How to quickly & easily fix coil-whine(coil choke noise)

[CoilWhineMakesMeWhine]

Hello,

I've recently purchased a Asus Rog GL752vw (i7-6700hq, 16 gb ddr4, gtx 960m, samsung evo 860, no hdd) laptop for a good price. The paste was hard as a rock, therefore the heat didn't dissipate at all, plus the heatsink was full of dust. I think the laptop hasn't been cleaned since it was bought in 2016, so the fan was working all the time. After cleaning and repasting there's a massive difference.

Now the fan rarely starting while browsing forums and watching youtube, lead to an unfortunate discovery. When I scroll with the mousepad or usb mouse, there's a "z z z z z z z z z z z z " noise, sort of like a faint continuous buzz with a 1/5 of a second in between. Since the laptop is silent, it's sort of annoying. Especially while trying to read something/browsing a forum. The noise also starts and continues indefinitely while I play a video or hover over taskbar icons. I opened task manager and it seems the noise starts when the integrated gpu is being used and stops right away when it goes to 0%.

I tried:
- lowering maximum processor state down to 20%
- on battery and also connected to power cable
- power saving mode and high performance mode - it's the same, with the difference of another random screech while using the latter, but that sort of disappears while using power saving/better performance/balanced mode
- reinstalling windows
- a few gpu drivers
- checked and it's not coming from the speakers
- with and without anything connected to it like an usb or external screen
- using touchpad and using usb mouse
- making the laptop run on the dedicated gpu by changing to high performance processor in Invidia settings (also choosing it for chrome), but it still uses the integrated one (maybe because these are light tasks)
- disabling hardware acceleration in chrome
- muting - unmuting
- switching ssd


The only thing that worked was uninstalling gpu drivers.

The noise seems to come from the middle of the keyboard, which would be around the cpu socket. When I cleaned it I noticed that the choke coils are upside down and later I tested it with the back panel off to try to locate the source of the noise. I think it comes from the opposite side, not the choke coils, although I'm not 100% certain.

Maybe I should pick it apart and try to run it with the motherboard out so I can pinpoint the sound source exactly. Although I'm slightly afraid I might ruin something.
This is also different than gpus making coil whine while being benchmarked to death or during intense gaming sessions so I'm not sure what to think. It's noisy while barely doing anything. Also the noise doesn't ramp up during more intensive tasks, it stays the same and it's drowned by the fan and speakers.

Any help is appreciated. Thank you !

 

[Mussels]

That definitely sounds like the power circuitry for the GPU is making the noise

Images from a youtube video of the laptop being disassembled:
There's not a lot on the 'top' of the motherboard, near the middle of the keyboard.

But when flipped over?
Good old coil chokes

These are the perfect example of what you can use superglue on to reduce that noise

The chokes may have their hardest work at low loads - not idle or maximum, but low values of light load, because they have to work their hardest to send just the right amount of power.

Educated guess that green is CPU, red is GPU, blue is VRAM
Those are the likely sources of noise, with your descrpition it's extremely likely to be from the red especially since they were covered by a heatsink - they're the most likely to run hot and break their factory seals over time, with greens second most likely
(Don't forget to replace the thermal pads on them with something new, after you glue em)

 

[CoilWhineMakesMeWhine]

Firstly, thank you for the tip about the thermal pads Mussels, I didn't even think of that.
I will try to take out the motherboard and turn on the laptop like that to locate the exact source of the noise.
One thing I haven't seen being discussed here is - how to clean between the coils really well when there's almost no space? I think if there's dust that sticks to the pcb between them it might be a problem for the superglue to seal well. Using a spray can doesn't seem like enough and I can't get in there with cotton swabs dipped in isopropyl alcohol.
The superglue I'm gonna use has ethyl 2-cyanoacrylate listed on ingredients and it's made by Bison. It also says liquid on it and it's a 3g container. I wonder - will be enough for the job?

 

[Veseleil]

Firstly, thank you for the tip about the thermal pads Mussels, I didn't even think of that.
I will try to take out the motherboard and turn on the laptop like that to locate the exact source of the noise.
One thing I haven't seen being discussed here is - how to clean between the coils really well when there's almost no space? I think if there's dust that sticks to the pcb between them it might be a problem for the superglue to seal well. Using a spray can doesn't seem like enough and I can't get in there with cotton swabs dipped in isopropyl alcohol.
The superglue I'm gonna use has ethyl 2-cyanoacrylate listed on ingredients and it's made by Bison. It also says liquid on it and it's a 3g container. I wonder - will be enough for the job?

I think I've spent almost three 2g tubes of super glue on my 6900xt.

