Keithley2001 transformer failure （Need help）

[BD1QMP]

Dear friend
I am from China.
I got a broken Keithley2001 recently, display overflow error.
When I open the cover, I found that the transformer is fail. the second windings are short (the windings for +-15V and 5V of analog board) .

Do you have the specs of the transformer?  such as the AC voltage and the current?


thank you very much!

DE BD1QMP

[alm]

Are you sure it's the transformer that's shorted, and not something on the board like a rectifier? Did you measure the AC voltage from the windings without anything connected?

See the service documentation available at TiN's Keithley 2001 page to see if it might be any help. I'd try to buy an original transformer (from a donor unit or from Keithley/Tektronix), because the transformer contains an electrostatic shield to improve isolation between the DMM and mains. There are some Keithley transformers for sale on eBay, but I don't know if any of them are for the 2001.

[inse]

Displaying overflow error and having shorted the secondary windings somehow does not line up…

[Swake]

I assume the controller and display still working because the digital part has a separate power supply.

[MathWizard]

If it actually is just that transformer winding, I'd cut that connection on the PCB, and just get some other little transformer with the correct voltages, and wire it in. It would be a lot cheaper, and it's rating should be pretty standard stuff.

[Kleinstein]

If the meter shows overload, it is not that likely that the transformer is bad.  A bad transformer more likely would cause an error for the power on self test.

The transformer for the DMM is a bit special:
 there are extra electrostatic shield(s) ideally 2, but not sure.
 the secondaries for the digital part and analog part also have good isolation - with many normal transformers multiple secondaries are not that well isolated (e.g. mains rated or better).

In the K2001 there is an additional preregualtion before the transformer. So the primary side likely is not the usualy 115V/230V , but a little lower. This would at least effect the turns ratio - the transformer is fine to run at a lower than rated voltage, but the output of cause is than also lower.
This extra preregulator circuit has shown failure in several cases and would be worth checking before exchanging the transformer.

For a test (to locate the error) a 2nd normal transformer is definitely a possibility. For the final use not so much because of the usually missing shields.

[BD1QMP]

thank you Alm,
I read some threads about 2001 repair.
I remove the transformer from the meter, then power the transformer directly, the output to digital board is AC 6.3V, all 3 AC outputs to analog board are 20-30mV.   
" alt="" class="bbc_img" />

the second windings over heat with something come out
" alt="" class="bbc_img" />

[BD1QMP]

If it actually is just that transformer winding, I'd cut that connection on the PCB, and just get some other little transformer with the correct voltages, and wire it in. It would be a lot cheaper, and it's rating should be pretty standard stuff

this is good idea, I will do it after China new year holiday. holiday is good for repair(a lot of time), but is bad for buying parts( every one go home, stop all the business )

[BD1QMP]

I assume the controller and display still working because the digital part has a separate power supply.

Yes, the output to digital board is normal, AC 6.3V

[Kleinstein]

Due to the preregulator the transformer may not actually be for 230 V, but maybe only 180 V or so. So the overheating could come from driving the core tto far into saturation.
It would be better to test the transformer at a lower voltage - e.g. 150 V on the full primaries and there measure the no load consumption (not just the current).
For the size of the transformer I would expect a no load loss of some 2-4 W, going down with lower voltage slightly faster than V².

One should get away with a standard 230 V primary as running on lower voltage is not a problem. One just needs to keep the voltage ratio.
Keep in mind that transformer voltages are usually given for nominal load and under no load the voltages will be slightly higher (e.g. some 10% maybe 15% for this size).
If in doubt a slightly larger core would give more space for the extra isolation.
In China there is chance to get a custom wound transformer for a reasonable price.

[BD1QMP]

Due to the preregulator the transformer may not actually be for 230 V, but maybe only 180 V or so. So the overheating could come from driving the core tto far into saturation.
It would be better to test the transformer at a lower voltage - e.g. 150 V on the full primaries and there measure the no load consumption (not just the current).
For the size of the transformer I would expect a no load loss of some 2-4 W, going down with lower voltage slightly faster than V².

