Keithley 181 - Nanovoltmeter

[branadic]

Hi,

we have a Keithley 181 Nanovoltmeter at work, which was turned off for a while. I didn't even know that this gear is in one of our lockers. So when I found it I went to my desk and turned it on, still functioning 
But after a few weeks of working without problems I noticed, that the 7 segment display turned of. Powering it on and of showed only a short sign of life. So what is your suggestion, what could be faulty? Only some caps are dry and recapping is necessary?
Hints are welcome.
I already downloaded the manual with the schematics.

-branadic-

[TiN]

I'd start with usuals, PSU checks, recap and contact checks on cables/switches.

[branadic]

Had the time today to take the thing appart and indeed I have a capacitor issue. Time for some complete recapping.
Original caps are from Richey any suggestions what to replace it with? Nichicon, Rubycon or are Yageo enough the effort?

-branadic-

[pelule]

I recommend a good quality brand like Nichicon to get long lifetime.
Absolute value is not critical, as they hav usually ~10% higher values anyhow to keep them in spec during aging.
/PeLuLe

[branadic]

Maybe someone can help, what brand is this capacitor inside the Keithley 181?

-branadic-

[GreggD]

That yellow cap is metalized poly(something).
It is unlikely to ever go bad. Should last forever.

[TimFox]

That film capacitor is from Electrocube.  The 650B series is still in production, but I don't know the exact meaning of the second line on the package.

[branadic]

Thanks TimFox, no I do know where the capacitors on my LT-LTZ1000 board are coming from. Seems like these capacitors are not available at digikey or any other distributor in germany

-branadic-

[pelule]

Hi branadic,
I have checked the distributors list of Electrocube (www.electrocube.com).
In the US they use:
   FUTURE ELECTRONICS
   www.futureelectronics.com
Future sells also in Germany and offers caps from Electrocube:
http://de.futureelectronics.com/de/Technologies/Product.aspx?ProductID=650D1A473JELECTROCUBE1063365&IM=0
/PeLuLe

[Shock]

1.0 MA = 1uF "A" 0.1% tolerance?
What does it say on the parts list?

[RobK_NL]

1.0 MA = 1uF "A" 0.1% tolerance?
What does it say on the parts list?

"A" denotes 50V working voltage.

[branadic]

As I said, I had some capacitor issue, some big 10.000µF inside was blown, probably due to many years without any power supplied to it. However, I ordered caps to replace all of them. I received some new Nichicon caps today, soldered everything to its place and what to say, the kit is now working again. 

To make it easier for all of you out there planing to replace the caps on a Keithley 181 here is a list of caps that are installed and the Digikey number I replaced it with:

• 1x 10.000µF/25V: 493-1072-ND
• 3x 1.500µF/25V: 493-1833-ND
• 2x 100µF/35V: 493-4501-1-ND
• 2x 6.8µF/25V: 493-13850-1-ND

As you can see, they perfectly fit their place. The 10.000µF is somewhat bigger in diameter, but no problem whatsoever.

-branadic-

[branadic]

The sad thing about this repair is, that Keithley part of Tektronix does not longer support this beautiful piece of gear, neither repair nor calibration. That's why it is now on my privat desk.
I don't have the possibilities to check whether it is still within spec or if it needs some calibration. And I don't have the gear to calibrate it

-branadic-

[Neomys Sapiens]

There are other calibration services that support it over here, such as
https://www.esz-ag.de

Helmut Singer might do it as well.

[branadic]

Thank you for the hint. I'm currently asking myself if I should sell the Nanovoltmeter with its accessories (original manuals and original test cable).

-branadic-

[Johnny10]

I have used mine twice in three years, I wonder the same as you.

Makes you wonder just how long to keep items you aren't using???
I believe my HP3457A which I use daily is accurate enough for my needs.

[branadic]

There are other calibration services that support it over here, such as
https://www.esz-ag.de

Helmut Singer might do it as well.

Contacted both of them and was told that HS uses an external service for calibration and esz doesn't offer their service for privat persons and the last time they calibrated such a unit was in 2009.
However, someone on this board will now help. The good thing is, I have the maintenance kit 1815 and he seems to have most of the necessary equipment to do the calibration. On the other hand he will use the unit for some own measurements. Quid pro quo. 

-branadic-

[Neomys Sapiens]

Yep, ESZ requires a bit of posing.
Make up a 'Meier GbR' or whatever is sufficient. Its about liability toward private end customers. They don't insist on fiscal documentation or anything like that.

[branadic]

Hi,

after the caps have been replaced I found, that the meter shows quiet high noise in 2mV range with input shorted.
The unit was rarely used before and powered of for at least 10 years. Dr. Frank adjusted the ranges with equipment he had at hand, but noise is still high.

I've heard that the matched input FET transistors pair could be the reason, but as mentioned the unit was rarely used and turned off for a while. My imagination is that this matched pair can only break when unit is often used, not when turned off.
I don't have the 100:1 divider, which is less of a problem as I can copy that with some Hammond aluminium die cast, some selected PWW (10k : 100R) and Pomona connectors.

