Keithley 155 Null Detector - Repair and Restauration attempt

[HighVoltage]

I got this Keithley 155 Null Detector really cheap, because it was not working and not in the best condition.

Opened it up and found leaked 9V batteries and the nasty tar based foam (to hold the batteries in place) everywhere.
(What a nasty stuff to clean.)
 
I found 4 pieces oversized 9V block batteries inside.
Does anyone know, if these are still available or what to use instead.

The Manual with circuit diagram can be found easily with a google search.

Any hints of what to do and what not to do in regards to repair with such a Null Meter?

OK, I will take it apart and do some cleaning first.

[ManateeMafia]

I have a couple of them. They had the same dry rotted foam that was everywhere around the batteries. Only one of them is functional.

The bad one pegs to one side but I have only done some of the checks in the service manual. The other one is fine and I put in regular 9v batteries. The foam needs to be cleaned up in it.

The foam is nasty stuff as it took some hours to get every last bit of it out. I haven't come up with a permanent solution but I have some 9v battery boxes and I think some double sided foam tape will work.

[Kleinstein]

The type of 2N6 battery is still available, but rather expensive.

If space is sufficient you might get holders for 6 AA (or maybe AAA) cells inside. Some even come with the same battery clips.

[TiN]

Subbed, waiting for good repairs. 

Hopefully this will kick me into getting dat 845AB fixed (already have all the replacement resistors for every carbon type on in meter).

[Testtech]

Regular 9V batteries work fine and last for years. I made a foam block to hold and isolate them.

[HighVoltage]

Thank you all for suggestions, I followed the advise and installed standard size 9V block batteries.
I found some self adhesive foam in the drawer and it works perfectly to keep the 4 batteries in place.

The most work went in to cleaning up and replacing some of the broken cables.

Interesting to note, the metal battery holder is floating on some PTFE (Teflon) spacers.
It looks like Keithley put a lot of thought in to the details of this instrument.

But so far, it seems to work.
Basic functions are alright but it also seems far out of calibration.
1 V applied shows about 0.9 V on the analog dial.
Looks like I have to read the manual in detail.

[quarks]

Bookmark

[pelule]

Hi HighVoltage,
it is a high sensative null-meter (1µV, 10x Scale).
Thus I would highly interested in a top view picture of the mounted tool (but with open lid). I am interested into the shielding construction in general.
I assume it is battery powered only - correct?
I did restauration of two a not that sensitive (3µV)  HP149A, which could powered also from AC. Hee I found the internal wiring (where are the wires going aside) had some major effect in noise sensivity.

[dr.diesel]

Can I join the club?  Mine has been sitting on the repair bench for too long!

[HighVoltage]

Hi HighVoltage,
it is a high sensative null-meter (1µV, 10x Scale).
Thus I would highly interested in a top view picture of the mounted tool (but with open lid). I am interested into the shielding construction in general.
I assume it is battery powered only - correct?
I did restauration of two a not that sensitive (3µV)  HP149A, which could powered also from AC. Hee I found the internal wiring (where are the wires going aside) had some major effect in noise sensivity.

Pelude
I will take those pictures tomorrow for you ...

Yes, my meter is battery powered only!
But ... I found the 230V PSU part of it separately and just bought it. That one is dead and needs repair as well.

My 155 has been on a dual power supply and turned ON for about 10 days now.
I did a first calibration 10 days ago, based on the instructions in the Service Manual.
All the POTS needed a few good movements before they got stable and reduced the noise.
It has applied 1.0000 V now and since then it seems the needle has not moved at all.

In the uV range it gets totally over sensitive. (That might be normal)
Getting the hands close to the instrument moved the needle right away.
It definitely needs a guard plugged in and probably some quality and shielded cables.

It is a surprise, that this instrument does not have a triaxial connector, as the Keithley Electrometer have, for instance.


dr.diesel
What is wrong with yours ?
I am glad we have the full schematics and probably can find all problems.

[dr.diesel]

dr.diesel
What is wrong with yours ?
I am glad we have the full schematics and probably can find all problems.

Unknown, I bought it cheap and just haven't gotten around to messing with it.  I'll poke at it in the next few days.

[HighVoltage]

I just finished another calibration cycle on the instrument and it seems to work really good.
But I think my Advantest R6144 is reaching its limit at the lower uV sources.

+ 10V shows exactly +10V on the scale
- 10V shows exactly -10V on the scale
So do all values in the mV range

+ 300 uV is full scale perfectly
- 300 uV seems to be off by the width of the needle.
+ 10 uV seems to be off by two little lines

Ooops .... I need a better source for this little Null-Meter calibration

The restoration of the case and handle came out good.
I exchanged all external rusty screws with stainless steel screws, all in American thread size.

