S2600cp2j + p4000M fan issue

[Nik]

Hey all,

I'm a regular lurker on here but time has come to say hi and whine about an issue

I recently bought a used Intel S2600cp2j + P4000M (specifically p4308xxmhjc chassis with 5 sys fans, shroud and passive cooling). The box was working when tested. When I brought it home I did the thing that I tell myself not to do and that's to update the bios to the latest 2.06.0005 together with the included BMC, ME and FRUSDR updates. The system had the original launch bios 1.02.0003 and presumably bmc/me.

Since then the machine will not come off 100%rpm on the sysfans. PSU fans spin down properly. I've tried a number of things including:

1. Reflashing everything again to latest version
2. Pulling CMOS battery/load defaults
3. Downgrading bmc, me and frusdr independently of bios
4. Downgrading bios to 1.02.0003 and associated bmc/frusdr/me (which is what it came with) - this was done in recovery mode

The box is picked up correctly by FRUSDR as P4000M with redundant fan sku. I've also unplugged and pulled the secondary psu as that's not usually plugged in and reflashed FRUSDR after to no avail.

Using SELVIEW I'm not seeing any critical issues at all. Occasionally see chassis intrusion closed then open regular messages (albeit the case was never opened). I've even unplugged the chassis switch thinking it had gone bad along the way.

One odd thing is I'm not sure what sensor I should be looking at for CPU temp in BMC, I have a few reporting negative values with an OK next to them which I assume are normal parameters.

Honestly no idea why this started happening since the update and even after the rollback. Would greatly appreciate any help if anyone's some across this or has ideas.

 

[marcoi]

you may need to run the intel profile tool, i forget how to do it now since its been a while. but it updates the bios with setting about the case, psu, fans etc. assuming this is the issue, that should get it back to normal rpm levels. I see if i documented how i did it but no promises.

 

[Nik]

you may need to run the intel profile tool, i forget how to do it now since its been a while. but it updates the bios with setting about the case, psu, fans etc. assuming this is the issue, that should get it back to normal rpm levels. I see if i documented how i did it but no promises.

Much appreciate the quick response. No idea what this profile tool is so appreciate any info on it.

 

[marcoi]

I found reference to the tool, its FRUSDR update process. but i havent done it in a while so i dont recall the steps. You should be able to find info by googling that though.

there may be other on here that used the tool more recently and can give a better procedure on how to run it.

 

[Nik]

I found reference to the tool, its FRUSDR update process. but i havent done it in a while so i dont recall the steps. You should be able to find info by googling that though.

there may be other on here that used the tool more recently and can give a better procedure on how to run it.

I've done the FRUSDR + SDT update already several times already using the stock sdt files so that's not it. Thanks though.

 

[ecosse]

There's a PMW offset figure in the BIOS that can be used to tweak the threshold. There's also lots of guides on the web (and this forum) about changing the values within the config files used as part of the whole BIOS/FRU/etc etc etc upgrade. Maybe that is what the previous owner did? I'll see if I can get some links later but this was some time ago so I might have got rid.

 

[Nik]

There's a PMW offset figure in the BIOS that can be used to tweak the threshold. There's also lots of guides on the web (and this forum) about changing the values within the config files used as part of the whole BIOS/FRU/etc etc etc upgrade. Maybe that is what the previous owner did? I'll see if I can get some links later but this was some time ago so I might have got rid.

Thanks, yes PWM is set to 0 in bios, I've changed it to a random number to see if it makes a difference (it doesn't) but it's described as an adder to ramp up fans even more.

I've seen some guides here to edit sdt files but I was under the impression that this wasn't necessary if you were rocking an regular intel chassis and intel fans.

 

[marcoi]

Not sure if this will help but here is some info on my system to compare too. The P1/P2 therm margins do show in neg numbers.
I would also double check the open case switch as that is known to trigger 100% fans mode when the case is opened.

 

[ecosse]

Thanks, yes PWM is set to 0 in bios, I've changed it to a random number to see if it makes a difference (it doesn't) but it's described as an adder to ramp up fans even more.

I've seen some guides here to edit sdt files but I was under the impression that this wasn't necessary if you were rocking an regular intel chassis and intel fans.

Yes, but if I remember correctly if the fans spin below the slowest figure set in the config file they would assume they were dead / sub-optimal / whatever and go back to the 100% wind tunnel setting.

That's a good point - I was tweaking these for a custom chassis with slow turning Noctua CPU fans. Hopefully the settings posted above help more!!

 

[Nik]

Well, not a great update. This quickly escalated into a more serious problem.

After rebooting/trying a few settings on the old bios that I recovered to I'm now facing a situation where the BIOS posts but I cannot boot to any drive or EFI.

I also cannot use recovery mode any more, it hangs on a blank screen and the diagnostic light turns to amber on the back.

I've tried stripping the memory down to two pairs only (one for each CPU) but I get the same issue. Clearing CMOS via battery pull and jumper also does not yield results. I can get into the BIOS just fine and change/save settings but alas it will not move beyond the bios post screen after. Just hangs forever.

I have a strong feeling the mobo is toast at this stage (maybe likely related to the fan issues above) but if anyone has a method that doesn't involve throwing it out of the window I'm all ears.

 

[marcoi]

At this point it doesn't hurt to try the standard type of troubleshooting. IE take the board out of case. Reseat both CPUs. Etc..

While the MB is out of the case you can look for any damage to components. Then test with a separate PSU. Test with one CPU. Also check if the battery still hold enough of a charge, you may need to replace it.

Good luck, hopefully troubleshooting will lead to a fix for you.

 

[Nik]

At this point it doesn't hurt to try the standard type of troubleshooting. IE take the board out of case. Reseat both CPUs. Etc..

While the MB is out of the case you can look for any damage to components. Then test with a separate PSU. Test with one CPU. Also check if the battery still hold enough of a charge, you may need to replace it.

Good luck, hopefully troubleshooting will lead to a fix for you.

Well after no fewer than 3 cmos battery pulls the board eventually booted and I was able to incrementally flash the bios up to 1.08. No change to the sound profile so it still sounds like a jet fighter.

I played with pwm in bios and it seems to ramp up the psu fan from what I can tell or it's a universal setting and it ramps up the only fan that isn't running 100%.

 

[Nik]

Not sure if this will help but here is some info on my system to compare too. The P1/P2 therm margins do show in neg numbers.
I would also double check the open case switch as that is known to trigger 100% fans mode when the case is opened.
View attachment 4914
View attachment 4915
View attachment 4916

View attachment 4917

These are almost identical other than dimm thermal options missing as I don't believe I have ram that supports this.

One oddity I see in selviewer is a chassis closure and chassis intrusion listed at the same time with usually chassis intrusion shown after in the list. Sometimes without any actual chassis open events. I've tried disconnecting the switch and refreshing frusdr as I've seen a section on chassis intrusion in there but no dice yet.

I'm attaching the sysinfo log file which has it all, everything looks ok to me, again, the only thing that's odd is the chassis intrusion trigger. Is there a way to disable chassis intrusion entirely? I've tried disconnecting it but still shows up.

 

[Bill1950]

I've tried changes to BIOS and have a lingering full-speed fan issue. HWiNFO64 CPU temps are 27C to 32C at idle, 39C under load. But so far, the CPU fans continue to run full speed.