Fuel - Mainboard Monitoring Failure -
Noguri - 04-28-2019
Hi all,
Since last week, my Fuel unfortunately caught the "Environmental Monitoring" disease...
The issue seems quite common and the known bypass is to disable the environmental monitoring (through the l1cmd command)
My problem is that the Fuel now even refuses to boot up (it automatically shuts down 60secs after being powered on) so no chance to perform the l1cmd command under IRIX...
Therefore my questions:
- Is it possible to access L1 controller even if the computer is switched off?
- If yes, how to actually access this controller?
- Last but not least: do I have to change of motherboard? :(
This is what it looked like the last time it did not shut down automatically:
https://i.imgur.com/2CPTnGq.png
Thank you for your lights!
Cheers!
RE: Fuel - Mainboard Monitoring Failure -
computron - 04-28-2019
first of all
Check the PSU voltages with a real meter, because PSU can be faulty
Cheers
Eve
RE: Fuel - Mainboard Monitoring Failure -
Raion - 04-29-2019
The fuel L1 can be accessed while the unit is off. You can do it using the first serial port. I don't know the settings sadly for serial.
You can turn off env monitoring but this is likely to be risky and destroy your unit. I've said this before but the fuel is a piece of shit design. The VPro cards will cook themselves and the boards are going to fail eventually.
You can replace the env monitoring chip but it's not socketed, so you will need to either solder it off or send it to someone.
Replacing the motherboard, be expected to pay hundreds now. The fuel prices have skyrocketed due to exponential part failures.
I myself just scrap fuels for parts now. They're too expensive to refurbish and I don't want to risk more wild Vpros dying.
RE: Fuel - Mainboard Monitoring Failure -
Trippynet - 04-29-2019
Yes, L1 can be configured via Serial with the computer off (although the PSU has to have power fairly obviously).
Use the 9 pin serial connector in the centre of the motherboard, 38400 baud using a program such as PuTTY should work fine (8,N,1 if you need the extra settings). You should then be able to issue the "env off" command then power it up fine.
If the PSU is still acting strangely, the modern fix is to use one of Kuba's ATX adapters with a normal ATX power supply.
RE: Fuel - Mainboard Monitoring Failure -
Noguri - 04-30-2019
Hi everyone,
Thanks a lot for your lights, really appreciated!
To be more specific, my issue is with the "PIMM 5V aux" which value is getting detected higher and higher over time: 2 weeks ago it was still around 5.5V and now it can exceed 6.4V (at least as reported by the L1 controller).
I also checked the PSU. All 5V outputs measures were OK. The tests were done without load anyway so maybe it is not conclusive whether the PSU is actually defective or not...
I also found an old thread on nekochan in which someone reported exactly the same issue as mine:
http://archive.irix.cc/apocrypha/nekonomicon/forum/3/10849/2.html
The PSU was defective in that case.
So probably I'll try to change the PSU first (ATX+adapter) before deactivating the environment monitoring if the problem still persists...
RE: Fuel - Mainboard Monitoring Failure -
kubatyszko - 05-01-2019
It's also possible to use L1 via the USB-looking port on the back, needs a PC with old redhat linux (9) installed and SGI L2 or L3 software (I personally use a virtual machine for that).
If the machine turns on and then off after 60 seconds, even with env monitoring disabled that could indicate many other issues - something wrong with CPU, peripheral device such as video card, or I2C bus issues (that's likely given monitoring chips problems).
Thanks
RE: Fuel - Mainboard Monitoring Failure -
Noguri - 05-01-2019
Hi Kuba,
Thank you for your reply.
If I may submit some additional details to you expertise:
- The problem appeared some time ago. First there were just warning (PIMM 5V aux warning, voltage around 5.5/5.7V then stabilizing below warning threshold after some time). I could see them on the console.
- Then one day the fault threshold was reached (>6V) and the computer powered down automatically (I could see this again on the shell).
