Ask HN: Why is coil whine in electronics so prevalent? | Hacker News Reader

Ask HN: Why is coil whine in electronics so prevalent?

45 pointswhyoh64 comments

I'm talking about the whine, buzzing and similar noises. Here's an explanation of how they happen: https://www.youtube.com/watch?v=4D6PKusyvUU.

Now, I've heard of some simple DIY remedies that work, like putting glue in/around the inductors.

So, I want to know why this doesn't get properly fixed in the manufacturing process. Because nobody cares? Is it too costly? I wouldn't mind paying a small price premium for a product if there was a guarantee of no coil whine.

View on HN

Mostly people don’t care.

When I was a kid I asked adults about that whine from TV sets and none of them knew what I was talking about. The conventional answer was that adults have a hard time hearing the 15khz whine due to age related hearing loss but I am older than my parents were back then and I still hear it just fine on those rare occasions I get near a CRT.

Right, most people aren't bothered by it or don't even hear it. I see a lot of complaints online, though, so there might be a market for whine-free equipment. It's quite annoying when you have to record some audio with sensitive mics. Even LED bulbs and PC monitors can emit audible noise.

It’s safe to assume that the viable market demand is already met by existing products.

By ‘viable’ I mean that companies can charge enough to remain in business by catering to a market segment that is willing and able to pay for the differentiation at margins sufficient to maintain the boutique sales channels necessary for low volume specialty goods.

Basically it is analogous to the audiophile market where discerning listeners are expected to pony up thousands of dollars on each piece of gear.

That’s my take.

YMMV.

My experience too is that you face a lot of gaslighting whenever you have non functional requirements.

Back when the Athlon 64 was new I was very annoyed by noisy desktop computers and went through great expense and trouble to pick out quiet parts for a PC, frequently I found the reviews of quiet PC parts from sites like Tom’s Hardware were just plain wrong and that from the power supply to fans and hard drives I had to try multiple parts before I got something really quiet.

I am thinking about building a new PC but I haven’t done it for years since I was so traumatized.

Your experience is exactly what I am describing. Exacting customers imply boutique pricing, and even customers of relative means aren’t in thousands per component to get what they want market segment…I mean for sixty grand you can probably get someone to build the $2500 PC you want.

A note on gaslighting:

If Bob hears something and Alice doesn’t, it is normal for Alice to doubt Bob’s claims…though that isn’t ideal, it isn’t really gaslighting because when it comes to sound quality some. people can hear a difference in a demagnetized audio CD and others on gold plated conductors for digital audio.

Without a relationship with some degree of intimacy, defaulting to the predominance of one’s own experience reinforced by the consensus of public opinions is not unreasonable.

YMMV.

If you’re planning on building a PC soon - GamersNexus has reviews of fans with measured dB ratings (some taken in an Anechoic chamber) and compares heatsinks with noise normalized performance. It may be very useful for you.

As a general guide though - SSDs are silent, and be quiet! (bequiet.com) has PSUs, CPU coolers, cases and fans optimized for silence (including coil whine)

If you record with a microphone you quickly realise just how noisy an untreated room is, even though your brain filters most of it out.

People with GPU on PC do care, it can be pretty loud.

or they put on headphones and then try to live with it.

Oh that was the worst. I don’t miss that past. CRT televisions would SCREAM at me. I hated movies at home.

We had a TV that you had to smack on the side after turning it on to make it stop screaming. It would still whine a little, but not as much.

> but I am older than my parents were back then and I still hear it just fine on those rare occasions I get near a CRT.

That only means that they had more hearing loss than you. If I was blindly guessing, you listened to less overly loud music in your youth than did your parents.

Edit: I guess my point is that it's still hearing loss, just not age related.

‘Age related’ hearing loss is an interaction between age, noise exposure, and other factors.

Yeah, I have some experience as a sound engineer in the sense of wiring up instruments and the board so people can do a musical performance but I can’t stand doing it much because I think almost every concert or bar is way too load. I think all the other sound engineers are deaf.

The navy found out that sailors working on aircraft carriers with jet aircraft were losing their hearing in a matter of months and did the initial research on the subject which revealed that different people have different threshold for how much noise accelerates age related hearing loss and that for some people this threshold is as little as 65db so there is no practical safe noise level.

the 15625Hz is a transformer not a regulator coil - rather hard to make it not 'vibrate'

I used to be able to hear that until about my late twenties.

Now my wife complains that I can't hear the (lot lower tones of the) beeps when the washing machine has finished.

Yeah, I could hear that into my late 20s, but at some point, I lost that range and it's been replaced with tinnitus. If I'm not listening to music or talking to somebody, I hear a constant tone around 10-20kHz in each ear. Not the same tone, of course. For the longest time I'd hunt for the damned coils whining away, with no luck, and I can't actually hear the ones in that frequency band anyway.

It's because cheap.

It doesn't bother everyone so they can get away with not stabilising the inductors.

The same way they put cheap capacitors with too-low temp ratings in PSUs that burn out prematurely - the consumer just buys a new one so why not.

We (this community) probably have the power to stop it. We have money, we care. We won’t tolerate coil whine. If you buy a laptop and it has coil whine and the manufacturer won’t replace it under warranty, come here, name and shame. We all stop buying manufacturers that do it. It quickly becomes worth the extra $5 to make it not happen.

