FYI I made this video demonstrating the problem. It is 100% that the PWM functionality on CPU_FAN header is broken. When connected to CPU_OPT (literally right next to it) it works properly and is controlled by the same BIOS settings.

Board was sent in for RMA... Let's hope it's not another RMA nightmare story.

https://www.youtube.com/watch?v=FWV1vwE5gV0

Z370-E returned from RMA, will test tonight... Paperwork says they replaced the CPU socket (which I wouldn't be surprised if I nicked it ONLY WHILE PACKAGING FOR RMA while trying to put on the CPU socket cover...).


I seriously doubt my issue was repaired... I bought a replacement Z370-E to use to hold me over with the same fans/CPU/PSU/etc. and have had no issues.... As a hardware electrical engineer, I'm highly suspicious it's fixed since it was only the PWM functionality on the CPU_FAN header that was not operating at all.



What do I do if the RMA did not repair the issue?

Raja@ASUS wrote:
If the RMA is unsucessful, there is a customer loyalty rep in North America that can help.

I misunderstood the paperwork; they replaced the CPU socket but it seems they still could not fix the issue (as I expected; I probably knicked a pin while packaging up for RMA...).

They shipped me a new Motherboard; so customer complaint "resolved". Thanks Asus! Temporary board sent back for return, and Windows OEM license xferred to my new Mobo..



@Raja a few recommendations from a fellow Engineer if you wish to pass onto your ME/EE team

1) Since I knicked the CPU socket on RMA packaging, I'll tell you where I went wrong when re-packaging the MOBO for RMA... The CPU-Socket Cover (black plastic cover) needs to be installed on the OUTSIDE of the CPU-lever bracket. Intuitively this was backwards to my expectation (as the CPU gets installed first, INSIDE, and then the CPU-lever bracket holds it down). The CPU cover has extruding plastic towards the CPU socket. If you try to put the cover on first and THEN try to hold it down the the CPU-Lever backet as I did you run risk of the extruding plastic touching the cpu socket. May sound stupid to someone who does this frequently, but I doubt I'm the first to make this dumb mistake (and how often is the CUSTOMER putting ON the cpu socket as opposed to removing). In my industry we call this Poke-Yoke design (idiot proof so even if someone temporarily tries to do it wrong they can't break it). I think it could be easily fixed with some ME tweaks to the CPU socket cover (e.g. the extruding cpu-socket side plastic removed). Just a recommendation


2) The PWM functionality did truly fail on my board I returned after 2months of use; I confirmed no short and same fans/cpu/etc. are still working fine. I had 2x NHD-15 PWM wired through Noctua's fan 2:1 PWM fan cable (only routes the sense signal from 1 fan, but both fans are driven by PWM/PWR/GND) signals. Since DC mode worked fine I know it's the PWM functionality that failed. Not that it's my design but I would check that the PWM control pin drive strength is enough to handle that perhaps its driven by a transistor or something. The Connector specs could certainly handle the current/Power of both fans, but since it's the PWM control that failed I'd start there. I'm now using both CPU fans split between CPU_FAN and CPU_OPT headers since I suspect there is a hardware flaw that maybe needs some investigation.

Not sure how much you guys investigate your field failures.


Thanks Asus! RMA experience successful.