Mains transformer for Keithley 195a

[giosif]

Hi,

I am trying to repair a Keithley 195a which is tripping the mains breaker in my house when trying to power it on.
After some investigation, I think I've found the culprit: the mains transformer in the meter is busted.
The proof: resistance between one end of one of the primary windings (pin 12 in the schematics) and the shielding on the primary side (pin 6) is about 28 kOhms (measured outside of the circuit).
I am not that familiar with transformers and shielding, but my thinking is that that resistance is way too low and it should, in fact, measure "infinite".
Would you agree?

[coromonadalix]

I thnik you're measuring wrong  28k ohms is 28000  ohms,  way to high for a short ??

[giosif]

I am measuring correctly and it is 28 kOhms.
And, while this is not a short, I think the resistance is too low for something that, at least in my mind, should be completely isolated from one another: primary winding vs shielding.
And the reason the mains breaker trips, I am guessing, is because of the leakage current between Live and Earth due to the above (i.e. 230V / 28,000 Ohms = 8.2 mA).

But this is all (educated?) guesswork on my side and I would appreciate it if someone could confirm or challenge it.

In the meantime, I was able to turn on the meter by providing the required DC power to the 4 rails (digital 5V, analog 5V, analog 15V and analog -15V) using my bench power supplies and the meter is working (at least DCV readings seem to be ok).
In any case, this proves there is no short past the voltage regulators.

[coromonadalix]

Do you have intelligent sensing  breakers, if they sense a leak they will trip for sure

If i recall you have a ground wiring going to the transformer shield / casing  pin 6  (just unsolder the hole pin, and adjust the pin to not touch the pad) once isolated,  is everything still works fine ?


Not sure, was it copper shielded ?   can you remove the foil, cant recall if it was internal ....   and sadly  the last(s) option(s) would be to rewire the transformer and or get an used 195a   to rebuild your meter.

The 5v digital supply is the one who work at 1 or 1.5 - 1.8 amps

[giosif]

Apologies, I should have included pictures from the beginning (please see attached).

Pin 6 does indeed have a wire to the metal core of the transformer and is eventually connected to mains Earth.
Now, I also thought about desoldering that wire but, then, if there is leakage somewhere inside the transformer, it means the metal core might become energized.
I am not sure I would want to try that.

On top of that, I noticed that pin 10 is not connected to anything and, at least according to the manual, that should not be like that.

With regards to the shielding, it is internal.

Yeah, rewinding is something which crossed my mind, but I wouldn't do it myself and using a professional service would likely be non-viable financially.

[bdunham7]

28K DCR to the shield indicates some sort of failure, and unless you know what it is, the meter should be presumed unsafe--and your RCD apparently agrees.  The actual leakage current may be more due to capacitive coupling, depending on the nature of the failure.  One thing you could try is baking the transformer at 100-150C for a while and then checking to see if that leakage resistance changes.  If moisture has seeped in somewhere, that may fix it for a while, although you would need to do more for a permanent fix.  Definitely don't go disconnecting the shield, things may get worse and fry your meter or you. 

How much room is in the meter?  You might be able to replace the transformer with two, a larger 5V for the digital and another for the analog stuff.  A modern double-bobbin variety would not need shielding.

[coromonadalix]

Sure    seeing theses values between the shield and a ac pin is the beginning of a failure of some sort,  heat  failing insulation etc ...

Maybe in a near future the xformer will surely die ...

[macboy]

The fault may measure 28 kOhm at a few volts that the meter uses to measure resistance, but with mains voltage, it may arc over and become effectively a short circuit. An insulation tester capable of measuring resistance with high voltage would tell you much more than a multimeter could.

[giosif]

Thank you, everyone, for your input!
I think it is quite clear that the transformer has at least one fault, even if only based on the fact the rest of the meter is functional while, with the transformer installed and powered through it, it trips the breaker.

At this stage, I see the following options:
1. Pay a company to rewind the transformer - I expect this to be costly (but happy to be contradicted).
2. Try to find the same transformer as spare (either NOS or from another broken 195a EDIT: or 196, by the looks of it /EDIT) - this would be ideal, but is likely to take a while, especially here in Europe where old Keithleys don't seem to be that common.
3. Look to find an equivalent replacement, either as one transformer or multiple, like bdunham7 mentions (I believe there should be enough room for two) - this is a more likely scenario and probably not as costly.

So, right now, I'm leaning towards option 3.
On with the search...

[bdunham7]

So, right now, I'm leaning towards option 3.
On with the search...

 
Now you just need to know the AC current requirements.