Lecroy Power Supply (Cherokee / Lineage Power SP651)

[darkstar49]

I've got several dead PSU's from various Lecroy suite-case format scopes (Xi, Xs, Xi-A, Xs-A, Xs-B, MXi, Mxi-A, MXs-A, MXs-B)
This one is from a WaveSurfer Xs-A (SP651-Y05A, Rev B), it did start and run for a moment, but shut off randomly after a few minutes.

I'm absolutely NOT versed into PSU troubleshooting, so ALL comment/hints/ideas/pointers/... are more than welcome.
I created this thread as I've got several of those, and I'm by far not the only one having PSU issues with those models, so I thought it would be nice to 'centralise' the ideas, experiences & solutions at a single place...

Just posting some pics, as I just opened the PSU a few minutes ago...
At the first sight, nothing obvious... but that's rather in line with the symptoms, it wouldn't run for several minutes if something major was blown/broken.

[darkstar49]

and the second pic, as size is limited to 5MB.

Next step is obviously cleaning up things...

[GertKruger]

Did you have any success with the troubleshooting? I recently had the same problem with a LeCroy WaveSurfer 44MXS-B. There is a forum https://groups.io/g/LeCroyOwnersGroup/topic/7707066 regarding the topic on fixing this particular oscilloscope. Apparently the PSU is a weak-link in the system.

Unfortunately no service manual for this model is floating on the net and lecroy.service.europe replied: "We don't sell this kind of spare parts." There is a UKcompany, Advance Product Services Ltd, that advertise repair of this PSU model, but no quote for repair could be given until they inspect it. Guess it will still be cheaper than sending the scope for repair at LeCroy.

I did manage to find a problem with one of the Schottky diodes (40CPQ045) on the -3.3V rail. It had an odd failure. No signs of smoke or charring. It no longer had diode blocking capability. When cool the resistance measured a few hundred ohm and closer to a short when warm. I guess the current mode PWM control protected the circuit from destructive failure. I couldn't figure out the switch-mode topology used, but I hadn't spend enough time reversing the schematic from the measurements I made.

The form factor of the scope seems to be a bit tight for the amount of power consumed. I remember the fan exit air to be quite hot. Another problem I see is the sandwiched layer structure of the boards in the enclosure. The power supply and motherboard is at the back and the acquisition board is on the front. The fans on top of the acquisition board gives way too much preference for airflow. The air flow through the PSU has to be done with the system fan, but given the layout structure of the fans on the aquisition board, most of the air forced over the acquistion board will have preference to exit the system fan, hence too little air flow through the PSU.
Another problem with the PSU is the choice of Kapton for electrical isolation of components from the heatsink instead of a thermal pad. Kapton has poor thermal conductivity. Tsk, tsk.

Update: I replaced the diode and the scope is working again. I added some more thermal paste between the heatsink and the aluminium shell/case to improve heat dissipation.