[rescue] Sun Ultra 1 Creator 3D - Power On Problems

Ahmed Ewing aewing at gmail.com
Sun Mar 11 11:19:34 CDT 2007

On 3/11/07, Lionel Peterson <lionel4287 at verizon.net> wrote:
> >From: Ronald Gilchrist <unixrescue at gmail.com>
> >Date: 2007/03/11 Sun AM 01:17:45 CST
> >To: rescue at sunhelp.org
> >Subject: [rescue] Sun Ultra 1 Creator 3D - Power On Problems
> >Hi all,
> >
> >I recently rescued a Sun Ultra 1 Creator 3D from being thrown out. After
> >bringing it home, and having a look inside (256 MB RAM, 9GB Disk, CD-ROM,
> >Floppy a 13W3 graphics card and an extra SCSI + 10/100 NIC SBUS card) I
> >tried to power it up.
> >
> >The CPU, PSU and Disk fans all spin up, for about 5 seconds and then spin
> >down. If I "power the machine off" and then "power it on" it does the same.
> >The power light flickered initially, but I don't see it anymore.
> >
> >Is this machine likely to be dead? Is it stopping because it doesn't have a
> >keyboard / mouse pulgged it? Could it be the lithium battery in the TOC is
> >dead and so is the machine (ie a new TOC would bring it back to life)?
> >
> >I've switched the serial ports to RS-232 (they were in their RS-432 default
> >state) and I plan to connect it up to an x86 linux box just as soon as I
> >find a F<->M 25-pin adapter, but in the mean time, I'd like to know, should
> >I bother?
> Anyone know how an Ultra 1 presents when the CPU is fried by a failed CPU
> fan? I wonder if that is the problem here?

If the CPU is truly "fried", then it's more likely that it won't even
POST without a valid CPU available to run the test, and the box would
stay powered on indefinitely in a braindead state (if you were lucky,
you might get some RED state registers dumped but probably not). It's
worth noting that only the earlier rev Ultra 1 system boards feature a
fan on the CPU--the later ones switched a CPU heatsink only.

However, that thinking is probably on the right track: an
environmental shutdown due to a slow/failed/undetected/physically
missing fan or a bad system board temperature sensor stuck registering
high is a more likely story, given the behavior of staying powered on
shortly then powering itself off. IIRC, besides the
board-rev-dependent CPU fan, there's one other fan in the chassis,
located towards the front of the box. And none of the U1s feature an
open chassis detection microswitch which would cause this
environmental shutdown behavior on boxes like E450s, so that would not
be a concern here. Unmodified, they'll run all day with the cover off.

All told, the POST output remains the best way to tell exactly what's
going on...

> I've already offered Ronald a replacement Ultra 1 base [...]

...assuming he's still even interested in a root cause analysis at
this point, given that generous offer :-)

Hope that helps,


More information about the rescue mailing list