Steve (and others who might want to tamper with switch mode power supplies) This fiddling with the PC power supply reference IC strikes me as dodgy for these reasons:- 1. What about the 6.3V or 10V rated caps on the 5V about?? They go _bang_ with continuous overvoltage. 2. Similarly on the 12V (which would be about 30V) would have 16V or 25V caps. (That would be the second _bang_). 3. By fiddling with components around the TL431, you may be changing loop compensation, although it should be fairly tolerant of gain or pole shifts that would occur of that magnitude. 4. Most PC power supplies are half bridges and therefore work in the forward mode, and so there is no guarantee that they would be able to reach the 12V on the 5V secondary, even at full duty cycle. In fact they probably wouldn't, so you might have an unregulated 9V supply. 5. If there are any overvoltage shutdowns sensing on the secondary side, they would, well, you know ... shutdown. 6. Playing with mains powered circuits is tricky (read dangerous) unless you know what you are doing. Switch mode power supplies are usually highly optimised for the function they are designed for, changing that function requires a bit more than Ohm's Law. Cheers, Don Peat, don@central.co.nz Saturn Telesystem Ltd, Wellington, New Zealand. XYGAX wrote: > In a message dated 22/12/97 22:17:44, you write: > > << Up to a point, the larger the load on the 5V line, the better the 12V > output will work. If your application needs primarily 12V, and very > little or no 5V, it's possible to modify the supply so it regulates via > the 12V output. This is rather advanced though. > >> > Now come on the regulation in a PC power supply is nearly always controled > with a TL431 programable zener its simply a matter of tracing the resistor > from the control pin to +5v read its value and scale it to 12v (the pin needs > 2.5v) Let ohms law apply.......... > Componants required 1 0.25w Resistor cost approaching nil. > > Cheers Steve............... > Merry kristmooose (oops beer again)