<transformer/rectifiers. So if you're going to have on-board linear <regulators anyway, there's little to be gained in efficiency by using an <SMPSU rather than a simple transformer/rectifier/cap supply. And the <latter is a lot simpler. Actually that's not true. having done it using DEC H74x power bricks on one system to regulate the +9.4V(@18A load) to 14V(unloaded) to +8. The reduction in onboard regulator heating was enough to make me investigate a simpler way. I ended up using SCRs and much cruder regulator (still switchmode) to narrow the wild swing to 8.5 to 9.3V over a 1-18A load swing. <I agree that if you were going to make a 8080/Z80 machine _now_ you'd not <copy the Altair or the Imsai. And you'd not use the S100 bus, most <likely. But that's not what they're trying to do here. Actually My comments were narrowly aimed. the IMSAI was a far better example of S100 and worth copying. Only ONE change I'd do... NO MAINS POWER ON THE FP! Back when I'd seen one altair and a IMSAI fried by someone comming in contact with that. It had no business there. <Oh come on. If you can't design a linear PSU, you probably can't design <the rest of the computer... And you test on dummy load (for considerable <periods). Linear is much easier than switchmode. Especially HV switch mode! <In any case, in an S100 system, the regulators are on the cards. If you <get the raw supplies slightly too high you're not going to damage <anything. And of course you test them before inserting any cards. True, but a common problem in S100 is excess heat, removing it and hte regulators if over dissapating shutting down (undesired due to when they might!). Simple switch mode and CVTs were a good solution as the regulation reqirements of 10-15% were acceptable for bus raw power. The solution used by some and I hold the TI peripheral bus (also used local on card regulation) was to use higher voltages, smaller caps to beat the effects of ripple. Again another bix that really needs the PS tamed down. Allison