Now will keep it running for a while and see if I notice a drift.
And next is to fix the 230V external power supply.

[pelule]

But I think my Advantest R6144 is reaching its limit at the lower uV sources.

I wouln'd trust you Advantest R6144 in this specifc range/value either.
Better to use a higher output voltage (10mV ?) and feed via 1000:1 resistor devider, I guess.
Input impedance of the 155 should be high enough.
By the way, I have searched for the schematics, t ohave a look into the differences to my HP419A (f.e. FET chopper).
But just found the instruction/service manual without schematics.
Any tip?

[dr.diesel]

Any tip?

Please see below:

[pelule]

@ dr.diesel
Great. Many thanks for the schematics.

[guenthert]

[..]
In the uV range it gets totally over sensitive. (That might be normal)
Getting the hands close to the instrument moved the needle right away.
It definitely needs a guard plugged in and probably some quality and shielded cables.

It is a surprise, that this instrument does not have a triaxial connector, as the Keithley Electrometer have, for instance.

Count your blessings!  Triaxial cables are astonishingly expensive (at least the fully assembled ones).  Here they aren't needed, as the input impedance isn't actually all that high (specified as 1MOhm in the lower ranges, same as Fluke's "high impedance" Null Detector Model 845AB -- the older HP419 specifies even only 100kOhm, iirc).  Further, a Triax connector would be impractical, as Null Voltmeters are often used to compare (measure the difference of) two voltage sources.

I rather wonder why they don't use BNC connectors for (simple shielded) coax cables, one for each input (with shield to chassis).  Hmmh, both the Keithley 155 and the Fluke 845AB specify an input isolation of 10^12Ohm (impedance between either input and chassis/earth).  Well, that wouldn't do much good with coax cables ...

[TiN]

BNC ain't good for thermal EMF, which are big error factor on uV ranges.

[HighVoltage]

Having used the 155 for a while now, I find the banana plugs working well. And for very low voltages I have been using a self made cable with copper/gold-plated spade connectors and PTFE insulation. It works very well. No need for Triax or BNC cables.

Before I had a Null-Detector I had no idea what I would be missing.

[Bryan]

Ok, I got to ask, but what is a NULL detector?

[HighVoltage]

It is like a comparator to a known voltage value with a very high accuracy.
The Null Detector can identify very well, when the difference to the known value is ZERO.
Or, in other words you can find a precision deviation between a known value and the DUT.

This is a really great instrument for calibration purposes.
Many voltage or current calibration procedures require a Null-Detector for exact settings.

[Kleinstein]

The null meter is a special kind of voltmeter, for low DC voltages. Often they are battery operated or at least have very good insulation at the supply side.

They are made for high precision near zero, not high accuracy:
So they are very good in finding zero, but the scale factor is not specially exact and linearity is not important. So old style analog meters are still good (arguably better than digital) for this - no need for a large number of digits. Often there are also ranges for higher voltages (like 10 V), but the main thing is performance in something like the 1 µV to 1 mV ranges.

[jeroen79]

Isn't the benefit of a null meter that it draws no current from the DUT once nulled with another reference voltage?

[Kleinstein]

The zero current at zero voltage come somewhat naturally, unless you have an bias current source in the meter. This also applies to old style analog meter.

The normal use is to have an adjustable voltage source (e.g. KV divider and stable source) to bring the null meter to zero. So often you find the null meters combined with such a adjustable voltage source. But you can have a null meter without such a source.

In old days they often used just a sensitive moving coil galvanometer without any extra series resistor for the sensitive part. The one with optical readout and light pointer are really sensitive - I remember some thing like 1 mV full scale, and this was not an extra sensitive one. Due to the copper winding resistance an primary current reading the gain was not temperature stable, but the zero is.

[HighVoltage]

Two of the Keithley 155 were just sold on ebay Germany yesterday, for Euro 479
That was above my limit.

Did anyone here buy them?

[guenthert]

Two of the Keithley 155 were just sold on ebay Germany yesterday, for Euro 479
That was above my limit.

Did anyone here buy them?

Holy moly!  Contrast that to most recent electronic fleamarket in the south (S.F.) bay, where a HP419A went for $5 (i hesitated for a moment as I have already two and foresee a move in the near future and that was a moment too long  ).

Sure, a Keithley 155 is more desirable (not the least for its much higher chances of actually still working), but still.