One should get away with a standard 230 V primary as running on lower voltage is not a problem. One just needs to keep the voltage ratio.
Keep in mind that transformer voltages are usually given for nominal load and under no load the voltages will be slightly higher (e.g. some 10% maybe 15% for this size).
If in doubt a slightly larger core would give more space for the extra isolation.
In China there is chance to get a custom wound transformer for a reasonable price.

Thank you.
Follow your method, I test the transformer with 110V, the no load power is 0.4AX110V= 44W.      Obviously, it is broken.

Today, I re-wound a 100W transformer for 2001.  AC 220V-- AC ~15V---0---~15V ; ~6.56V; ~6.56V. After replacing the cap and fix the burn PCB, I will try it.
 

[Kleinstein]

With a transformer 0.4 A x 110 V is not 44 W but 44 VA. The voltage and current are often not in phase, especially not for the no load case of a working transformer.  0.4 A at a reduced voltage is still quite a bit on the high side. And the transformer likely has an internally shorted winding.

Besides the preregulation the K2001 also has some automatic detection of 110 / 230 V supply. If this fails it may cause damage. In a 230 V country one could opt. to remove that relay and wire is for a fixed 200-250 V. 

[BD1QMP]

With a transformer 0.4 A x 110 V is not 44 W but 44 VA. The voltage and current are often not in phase, especially not for the no load case of a working transformer.  0.4 A at a reduced voltage is still quite a bit on the high side. And the transformer likely has an internally shorted winding.

Besides the preregulation the K2001 also has some automatic detection of 110 / 230 V supply. If this fails it may cause damage. In a 230 V country one could opt. to remove that relay and wire is for a fixed 200-250 V.

Thank you， my plan is to remove the 110/230V selection circuit.

[BD1QMP]

After the holiday, I am back.

I replaced U505 14094 and U506 LM393. Unfortunately, when I remove the IC, 2 pads is off, and one track is open.
Burn CR110 diode cause PCB base burn out, with the help of a camera, I can zoom in and use a knife to get rid of the burn FR4 material and avoid to hurting the track of the 4 layer PCB.
 
I replaced CR109.110,111,112 and CR513,514 . because all the pad of cR110 is burn out , I have to use flying wire to re-connected the tracks.

+5V track under the CR110 is burn out, I use flying wire recover it.

after double check the connection, I connect the transformer (self made)  to the PCBs.

when I power on it. the 2001 meter come back.

Next step is to find a chinese transformer supplier to make a new one for me. and re-calibrate the meter.

[BD1QMP]

Now, I tear down the transformer to see what is wrong with it.

put the transformer in boilling water for 20 - 30 m. then take out the magnetic core piece by piece.

after removing the black plastic cover , we can see the second windings.  the layers of +-15V and 6V became black, the wires became short circuit.

the 6V windings for digital board is good.

[openloop]

Just rewind it yourself. Much fun actually (I presume it's a hobby for you).

I believe that's EI66 transformer iron - finding some (and for very little money) should not be a problem, especially in China.
Wire diameters and number of turns you can get from what's left of your transformer.

Just get one of those cheap, manual transformer winding machines and go for it!
Excellent intro into winding own transformers is here: https://ludens.cl/Electron/Magnet.html and https://ludens.cl/Electron/trafos/trafos.html

[BD1QMP]

Just rewind it yourself. Much fun actually (I presume it's a hobby for you).

I believe that's EI66 transformer iron - finding some (and for very little money) should not be a problem, especially in China.
Wire diameters and number of turns you can get from what's left of your transformer.

Just get one of those cheap, manual transformer winding machines and go for it!
Excellent intro into winding own transformers is here: https://ludens.cl/Electron/Magnet.html and https://ludens.cl/Electron/trafos/trafos.html

thank you very much. I am not familiar with transformer. you provided an important information of the transformer (EI66). with the information, I just find a supplier in Taobao who can make the transformer for me. 2 for 170RMB(about US$25).

[alm]

Can the supplier also do the double electrostatic shield as shown in the schematic? I think this is key to performance specs like common mode rejection.

[BD1QMP]

Can the supplier also do the double electrostatic shield as shown in the schematic? I think this is key to performance specs like common mode rejection.

I told supplier about the electrostatic shield, he said he can add it. I need double check it when I receive the transformer.