I've powered up the SOL-20 and nothing smoked :) First, I pulled the 17,000uf and 54,000uf electrolytic cans, and worked them up to 15V through a 10K resistor. I later stepped this down to 1K. I let each cap "cook" for about 48 hours. I did this while I was waiting for an ESR meter I ordered to come in. <URL: http://clientes.netvisao.pt/greenpal/evb1.htm > The meter arrived today, and I checked the two big caps and the two 2500uf caps on the power supply board. They were right in spec. I then reinstalled the caps and disconnected the supply from the motherboard and the riser card, then powered it up through a variac. The supplies looked good, so I plugged them into the motherboard and riser, attached a monitor (which I first had to get working...) and it sorta came up. I appear to have a bad 2102L in the video memory. Bit 0 was always on, and the problem followed moving the chip. The keyboard is dead, except for the LEDs. The next post has a question about that. --jc Doc Shipley wrote:

Doc Shipley wrote: smoked outlet? wtf? Am I right in guessing you mean a mains socket, and if so, how do you actually manage to break one? Gordon.

Part 1 answer: You should stop when the line voltage to the equipment reaches a normal value - with an unloaded linear PS, this should be LONG before you get to a 200w bulb. You should perform the procedure with the power-supply under no load. (ie: remove all the cards). If it is truly under no load, then you won't need to move "up" in lamp sizes, because as the caps charge, the current drawn by the supply will fall off toward zero, and the voltage dropped across the lamp will drop toward zero. As the current drops, the lamp filiment will cool causing it's resistance to drop as well. If there is no load at all, then the line voltage to the system would reach "normal" even with a low wattage lamp in series. In practice this may not happen with a really low wattage bulb due to some fixed current draw by such things as the chassis fans and a bleed resistor on the supply (if present) - You can disconnect these things and still bring up up to normal voltage through a low-wattage bulb. I prefer to do this, because if a component fails (like a cap shorting), the lower wattage bulb you have in series, the lower the current that will flow in the "short" circuit - I like to let long out of service linear power supplies "idle" thought a 25 or 40w bulb for a while before I go to higher power/load tests. Part 2 answer. Assuming you have the power supply under no load, even with a small bulb, the input voltage will climb toward line level rather quickly. It is best to reform caps at a slower rate. Best thing to use is a variac, so you can actually control the voltage being applied to the caps, and bring it up in steps, allowing the caps to form for a good long time at each step. If you don't have a variac, you can fairly easily divide line voltage down to 1/3 and later 1/2 by using 3 and then 2 same wattage bubs in series, and connecting the equipment across the bulb closest to the neutral side of the line - again is is assuming that the equipment will eventually charge to virtually no power draw allowing the line voltage to drop evenly across the bulbs. Note that using different wattage bulbs to control the division does not usually work as well as you would like, as the higher watt bulb will stay "cold", dropping most of the voltage across the lower watt bulb. -- dave04a (at) Dave Dunfield dunfield (dot) Firmware development services & tools: www.dunfield.com com Collector of vintage computing equipment: http://www.parse.com/~ddunfield/museum/index.html