EEVblog #401 - Lecroy 9384C Oscilloscope Repair - Part 2

[notsob]

Dave versus Lecroy 9384C

I smell a new T-Shirt coming on

[jancumps]

...  I know at this point you must be having serious regrets at making this repair video ...

it has been one of the biggest successes. Two follow up videos, two lively forum threads. Fun.

[amyk]

ASICs not happy.

Low voltage high current devices like CPUs can look shorted (e.g. 1.2v core voltage, 20W in stopped state = ~ 16.7A or 72mohm resistance) but I don't think that's the case here.

Unless I'm missing something, from the posted schematic fragments it looks like they have neither power nor ground pins shown...

-5.2V and Vee means ECL logic.

The 9384M manual that was posted shows the MDX622 as an "2:8 ECL to TTL demultiplexer".

So officially is the LeCroy now LeCrisp?

I think it has LeCroaked.

[itdontgo]

Number 2 might just be hotter than the others because there is radiation heat transfer to it from 1 and 3.  Number 1 radiates into a cooler area and so should always be cooler than two.

Why not force it to run with your power supply connected and see what doesn't work on the scope?  If one channel is faulty presumably it's that channel's ASIC.

[itdontgo]

Number 2 might just be hotter than the others because there is radiation heat transfer to it from 1 and 3.  Number 1 radiates into a cooler area and so should always be cooler than two.

Why not force it to run with your power supply connected and see what doesn't work on the scope?  If one channel is faulty presumably it's that channel's ASIC.

To isolate it you would need to put something like Al foil between the heatsinks to reflect the heat from the other two - but the you would also need that on one to reflect it's heat back at it like number is getting...