Help to identify the triax connector on an ADCMT 5450 electrometer.

[aronake]

I bought an ADCMT 5450 electrometer.

https://www.adcmt.com/en/products/dc/5450

This may be more a test equipment question that a metrology question, but due to the metrologyesque nature of the ADCMT 5450 I am posting here.

Problem is that it did not come with any measurement cable, which I was aware of.

The meter have a somewhat "non standard" triax connector which ADCMT call S.TRIAX. It seems it has a bit more insulation than a normal TRIAX connector.

I would need help to identify this connector so i can make my own cable, or find an adaptor from this connector to a normal triax so i could use my Keithley triax electrometer cable. Anyone have any idea what is the name of this kind of triax or any part number or supplier for this triax connector?

Attached picture of the connector on the electrometer and an adaptor ADCMT sell.

I am trying to contact ADCMT to see if they have a cable or an adaptor to sell, but not too sure I will hear from them, and/or suspect the cable is rather expensive so trying here.

[aronake]

Page 4 on this document has more information about this "much safer input connector":

https://www.n-denkei.com/taiwan/wp-content/uploads/sites/11/2021/08/E-5450-5451.pdf

[aronake]

It seems its the same connector as on the Agilent/Keysight 4339B high resistance meter.

Some info here, but did not help me to be much more certain on what coonector i am looking for.

https://electronics.stackexchange.com/questions/396471/unknown-triax-connector

[Conrad Hoffman]

I can find what looks like the connector on Digikey and other places, but the data sheets never disclose the thread size. If you knew that you could measure what you have and know the mate would fit. I just did a Google image search on TNC and triaxial and looked for something that looks like it can mate.

[aronake]

I can find what looks like the connector on Digikey and other places, but the data sheets never disclose the thread size. If you knew that you could measure what you have and know the mate would fit. I just did a Google image search on TNC and triaxial and looked for something that looks like it can mate.

Thanks! Thats some good suggestions. The electrometer is still in shipping so will measure and try to figure out when I get it. But hope is that someone may have some suggestion what connector it is.

[Sparks-Components]

Based on my experience in making HP4339B measurement cable this connector is

Triaxial TNC
Teflon needs to be cut

[aronake]

Based on my experience in making HP4339B measurement cable this connector is

Triaxial TNC
Teflon needs to be cut

Great! I managed to figure out this. I cut a bit less of the teflon as per below picture.

Cable works very well, but I have not taken it higher than 100 V yet. But would not think 1000 V as the 5450 can output would be any problem.

[aronake]

Based on my experience in making HP4339B measurement cable this connector is

Triaxial TNC
Teflon needs to be cut

Hi Sparks-Components,

Since you seem to be an expert on high resistance cables, how do you do on the DUT side of your cables?

To be able to run the meter with only the triax cable, I plan to put connectors on both the central wire of the triax and she outer shield. My suspicion when making the cable that normal shrink tube had too low resistance to make a good cable for dividing wires, which showed out to be true. My not so elegant solution on the picture below. I took the outer shield to one connector, then gap with the middle protection layer. Ideally I would want to pack this nicer, but would need some tubing (which ideally can shrink) with high enough resistance.

I will also do one cable where only the central wire goes to a banana or crocodile clip.

[alm]

To be able to run the meter with only the triax cable, I plan to put connectors on both the central wire of the triax and she outer shield. My suspicion when making the cable that normal shrink tube had too low resistance to make a good cable for dividing wires, which showed out to be true. My not so elegant solution on the picture below. I took the outer shield to one connector, then gap with the middle protection layer. Ideally I would want to pack this nicer, but would need some tubing (which ideally can shrink) with high enough resistance.

I've used PTFE heat shrink for high impedance (electrometer) applications.

[Sparks-Components]

You can use anti-leak tape which is made of PTFE
First wrap some anti-leak tape and then shrink it with ordinary heat shrink tubing.

Anti-leak belt or raw material belt. The translation seems wrong.

