[cctalk] Re: Identifying a Failed Diode in a Rainbow H7842 Power Supply

I'm sure I recall we had a discussion about this a some time ago and I think we came to the conclusion that a variac is not a good tool for this sort of work, mainly because it doesn't limit current. A series light bulb bulb on the other hand will not allow the power supply to draw more current than the maximum current drawn by the light bulb, even if the power supply is a complete short circuit. This avoids the need to monitor an ammeter and react quickly to switch off if a high current is noticed. It also avoids having to aquire an AC ammeter which can accurately measure the odd current waveforms likely to be drawn by a switch mode power supply. An isolation transformer can be added if it turns out to be necessary to probe the primary side of the chopper transformer. Usually this won't be necessary as the light bulb will give a good indication of what is happening there. If the bulb lights at normal brilliance and the power supply does nothing, there is likely to be a serious short on the input side. If the bulb barely glows and the power supply produces reasonable output voltages into small loads, then all is probably well. If the bulb glows somewhat and one or more outputs are missing, then check the associated output rectifiers and smoothing networks for shorts. If the bulb pulses, the power supply could be tripping in response to an overload somewhere. If the bulb doesn't light and the power supply does nothing, either something is open circuit or the chopper control circuit is not working. I should probably add that the bulb needs to be a mains bulb. A 100W bulb will limit the current to less than half an Amp (in this part of world anyway). Remember to check that the reservoir capacitors for the rectified mains have discharged before handling that part of the power supply. Regards, Peter Coghlan.