 

[CoilWhineMakesMeWhine]

I think I've spent almost three 2g tubes of super glue on my 6900xt.

Damn. That's a lot. I guess I'll buy 1 more to be safe.

Also I'm having difficulty with those 2 coil thermal pads. I'll measure thickness (because I can't find that either) when I open it but I'm not sure what thermal conductivity I should buy. Is 6W/mK enough? Those are standard around here and there's a few Grizzly Minus ones that are a lot more expensive that are listed at 8W/mK.

 

[Anotherday]

So, decided to attempt this with MG Chemicals Acrylic conformal coating, since I did not want to use glue, and did not have the right type laying around. I'm sure most of you considered or thought about using conformal coating so here's an update. Honestly It did really not have that much of an effect. It may have lowered the sound a little bit, but that might just be a placebo effect.

Kinda sad it didnt remove it completely so I undervolted my 4090 to 875mv @ 2505 Mhz instead, and at that voltage its less audible and that's a noise level I can live with. My GPU is in a fractal north mesh, so sound travels easier.

Yes I cleaned the tops off before putting the card back together. Yes it's a messy job but I don't care, its conformal coating so its completely fine. End result is meh, wish it had a bigger impact.

 

[Mussels]

how to clean between the coils really well when there's almost no space?

It's not possible - you'd have to break off the ceramic shell without damaging anything, and that's impossible to do

So, decided to attempt this with MG Chemicals Acrylic conformal coating, since I did not want to use glue, and did not have the right type laying around. I'm sure most of you considered or thought about using conformal coating so here's an update. Honestly It did really not have that much of an effect. It may have lowered the sound a little bit, but that might just be a placebo effect.

Kinda sad it didnt remove it completely so I undervolted my 4090 to 875mv @ 2505 Mhz instead, and at that voltage its less audible and that's a noise level I can live with. My GPU is in a fractal north mesh, so sound travels easier.

Yes I cleaned the tops off before putting the card back together. Yes it's a messy job but I don't care, its conformal coating so its completely fine. End result is meh, wish it had a bigger impact.
View attachment 302776View attachment 302777

Oh god why?
You arent meant to put anything on the top or sides, that's going to ruin their ability to transfer heat
You've used the wrong product on the wrong components, on the wrong part of the components... of course it didnt help
In the long run that's going to damage that GPU, if it hasnt already

 

[Anotherday]

It's not possible - you'd have to break off the ceramic shell without damaging anything, and that's impossible to do


Oh god why?
You arent meant to put anything on the top or sides, that's going to ruin their ability to transfer heat
You've used the wrong product on the wrong components, on the wrong part of the components... of course it didnt help
In the long run that's going to damage that GPU, if it hasnt al

It's not possible - you'd have to break off the ceramic shell without damaging anything, and that's impossible to do


Oh god why?
You arent meant to put anything on the top or sides, that's going to ruin their ability to transfer heat
You've used the wrong product on the wrong components, on the wrong part of the components... of course it didnt help
In the long run that's going to damage that GPU, if it hasnt already

Haha I understand your worry. I cleaned the tops before putting it back together. I'm not really worried about heat transfer. I figured if glue works then why would conformal coating not work as it's also something that could get in between the edges and becomes harder, except it was actually made to be used on electronics.

Also what do you mean on the "wrong part of the components?", Maybe it is hard to see due to the transparancy and light but it was actually used mostly on the edges where it could seep in.

Honestly I don't mind, not worried at all, worth trying. Don't underestimate how hard it is to do it when they are so close together, you are bound to get some on th sides and top that you have to clean off.

How warm do these parts usually get? If needed I will open it up and clean the sides a bit more.

 

[CoilWhineMakesMeWhine]

Update:
So, I opened it up, turned it on and listened -

Strangely enough, this is where the noise comes from. I've listened before plugging those two screw holes because I thought that maybe the noise is coming through the first hole from the left. But no, even after plugging the holes with gum (I know, not the best idea, but that's what I had), I could still isolate the sound coming through that area I circled. It's like it was coming through the motherboard pcb. I've spent a lot of time going with both ears over it and around the board, even used a toilet paper cone.

After taking out the motherboard and turning it upside down -

The yellow circle is where I've heard the noise on the other side. Still can't tell why it was so loud on the opposite side of the coils. Even before and after I plugged the holes it seemed to have the same loudness and I could definitely tell it was from that area only. From near the other hole it was faint.
Now, after turning the board, the red circled coils seemed to be the ones making the annoying noise.

I applied generous amounts of superglue around them which you can see in the next pic -

Sometimes while listening I thought it might be the inductor (I think?) I circled in pink -

 

[lexluthermiester]

Sometimes while listening I thought it might be the inductor (I think?) I circled in pink -

That is a capacitor, not an inductor. The following(highlighted in red) are chokes and are likely to be a source of noise.

