Electronics > Repair
Keithley 2002 repair help
<< < (9/43) > >>
nikonoid:
Now that I am on the way of fixing the 2MOhm range and waiting for parts, I am looking for the reason for overheating front panel.



--- Quote from: Le_Bassiste on June 02, 2017, 07:47:44 pm ---hmm, after another look at your photos, i now see that it is not the transformer that caused the brown spot on the pcb,   :palm: it's one of the two push-pull transistors on the primary side.
given the age of the instrument and the fact that the supply of the display is still giving the right voltages for filament and anode, you probably shouldn't be too worried about it.
the only thing that you could do is to exchange the transistors with some better ones. don't know from top of my head whether these are BJTs or MOSFETs, have a look at the KEI2001 schematics (VFD supply should be same as in KEI2002, but is located on digital board) on TiN's site to find out.

--- End quote ---


I took an infrared image of the front panel after the meter was on for a while and it shows very noticeable overheating in the same area where the board is discolored. See attached IR photo.



I do not have a high resolution thermal imager, so the photos are somewhat blurry. Nonetheless I can see that the VFD transformer gets hot, but only to 60C; At the same time 5 transistors in the vicinity and 4 somewhat large 1/4W resistors get to about 80C. This is with the panel open i.e. some ventilation. The temperature is probably much worse when panel is closed. In overall thermal shot of the front panel I can see the temperature bleeding into area of banana terminals and can attribute to EMF voltages.

I checked resistors and they have proper values and 4.7k resistors had a voltage drop of 31V DC, corresponding to dissipating 200mW for a resistor specked at 250mW. Maybe what I have is completely normal.

Can someone please take a thermal photo of the front panel of K2002 after it had been running for a whole? No opening of the unit required. Thanks.
 
Samogon:
Tomorrow will make thermal images, going to calibrate some agilent 34401a DMMs.
Le_Bassiste:

--- Quote from: nikonoid on June 12, 2017, 03:06:24 am ---Now that I am on the way of fixing the 2MOhm range and waiting for parts, I am looking for the reason for overheating front panel.


I took an infrared image of the front panel after the meter was on for a while and it shows very noticeable overheating in the same area where the board is discolored. See attached IR photo.
I do not have a high resolution thermal imager, so the photos are somewhat blurry. Nonetheless I can see that the VFD transformer gets hot, but only to 60C; At the same time 5 transistors in the vicinity and 4 somewhat large 1/4W resistors get to about 80C. This is with the panel open i.e. some ventilation. The temperature is probably much worse when panel is closed. In overall thermal shot of the front panel I can see the temperature bleeding into area of banana terminals and can attribute to EMF voltages.

Can someone please take a thermal photo of the front panel of K2002 after it had been running for a whole? No opening of the unit required. Thanks.

--- End quote ---

here you go: my 2002 after approx. 120 hours uptime, sitting between a cooking 2001 and a reference box, ambient at 30°C

nikonoid:
Le_Bassiste,

Thank you for posting. Your meter shows 22C degree rise over ambient (great point to bring ambient temperature into this).
I was getting 41C max with ambient at 22C to amount to 19C rise over ambient. Give or take the same result as yours.

My heat center is also exactly at the same spot as yours. It looks like it is not the transformer but resistors and transistors that heat up the most.

This is bound to create temperature differential in connectors. How could this be on such a precision instrument?
Do I now have to use low EMF cables?

Samagon,
Can you please upload your thermal photo?
Kleinstein:
I agree that for a precision instrument so much heat is not a good idea. But on the other side low noise amplifiers need a certain current and thus generate heat. The heat is one reason I do not like the rather power hungry VFD or even Nixi displays. So a simple B&W LCD also has advantages.
Still I don't see a need for the power supply here likely the driver for the small transformer needs to get so hot. At least today there should be more efficient DCDC converters - even if it has to be low noise / low EMI. It looks like the transistor is the main heat source - except in a linear supply this should not happen so much.  The temperature really seems to be on the high side - the discoloration of the board is not a good sign, even if this might have happened in a hot environment. So there might be a fault causing too high a consumption.

However the main heat is not so much near the terminals, but more to the other side. Just plain copper wires are already not that bad with thermal EMF. This is also just a regular DMM, not a nV or special low level meter.
Navigation
Message Index
Next page
Previous page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod