Keithley 181 - Nanovoltmeter

[Kleinstein]

I simply checked  the  noise on your 181A with the 3458A, in its 10mV A.C. range.
This did the job very well , if I remember right, it was about 0.5 mV rms, at the 1000x amplifier output
Frank


P.S.: That measurement was the 'pump-out', as Keithley used to call it, it's nothing else but the chopper spike suppression, which has to be trimmed.
These 520µV, measured easily with the 3458A, were well within  the limits, but that has nothing to do with the amplifier noise, and also a pre-amplifier is not necessary for fault-finding.

The error is definitely located in the pre-amp, as the 2V .. 1kV ranges were ok.

What's really strange is that the excessive and sporadic noise, of 500nV up to several mV is in the 2mV range only (resolves 1nV or 6 digits), but it's not visible at all in the 20mV range.
As the 20mV resolves 10nV, it should be visible on the 2nd last digit, but there's nothing.
And the only difference in the setting is a reed relay, which is switched off (inactive) for the 2mV / 1000x range of the pre-amp.
Therefore, all semiconductors might probably be ok.

Either this reed relay is not ok ( so de-solder it in first place).
Or the whole pre-amp might oscillate at 1000x, which I didn't check with an oscilloscope, only with my 34465As digitizing function, i.e. at 50kHz sampling rate / bandwidth.

This description already strongly points towards the relay used to switch between the 2 mV and 20 mV ranges.  The pot R409 might be a potential problem too.
The good thing it, the input FETs are very likely not the culprits.

[Echo88]

Indeed the modules from E.M. Electronics leave the standard nanovoltmeters in the dust. A pity that they are pretty much unobtanium on the used market. Seems theres only a real small market for excellent nanovoltmeters.

[branadic]

I will work through the nanovolt preamp step by step: I have the newer version with minor changes within the circuit. This starts with U404 which is a LF351 with output at pin 6 instead of the original pin 7.

1. Checks of Preamp Zero (pin 6 of LF351N with 2mV range selected and input shorted)
As found: ++7.964V --> this is way out of spec given a maximum of +/-2V

Time to desolder R423 and R424, insert some 20k pot, find the new ratio where pin 6 is within +/-2V, measure it and replace R423 and R424 with the new ratio. I did it the fast way and paralleled some decade resistor box to R423.

original ratio: R423:R424 = 9.5458k : 8.5707k --> +7.964V
new ratio: R423:R424 = 9.5458k|| 8.2k : 8.5707k --> +425mV or R423:R424 = 9.5458k|| 6.8k : 8.5707k --> -356mV

So we need R424 = 2* R423 to get close to zero. So something like 6.8k and 12k could do the job quite well.

[branadic]

Either this reed relay is not ok ( so de-solder it in first place).
Or the whole pre-amp might oscillate at 1000x, which I didn't check with an oscilloscope, only with my 34465As digitizing function, i.e. at 50kHz sampling rate / bandwidth.

Meanwhile I'm sure it has nothing to do with the relay, which is only active in 20mV range and off in 2mV and 200mV range. I guess I need to power the preamp by a lab power supply to make sure it's not a power supply issue.
By the way, the preamp is based on of the shelf parts with original manufacturer and not Keithley labeling. 

-branadic-

[Alex Nikitin]

I'm slowly restoring my Keithley 181. I've replaced all electrolytic caps as some of them already failed, and Volts ranges are working fine with the exception of 20V range which had a very high offset drift (about 1mV over 20 min after switching to the range). I've replaced the relay K302 with a small 5V REMtech reed relay I had in a draw, the drift now 10 times better (under 100uV over 20 or so min) but still quite significant    . I am waiting now for some low thermal Pickering reed relays, hopefully these will resolve the problem more or less completely. On the mV ranges I have a sort of a popcorn noise with random shifts so have to investigate the preamp section later. The ADC in the K181 is quite interesting, simple 6.5 digits design with 10ppm linearity and reasonably low noise (below 5ppm p-p on 2V range). The reference is LM399. One interesting feature is an isolated (referenced to the GPIB ground) analogue output with a 12bit DAC, which can be set to show the top 3.5 digits or the bottom 3.5 digits.