You might also want to do these(also highlighted in red);

Remember, any choke can be a source of coil whine. So if you do a super-glue job, save yourself some hassle & time by doing ALL choke coils on a given component, in this case a laptop motherboard.

 

[CoilWhineMakesMeWhine]

A reply from the gentleman himself!
I went all out and sealed every last coil I could possibly find on the mobo.
Now let's hope for the best!
I'll update later.

 

[lexluthermiester]

A reply from the gentleman himself!
I went all out and sealed every last coil I could possibly find on the mobo.
Now let's hope for the best!
I'll update later.

Cool, please do let us know how it went!

 

[Mussels]

Just remember the goal is to seal around the base of the chokes, and hopefully have some go inside - it's not to flood the area, or coat the sides or tops of any of these devices. You don't want to make their thermals any worse, as that can only lead to earlier failure or more noise over time.

 

[lexluthermiester]

Just remember the goal is to seal around the base of the chokes, and hopefully have some go inside - it's not to flood the area, or coat the sides or tops of any of these devices.

True! Keeping things clean and tidy is always a good idea.

You don't want to make their thermals any worse, as that can only lead to earlier failure or more noise over time.

Chokes don't create much heat(if any). It's not something to be worried about.

 

[Mussels]

True! Keeping things clean and tidy is always a good idea.

Chokes don't create much heat(if any). It's not something to be worried about.

Some of the above images went far beyond the chokes and onto other components, less is more with this

 

[CoilWhineMakesMeWhine]

So, I've done two 24 hour cycles because after waiting for the first round to dry, when I plugged it in and tested, the noise was still there.
I looked at all the edges of the coils, mainly the two I located the sound to be coming from and I decided to go all in. When I say all in, I mean all in. This would probably not be allowed on most adult sites.

The result:

The white stuff is the result of the cyanoacrylate curing over night, as most of you already know. I only noticed it in the morning. If I wipe it with a cotton swab soaked in alcohol, the coil looks like new. However, after it dries, it goes white again.

A few close-ups from the two coils that produce the sound:

Unfortunatelly, even after the second coating, the sound is still there. If it did something, it's probably around 1 - 25 % improvement, which is almost indistinguishable to my ears.

This time around I made two videos as well, so you can hear the noise:

This is before putting it together, with the board upside down and the coils exposed. I plugged a hdmi cable to my monitor and I played a video with no sound.
I switched between the two coils while filming with the mic down. The first one from the "5 group" is making the loudest noise.
Also, don't mind the MX-4 container, I put it there cuz my ugly mug was getting reflected in the left speaker shiny thingy.

And after putting it all together. You will notice that when I scroll it buzzes and when I stop scrolling, the noise stops as well. The noise seems to be more pronounced around the d - v keys, but I can also hear it faintly from up to 1-1.5 meters away when it's silence.
In my previous post. When I had the keyboard removed and I was listening, with my ear going over the mobo, with the screw holes plugged. It felt like the sound was coming through the pcb from a pretty specific spot.
From here -

Now that the coils are sealed (an understatement), I think my hunch is correct, that the noise is going through the coil casings and the pcb.


PS: Is it dangerous that some of the superglue engulfed other nearby components? Should I use something to try to dissolve it? And if yes, what?

 

[lexluthermiester]

So, I've done two 24 hour cycles because after waiting for the first round to dry, when I plugged it in and tested, the noise was still there.
I looked at all the edges of the coils, mainly the two I located the sound to be coming from and I decided to go all in. When I say all in, I mean all in. This would probably not be allowed on most adult sites.

The result:

View attachment 303112
The white stuff is the result of the cyanoacrylate curing over night, as most of you already know. I only noticed it in the morning. If I wipe it with a cotton swab soaked in alcohol, the coil looks like new. However, after it dries, it goes white again.

A few close-ups from the two coils that produce the sound:

View attachment 303113

View attachment 303114

Unfortunatelly, even after the second coating, the sound is still there. If it did something, it's probably around 1 - 25 % improvement, which is almost indistinguishable to my ears.

This time around I made two videos as well, so you can hear the noise:

This is before putting it together, with the board upside down and the coils exposed. I plugged a hdmi cable to my monitor and I played a video with no sound.
I switched between the two coils while filming with the mic down. The first one from the "5 group" is making the loudest noise.
Also, don't mind the MX-4 container, I put it there cuz my ugly mug was getting reflected in the left speaker shiny thingy.

