[mrthinlysliced]

So I've been busy compiling GCC (4.7.1, nothing further right now, but as least it's working!) and the FAN 1 has given up the ghost:

Code:

001e02-L1>
001e02 ATTN: FAN 1 warning limit reached @ 0 RPM.

001e02 ATTN: Environmental redundancy lost.

I've also got a disabled ODYSSEY - maybe intentional when the FAN 1 is out?

Code:

Description       State       Temp       Temp       Temp       Temp       
----------------- ----------  ---------  ---------  ---------  ---------  
 0 INTERFACE 0       Enabled   50C/122F   54C/129F   58C/136F   24C/ 75F
 1 INTERFACE 1       Enabled   50C/122F   54C/129F   58C/136F   23C/ 73F
 2 INTERFACE 2       Enabled   50C/122F   54C/129F   58C/136F   22C/ 71F
 3 PCI RISER         Enabled   50C/122F   54C/129F   58C/136F   22C/ 71F
 4 ODYSSEY        <not present>
 5 NODE              Enabled   50C/122F   54C/129F   58C/136F   22C/ 71F
 6 BEDROCK           Enabled   70C/158F   75C/167F   80C/176F   26C/ 78F

Q1) Which fan is FAN 1 - Far as I can tell the dual-fan in a brick behind the hard drives = spinning, OK - fan infront of the power supply enclosure = spinning, OK, Fan at the rear over the power connector = spinning, OK

I didn't look under the GFX card, figured I'd ask advice here before needlessly dissembling old hardware.

Q2) The ODYSSEY not present - related to disabled FAN 1 - or I've possibly got a dead GFX card?

Thanks for any help.

Mr TS

[HAL]

Hmm,

I have never had such a problem on my Onyx350 or RM Tezro so far.
So you say the massive blower next to the ODY sucking air from inside the chassis blowing it out on the back is ok and the fan between
the SCSI-backplane and the pci-x compartment is also running, then I assume you are right that the fan not working must be the one
underneath the V12 case. It would also make sense that the ODY is kind of disabled, probably in the PROM (console=d), without the
required cooling.

[jan-jaap]

Is that ODY fan attached to the gfx board or the mainboard?

The disappearing act of the V12 can be as simple as it's Dallas I2C chip gone wrong -- they have a reputation for it. The I2C bus is handled by the L1 and -- guess what -- monitors the fans also.

[mrthinlysliced]

So you say the massive blower next to the ODY sucking air from inside the chassis blowing it out on the back is ok and the fan between
the SCSI-backplane and the pci-x compartment is also running, then I assume you are right that the fan not working must be the one
underneath the V12 case. It would also make sense that the ODY is kind of disabled, probably in the PROM (console=d), without the
required cooling.

Yep, massive blower, massively blowing .-)

Your comment sounds reasonable about why the ODY is disabled. Time to unmount the GFX bits and take some photos tomorrow.

Is that ODY fan attached to the gfx board or the mainboard?

The disappearing act of the V12 can be as simple as it's Dallas I2C chip gone wrong -- they have a reputation for it. The I2C bus is handled by the L1 and -- guess what -- monitors the fans also.

Aha. With the info from you both it starts to make a little sense. A little deconstruction for tomorrow, maybe try a boot without the ODY attached to see what the PROM says and if the machine works headless while I work out a plan of repair.

[indy99]

As you've probably already discovered there should be a fan attached to the underside of the graphics board. It connects to a standard three-pin fan header on the system board, but uses the non-standard IP35 SGI fan pin-out.

If your graphics board is disabled due to mechanical failure, if an original O350 rack-style graphics board isn't readily available you should be able to temporarily replace it with a Fuel V12 or V10 until an original replacement becomes available.

[mrthinlysliced]

Here's the sick child (quad 700mhz, 8GB RAM, V12):

   

And here's the fan that seems to be dead:

   

This is the blower details in close up:

   

And from a little hunting (manufacturer website is still there!) - it seems to be very similar to their current products listed on this page:

http://www.pscmf.com/product/showproduct...15&lang=en

From the product number on my dead fan - BD1275307B1

BD12 - Blower DC 12V
7530 - Seems to be the product housing / dimensions
7?
B1 - Ball bearing

Which matches up to the RBD7530B1, which is rated 12VDC 0.5A, 3800 rpm.

So my first attempt I guess will be to try and source a replacement fan (taking care to respect SGI wiring) and see if that brings the V12 back to life.

I tried booting headless, but unfortunately it doesn't like that - so no go till I resolve this.