Tracking down the source of a high pitched sound at the edge of hearing

[cdev]

I have one of those high pitched whines coming from my PC, Its probably around 16-20 KHz judging by the characteristic that its literally right at the edge of hearing..

From what Ive read in the past, its likely to be an inductor, right?

Hmm, maybe I can come up with a frequency using sensors, plus Baudline or some similar FFT program (I probably have a bunch of programs that might be able to manage this)

Using and a piezo button or condenser mike.

Or maybe just a coil or two coils at right angles.. to pick up the magnetic field. Attached to a coil, or two coils and my scope. (that should pick up a magnetic field)

Or an RF probe, since the frequncy is in the VLF range..

Will try to see if that can give me some insight and attach what I can find here. .

What on a motherboard could/would be causing this?

[edpalmer42]

How old are you?  If you're out of your teens, you probably aren't hearing 16 - 20 KHz.   

The easiest way to find a sound like that is to take a rubber or plastic hose and stick one end on your ear and go looking for the sound by moving the other end across the board right on top of the components.  If you can hear the sound from a distance, it's guaranteed that you'll find it with the hose.

[IanB]

Something to do with the video in my PC makes a high pitched sound. When I rearrange windows on my monitor there is an audible sound track. I guess it could potentially be a variation in load on the power supply, and I'm hearing a variation in the switching frequency. Or it could be something to do with the video refresh circuitry. But I would assume that many digital electronic devices with clock driven circuitry will tend to make audible noises.

[cdev]

Ian, I get that too,

and I can also see changes in the waterfall display - caused its likely by RFI on my SDRs. But thats not the same as the sound that I'm describing, although it may be caused by the same family of changes, just on the audible domain. Maybe thats the RF part of it.

Something to do with the video in my PC makes a high pitched sound. When I rearrange windows on my monitor there is an audible sound track. I guess it could potentially be a variation in load on the power supply, and I'm hearing a variation in the switching frequency. Or it could be something to do with the video refresh circuitry. But I would assume that many digital electronic devices with clock driven circuitry will tend to make audible noises.

Also, although I'm older I may still have managed to keep more high frequency hearing because I take things like NAC which help prevent hearing loss quite a bit. (n-acetyl-cysteine, preursor of glutathione- its an antioxidant and presby-(I forget the rest, so I guess I am not immune to some of the effects of aging! ) aging associated hearing loss is caused by pro-oxidant changes that are response to interventions like manipulation of precursor availability )

[cdev]

Maybe you are right but when I adjust the highest bands on my equalizer with audio programs I can tell its being adjusted. So i still have some hearing up there.

How old are you?  If you're out of your teens, you probably aren't hearing 16 - 20 KHz.   

The easiest way to find a sound like that is to take a rubber or plastic hose and stick one end on your ear and go looking for the sound by moving the other end across the board right on top of the components.  If you can hear the sound from a distance, it's guaranteed that you'll find it with the hose.

TO make a long story short, I just cant do it that way because the sound really bothers me

[SilverSolder]

Something to do with the video in my PC makes a high pitched sound. When I rearrange windows on my monitor there is an audible sound track. I guess it could potentially be a variation in load on the power supply, and I'm hearing a variation in the switching frequency. Or it could be something to do with the video refresh circuitry. But I would assume that many digital electronic devices with clock driven circuitry will tend to make audible noises.

Ground loop issue with the audio circuitry...

I've seen the same thing in some PCs.  Usually curable with an external USB sound card.

[cdev]

I shuld mention that the very high pitched sound isnt coming through or out of the audio system, its coming directly outof the computer, so it may be coming from the power supply, motherboard or video card.

I think its likely the motherboard..

Something to do with the video in my PC makes a high pitched sound. When I rearrange windows on my monitor there is an audible sound track. I guess it could potentially be a variation in load on the power supply, and I'm hearing a variation in the switching frequency. Or it could be something to do with the video refresh circuitry. But I would assume that many digital electronic devices with clock driven circuitry will tend to make audible noises.
 