Cheers

Alex

[branadic]

I have a similar problem on the mV range. I probed around today and the frist thing broken is in the bootstrap part of the nanovolt preamp. Either the 6.4V zener diode VR404 = 1N4571 or the fet CR406 = J505 is broken. I could find 1N4571 at Mouser, but any advise for replacement of the fet?

-branadic-

[BU508A]

Hi branadic,

I have a similar problem on the mV range. I probed around today and the frist thing broken is in the bootstrap part of the nanovolt preamp. Either the 6.4V zener diode VR404 = 1N4571 or the fet CR406 = J505 is broken. I could find 1N4571 at Mouser, but any advise for replacement of the fet?

How about the LS845?
http://www.linearsystems.com/product-search-result.html?type=products&partnumber=LS845

This site is saying, the LS845 could be a replacement for the J505
http://www.allxref.com/search.htm?part=J505

Update/edit:
possible source in Germany:
http://ib-fluck.de/produktlinien/

HTH

Andreas

[Alex Nikitin]

I have a similar problem on the mV range. I probed around today and the frist thing broken is in the bootstrap part of the nanovolt preamp. Either the 6.4V zener diode VR404 = 1N4571 or the fet CR406 = J505 is broken. I could find 1N4571 at Mouser, but any advise for replacement of the fet?

-branadic-

J505 is specified at 1mA, so most JFETs with ~1mA current will do, of a JFET with a higher current and a resistor to get 1mA.  VR404 can be replaced by 6.9V LM329 I suppose.

Cheers

Alex

[branadic]

I found some J505 and ordered them. Replacing the temperature compensated zener 1N4571 (6.4V, 1mA, 50ppm/K) by LM329 needs some further tweeking of the resitors too, as the output of the bootstrap circuit should deliver 10V ±10%.
If someone on this board has a few 1N4571(A)/1N4572(A)/1N4573(A)/1N4574(A) laying around I would be happy to buy them. Otherwise I have to order at Mouser. They stock them fabricated by Microsemi.

-branadic-

[Alex Nikitin]

I found some J505 and ordered them. Replacing the temperature compensated zener 1N4571 (6.4V, 1mA, 50ppm/K) by LM329 needs some further tweeking of the resitors too, as the output of the bootstrap circuit should deliver 10V ±10%.
If someone on this board has a few 1N4571(A)/1N4572(A)/1N4573(A)/1N4574(A) laying around I would be happy to buy them. Otherwise I have to order at Mouser. They stock them fabricated by Microsemi.

-branadic-

With 7.0V from LM329 you'll get about +/-11V without changing the resistors, so should be OK.

Cheers

Alex

[branadic]

Received some 1N4571A and NOS J505 that I replaced today, giving 9.77V and -9.70V, which is well within 10V +/-10%. This replacement also requires an offset adjustment of U404, thus a replacement of R423 and R424. Currently I have a 20k pot at the Berg pins installed and observe, if I still have large jumps.

-branadic-

[branadic]

I have put everything together, even though the board isn't cleaned yet. I've put a short directly at the screw terminals of the nanovolt preamp input under the shielding and put everything together with the 20k pot installed back and closed the chassis. The meter is now running a while and all jumps in reading have vanished. Instead the display is running up and down a view nanovolts around an offset of a few 10nV. Looks like a decent repair by now 

The behaviour is now similar to what is shown here:

https://xdevs.com/cm/RandomPics/K181_Cable/20171009_194206.mp4

-branadic-

[branadic]

It's done.
After binning some 9k and 11k 1% resistors today I found a good combination to replace the temporary 20k pot at the Berg pins. Pin 6 of U404 now delivers an offset close to zero in the 2mV range, which is well below the required +/-2V.
I carefully cleaned the whole preamp with Aceton and IPA with some cotton buds, removed the temporary short at the preamp input, reconnected the cable at the input, installed the shielding box and put everything back together.

Is someone interested in a working Keithley Model 181 Digital Nanovoltmeter with Model 1506 input cable, Model 1815 Maintenance Kit and another set of spare parts for CR406, CR407 (2x J505) and VR403, VR404 (2x 1N4571A)?

