Bug in Z240-SFF Fan-Control

[Faisal65]

I also found old ME V11.6 but we need Override Chip Security to install. I found FDO Pins on motherboard when I installed jumper pc did not power on. Unfortunately Only HP can now solve this problem. hope HP encounter this problem.

[PeterScbr]

Hello Marc,

I had a similar situation with a Z240 mini tower
Before I killed my mainboard (with a bios upgrade) it was running fine, also with long term full load, with a i7-6700K processor. Then I obtained a donor PC (Z240 mini tower) with an i7-7700K. Both are 91W TDP. Mainboards are identical, CPU coolers are a little different but they have the same HP part number. I transplanted the mainboard with CPU and CPU cooler to my own case.

And then: running the 7700K at full load, I got the same problem as you describe! CPU temp around 90 degrees Celsius, Low CPU RPM (+/- 1000 rpm - 1200) , sometimes but regularly a huge CPU fan surge of up to 3000 rpm and a temp peak in the CPU of up to 94 degrees Celsius.

I have been searching and trying but could not come up with an answer until I started to question the functionality of the CPU cooler. It was cooling, because CPU temp would normally be very close to 90 degrees, but not over it. And when a heat surge occurs, the CPU fan speeded up a lot which draws the heat away from the CPU. Obstructing airflow through the CPU cooler made the CPU fan speed rise nice and easy. I asked my self 'do the heat-pipes work as they should?´ But since with high CPU fan speeds cooling works, the heat-pipes must be OK. Right?
So I decided to put the CPU cooler from the 6700K onto the 7700K. When I took the cooler off the 7700K I saw that there was enough thermal paste but that it was not evenly thick distributed over the entire surface on the CPU, nor on the foot of the cooler. It looked like one heat-pipe had good contact with the CPU , the second one only for maybe 15%. Thermal paste in other regions was far too thick. After cleaning the foot of the CPU cooler, you can see and feel very well that the foot is not flat and the heat-pipes stick out of it.
After installing the 6700K CPU cooler (which has a flat foot) onto the 7700K,  CPU temperature rises at full load close to 90 degrees, but there is less fluctuation then before. CPU fan speed rises slowly starting at +/- 80 degrees from +/- 1000 rpm to +/- 1500 rpm. Case fan and Power fan also react to temperature change, but more relaxed. I have not had one CPU temp peak or CPU fan speed surge for more than an hour at full load.

I hope you find an answer to your problem and that what happened with my system might be helpful in your search.

P.S. Personally I think running a CPU at those high temperatures, so close to the limits is not right. So I decided to look for an option to lower the CPU clock a bit at full load. I found a CPU clock throttling script (for Linux) which works very well. You set a "desired max CPU temperature" and the script throttles or speeds up the CPU clock. You might want to take a look: [content removed]

Kind Regards...

[marcmeyer]

Hello Community,

During the holidays, I had the opportunity to test and analyze further and can now confidently say that this issue is a bug, likely within the BIOS:

I purchased a second Z240 SFF workstation (i5-6500, 16GB):

• In this used workstation, all fans operate as expected and accelerate significantly at 80 degrees Celsius, ensuring sufficient cooling for the CPU.
• I also tested an i7-6700 in this used workstation, and all fans worked flawlessly, accelerating as expected.
• Next, I installed an E3-1270v6 in this used workstation. The PC runs properly with this CPU, BUT the fans no longer accelerate. Suspecting a CPU defect, I acquired another E3-1270v6, but it exhibited the same issue – no fan acceleration. Finally, I tested an i7-7700, which also started the PC correctly, but again, the fans did NOT accelerate with this CPU.
• I was able to replicate this exact behavior with my original Z240 SFF workstation.

Having tested various CPU coolers, different case fans, and two independent Z240 SFF units, I can confidently confirm that the issue ONLY occurs with Kaby Lake CPUs. (The i5-6500, i7-6700, and E3-1245v5 have NO fan control issues. The E3-1270v6 (both units) and i7-7700 have the fan control issue where the fans do NOT accelerate at higher CPU temperatures.)

Does anyone have any tips for me, or could the HP support team please look into this (even though the Z240 is EOL)?! Thank you.

Wishing everyone a Happy New Year.

P.S. User Faisal65 has exactly the same problems with a Kaby Lake CPU at the link below: 
https://h30434.www3.hp.com/t5/Business-PCs-Workstations-and-Point-of-Sale-Systems/HP-Z240-SFF-CPU-Fa... 

[NonSequitur777]

@marcmeyer,

Interesting, because when I upgraded my HP Z240 SFF to the very max -with either a Xeon E3-1280 v6 or an i7-7700K, I did not experience the fan control issue you documented:

Link: https://www.userbenchmark.com/System/HP-Z240-SFF-Workstation/23033.

Have you installed the latest BIOS update (version: 01.92 Rev.A - August 19, 2024) yet:

Link: https://support.hp.com/us-en/drivers/hp-z240-small-form-factor-workstation/8354567.

Kind Regards,

NonSequitur777

[marcmeyer]

Thank you for your answer.

Yes, I know: you are "the crazy one" with the 3080 in the Z240. 🔥🤘😎

But as described, I am not the only one with this bug (see: https://h30434.www3.hp.com/t5/Business-PCs-Workstations-and-Point-of-Sale-Systems/HP-Z240-SFF-CPU-Fa... )

BIOS and IMEF are of course up to date (1.92 / 11.8.94.4494).

[toolboy666]

In favour of HP's official modification of the BIOS to control the fan speed, similar to curve adjustment.