My TRIAX cable shielded box project for Keithley Instruments

[HighVoltage]

Project 1
This is my first attempt to build a shielded box with a TRIAX Cable to measure very high resistance values with my Keithley 614 Electrometer. This meter has the dual pin TRIAX connector and I found a fitting cable.

These are the highlights of the materials used:

- Metal box with metal lid
- Bottom and side of the box covered with 5 mm thick PTFE (Teflon) plates
- Original Keithley TRIAX cable
- Outside shielding connected to the metal box (blue cable)
- Inside shielding extended with copper/silver-plated cable and Teflon insulation (orange cable),  (crimped)
- Inner core extended with copper/silver-plated cable and Teflon insulation (orange cable),  (crimped)
- Crocodile clips pure copper (crimped)

This Russian military glass resistor inside the box is marked with 100 GOhm and reads very stable 95.5 GOhm after a few minutes of not touching anything. So far I am happy with this box for this older Keithley 614 Electrometer.

OK, I am looking for improvements and hints of what to change or do differently for the next project.

My next box will be for the DMM7510 and 2450 SMU (3 pin TRIAX)
One improvement I would probably use are Triax Panel Mount Connectors.
But they are sooo expensive, same as withe the TRIAX cables.
 
Thanks for any suggestions.

[VintageNut]

Remove the resistor from the box. Leave the alligator clips where they were. Close the box. Force maximum voltage and measure current. This will tell you if you have any leakage.

[plesa]

Fisher make triax which are cheaper than BNC ones. Or SHV or MHV ( high voltage BNC) are another option.
White panels are from PTFE?
If you want to make experiments with photodiodes I propose to put some black paper to panels.
And add temperature and humidity measurement.
If you are going to make some HV experiments add protection to your electrometer (GDT, 1M-100k resistors and two antiparalell diodes 1N3595 or DPAD1 are good).

[manganin]

OK, I am looking for improvements and hints of what to change or do differently for the next project.

I did something similar last year.

A cheap cash box (from Lidl) with a row of BNC feedthroughs (one triax) and a flexible copper strap connecting the top and the bottom.

Two of the connectors with fixed crocodile clips for measuring resistors. No wires or components touching the box = no leakage path.

[quarks]

very interesting, thanks for sharing

I just saw this Video

in there I saw Agilent box (around 4:06min it is large in the frame)

[HighVoltage]

Remove the resistor from the box. Leave the alligator clips where they were. Close the box. Force maximum voltage and measure current. This will tell you if you have any leakage.

VintageNut
Good idea but first I need to get a TRIAX to "something else adapter", so I can apply the voltage to the same input cable.
My Keithley Pico Amp Meter will probably be helpful for this test.

plesa
Since the Keithley 2450 has already the 3 prong TRIAX connectors on the back, I think I will stay with them.
One idea is to buy two cables and cut them in 1/2 to have 4 cables
But I think I will make the box really nice with TRIAX Panel Mount Connector  like in this pictures

[HighVoltage]

in there I saw Agilent box (around 4:06min it is large in the frame)

quarks
That was an interesting video, thanks.
The box shown is an Agilent 16442B, its probably $10k, I would guess and looks really nice.
Interestingly, they are using 2 mm banana plug connectors inside.
 

[ManateeMafia]

HP test fixtures like the 16058A are available on ebay though the quality and condition are not as nice as the one in the video. It looks like the enclosure is coated with a conductive shield.

The 16058A has triax and BNC connectors already installed. It also uses modules for different components.

[zlymex]

Looking very nice.
I got a Keithley 6104 test enclosure, and modified for triax connection, just in case you haven't seen that.
https://www.eevblog.com/forum/metrology/i-plan-to-make-an-100-meg-standard-resistor/msg917093/#msg917093

[VintageNut]

I acquired a KE 8002 high resistance test fixture. It matches perfectly for a KE 487 picoammeter that I also acquired.

[macboy]

I guess I must have a little bit of OCD or something, because I cringe when I hear "Triax connector". Triax is a type of cable, not a specific connector. I suppose that any connector that fits Triax cable is a triax connector. By the same token, SMA, BNC, and RCA are all "Coax" connectors.

The connector we are referring to is a TRB, also occasionally called Trompeter after a common manufacturer (like Kleenex or Q-tip). TRB connectors usually have 3 lugs rather than 2 these days to help prevent damage due to dummies mating them onto BNCs. To make it even more complicated, there are keyed versions where the angle between the lugs is not all 120^o.

The TRB connector is also commonly used with Twinax cable in addition to Triax. There is only a slight internal difference to ease termination of the cable, but the mating surfaces are identical; it's still a TRB connector.

On a helpful note, you will find more hits for used (and especially new) cables and connectors if you expand your search to include TRB and Trompeter, since merchants often do not refer to them as Triax.

[VintageNut]

Good points. My instruments have 2-lug triax and 3-lug triax connectors. Most have the 3-lug. Keithley discontinued the 2-lug connectors because you could insert a 2-lug coax plug which destroyed the 2-lug triax jack. The coax center pin is much larger the triax jack receptacle. The coax plug center pin forces the jack receptacle to expand and is ruined. I have witnessed this with a model 7065 Hall Effect switch card. It is a very expensive card and is easily ruined by unsuspecting users not aware of the difference between triax and coax.