-branadic-

[Neomys Sapiens]

Is some interested in a working Keithley Model 181 Digital Nanovoltmeter with Model 1506 input cable, Model 1815 Maintenance Kit and another set of spare parts for CR406, CR407 (2x J505) and VR403, VR404 (2x 1N4571A)?

-branadic-

Hmmm, what's your asking?

[Alex Nikitin]

At last I've had a chance to fix my Keithley 181. At the end the main reason for the excessive noise turned out to be an incorrect resistors ratio in the offset compensation network. It looks like at some point input JFETs were replaced and the offset resistors were not changed. After re-setting the offset everything is quiet, here is the 100s trace from the analogue output , a shorting plug on the input and zero ON (without zeroing the offset is about 25nV). 4nV per division vertical scale.

Cheers

Alex

[HighVoltage]

I bought a Keithley 181 for a low price and it arrived dead with lots of loose parts inside.
It must have been thrown around during shipping.

I have not started the trouble shooting. Probably best to start with the PSU and electrolytics.
The analog output binding posts are broken.
And a loose brass nut was inside that does not have a place of origin as far as I have seen, hmmm...

Interestingly, the loose bindings posts have no threads on them.
But the two loose cables have lugs on them.
Any advise before I start?

Here are some pictures:

[Shock]

A good inspection to check safety and see if any physical problems are going to cause more damage. If it's been worked on, what may have been changed. Then confirm what the no power issue is.

I generally work on the fault before reconditioning anything unless it's dirt or bad controls. These type of models are known for stuck button issues and the 170 series is dirty ganged switches. Anyway the service manual is the way to go as it should have a test/check procedure.

Any tantalums and the electrolytics are probably mandatory replacements if it's a keeper. 

[Neomys Sapiens]

That red jack (not a binding post) doesn't seem original. The original had binding posts and the brass nut would fit there. That other part I cannot recognize.

[lowimpedance]

Those leads go to standard looking 4mm binding posts (gold plated ?). Any way google images turn up a picture on TiN's archive showing an inside photo which includes the binding posts on the rear.
As stated above the only original bit appears to be the nut  .

[tooki]

I bought a Keithley 181 for a low price and it arrived dead with lots of loose parts inside.
It must have been thrown around during shipping.

I have not started the trouble shooting. Probably best to start with the PSU and electrolytics.
The analog output binding posts are broken.
And a loose brass nut was inside that does not have a place of origin as far as I have seen, hmmm...

Interestingly, the loose bindings posts have no threads on them.
But the two loose cables have lugs on them.
Any advise before I start?

Here are some pictures:

This model shipped with proper binding posts (Polklemmen) on the back, not the press-in banana jacks (Bananenbuchsen) shown here. So definitely an aftermarket modification.

[HighVoltage]

Thank you all for the feedback.

It is powering ON now.
The mains switch was broken and I just fixed it.
The instrument is working now and not far off in the 10V calibration.

Ok, now I will be looking for some fitting binding posts.

Any idea of where to get a fitting shorting block for the zero calibration?

[HighVoltage]

Pomona 3770 binding posts fit perfectly
Instrument is running perfectly now.

[Dr. Frank]

Hello,
maybe you also need these Amphenol input connectors ?
I bought them quite cheap at TME, silver surface only, but working.
That's 97-3106A-16-11P and 97-3057-1008-1
You might use one of those to build a zero connector.

For shorting, simply get a thicker copper wire, from a stiff power cable, isolate, bend and plug it into the holes. Add a small, thermally isolated box around, ready.

Frank

[HighVoltage]

Hello,
maybe you also need these Amphenol input connectors ?
I bought them quite cheap at TME, silver surface only, but working.
That's 97-3106A-16-11P and 97-3057-1008-1
You might use one of those to build a zero connector.

For shorting, simply get a thicker copper wire, from a stiff power cable, isolate, bend and plug it into the holes. Add a small, thermally isolated box around, ready.

Frank

Thank you, Frank.

I made my shorting block from one of these Amphenol connectors with a little custom aluminium cap fitting the end of the connector for shielding purpose.

The copper wire I used is 99.999 % pure copper at 2.7 mm diameter, fitting snugly in to the openings of the Amphenol connector and then soldered tight.

[Kosmic]

Nice work, look really good 

I also need to make one of those for my 181.