But I don't have the low noise preamp (Ithaco 1201) too, that's needed for troubleshooting compared to 181 service manual.
What I could use for first indication is a Keithley 2002 togehter with a live data plotter software as a noise analyser to locate the source of noise. On the other hand I found the manual of the low noise preamp and the first stage looks like some INA103 + DSO could do the job as well.

What is your suggestion on that and does anyone of you have any experience on what to look at? Thanks for your help.

EDIT: Here is a video I got from CalMachine how the noise can look like on a 181, but I do have intermittent jumps of 500nV in the 2mV range.

-branadic-

[Dr. Frank]

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.

[Kleinstein]

The low noise amplifier (Ithaco 1201) is not that special. It is a low noise high impedance amplifier. The input of the K2002 might be about as good. Not many places in the circuit would actually need the high impedance.

There are more parts than the expensive input FETs that can get damaged or turn bad over time.

Quite some test could be done without an extra meter. One of the first and more obvious test would be to check if the noise is coming from the preamplifier or the rest of the instrument. So check for noise in the 2 V range  - this should be reasonably low. One could also measure the preamplifier output  with an external meter, so only use the preamplifier of the 181, but not the ADC.

With the preamplifier, one of the first parts to check would be the supply, including the bootstrapped one.

One could also check the input amplifier before the output buffer.

[VintageNut]

The 181 and 182 are one of the few KE vintage pieces that I have not had the itch to buy. I have two of the 147 which have a 30nV full-scale range. I have a 148 which has a 10 nV range but it is difficult to actually make a stable <10 nV phenomenon in open air.  I have had to measure 20nV drop across a tetrajunction. None of the digital brown boxes can do this measurement.

I would use a KE 260 to calibrate the 181. I have two of the 260 nanovolt source boxes. One I re-capped and checked it against one of my 147 nanovoltmeters.

I checked the claimed <=4 nano ohms of my Ohm Labs 100-0 with the 147 nanovolmeter.

Personally,I would call the 181 a microvoltmeter. You can actually measure a microvolt with it.

Good luck with the troubleshooting.

[Echo88]

I own both a K181 and K148 and while the 148 is indeed more stable and has less noise, i value the K181 for its digital zero button instead of the finicky K148-potentiometer, the faster thermal equilibrium and the silence compared to the buzzing of the K148. With averaging in my software i get about 6nVpp noise with the K181, so it is indeed a nanovoltmeter.

But its really an excursion in the world of thermal stabilization, patience and error source avoidance.   

[VintageNut]

I own both a K181 and K148 and while the 148 is indeed more stable and has less noise, i value the K181 for its digital zero button instead of the finicky K148-potentiometer, the faster thermal equilibrium and the silence compared to the buzzing of the K148. With averaging in my software i get about 6nVpp noise with the K181, so it is indeed a nanovoltmeter.

But its really an excursion in the world of thermal stabilization, patience and error source avoidance.   

I agree that it takes patience and knowledge to measure below uV. I have spent days learning how to measure small resistances and small voltages.

Your 6nV p-p noise is only one source of error. There is also the range error and/or offsets and/or nonlinearities.

My point remains. I do not have a compelling reason to acquire a 181 or 182. If I needed to measure small voltages or small resistances for daily work with lots of digits of precision, I would either acquire a KE 1801 or create a similar solution from an EM Electronics A10 or A20. The A10 is what Keithley used as basis for the 1801.

I had occasion to use a KE 2001 + 1801 for a couple of hours. If I could buy one, I would snap it up in a moment.

[branadic]

I own both a K181 and K148 and while the 148 is indeed more stable and has less noise, i value the K181 for its digital zero button instead of the finicky K148-potentiometer, the faster thermal equilibrium and the silence compared to the buzzing of the K148. With averaging in my software i get about 6nVpp noise with the K181, so it is indeed a nanovoltmeter.

But its really an excursion in the world of thermal stabilization, patience and error source avoidance.   

I agree that it takes patience and knowledge to measure below uV. I have spent days learning how to measure small resistances and small voltages.

Your 6nV p-p noise is only one source of error. There is also the range error and/or offsets and/or nonlinearities.

My point remains. I do not have a compelling reason to acquire a 181 or 182. If I needed to measure small voltages or small resistances for daily work with lots of digits of precision, I would either acquire a KE 1801 or create a similar solution from an EM Electronics A10 or A20. The A10 is what Keithley used as basis for the 1801.

I had occasion to use a KE 2001 + 1801 for a couple of hours. If I could buy one, I would snap it up in a moment.

Well, for precise and low noise measurements we do have a Keithley 2182A at work, no problem whatsoever. But the linked video above indicates, that you can make stable and low noise measurements with Keithley 181 as well if everything is in perfect condition.

Keithley 2002 is now warming up to start some poking around to find the source of noise and thus the faulty part.

-branadic-