Pay attention to pollution

Pollution will leak electricity

你可以利用 止洩帶 它是PTFE製作的 
先纏繞一些止洩帶後再用一般的熱縮套管收縮即可

止洩帶或是生料帶  翻譯好像不太對

注意汙染
污染會漏電

[Sparks-Components]

The photos from 4 years ago are so nostalgic
I wrapped my clips in PTFE
At 100V output voltage
You can see that 0.5XXnA dropped to 0.02XnA (instrument limit/no zero current compensation)

[aronake]

Thanks to Alm and Sparks-Components for input!

I will buy some different PTFE heat shrink, leak tape, etc and do some experimentation.

I think i am on the right track now. The banana plugs I use are way to conductive, the better way seem to but on crocodile clips at the end.

[aronake]

I now got some PTFE heat shrink, some PTFE tube and some PTFE tape.

Having measured them, they all seem good with extremely high resistance.

One observation, the PTFE heat shrink I got is a bit difficult to work with. Requires very high temperature to shrink, like quite some blasting with the hot air station at 300 degrees. Is this normal?

The ADCMT 5450 show around 7 peta ohm at 100 V with no cable. Running cable with only triax connector, it show around 4 peta ohm. With my banana plugs and PTFE heat tube hack it shows around 1 peta ohm. What do other people here see for these kinds of measurements with electrometers and high resistance meters?

Humidity around 70% so pretty bad environment of high resistance experimentation.....

[Alex Nikitin]

Keithley 617 here, with 1m low noise PE (Huber and Suhner) triaxial cable on the input and 100V bias the leakage is less than 5fA, so over 20 Peta Ohm. The humidity is only 35% though with 24C in the lab.

Cheers

Alex

[Sparks-Components]

10 to the 15th power. This impedance value is very good. Furthermore, there are many details that make it possible to measure smaller currents and higher impedances.

For high impedance measurement, all operations and open wiring in the final section must be in a shielding box, and the gaps are best plugged with aluminum foil. Biscuit boxes, large cans of corn cans are acceptable, and glass bottles wrapped in aluminum foil are also acceptable.

No light, no magnetic field, no electric field
Keep high-impedance instruments away from everything else
Contains the operator (seriously

Very high impedance usually requires more average measurements depending on instrument characteristics.

For G ohms and above, it is best to do double masking. If T ohms and above are to be stable, it must be done. For P ohms, special test fixtures and compensation corrections are almost required.

see low noise cable
You can consider purchasing test cables with graphite conductive layers. Also known as: low-noise cables
PTFE cable, you can shake the cable slowly and you will find interesting phenomena (capacitive deformation/charge movement: current flow (pA level))
If the cable is fixed, it is not necessary, but there should be no vibration (fan/motor/operator breathing)

[aronake]

After some experimentation I figured out what was wrong with the cables I made. On the triax cable I have the innermost conductor have two layers of insulation to the guard. The inner one is the PTFE (white/transparent typical PTFE colour), the outer one  is quite conductive (black). I thought there was only one (the black) and tough that it was black PTFE. Not removing that made terrible amount of leakage.

Once i got the black insulation removed and only the white PTFE left, it was easy to make cables that max out the meter. Even at 1000V the leakage current is below measurable threshold and hidden in noise. The meter jumps around between around -10 to +10 fA.

Probably many people here knew already about this double insulation, or it may be from triax cable to triax cable, but this may be useful information for someone.

And thanks to all here on input on how to find (or make) the correct triax connector and on leakage currents.

[aronake]

I realised I was cheating a bit in my previous measurement. I was running in "two cable mode" i.e. voltage on "normal" cable and measurement on the Triax. So no wonder there was very little leakagecurrent.

When running in "single cable mode" feeding voltage though the triax, i got around 0.1 pA at 1000 volt.

I tested my original ADCMT cable for my ADCMT 8252 electrometer with an adapted. That comes down to noise floor of the 5450 in single cable mode. So I have some more tweaking to do.