A power supply may or may not be efficient power converters based on design, etc. Watts on the primary side (the 110v side) may not say much about anything on the secondary side. Making a tally of actual secondary-side demands on each voltage is necessary any time something might be close or "loaded" or "big" or "fast". Best not to get close to 80% of capacity in any case. Sometimes not good to be below 10% of rating either. Some supplies are sensitive to power line glitches and some are not. Some really crummy supplies even amplify glitches and may fry circuits for that reason. Any supply gets more sensitive as load approaches the maximum capacity. Your conclusion is a good sanity check. Chuck > -----Original Message----- > From: tclug-list-bounces at mn-linux.org > [mailto:tclug-list-bounces at mn-linux.org]On Behalf Of Wyl Newland > > On Tue, 1 Mar 2005 11:13:10 -0600, Jack Surek <jsurek at mn.rr.com> wrote: > > well-designed supply would not fail over the life of a computer--it would be > > designed with parts rated at twice the voltage stress seen in worst-case > > operation. > > One of my machines started to reboot periodically. Long, painful, > multi-day story made short. By opening box and adding up amps by > voltage, from each disk drive, fan, etc., I found I had for a long > time pulling 18 amps 12v on a supply rated at 16 amps 12v. A modern > Antec power supply with 28 amps on 12v solved the problem for me. > > I decided that amps by voltage rather than watts seemed the safer way > to size a power supply. If my conclusion is incorrect, please > explain.