[comp.sys.ibm.pc.hardware] UPDATE: My 386 hangs at 25MHz, but not at 20MHz

evil@arcturus.uucp (Wade Guthrie) (08/16/90)

I've gotten lots of replies (T*H*A*N*K*S), but (alas) none have worked
yet.

So far, the replies have said that it may be a bus speed problem
(I checked the motherboard jumpers -- all are set for maximum wait
states on both the 8-bit and 16-bit busses), or that the Adaptec
is pumping the disk/board too fast (I set the jumpers on the SCSI
board for maximum wait states -- goose egg).

The current theories go to the following:

	- my memory is 80ns, and that may be too slow for 25MHz 
	  operation,
	- my hercules-compatible may be defective,
	- I don't have the double-sigma 386 (that is going to
	  be fixed).

I've tried booting from floppy with no autoexec or config and running
a non-windows application -- it still doesn't work, so I've ruled out
a windows driver that is not compatible with my hardware config.  The
next thing to try is writing some specific programs (My turbo C++ prof.
should be arriving any day now :->) to try to pinpoint the problem:

	- allocate a huge array, and write quickly to sequential 
	  memory locations for around 20 minutes, if it doesn't 
	  hang, try random spots in the array,
	- open a file (not fopen -- I don't want anything getting
	  in the way of fast file access) and continue to write 
	  until I fill the disk, if this doesn't hang it,
	- write directly to the hercules memory to fill the screen,
	  if this doesn't hang it,
	- use bios calls to fill the hercules screen, if this doesn't
	  hang it,

	ASK FOR MORE HELP ! (or ignore the problem)

Just thought you-all'd like the update.
-- 
Wade Guthrie (evil@arcturus.UUCP)    | "He gasped in terror at what sounded
Rockwell International; Anaheim, CA  | like a man trying to gargle while
My opinions, not my employer's.      | fighting off a pack of wolves"
                                     |                Hitchhiker's Guide

smosher@big-bass.East.Sun.COM (Steve Mosher - Sun BOS Manufacturing Engineering) (08/18/90)

a low vcc can also cause the same problem. trying measuring the +5 on the board
and tweek it to 5.00, and see what happens. my guess is that it would be less
than nominal.


Steve Mosher
Sun Microsystems               smosher@Sun.COM