And after putting it all together. You will notice that when I scroll it buzzes and when I stop scrolling, the noise stops as well. The noise seems to be more pronounced around the d - v keys, but I can also hear it faintly from up to 1-1.5 meters away when it's silence.
In my previous post. When I had the keyboard removed and I was listening, with my ear going over the mobo, with the screw holes plugged. It felt like the sound was coming through the pcb from a pretty specific spot.
From here -

View attachment 303118

Now that the coils are sealed (an understatement), I think my hunch is correct, that the noise is going through the coil casings and the pcb.


PS: Is it dangerous that some of the superglue engulfed other nearby components? Should I use something to try to dissolve it? And if yes, what?

Acetone will remove excess cured superglue. The strong stuff from a hardware store is best, but fingernail polish will work too, it'll just take longer.

 

[Mussels]

You can use acetone on a cotton swab to remove it from places you dont want it, like the tops of the components
Just don't leave it on the device, as it can slowly melt plastics - you use a second clean swap or lint free cloth to wipe up the excess (which takes the partially dissolved glue with it, too)


Perhaps try something temporary while the systems open like a huge blob of bluetack or whatever its called in your country, to be sure you've found the right component?


It even comes in colours now - but this is only for short term use, it can't handle heat without going soft and gooey... but its malleable and dense, so you can use it cover up a small source of noise and move it around easily

I've seen that mouse cursor movement issue before on older intel systems, usually with weak/old PSU's - disabling C-states in the BIOS may reduce your idle efficiency slightly but could solve the noise issue, as could a new/better power brick or even ferrite bead(s) on the power cable itself.

 

[CoilWhineMakesMeWhine]

Acetone will remove excess cured superglue. The strong stuff from a hardware store is best, but fingernail polish will work too, it'll just take longer.

I see.
My main question is - should I remove it or it's fine to leave it like that?

 

[Mussels]

I see.
My main question is - should I remove it or it's fine to leave it like that?

I'd wipe it off the top of the components its on so they dont end up sticking to anything (it's possible even if its just something like hair or dust catching on them later), and for thermals if they're attached to any cooling whatsoever

It's not something you'd need to do this very second, but since you're still working on the issue buy some acetone before you open it up next

 

[CoilWhineMakesMeWhine]

You can use acetone on a cotton swab to remove it from places you dont want it, like the tops of the components
Just don't leave it on the device, as it can slowly melt plastics - you use a second clean swap or lint free cloth to wipe up the excess (which takes the partially dissolved glue with it, too)


Perhaps try something temporary while the systems open like a huge blob of bluetack or whatever its called in your country, to be sure you've found the right component?


It even comes in colours now - but this is only for short term use, it can't handle heat without going soft and gooey... but its malleable and dense, so you can use it cover up a small source of noise and move it around easily
View attachment 303233

We actually do have it here as well. I'll try it next. Thank you for the suggestion.
If my theory is correct though and the buzzing goes through the coil casing and the pcb, what could realistically be done, apart from changing the coil?
Also, maybe it's hard to dampen it because it's a lower pitch noise?

 

[Mussels]

We actually do have it here as well. I'll try it next. Thank you for the suggestion.
If my theory is correct though and the buzzing goes through the coil casing and the pcb, what could realistically be done, apart from changing the coil?

Thats why you wanna see if something like blutack works - does dense material help cover the noise? Does pressure pushing it down to the PCB help?

definitely look into the power brick as it could easily be the cause of the noise if it's going out of spec due to age, see if you can hear it whining against your ear in any of these situations

 

[CoilWhineMakesMeWhine]

Thats why you wanna see if something like blutack works - does dense material help cover the noise? Does pressure pushing it down to the PCB help?

definitely look into the power brick as it could easily be the cause of the noise if it's going out of spec due to age, see if you can hear it whining against your ear in any of these situations

I've tried it on battery and it's the same.
Also tried on power saving mode vs balanced vs performance, it's exactly the same - I tried this both on battery and connected to the power brick.
Tried with the battery removed as well - same.
In the first video I made, the speakers, the touchpad, lateral ports and keyboard lights were disconnected as well.
Even with the laptop screen off and using an external monitor, it still does it.

 

[Mussels]

I've tried it on battery and it's the same.

Good thinking on that set of tests

C-states are a bios setting, not a windows setting

 

[InVasMani]

I used nail acrylic in the past on a 8800GT. It was easy to apply and worked well. Not sure if acrylic has any degradation concerns though though it never gave me any issues and card was inexpensive enough that it wasn't too concerning at the same time. I'd probably do it again on a relatively inexpensive GPU unless there is a clear and obvious reason to avoid it and use something else instead that's just as simple and easy. That card had terrible coil wine under load before and was perfectly quiet after. It was easy to apply too due to the brush.