Author Topic: Keithley 616 Electrometer input stage replacement  (Read 1247 times)

0 Members and 1 Guest are viewing this topic.

Offline muvideoTopic starter

  • Frequent Contributor
  • **
  • Posts: 418
  • Country: it
Keithley 616 Electrometer input stage replacement
« on: March 03, 2023, 07:28:46 pm »
Finally, after several years of service in my home lab, my 616 is dead.
Or to be more precise, I have killed it.

[Boring explanation starting]
Everything started with a stupid accident involving a broken keithley 220
with killer attitude.
Anyway after the first accident I repaired it (fortunately no sensitive electronics were involved)
and I was curious to track some small leackage problems
of the 616. The first was the 10-10A range that was way off.
This problem was solved after cleaning the seemingly already clean 1010Ohm resistor.
Then I started playing with the residual leackage read by the meter, it was measuring
0.3pA with open leads, aiming to understand what was the source of it.
[Boring explanation ending]

Well, I have played too much, I managed to kill Q101 mosfet with a gate perforation.
So I have ordered some 3N163 with the hope to be able to fix the damage


Waiting for them to arrive, I'm playing with some opamps to try to replace all the above circuitry
with something more modern.
This is the schematic I'd like to try, for now I've only tested the principle with bare minimum cimponents,
but I need to test and verify the behavior with the voltage regulators and the zero circuit.
Is there any hope to be able to create a decent electrometer input this way?

« Last Edit: March 03, 2023, 07:40:14 pm by muvideo »
Fabio Eboli.
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 14076
  • Country: de
Re: Keithley 616 Electrometer input stage replacement
« Reply #1 on: March 03, 2023, 08:14:38 pm »
The CMOS OP-amp should be OK. The offset trim circuit may not be ideal with unnecessary much dependence on the balance of the +-5 V regulators. Chances are there should be additional resistors (e.g. 1 K) from the pot to ground.
 

Offline muvideoTopic starter

  • Frequent Contributor
  • **
  • Posts: 418
  • Country: it
Re: Keithley 616 Electrometer input stage replacement
« Reply #2 on: March 08, 2023, 11:44:54 am »
The CMOS OP-amp should be OK. The offset trim circuit may not be ideal with unnecessary much dependence on the balance of the +-5 V regulators. Chances are there should be additional resistors (e.g. 1 K) from the pot to ground.

You are right about the offset.

I've redrawn the circuit to make it balanced around Common,
added also a guard ring around opamp inputs.
I still didnt had a chance to test it, I'll tune C5 and R6 manually.
Maybe this thing needs a proper pcb with a decent layout, I was hoping to
get away with a perf board, this will add some time to it.
I'll report back any result, thanks.
Fabio Eboli.
 

Offline Swainster

  • Frequent Contributor
  • **
  • Posts: 258
  • Country: sg
Re: Keithley 616 Electrometer input stage replacement
« Reply #3 on: March 09, 2023, 11:30:21 am »
I replaced my 616's input amplifier with an LMC662 'bodged' onto the original board - the inspiration came from the popular 617 thread: https://www.eevblog.com/forum/testgear/fun-with-low-leakagebias-current-femtompere-electrometer-keithley-617. I forget exactly how I wired it now, but I recall that the supply was dropped down using a zener shunt regulator or 2. It seem to perform OK at the time, except that the zero adjust was not very friendly, having to crank the knob like crazy to get anywhere (and only just having enough range to zero). I meant to go back and address this but never got around to it.

More recently I wanted measure some multi gigaohm resistors and my first 2 instruments wouldn't cooperate so I pulled out the 616. Well, it is said that trouble comes in 3s and sure enough my 616 seems to have developed significant leakage, though possibly before the input amp as it is relatively constant across the ranges (i.e. when measuring resistance it is always 2 or 3 Gigaohm, no matter what range). I guess that a likely suspect would be the 3 lug triax connector that I fitted (i.e. dodgy workmanship on my part), or possible the zero check relay. Anyway, too many things on the to-do list at the moment, though the 616 might be worth bumping up the list if I'm going to be playing around with more tiny current measurements... got 2 picoammeters on the same list though :palm:

After much searching I found a pic of the rework... Looks like I kept the original op-amp and just replaced Q101 with one amp from the LMC662, and shorted across Q107... No wonder the zero adjust was so feeble. Ugly bodge indeed!

 

Offline muvideoTopic starter

  • Frequent Contributor
  • **
  • Posts: 418
  • Country: it
Re: Keithley 616 Electrometer input stage replacement
« Reply #4 on: March 17, 2023, 01:19:03 pm »
Well, finally I had a little time to test the circuit.
This is my prototype, ugly as few:


Zero is decent (input Volts, normal, shorted clips on the cable):
https://photos.app.goo.gl/X6KuVJP2kUKAxUt16

Open terminals on lowest current setting:
https://photos.app.goo.gl/rCng2ozZPkXNt7GJ7

Set 220 to 5pA:



and -5pA:



Not bad for a few euro opamp.
Now a proper pcb is needed.


« Last Edit: March 17, 2023, 02:05:35 pm by muvideo »
Fabio Eboli.
 
The following users thanked this post: alm

Offline muvideoTopic starter

  • Frequent Contributor
  • **
  • Posts: 418
  • Country: it
Re: Keithley 616 Electrometer input stage replacement
« Reply #5 on: March 17, 2023, 03:29:09 pm »
I have uploaded pcb files (Kicad) in this repository:
616 Buffer on GITHUB

PCB ordered, now back to other repairs :)
« Last Edit: March 17, 2023, 03:32:42 pm by muvideo »
Fabio Eboli.
 

Offline MathWizard

  • Super Contributor
  • ***
  • Posts: 1377
  • Country: ca
Re: Keithley 616 Electrometer input stage replacement
« Reply #6 on: March 13, 2024, 10:47:38 pm »
I had my eye on a Keithley 616 electrometer, sold for parts or untested anyways, for only ~$200, but it must have sold already. I see newer models are into the many $1000's.

I didn't look in detail at the schematic, but reading a few threads on here about some 1970's Keithley electrometer's, seems there's some pretty rare and custom parts involved in repairs.

But does anyone make DIY ones, based off the older design's that are out there, but just with modern off-the-shelf parts ? I've seen some expensive resistors, and op-amps, but even to make some 1950's 1960's quality electrometer, what might it cost ? Never mind how to design it to actually work, I'm just curious. Someday I need to try making some reference boxes, just with my random parts box, and see how good they are.
« Last Edit: March 13, 2024, 10:49:21 pm by MathWizard »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf