Anyone tried resolving PSU coil whine?

[Kilrah]

On 1/4/2021 at 9:31 PM, jonnyGURU said:

I'm telling you, you're incorrect.  At least in practical application.

 

There is a "balance".  You have to find the fine line between switching losses and losses in copper.  Currently, the balance requires a switching frequency just above audible noise level and slightly larger magnetics to obtain current efficiency requirements. 

Sure there's a balance but it's just not where you think it is anymore, things have progressed a lot in the past decade.

 

AFAIK >100KHz is common for silicon-based MOSFETs nowadays, and you should tell the guys who use GaN parts and copper amounts that are so tiny the transformer can be made with PCB tracks that they aren't allowed to switch at 1MHz even when that yields the highest efficiencies around 

 

[Guest]

34 minutes ago, Kilrah said:

Sure there's a balance but it's just not where you think it is anymore, things have progressed a lot in the past decade.

 

AFAIK >100KHz is common for silicon-based MOSFETs nowadays, and you should tell the guys who use GaN parts and copper amounts that are so tiny the transformer can be made with PCB tracks that they aren't allowed to switch at 1MHz even when that yields the highest efficiencies around 

 

Seems like you still haven't figured out who I am and what I do for a living. 

 

But you are correct about GaN. But that's because of lower switching losses.  Because of GaN primary FETs, the AX1600i is 100W more power, 20mm smaller size and slightly more efficient than its 1500W predecessor.  If you haven't seen one before, that's the picture I posted earlier.   But you'll note that the magnetics still aren't as ridiculously small as you assume they shod be, and the bud still needs a butt load of RTV to dampen Audible noise.  GaN still has a long way to go in the PC segment.  There's only one tier one working with it (Flex). Even Delta won't touch it for desktop PCs.  

[Guest]

I think it's important to understand context.  A 65W single voltage output adapter using GaN and smaller magnetics can be 40% smaller than its silicon counterpart and not produce Audible noise. 

 

An ATX PSU is typically at least 550W and therefore have to be efficient across a much wider range of loads.  To achieve this, switching frequency has to be dynamic, and in the most recent instance, be able to switch from resonant mode to a burst mode to achieve efficiency as low as 10W.  

 

While you're not entirely incorrect, you're applying text book to real life and not considering that not all power supplies can be created equally because they all have different applications.

[Mark Kaine]

4 hours ago, jonnyGURU said:

The cost increase I've seen is average 7%.  This equates to about a 3% cost increase on average per unit.

yeah, I really don't think it's worth sacrificing quality over this, and noise *is* quality for an industry that constantly advertises with being oh so "silent"... 

 

Of course I can only imagine how these thought processes go, but from my perspective this is a lot more damaging than rising prices a couple of bucks... 

 

A company basically has 1 chance to win over new customers (or even keep old ones) mess that up and they might not come back... 

 

As I said this is *my* perspective, I know full well a lot of gamers have full on Stockholm syndrome for certain brands... I still don't think it's the right thing to do here, sacrificing quality just isn't a good thing to do long term. 

 

Also, shoulda bought in May, copper was dirt cheap then! 

 

 

22 minutes ago, jonnyGURU said:

Seems like you still haven't figured out who I am and what I do for a living. 

Tbh, I read it, but I already forgot exactly... (it's corsair isn't it?) I know you have that website, which really isn't bad at all...  I do like the reviews a lot, but I think there isn't a lot of newer stuff sadly, and I think you said you don't have enough time for it anymore due what you do otherwise (could be remembering wrong) 

 

 

 

 

[Guest]

28 minutes ago, Mark Kaine said:

yeah, I really don't think it's worth sacrificing quality over this, and noise *is* quality for an industry that constantly advertises with being oh so "silent"... 

 

Of course I can only imagine how these thought processes go, but from my perspective this is a lot more damaging than rising prices a couple of bucks... 

 

A company basically has 1 chance to win over new customers (or even keep old ones) mess that up and they might not come back... 

 

As I said this is *my* perspective, I know full well a lot of gamers have full on Stockholm syndrome for certain brands... I still don't think it's the right thing to do here, sacrificing quality just isn't a good thing to do long term. 

 

Also, shoulda bought in May, copper was dirt cheap then! 

 

 

Tbh, I read it, but I already forgot exactly... (it's corsair isn't it?) I know you have that website, which really isn't bad at all...  I do like the reviews a lot, but I think there isn't a lot of newer stuff sadly, and I think you said you don't have enough time for it anymore due what you do otherwise (could be remembering wrong) 

 

 

 

 

Ok.  We'll just have to wrap this up here. 

 

I get your perspective.  I appreciate that you acknowledge that.  But keep in mind that your perspective essentially insinuates that every desktop PSU engineer at Corsair, beQuiet, Seasonic, etc. are wrong. 

 

[Mark Kaine]

14 hours ago, jonnyGURU said:

We'll just have to wrap this up here

Lol, ok! I just saw the GN video, I remembered right! 

 

14 hours ago, jonnyGURU said:

But keep in mind that your perspective essentially insinuates that every desktop PSU engineer at Corsair, beQuiet, Seasonic, etc. are wrong

... well, they're probably not technically wrong, but maybe have the wrong priorities... the more I look into this the more suspect things I find, like "single rail is fine" and "no one really cares about coil whine"... these, while maybe not completely wrong, are pretty skewed priorities imo, as these things should matter just as much as everything else too. 

(as it affects safety, user experience, etc) 

 

PS: I'm trying since several weeks to buy a PSU now, the experience is extremely awful and tedious as manufacturers often don't even have all the info on their website (yes I just figured out if something is multi or single can be seen on the label, but that's just one of many things that bother me) not to mention the experience I had with the prime gx which is supposed to be a high end, high quality unit... I've already said this elsewhere, the built quality was really bad, flimsy materials... like I'd expect a €20 ali express psu to be maybe... and of course it blew up the moment I plugged it in, thankfully I used the tester they ship them with (I wonder why.......? lol) 

 

I've seen so many reviews now and there's always something making me go nope pretty much instantly (usually it's "single rail for high reliability" or some such, huge red flag!) 

 

Quote

• Top-notch performance
• Quiet operation
• Excellent build quality
• Fully modular
• Individually sleeved cables
• SFX-to-ATX bracket adapter included
• 7-year warranty

AGAINST

• The semi-passive mode cannot be deactivated
• High inrush current with 230V input
• We would like to see 2x EPS connectors
• Turn-off overshoot at +12V needs to be addressed

 

^ I have no idea what the other two things are but for the bolded this is what prevents me from buying a ton of PSUs outright... "semi passive" "eco" even when there's a button to turn it off, I don't want it, it's just another part prone to failure... or "d f. ", like I want my PSU to blow dust on my components, like literally wtf... (cannot be turn off either of course) 

 

So it seems to me there are just a lot of features and stuff that engineers think of to differentiate their brand from others, without much thought to how useful the "invention" actually is, just another thing on the box that sounds good... 

 

 

And "inrush current" is apparently of no concern either (at least reviewers test it though), just buy another house, I guess! 

 

 

So yeah, really, engineers will know what they're doing, but if their priorities line up with mine is another question, it's also a management higher up thing though, I suppose. 

 

Hence I'll just stay with Bequiet, problem is they're basically constantly sold out or overpriced cause everyone wants one! 

 

So yeah, more options, and especially more straight forward options without shenanigans like "silent mode" and reverse running fans that blow dirt on your new and shiny GPU would be nice...

[Guest]

3 hours ago, Mark Kaine said:

Lol, ok! I just saw the GN video, I remembered right! 

 

... well, they're probably not technically wrong, but maybe have the wrong priorities... the more I look into this the more suspect things I find, like "single rail is fine" and "no one really cares about coil whine"... these, while maybe not completely wrong, are pretty skewed priorities imo, as these things should matter just as much as everything else too. 

(as it affects safety, user experience, etc) 

Yeah..  You really should learn more about how a PSU works, is designed, manufactured, tested, etc.  Seems you're way off base here.

 

Here is a good read about why you don't see smaller magnetics despite higher switching frequencies, BTW:  https://cpb-us-e1.wpmucdn.com/sites.dartmouth.edu/dist/c/87/files/2018/03/Scaling_Paper.pdf

 

I really can't state it any better.

[electropical]

6 hours ago, Mark Kaine said:

... well, they're probably not technically wrong, but maybe have the wrong priorities... 

 

One unwritten rule in business is that you do not sell what you like, but what the customer wants. 

 

The vast majority of PSU buyers don't give a heck about overall quality, they just want something that "works" for as cheap as possible. When you're in the PSU industry, you can't just put every imaginable feature in your product and call it a day, those features come at a price most customers are not willing to pay.

 

And engineers do not have the wrong priorities, for the simple reason that they actually do not have priorities of their own. Actually it is their employers that edict priorities for them to follow, and those priorities are usually geared towards making as much money as possible, not towards making the best possible product.

 

As for the idea that be quiet! is making the most popular PSUs out there, I don't think it's true. Their PSUs do not have a very good reputation and can't compete with brands like Corsair / Bitfenix / Enermax / Seasonic for any given price point (heck, even Cooler Master is more competitive in the low end for that matter!).

 

I do agree with you that there is room for improvements, but is the average consumer willing to pay more for those improvements?

 

[Mark Kaine]

36 minutes ago, electropical said:

And engineers do not have the wrong priorities, for the simple reason that they actually do not have priorities of their own. Actually it is their employers that edict priorities for them

Well, that was what I was trying to imply... 

 

36 minutes ago, electropical said:

you can't just put every imaginable feature in your product and call it a day, those features come at a price most customers are not willing to pay.

and I was saying I want different features than what's currently offered at large, not "all possible features". 

 

As I already said for example "silent mode" seems superfluous when there already exist PSUs where you virtually cannot hear the fan, especially when it's not under load, so for me the inclusion of a fanless mode is just another part that can fail, and that I actually have to pay for. I'd rather pay for some super glue to suppress coil whine, which would probably be cheaper and actually something that I *can* hear and that's extremely bothersome, and it keeps coming up to so I'm by far not the only one who hates it... 

 

yes there are use cases for a fanless psu, but those are extremely niche I would think... 

 

^ keep in mind this is just an example of how I don't think current feature sets make a lot of sense and are actually wasting money and resources - and I do not actually know if the idea with the glue to stop coil whine would work or not, it seems ingenious, and simple though! 

 

36 minutes ago, electropical said:

and those priorities are usually geared towards making as much money as possible, not towards making the best possible product.

well, yeah, that is probably true, doesn't mean I as a customer can't ask for improvements and features that would make more sense than the current ones. 

 

Seriously I have a very hard time to wrap my head around "fanless mode" when I already can't hear my PSU fan at all... my noctua fans are probably 20 times as loud, or "dust free" features where the fans spin at 100% in reverse to blow dust *into* your case, depending on orientation admittedly, no wonder they don't have a long life span lol (talking about enermax here obviously, only 30000 hours for a PSU fan seems really... kinda awful?) 

[Mark Kaine]

On 1/7/2021 at 6:32 PM, jonnyGURU said:

Here is a good read about why you don't see smaller magnetics despite higher switching frequencies, BTW

To be honest, I don't understand a thing, this is just a bunch of mathematical formulas, and I'm pretty sure that you know the average internet user won't understand it... What does it have to do with coil whine exactly? 

 

And I think you didn't adress if putting some "super glue" on the coils wouldn't help, or why that's a bad idea, it seems ingenious (if it then worked) you basically only said "copper is expensive deal with it". 

Thats not a very good answer when the topic at hand is how to avoid this annoying, and from user perspective unnecessary trait some psu / gpu combinations are showing, with all due respect. 

 

 

 

[Guest]

2 hours ago, Mark Kaine said:

To be honest, I don't understand a thing, this is just a bunch of mathematical formulas, and I'm pretty sure that you know the average internet user won't understand it... What does it have to do with coil whine exactly? 

Oh... You were talking about using GaN reducing magnetics sizes, so I thought you'd understand the paper.

 

2 hours ago, Mark Kaine said:

And I think you didn't adress if putting some "super glue" on the coils wouldn't help, or why that's a bad idea, it seems ingenious (if it then worked) you basically only said "copper is expensive deal with it". 

 

Thats not a very good answer when the topic at hand is how to avoid this annoying, and from user perspective unnecessary trait some psu / gpu combinations are showing, with all due respect. 

 

 

 

Umm.. never said "super glue". 

 

One part RTV (as opposed to two part), that is thermally conductive without being electrically conductive.  This can "dampen" the noise some.  This is called "potting".

 

I also said, "copper is expensive deal with it".  Never said that.  I said that "coil whine is often inevitable."

 

 

[Mark Kaine]

Oh, you were actually addressing this... I've actually seen this sort of "glue" (silicone) on 100 dollar PSUs... 

 

On 1/5/2021 at 7:13 PM, jonnyGURU said:

 

 

It is SOP in most factories.  Problem is, one part RTV that is non-electrically conductive yet not a thermal insulator is HELLA EXPENSIVE!!!!!

 

If you're paying a tier 1 to build you a $500 PSU, then the cost of RTV is a drop in the bucket:

 

 

If you're trying to build a $100 PSU at a China owned and operated factory, well.... not so much.

 

but I see now, from your perspective this is all about cost... 

On 1/6/2021 at 8:10 PM, jonnyGURU said:

 

 

There's no fixed number as the cost is increased based on the amount of copper used (PCB layers, trace size, copper windings, etc.), which varies from unit to unit.

 

The cost increase I've seen is average 7%.  This equates to about a 3% cost increase on average per unit.

 

It's not satire.  This is what I Googled for you:

 

https://tradingeconomics.com/commodity/copper

 

https://www.reuters.com/article/global-metals/metals-china-data-and-vaccine-optimism-lift-copper-price-idUSL1N2IV0JW

 

 

however, in this case 3% aren't the end of the world, and I suppose it's similar with this RTV stuff. 

 

So this is all about priorities, and what designers, or people who make decisions think the customer wants... which is apparently "silent mode" switches and RGB...

 

I'm just saying you could probably make a more straight forward option with a lot less risk of coil whine, with a good cooling unit that's silent without any "modes" (Bequiet does it too, I can't hear my PSU at all and that's a 500w unit that powers a 650w 'minimum' required GPU...) 

 

It's just not a priority and I think you're maybe exaggerating the costs - idk what does a silent mode switch with the corresponding tech cost, for example? I suppose it's probably cheap, but I don't know... 

 

Think about it, it could also be great marketing, a "new feature"... the new and super silent "anti coil whine PSU" 

 

(of course the unfortunate part is GPUs can have coil whine too, idk if a PSU can prevent that...) 

 

 

1 hour ago, jonnyGURU said:

Oh... You were talking about using GaN reducing magnetics sizes, so I thought you'd understand the paper.

Nah, I didn't? I have no idea what gan is... but I did learn there isn't much headroom for magnetic sizes in that paper... Must have been someone else... I just said manufacturers should use more copper, if that helps against coil whine and not less! : P

 

1 hour ago, jonnyGURU said:

One part RTV (as opposed to two part), that is thermally conductive without being electrically conductive.  This can "dampen" the noise some.  This is called "potting".

Ok, yeah I get that part now, I mean even if there aren't plans to use this more effectively currently, it's definitely interesting, and indeed seems like a simple solution. 

 

 

 

PS: I'm pretty sure I used this RTV stuff or something really similar at least, we were making prototypes for all kinds of companies like Microsoft (or Intel, forgot lol)

, infinion, etc, (including soldering etc) and we used this for so called "tags" to make them shock proof. And our boss also said it's expensive and we shouldn't use too much, but we ended up using a lot anyways, and a tube did last a good while too, although I think you'd need indeed more for a PSU... it's 20 bucks per tube or so maybe less, I think? 

 

 

 

 

 

[artuc]

On 1/8/2021 at 4:28 AM, Mark Kaine said:

As I already said for example "silent mode" seems superfluous when there already exist PSUs where you virtually cannot hear the fan, especially when it's not under load, so for me the inclusion of a fanless mode is just another part that can fail, and that I actually have to pay for. I'd rather pay for some super glue to suppress coil whine, which would probably be cheaper and actually something that I *can* hear and that's extremely bothersome, and it keeps coming up to so I'm by far not the only one who hates it... 

 

yes there are use cases for a fanless psu, but those are extremely niche I would think... 

 

Well actually industrial SMPS and LED drivers typically to a few hundred watts are fanless and it's usually preferred not to have a fan.

 

Fanless mode in and of itself is not really a point of failure, it's possible I guess but overall it both makes the power supplies quieter by not running the fan or running it at lower RPM.

 

It is extending the life of one of the most likely parts to fail. I would guess it has massively extended the lifetime of many PSUs that have it. It is not just some minor feature.

 

Cost for it would be a few dollars if that ? Dedicated IC or comparator with a thermistor and some other components.

 

I agree on that potting or coating the coils should be done and it is not worth the warranty and PR cost not to do it.

 

But then it all comes down to bean counting and you have far more egregious decisions like a device where a few cents was saved on a diode and the end user inserting the battery backwards destroys it. Electronics techs would see this all the time where a few corners are cut, components are skimped on and at the end lf the day an expensive device is brought down by saving dollars... or even cents.

 

It sucks but this is the reality. 

 

[Shawk]

Hi, not sure if I should have created a new topic or if its ok to attach to this one. 

Upgrading to a RTX 3080 FE has also introduced me to the world of unwanted noise / coil whine. I have, luckily, very good hearing. Unfortunately, that also irritates the hell out of me when I hear low level sounds continously (e.g. case fans). It is like low level tinitus. 

 

I tried the following 4 PSUs, one after another:

- Fractal Ion+ 760W

- BeQuiet Straight Power 11 Platinum 650W -> That was actually the worst. Could have made an artists sound performance out of those noises.

- Enermax Revolution D.F. 650W

- Seasonic Focus Platinum 650W

 

Only the Enermax Revolution did not show any form of coil whine. Unfortunately the overall noise level (its fan is always on, and not very silent) is as annyoing to me as the coil whine of the others.

 

I have read the comments above, and my take-away seems to be: The direction of the industry might make this phenomenon more prevalent.

 

Before continuing to buy and return units (which is not a money issue but more an ecological one): Do you think throwing "premium level" PSUs (MaxTytan, etc.) at that problem can make a significant difference? Should I try more Gold PSUs? Different brands? Is there any viable structured approach I can take, or is it just lottery?

 

Thanks a lot in advance!

 

 

[Guest]

5 hours ago, Shawk said:

Hi, not sure if I should have created a new topic or if its ok to attach to this one. 

Upgrading to a RTX 3080 FE has also introduced me to the world of unwanted noise / coil whine. I have, luckily, very good hearing. Unfortunately, that also irritates the hell out of me when I hear low level sounds continously (e.g. case fans). It is like low level tinitus. 

 

I tried the following 4 PSUs, one after another:

- Fractal Ion+ 760W

- BeQuiet Straight Power 11 Platinum 650W -> That was actually the worst. Could have made an artists sound performance out of those noises.

- Enermax Revolution D.F. 650W

- Seasonic Focus Platinum 650W

 

Only the Enermax Revolution did not show any form of coil whine. Unfortunately the overall noise level (its fan is always on, and not very silent) is as annyoing to me as the coil whine of the others.

 

I have read the comments above, and my take-away seems to be: The direction of the industry might make this phenomenon more prevalent.

 

Before continuing to buy and return units (which is not a money issue but more an ecological one): Do you think throwing "premium level" PSUs (MaxTytan, etc.) at that problem can make a significant difference? Should I try more Gold PSUs? Different brands? Is there any viable structured approach I can take, or is it just lottery?

 

Thanks a lot in advance!

 

 

You should have created a new topic.  This one already went way off the rails.

 

[Mark Kaine]

yes, I do did fix it, I also created it, but then I fixed it! uhhh... 

 

tldr: put 'Bequiet' pcie cables in 'Bequiet' PSU, get really nice, incredibly annoying coil whine as result, put back my old 'Bequiet' flat pcie cables, coil whine gone - the funny thing is I had those 'add on Bequiet pcie cables' for about a month before the coil whine occurred, so quite mysterious... at the least the old, original cables of the PSU fixed it! 

On 1/2/2021 at 12:17 AM, Ralphred said:

Back in the old days, coils were coated in wax to help combat this. If you know which coil is the culprit you can remove and dip it, if not some kind of conformal coating (available from electronics stores) may help if directed at the coils. 

Hot glue does sound like a good compromise, non-invasive to the point it doesn't really matter if it's there or not.

The sound comes from "loose" turns in the coil, the idea is to get some medium in to the gaps to prevent the movement and stop the noise.

 

Personally, there are some nasty voltages in those psu's tho, so would probe each coil with some plastic chopsticks I have, to keep a good distance. After identifying the offending coil remove it from the board and dip it in bees wax for a while (so the coil reaches the melting temperature of the wax, you want it to infuse in to the coil, not just sit on the surface).

There is a caveat with this though, the wax melts about 60-65 degrees C, as my CPU/GPU never go above this temperature I'm confident my PSU never does either, if I thought it did I'd have to use hot glue, which melts a lot higher (150-200) and would obviously require some slightly more elaborate (and less safe) heating rig than a bath of boiling water.

you could also use some kind of 'dual component' glue that's specifically used for electronics, and does not get completely hard when dried, a (big) tube costs around 20-30 bucks or so (and no, sorry, I don't remember the name of the glue) 

 

the advantage of this would be you don't have to remove anything, just put the glue around the coils, done... I guess!

 

 

Edit: oops, didn't see how old this thread is!

[Ralphred]

2 hours ago, Mark Kaine said:

Edit: oops, didn't see how old this thread is!

Thread necromancy is allowed when it's a solution to the original problem