[rescue] Serial console on an IBM RS/6000 3CT

[Andrew Jones]

Jonathan C. Patschke wrote:
>The machine is failing BIST.  Hunt around for a socketed EPROM chip
>on the motherboard and re-seat it.

I found three socketed chips.  One was microcode, which I left alone.
One was a regular DIP, labeled very strangely, on the motherboard.  I
reseated it.  The last was an EPROM in a ceramic package on the CPU
daughterboard, which I also reseated.  This one was labeled OCS, and is
presumably the firmware.  I reseated it as well.

No dice.  Maybe I'll just have to wait for a ruined 3[ABC]T or 39H
system to come across my path and try to use this 3CT for parts. :(

I'm still pretty hell-bent on getting a multi-chip POWER/POWER2 system
up and running. There's something innately appealing about the bridge
between modern VLSI chips and the discrete processors of yesteryear.
(And the other boondoggle multi-chip RISC, M88k, is rarer than hens
teeth and twice as hard to run software on.)

Sridhar Ayengar wrote:
>The switch might be broken or stuck.  One thing I've found in my
>experience is that an RS/6000 with no key is pretty close to useless

>The machine won't boot of the CD-ROM unless the key is in "service
>position.  If the reset button doesn't work, then the key is in
>"secure" position.

This is a much more appealing hypothesis than Patschke's, for obvious
reasons.  I'll have to try a multimeter and see if the switch is totally
nonfunctional.  It's definitely not stuck, as I can feel the thing
"click". (Also, the lock was originally in "OK" position before I popped
the switch off of it, hence my ability to open the case in the first
place. This would mean that the "reset" button /should/ work, no?)

Perhaps I should have mentioned that the display goes completely blank
after displaying the code '124' -- which makes Patschke's hypothesis
ring true.  I could have sworn the power-on tests were supposed to be a
lot longer, boot media or no boot media.