[dr.diesel]

But I think I will make the box really nice with TRIAX Panel Mount Connector  like in this pictures

That is how I made mine:

[HighVoltage]

The connector we are referring to is a TRB, also occasionally called Trompeter after a common manufacturer (like Kleenex or Q-tip). TRB connectors usually have 3 lugs rather than 2 these days to help prevent damage due to dummies mating them onto BNCs. To make it even more complicated, there are keyed versions where the angle between the lugs is not all 120^o.

The TRB connector is also commonly used with Twinax cable in addition to Triax. There is only a slight internal difference to ease termination of the cable, but the mating surfaces are identical; it's still a TRB connector.

That was some very helpful information to find the correct binding posts. Thank you.

[HighVoltage]

I found this really nice Keysight Video with some basic information to TRIAX / BNC wiring



https://youtu.be/N66Oot1nJaQ

[quarks]

Also have a look at the great robrenz "DIY Triaxial test leads and shield Box" video

[TiN]

Newcomers into high-impedance/low-current stuff don't understand why triaxial connectors cost arm and a leg? Well, opposite to regular cheap BNCs (high-frequency/high voltage version of BNCs are expensive as well!) triaxial connectors have significantly more bits and pieces into them and also machined to higher precision, use PTFE insulator (usually) and less qty product.

Was making a cable for my low-current box using CNMC low-noise wire (with PTFE outer and inner insulation, graphite coated inner conductor insulator, copper braid, inner silver-plated copper braid, etc.) and Cinch/Trompeter PL75-29 cable plug (23.5$) + Pomona 5219 (24$).

[HighVoltage]

I am just starting again with this work in progress and everything I have learned from my first attempt above.

Project 2
This is now my second attempt to build a shielded box with 4 TRIAX Cable Connectors and 2 low emf binding posts, to measure very high resistance values and ultra low currents.

This box is intended for the Keithley 2450 SMU and Keysight B2987A Electrometer.

These are the highlights of the materials used so far:
- Metal box with metal lid and hidges
- Bottom and sides of the box will be covered with 5 mm thick PTFE (Teflon) plates
- 4 pieces TRIAX connectors, modified to accept 2 mm banana plugs
- 2 pieces LOW EMF binding posts, modified to accept 2 mm banana plugs
- All inside wire connection will be made with gold plated 2 mm MC connectors
- May be I will add 2mm gold tubes to the guard connectors as well

Later I will add:
- Magnetic switch for the lid
- PT100 Temperature sensor inside
- Humidity sensor inside

Here are the first pictures:

Anything I am missing?

[HighVoltage]

This is the modified Triax connector to accept the 2mm banana plugs

[Conrad Hoffman]

What does the Teflon do? I'd think it would cause more problems with surface charge and storage than anything else.

[4CX35000]

This is the modified Triax connector to accept the 2mm banana plugs

I have of doing the same to for my Keithley 617, but in my case I will make do with screwing the lid as normal. So long as the instrument is kept in a reasonably warm and dry enviornment, i don't see any reason in using the shielded box more once than every year.

Farnell sell the Triaxial cable (Yellow outer), connectors can be bought from Farnell or RS Components and the rest inside my shielded box are standard gold plated components inside a stainless steel box.

I decided to replace the two lug triaxial connectors with a three lug type on the Keithley 617 and do the same on my Keithley 220 DC Current Source, so both will need alignment and checking with the resistance box when ready.

On the Keihtley 617 it has a voltage source facility which I checked yesterday and is working fine.

[HighVoltage]

What does the Teflon do? I'd think it would cause more problems with surface charge and storage than anything else.

The PTFE is for insulation purpose only, since I am planning to source up to 1000V in this box.
On my Project 1 box above I used already teflon and did not notice any surface charge  problems.
But I will take a closer look at your interesting advise, thanks.

[razberik]

Teardown of Keysight B2987A Electrometer please !
Sorry for hijacking your topic.

[HighVoltage]

Teardown of Keysight B2987A Electrometer please !

I just ordered the B2987A from Keysight a couple weeks ago, it will probably arrive in 2 more weeks.
Will see, if I have the guts to open it up.

OK, I added the TRIAX connectors to the box and two LOW EMF copper tellurium / gold plated binding posts.
Also added the PTFE plates inside. I even found a 3M double sided tape that sticks well to the Teflon.
The cables are original Keithley 7078-TRX-3
With the lid closed, this 100 GOhm resistor shows 95 GOhm very stable.
But if I touch the box, the reading moves more than expected.
Although the lid is grounded to the box, I need to add a screw to close it tightly, so touching the box will not influence the reading.
Hmm, I might need a closing mechanism.

[HighVoltage]

My largest resistor is one marked 1000 GigaOhm and it shows almost 1 TeraOhm.
It is interesting how slow it climbs up from 0.9 Tera Ohms
I am using only 50 Volts for this measurement, because I can not find my interlock adapter for the 2450.

Next to do:
- Install PT100 and Type-K temperature sensors
- Install magnetic switch for interlock.