Follow-up to "MOV voltage"
First of all, the equipment is not for computing, data integrity is not imp ortant. It is active for a few seconds over perhaps 30 minutes of standby. The equipment is labeled by a reputable OEM manufacturer. The boards are poorly designed, perhaps done by the intern.
The line input stage is similar to old AT power supply, with hard switch an d huge caps. ATX power supply would have the caps power up first (or alway s on?) and soft power switch. So, we would need to redo the line input wit h power cap sequencing or power inverting. There is a 1A fuse after the ca p and couples of 0.5A fuses further down the line.
Fixed one of the board by replacing the 0.5A fuse near a 1.5A LM317. Still looking for problems in other boards (fuses seem OK). Anyway, will by-pas s the fuse and swap out the LM317 with a 0.5A MK317something. The intern d id not google for a lower current version. Protecting MOSFET with fuse mig ht makes sense, but the 317 has internal current limitation. 317 costs not much more than the 0.5A fuse and is still protected by the upstream 1A fus e anyway.