant for 'copper thermocouples' which strikes me as odd but might be suitable for this application.
Presumably Deoxit is a good idea.
Mates in chassis mounting or inline are available.
PCB mounting is also available.
I have also seen a version with round pins which is larger.

Clearly it isn't shielded/guarded but on the other hand it is small so the coupling loop for EMI is small and the pins are in close proximity and so will be close in temperature (which shouldn't matter because it is all copper)

Those thermocouple-style connections are designed to connect copper wires used after a ice-point calibrator (real ice or electronic circuit) that connects the thermocouple materials into the world of copper connections.  Mechanically, they are the same as the connectors for J, K, etc. thermocouples that use the same alloys as the wires.

[martinr33]

An update on Pomona low thermal spade connectors.

I started testing with magnets, and found mixed results on individual cables. eg black spades magnetic, red not.

The difference is visible. Low thermal gold plating has a sort of matte look, while gold over nickel is mirror bright.

What I  cannot easily tell is if the nickel is a mistake, or the whole spade is a mistake (i.e. nickel over copper probably OK, nickel over brass is bad).

[mzzj]

Own measurements have shown that the plating (tin versus gold) does not make a measureable difference (at the levels discussed here), always provided copper is used as the base material. Gold has additional advantages of course.
Brass e.g. used for the spade material is making a difference though, you do see differences.

This is my limited experience also. Bog standard tinned spade connectors seem to work very well on 8508A and 3458A compared to bare  copper wire or the 8508A "official" zeroing fixture.
All of the fancy low-emf banana connectors  seem worse than 10 cent tinned spade connector. I'm thinkig this has to do with the thermal mass and conductivity: Adding a "big" hunk of copper to DVM terminals creates slightly different thermal conditions for the binding posts. This is probably even more important if your lab temperature is fluctuating.
Body material, smaller mass and thermal "disturbance" seem far more important than plating.
Probably a different situation with connectors that don't have as firm mechanical connection like binding post+spade as it provides sort of wiping action + ample of contact pressure.

Anyone attempted to chemically strip the ordinary tinned spade connectors to bare copper? Electrolytic stripping with NaOh (drain cleaner) and 6 volt power supply seemed reasonably straightforward.

[DavidKo]

...
Anyone attempted to chemically strip the ordinary tinned spade connectors to bare copper? Electrolytic stripping with NaOh (drain cleaner) and 6 volt power supply seemed reasonably straightforward.

I used HCl - it took a while, but the work was done after few weeks outdoors in the jar. The surface was than matt, so I polished it.

[krasimir.k]

I was going by pp.128 of the Keysight 34420a manual.
Here, Keysight suggests putting a drop of Deoxit in the contacts of the Lemo connector.
Those contacts are copper.
They don't say remove it afterwards. In fact they end with 'Engage and disengage several times to distribute...'
They also don't specify which Deoxit product to use which is a whopping drawback.

Taken from Agilient 34420A user guide page 273:
"You can obtain additional DeoxIT, part number D100L2"
Kind Regards,
Krasimir

[tatel]

Since I'm in Europe, this is what I got

https://www.conrad.com/en/p/cramolin-schutz-1151611-lubricant-400-ml-820553.html

400 ml, 10.99€

This schutz thing remains over the copper as a thin film, while cramolin contaclean mostly evaporates and cramolin wasche really evaporates after cleaning the contaclean thing.

Not affiliated in any way with cramolin nor conrad. But as I understand it, these products look really like the original DeOxit products, even in color. AFAIK original DeOxit was cramolin. And, 11€/400 ml looks way cheaper that what I found when I was looking for DeOxit

[alm]

I'm not that familiar with the Cramolin product line, though I have used their Contaclean. From what I understand, Cramolin Contaclean is like Deoxit D5 (with solvent / flushing action), so I guess Cramolin Schutz (also given the name) is more like Deoxit Shield, which is advised to be used after cleaning to protect contacts against future corrosion, and leaves more of a residue. I'm not sure if this extra residue is beneficial in this application. Given Keysight's recommendation of Deoxit D100, I'd think that Cramolin Contaclean would be the better match with a better cleaning action and less (but still some) residue.

[tatel]

You're probably right. Here you can see how the film looks over copper clad boards after a week in hot and humid summer with open windows. No the best pictures I took in my life but I think the difference in film thickness is clear.

https://www.eevblog.com/forum/repair/no-deoxit-but-cramolin-contaclean-cramolin-protection-and-cramolin-spraywash/msg4960474/#msg4960474

IIRC, original Deoxit was Cramolin, reselled or even produced under license in the US. Then there was some drama and looks as if Craig developed copies/new products.

Some more testing would be in order, I guess.