- Now the PIMM 5V aux is above 5.5V from the startup, triggering the warning. It cannot stabilize anymore and goes beyond 6V after some time. At that particular moment, the controller triggers the auto power down countdown (30sec).
Here is the "env" command output from L1 controller every 2/3secs from start til powered down:
Code:
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.85
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.85
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.85
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.85
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.85
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.82
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.82
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.82
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.82
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.82
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.82
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.85
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.85
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.85
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.88
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.41
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.41
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.85
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.41
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.82
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.82
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.41
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.77
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.75
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.77
PIMM0 5V aux Warning 10% 4.50/ 5.50 20% 4.00/ 6.00 5.75
PIMM0 5V aux Enabled 10% 4.50/ 5.50 20% 4.00/ 6.00 5.38
PIMM0 5V aux Fault 10% 4.50/ 5.50 20% 4.00/ 6.00 6.29
PIMM0 5V aux Fault 10% 4.50/ 5.50 20% 4.00/ 6.00 6.24
PIMM0 5V aux Fault 10% 4.50/ 5.50 20% 4.00/ 6.00 6.27
PIMM0 5V aux Fault 10% 4.50/ 5.50 20% 4.00/ 6.00 6.19
PIMM0 5V aux Fault 10% 4.50/ 5.50 20% 4.00/ 6.00 6.11
PIMM0 5V aux Fault 10% 4.50/ 5.50 20% 4.00/ 6.00 6.16
PIMM0 5V aux Fault 10% 4.50/ 5.50 20% 4.00/ 6.00 5.88
PIMM0 5V aux Fault 10% 4.50/ 5.50 20% 4.00/ 6.00 5.93
PIMM0 5V aux Fault 10% 4.50/ 5.50 20% 4.00/ 6.00 5.95
PIMM0 5V aux Fault 10% 4.50/ 5.50 20% 4.00/ 6.00 5.98
PIMM0 5V aux Fault 10% 4.50/ 5.50 20% 4.00/ 6.00 5.95
PIMM0 5V aux Fault 10% 4.50/ 5.50 20% 4.00/ 6.00 5.80
Another print screen from my last power on attempt:
https://imgur.com/IJEO0yP
Does it give more insight about what could be the problem?
Best regards,
RE: Fuel - Mainboard Monitoring Failure -
kubatyszko - 05-01-2019
I've been seeing similar warnings on my Fuel in my parents basement

In my case, it was a combination of: old power supply (via my ATX PSU adapter), and bad voltage regulation PIMM or bad motherboard - I wasn't able to complete troubleshooting it. I only swapped out the PIMM (CPU module) for another with no luck.
In your case it could be either of: PSU, PIMM, MB - unfortunately it's hard to say.
Motherboard supplies 3V3, 5V and 12V to the PIMM, which then gets regulated to respectively 1.5 , 3v3 and 5v. So the issue could be with the 12V rail.
Swapping out the PSU (original, or ATX with an adapter) or PIMM is the cheapest way to isolate the issue.
RE: Fuel - Mainboard Monitoring Failure -
Noguri - 05-02-2019
Hi,
Thanks for your very comforting reply :(
If I may come back on the PSU side of the problem:
My original PSU is from NMB (GM430WTXW01SSV model). This is clearly a WTX form factor PSU...
So is there anything preventing the use of another WTX PSU (with roughly the same specs. for 3.3/5V outputs obviously) as replacement?
Or, unless I find an original SGI-approved PSU, do I have to jump to an ATX one (with adapter)?
For example, the HP2-6460P from Ever Case Tech may be a nice (and quite cheap) option (they even claim it can replace the GM430WTXW01SSV)
Thanks for your lights!
Regards,
RE: Fuel - Mainboard Monitoring Failure -
kubatyszko - 05-07-2019
You'd have to verify whether that HP PSU supplies fan rpm signal to the socket to ensure Fuel doesn't complain (or shut down).