Ground loop issue with the audio circuitry...

I've seen the same thing in some PCs.  Usually curable with an external USB sound card.

No, I dont think that in my own case that its the audio card, It may be a coil.

The noise goes away when the computer is off or suspended.

[SilverSolder]

Rearranging windows on the screen means a spike of data getting moved over the bus into the video buffer - side effects of that spike in data movement is what is leaking through to the speaker.

I never found a cure for it on any of the PCs that I've owned which had this problem, other than using an external USB sound card...

[Rasz]

@SilverSolder hearing interrupts and data transfers over the speakers/headphones is something we solved around 2005

@cdev @IanB its called "coil whine", google will throw thousands of results under that phrase. Happens on crap hardware and top of the line GPUs. Usually the most successful solution is RMA

[SilverSolder]

@SilverSolder hearing interrupts and data transfers over the speakers/headphones is something we solved around 2005
[...]

The issue can still rear its ugly head if you use external audio amplifiers - ground loops / common mode effects between the external amp and the PC can be a real b!tch...

[cdev]

Yes, I'm trying to figure out a good way to isolate the component. Its become so loud its no longer on the edge of hearing, ouch.

I have a GTX-1070 video card and a quite old Gigabyte motherboard with a 6 core AMD processor. When I shut down the MB and turn it back on again, it takes some time to develop the full on whine when I turn it on.  Its there but not loud at the beginning.

Assuming its coil whine, would putting one of these two on a coil help, or hurt

nail polish or hot glue? 

Another idea, maybe under clocking the video card? 

[ambrosia heart]

When I shut down the MB and turn it back on again, it takes some time to develop the full on whine when I turn it on.  Its there but not loud at the beginning.

Maybe ATX power supply.  Borrow one from friends to exchange  with yours  to see if problem  solve.  There is some problems in ATX , causing the switching freq. to drop
down to audio freq.    The high pitch must come from SMPS transformers (trans) such as main trans,standby trans and transistors drive trans.   Open the ATX cover you will hear clearly.

If high pitch sound comes from motherboard, CPU must be SO HOT that triggers the buzzer.   

Enter BIOS to change the threshold temperature by setting  it higher.  Also remove CPU heatsink to see if thermal compound dried.

[Kerlin]

That sound can be quite a problem.
When I was in high school I built an oscillator at that frequency in a pen with a peizo earpiece on the output.
At that frequency it is directional, I took it to school and sat at the front of the class and used it to give teachers headaches.

[jonpaul]

We have seen many PC mobo, PC ATX PSU and Video cards with very bad current mode SMPS design and stability.

Result is intermittent sub harmonic or current mode instability, at a sub multiple of SMPS freq,   30K...1MHz. so in the audible range.

These oscillations are due to poor loop stability, and depend on the line, load.

https://electronics.stackexchange.com/questions/135325/smps-what-is-current-mode-instability-aka-sub-harmonic-oscillation

During various programs and displays the load changes dynamically so the oscillations are intermittent and changing.

There is no inductor or transformer to modify for silent op.

The culprit is usually poor SMPS design and test, by a Chinese mfg.

VERY hard to track down and fix is in the loop stability.

Check the drive pulses of the PSU , often you can sync on the oscillation.

We have even seen these problems even in very costly professional audio and broadcast equipment costing 5K..25K $.

From an old SMPS designer....

Jon

[cdev]

Thank you, this is what I want to try to do. Now, here is an interesting question. Previous computer SMPS was also subject to annoying whine and one day just blew up. After replacement with another power supply whine seemed much reduced, now its getting bad again. Is there any chance that this issue may be the result of a bad interaction between multiple components, bad combination of driver software and hardware, or failing component not in the power supply, but causing - perhaps it may have been the cause of the last PS failing?

Its a standard PC, with an old motherboard, two not too old video cards, and a new (Corsair) power supply. The blown out power supply was a Rosewill "Green" supply.

It gets better sometimes, its better, almost inaudible, when its first turned on.

the motherboard is old.

