Is a CPU required in motherboard that supports IPMI to control fans?

[jbod_or_die]

I recently acquired a Supermicro X9SPU-F motherboard for purposes of powering a JBOD. It supports IPMI and is connected to a Supermicro PWS-1K28P-SQ PSU in a Supermicro CSE-847 chassis. I don't have a CPU or RAM installed. I've connected all the fans to each of the headers but the fans run at 100% speed. I've also tried going into the IPMI and setting the fan mode to "Optimal" but there's no change in RPM or fan noise. I've also noticed that there's no diagnostic information like fan RPM speeds and temperatures. Is this because I'm not using a CPU or could I have something not connected properly? I've attached a picture.

Thanks, appreciate any insight on this.

 

[RolloZ170]

check the health in the IPMI, you should see the reason of the FAN panik mode.

 

[jbod_or_die]

Thanks, how can I see that? I'm in IPMIView and not seeing any section that lists this, as well ipmitool. Here's the output of ipmitool sensor for example:

System Temp      | na         |            | na    | na        | na        | na        | na        | na        | na
Peripheral Temp  | na         |            | na    | na        | na        | na        | na        | na        | na
CPU Temp         | na         | discrete   | na    | na        | na        | na        | na        | na        | na
FAN 1            | na         |            | na    | na        | na        | na        | na        | na        | na
FAN 2            | na         |            | na    | na        | na        | na        | na        | na        | na
FAN 3            | na         |            | na    | na        | na        | na        | na        | na        | na
FAN 4            | na         |            | na    | na        | na        | na        | na        | na        | na
FAN 5            | na         |            | na    | na        | na        | na        | na        | na        | na
Vcore            | na         |            | na    | na        | na        | na        | na        | na        | na
3.3VCC           | na         |            | na    | na        | na        | na        | na        | na        | na
12V              | na         |            | na    | na        | na        | na        | na        | na        | na
VDIMM            | na         |            | na    | na        | na        | na        | na        | na        | na
5VCC             | na         |            | na    | na        | na        | na        | na        | na        | na
-12V             | na         |            | na    | na        | na        | na        | na        | na        | na
VBAT             | na         |            | na    | na        | na        | na        | na        | na        | na
VSB              | na         |            | na    | na        | na        | na        | na        | na        | na
AVCC             | na         |            | na    | na        | na        | na        | na        | na        | na
Chassis Intru    | na         | discrete   | na    | na        | na        | na        | na        | na        | na
PS1 Status       | na         | discrete   | na    | na        | na        | na        | na        | na        | na
PS2 Status       | na         | discrete   | na    | na        | na        | na        | na        | na        | na

 

[jbod_or_die]

Could it be that because the CPU doesn't exist, the motherboard is freaking out and blasting the fans?

 

[RolloZ170]

Could it be that because the CPU doesn't exist,

its possible, i would have never powered on a board without cpu.
i call it Panik mode. same if some FAN is running too slow, all FANs 100%.
check the LOGs (BMCWebGUI)
you can press the power button longer to switch OFF, the FANs should stop.

 

[zac1]

On Supermicro boards with CPUs and RAM, I've noticed sensors don't report and fans run at 100% until after POST. With no CPU or RAM, the system is unable to POST, so it seems the behavior you're seeing is consistent with my experience.

 

[RolloZ170]

On Supermicro boards with CPUs and RAM, I've noticed sensors don't report and fans run at 100% until after POST.

with EPYC(SoC)this is normal. Intel have chipset(FW=iME/CSE/SPS) which polls the power button or powers ON after AC loss if set so in BIOS before.

 

[RolloZ170]

but just plug CPU/RAM in, and see.

 

[gea]

Google for a "DC 12V PWM Fan Temperature Controller".
They are cheap and you can control fans either manually or depending on temperature.

 

[Stankyjawnz]

For the 847 chassis you need high static pressure because the backplane has terrible airflow. So running the 80mms fans slower may result in too high hdd temps. I would recommend doing a 120mm fan wall to replace the 80mm near the front. You can zip tie 3 together. With 120mm fans you can plug them into the backplane and run at 100% anyway and noise is still good. These offer much higher static pressure per db than 80mms.

The Best 120mm Fans

For PC cases, heatsinks, and radiators

www.hecka.tech

I don't know the best solution for the rear fans but you could do an external 120mm fan wall in front of the rear bays or maybe even see if the rear temps are ok with just the middle 120 fan wall.

 

[rtech]

In 847 you can go for 3x 140mm fanwall. I would go for fans with high static pressure such as Sunon industrial fans or Noctua industrial line

btw. i do have P12 3-pin they run at maximal speed and theres no whining or any other noise apart from aerodynamic noise from moving air.

 

[jbod_or_die]

Thanks guys, will look into changing out the fan wall. However, it seems that throttling down the stock fans can do the trick according to this thread reply: https://forums.servethehome.com/index.php?threads/making-sc846-quiet.8149/post-74283

The issue is that CB3 is almost impossible to find so maybe in the meantime I manually throttle the fans down until they are back in stock.

 

[mattventura]

The CB3 is pretty expensive considering that you can often get an actual motherboard with IPMI for less. The only issue with doing that is that the main 24-pin power cable won't always reach a smaller case.

If you don't need to be able to monitor every fan or control them independently, you can also do a very rudimentary fan quieting hack:
1. Buy a fan splitter that pulls power straight from the PSU via a 4-pin molex connector like this one (this is important - it has to be a molex or floppy connector, not a SATA power connector).
2. Carefully pop the 12V pin out of the Molex connector.
3. Pull the retention tab back out a bit so it will seat tightly once again.
4. Re-insert it into the 5V position.
This will let you power your 12V fans off of 5V, while still retaining PWM control and RPM reporting of one of the three fans.