Fix to coilwhine on Maximus Hero XIII

[Dave2150]

EDITED - now completed fixed. Update BIOS to latest, go to Advanced in UEFI, find acoustic management options, and try the following settings:



Enjoy!

[EncodeGR]

Hmm, very interesting, i'll give it a try.

However i've already fixed my coil whine by disabling the C-States of the CPU.
You see, i noticed that my problem was actually the *low* VCore voltage that was applied to the CPU, while idling, especially when idling + moving the mouse.
Maybe your coil whine had a different cause than mine, who knows.

It's good to see that it resolved your problem though, i will re-enable C-States and try it.
Thanks for the update on that topic !

[BigJohnny]

[Ataemonus]

I had this coil whine as well, but the AURA or Armoury Crate or whatever was not installed. So that was not the case with my system at all.
Disabling C-States worked, but I did not want to leave that disabled.
For me, I was using a 10900K, and the whine was only present when playing some video or audio file via some player, otherwise, the YT or whatever, did not produce the same result.

[EncodeGR]

I'm afraid Lighting Service is not the cause of the problem for me either. I don't even have that service running.
It has to do with low VCore in relation to VID.

[EncodeGR]

This is clearly a VRM problem related to VCore changes.
I've tested it multiple times. Vcore fluctuations = Coil Whine.

Instead of disabling *all* C-States (which works), i tried a different approach.
I switched from Auto / Adaptive Voltage for VCore to Manual.

As expected, applying a fixed cpu voltage completely removes the coil whine issue.
I will also try to change the "VRM Switching Frequency" in Bios, as many people have suggested that it may solve the problem too.

[Silent Scone@ROG]

[EncodeGR]

This has nothing to do with the Lighting Service and is caused by the change in CPU usage and current demand. Power saving modes can exacerbate it, but you will often always get some noise when there are rapid changes in current.

Thank you for clarifying it Silent Scone. It makes sense.
So, let me see if i got this right. If for example i set an OC "By Core Usage" with Auto Voltage + V/F (and C-States Enabled) as follows:

1-core : 53x
2-core : 53x
3-core : 52x
4-core : 52x
5-core : 51x
6-core : 51x
7-core : 50x
8-core : 50x

(ex V/F: 0.60v - 1.45v

i should expect some high coil whine, if the application core-usage jumps all over the place, simply because the frequences (and voltages due to V/F Curve) jump all over the place.

However, if an "Sync All Cores" Manual/Fixed OC is applied, Vcore will be fixed, and thus no coil whine should be observed. Am i correct ?
Same thing applies when all C-States are disabled, since VCore stays max.

I wouldn't mind disabling C-States, however the problem is that when i disabled C-States, my "By Core Usage" OC was messed up, locking the boost to the lowest core-ratio. Should have i disabled "specific" C-State(s) or changed "Package C-State Limit" ?

[EncodeGR]

** UPDATE **

I think i can confirm there is a bug in Z590 Hero Bios, with power saving when OCing using "By Core Usage".
After extensive testing i found that:

• "Sync All Core" 50x + Auto Voltage = NO Coil Whine (C-States = Auto).

• "By Core Usage" 50x-50x-50x-50x-50x-50x-50x-50x + Auto Voltage = CONSTANT Coil Whine. (C-States = Auto)

The above two OC configurations should result into the same thing, same clocks, same voltage applied. Only "By Core Usage" though results in constant coil whine.

Few Things About "By Core Usage":
It's weird that while all "high performance" / "balanced" / "power save" modes in Windows behave as they should in term of clock speeds (meaning while idle, either: Full Speed / Base Clock / Minimum Clock ), in terms of voltage there is some strange behaviour, resulting in voltage drops to 0.666v in EVERY profile.

[Saltgrass]

I was just wondering, since the board has two 8 pin CPU power connectors, are you using one or both of those?

A month or so ago I complained about the Lighting service keeping the processor running at high speeds. That was eventually fixed, I believe, but I haven't checked lately.

Coil whine can be caused by multiple events. I have seen descriptions of it being so bad on a Dell system that the student could not stay in class when it was happening..

My board will be here maybe Thursday so I will check when I get it set up.