[jonpaul]

Sub harmonic osc is had to track, any load or line changes affect the control loop and the oscillation feq, ampl, etc.

To exonerate the PSU, I would substitute a high quality known brand.

The blowup has little to do with sub harmonic.

Rather than rescue ancient WS builds with cheap Chinese PSU, we found it much easier to replace entire WS with modern HP, you can find used HP for example Elitedesk 800.

Jon

[cdev]

What do people think about using epoxy if the inductors that are whining can be identified?

https://forums.tomshardware.com/threads/a-story-of-how-to-eliminate-coil-whine.3663256/

[magic]

I blame it on piezoelectric effect in MLCCs when their DC voltage is varied.
It tends to go away if you disable CPU power saving features.
Same thing may occur on GPUs.

[Mr. Scram]

I'm 28. I can hear 16k CRT noise easily across the room.

There are quite a few charts which show what can be expected. Between 20 and 30 hearing loss is comparatively benign, every ten years past that means more loss than mere cumulative effects. Past 40 things really start accelerating and that's before taking damage from external sources into account.

Considering the remarkably adaptive nature of our brain I'm also hesitant to draw any conclusions from self reporting. People love fooling themselves into being some kind of exception.

[SilverSolder]

I'm 28. I can hear 16k CRT noise easily across the room.

There are quite a few charts which show what can be expected. Between 20 and 30 hearing loss is comparatively benign, every ten years past that means more loss than mere cumulative effects. Past 40 things really start accelerating and that's before taking damage from external sources into account.

Considering the remarkably adaptive nature of our brain I'm also hesitant to draw any conclusions from self reporting. People love fooling themselves into being some kind of exception.

Thankfully, most music information is sub-10KHz so the enjoyment of music and sounds doesn't go away!

[Rasz]

What do people think about using epoxy if the inductors that are whining can be identified?

https://forums.tomshardware.com/threads/a-story-of-how-to-eliminate-coil-whine.3663256/

if you know exact coils why not replace them?

[cdev]

I don't know the exact source yet.. My computer is crammed in with a lot of other hardware and so when I work on it I always bring it to the kitchen table to have a bit of space. Ive been looking for a good program to use to generate a waterfall-like FFT or similar under Linux that I can use, with either a mike or a piezo disk (better higher frequency response, most likely).

But, the audio subsystem has been uncooperative. Something about my audio inputs is currently broken on my Debian install and inexplicably Baudline, which has served me well in the past for this kind of thing hasnt worked when Ive tried to use it. It isnt seeing my audio inputs from the computer mike, nor from an external USB audio dongle. I have soime time tonight so maybe I'll see if I can get one of my other audio programs to help me do this. 

The level of noise changes quite a bit during use, sometimes its almost imperceptible, other times its really loud. It never shifts rapidly.   

By the way, acetylcysteine quite possibly is a good way to delay noise induced hearing loss somewhat.

https://pubmed.ncbi.nlm.nih.gov/?term=acetylcysteine%20hearing

What do people think about using epoxy if the inductors that are whining can be identified?

https://forums.tomshardware.com/threads/a-story-of-how-to-eliminate-coil-whine.3663256/

if you know exact coils why not replace them?

Good question. I guess that would be what I would do if a less invasive procedure like putting some resonance damping material (maybe clear nail polish or epoxy) on the coil wasn't successful.

[magic]

Load all CPU cores to the max and see what happens.

[SilverSolder]

[...]putting some resonance damping material (maybe clear nail polish or epoxy) on the coil [...]

When I had a motorcycle alternator rewound many years ago, the technician dipped the whole assembly in epoxy "to make sure the windings don't move around and fail again" - he was right, it still works 30 years later!

[wraper]

Most likely it's not a coil whine but MLCC whine which happens way more often.
https://www.murata.com/en-global/products/capacitor/ceramiccapacitor/apps/notepc
https://e2e.ti.com/blogs_/b/powerhouse/archive/2016/08/09/how-to-reduce-acoustic-noise-of-mlccs-in-power-applications