rrsum@matt.ksu.ksu.edu (Rick Summerhill) (12/21/90)
I keep seeing the undercurrent of a thread lately: Someone has an IBM clone PC type machine with a standard ISA bus. DOS boots fine and all diagnostics (which are usually, inherently DOS oriented) check out fine. But when UNIX tries to boot, with some combination of hardware factors, it won't. Here's another example, one related to my machine: I recently purchased a huge new tower case with a 330w power supply (I want to add new drives, etc., and I have 18 Mbytes of memory, counting video etal.) My old power supply is 200w. If I sit the power supplies side by side, and boot UNIX on either, then it boots from the old power supply, but not from the new, larger one! One would think the new power supply is bad, right? Yet when I run diagnostics, everthing checks out OK on the new one. Moreover, if I put it on the bench and draw lots of current from the 5v lines - up to 25 amps - everything checks out OK! The voltages are clean as can be. In addition, it boots DOS just fine. Now, when it fails to boot UNIX, it does so in sporadic ways. Usually, it fails to get through checking the /usr filesystem, but sometimes it actually completes the boot procedure and waits for a login. Then, whenever I start X (using a Sigma Legend video card), it always reboots. Rebooting here means that it goes back to the bios and does a memory check as if it were just turned on - as if the 5 volt line dropped for an instant and then came back. Does anyone know what is going on here? Are there combinations of hardware that don't like each other? Can oscillators in the hardware effect each other? Etc., etc. Any help would be greatly appreciated. --Rick Summerhill -- Rick Summerhill Phone: (913)532-6311 CTA, Cardwell Hall FAX: (913)532-5914 Kansas State University Net: rrsum@hermzel.ksu.ksu.edu Manhattan, KS 66506