DM3068 failure above 23 degrees celsius

[aclosada]

After a failure on my lab´s  A/C,  I`ve found out that my Rigol DM3068 doesn´t like high temperatures. If the temperature is below 23 degrees celsius, It works ok (passes the auto test). Above this temp it becomes intermitent with displayed  measurement out of range and relay banging (in any mode : resistence, voltage or current). Also,  the NPLC indicator doesn´t advance continuously, but stop many times like as the microprocessor responsible for measurement is overloaded. If the temperature is above 30 degrees celsius, the boot takes forever to end (more than 10 minutes) or maybe it just locked up.

I opened the case in hope of something obvious, but nothing jumped to my eyes.  (yes, i put my finger on all heat sinks and big ICs searching for something overheating)
There is one funny thing, if the temperature is high, like 31 degrees, even after allowing the meter to cool to the ambient temperature, it fails imediately during power up. I mean funny because if there is something overheating it should require some time to fail...

 I don´t know if the problem is present since purchase, because I never faced it before A/C failure and the warranty is already expired. I contacted the rigol distribuitor (I purchased from them) and they say that ANY problem besides recalibration is beyond  economical repair. So, try to repair it by myself is my last resort. Does anyone has an schematic of DM3068 ? I will apreciate any help.

[Kleinstein]

There usually is no schematics for the more modern meters anymore. This is because a component level repair is difficult - with no schematics even more difficult  .

A problem at higher temperature is often a part getting to hot. A first step would be to check the supplies.
It is also possible to have a problem related to temperature induced stress form thermal expansion. So it might be sensitive to vibrations / board bending too.

There might be an option to find temperature sensitive parts with local heat.

[wraper]

If you have hot air station, I would suggest heating small areas one at a time to localize location of the issue.

[aclosada]

Thanks,    as soon as I get the A/C working again I will try some local heating. Any news I will post here.

[picitup]

I used to repair pcbs many years ago and we got cold faults all winter and hot faults all summer.  For temperature sensitive faults we used a hair drier and a can of freezer.  For a hot fault, you could usually make it appear by heating the whole board with the hair drier.  To find the dodgy component, you simply dribble some freezer on each component and typically a single component cooled would then repair the fault.  Change the component and hey-ho.

Although I'd like to think I can trace and diagnose faults formally, I've repaired many pcbs with 50+ chips with the hair drier and can of freezer lol.

I hope this helps.

Steve