dhedberg wrote: ↑Fri Jun 29, 2018 8:39 pm
I totally agree on this one. Most people will just replace an existing PSU anyways, and never run any experiments with it, so can't see why they would even try to run it without a load, unless they by mistake forget to hook it up to the motherboard. In that case, the fuse can be replaced. Keep the costs low, put a warning text somewhere on the PCB, and ship the PSU with 2 spare fuses.
Yeah, I'm thinking there's only so much "idiot proofing" I can really do without ramping up costs for almost nothing.
I was just testing the voltage clamp idea.. only had 18V zener to hand, so used 20VDC as the overload voltage.. pulls 320mA, so about 6.4watts... doesn't sound much, but burn your finger on the a TIP31 after 30 seconds.. a higher input overload of just 20.80V pulls 640mA, so about 13.31watts... So if someone left it running like that, it could get hot enough to start smoking without a heatsink.
Only option is to mount a SMT one on the board.. even so, it shouldn't be left running that way for long periods like that. Realistically it would only be able to give 1Volt head room anyway... I thought the peak voltages would limit easier than that, but seems not.
So think will just leave it as is. Its only really a precaution. The mains voltage would have to be higher enough to let the DC float up to 20V before it comes a problem.. here that worked out at 245VAC input to the transformer.
So *IF* the mains voltage was higher than 245VAC AND *IF* the PSU was running with no load, then it likely will blow the fuse.. but it shouldn't be run with both those conditions in the first place.