Here I’ll start. I have a Dell xps 13 that does it. I bought it used so there’s no warranty. It’s useless and that manufacturer is dead to me.

Now you go.

Hello Asus, my whiny friend.

I think the worst offender is the HDD which isn't made by asud

> Dell XPS

I have one that chirps at me while scrolling on the screen on Linux.

Dell XPS laptops are generally famous for horrendous quality control and coil whine. I don't know reviewers can ever not mention that.

> We (this community) probably have the power to stop it.

Probably yes in theory, gonna be hard in practice. Most consumers don't really care enough when they find out it whines when they already have it. Reviewers don't mention it, so there's no good way for you to know.

I too have somewhat non-standard requirements for products I buy (haptics, how does it sound when you tap the back, how much does it flex, can you open it with one hand, how much does the software glitch ...) and pretty much the only way to inform others about them that I can think of is starting to do reviews. But try breaking into that field ...

I've never seen it but I know it's extremely common in these. And worse, it's been mentioned in reviews since they started the ultra-thin line so they are clearly aware of it yet it's still a problem many revisions later.

It's kinda crazy my $180 Chuwi LapBook with about the same thickness doesn't suffer from this but a $1500 XPS does.

Especially those remote controlled light bulbs. Why is that a thing? Who wants to have a constant buzzing sound in the background when the light is turned off using the remote.

These days you also have ceramic capacitors that exhibit this same behaviour. Some physically small but high capacitance caps are acoustic noise sources.

> So, I want to know why this doesn't get properly fixed in the manufacturing process. Because nobody cares? Is it too costly? I wouldn't mind paying a small price premium for a product if there was a guarantee of no coil whine.

It is too costly. You want a low frequency to be good at EMC ( i.e. FCC regulations) and you want higher frequency to spare money on components. Easiest is low frequency and hope for the best. Nobody cares about your ears as long as there is no regulation specifying noise. And glueing is expensive and, in the long term, might not bring any improuvment.

Relatedly: why is flickering LEDs so common? I presume most people also just don't see/don't care?

If you want to have a range of brightness you have a couple of options. One option is to vary the voltage you supply to the led, leading to a reduced brightness at lower voltages. The alternative is to use pulse width modulation, basically turning the led on and off fast so that it is on for a percentage of the time at full power.

Going with lower voltage is less efficient as the LED will output less light per voltage but not linearly, it will still use most of the power of full brightness at half.

Choosing PWM allows you to skip this problem by keeping the voltage identical but by using human persistence of vision to get the ideal number of photons to your eyes per your perception frame. The problem comes from making a slow PWM signal, say 60Hz, or having two similar but not identical PWM signals near each other, such as two different TV screens with a different backlight PWM frequency. That can make you see flashing because of the out of phase brightness peaks and troughs lining up.

I don't think this is the flickering the original poster is referring to. Dimmers are a bit of a nightmare with LEDs, however. I usually buy high quality (pricey) dimmer switches and good quality LEDs that are listed by the dimmer as compatible.

The flicker that bothers me, and maybe the GP, is likely due to cost cutting in the LED power supply. Not enough filtering or ability to handle the normal voltage dips in a home electrical system. As above, the solution is usually to spend more $ on bulbs, but not a guarantee.

Triac based dimmers + LEDs are quite a dumb combo, effectively the dimmer cuts the sine wave and the LED driver tries to determine how much it was cut.

I wonder if a good solution would be to have a fairly hefty capacitor taking charge from the supply side and then have the system measure the capacitance to determine how it will light the bulb from there. It could check for a capacitance drop to know it is using too much power and drop the output brightness using voltage or PWM, then bring it back up as the capacitor fills more. You would definitely get fluctuations, but increasing the capacitor size would absorb most types of fluctuation and also allow a system agnositc approach, able to ignore whether it is a PWM, sine cutting, or other method of dimming on the power supply. A nice benefit would be the very slight curve in the light output when you turn it off, just extending the lighting a few parts of a second as the capacitor drains.

You need to be discharging the capacitor somehow; I suppose it was meant to measure the charge, even then measuring the cut off of the sine is not hard. It's just all retrofitting design - led into incandescent bulbs, LED drivers into sine cutters - all to preserve the existing lighting fixture.

I have some near ceiling lights that allow dimming via a remote control (infra red or 433MHz) or fast switching on/off (remembering their state afterwards).

There are other ways to communicate, e.g. using the zero cross - the dimmer can send whatever signal to the lights downstream - the latter would read it and set the brightness.

> Going with lower voltage is less efficient as the LED will output less light per voltage but not linearly, it will still use most of the power of full brightness at half.

I don't think this is true? Aren't LEDs more efficient at lower voltages/currents?

They consume less for sure but the light they emit decreases faster, thus per light the efficiency decreases.

This is because leds have a fixed voltage drop and just a bit above that your light efficiency is almost zero.

You're right -- look at any LED datasheet and you'll see the efficiency get lower at higher currents.

The real reason for PWM dimming is simplicity, expanse and size.

Turning an LED on and off is a lot simpler than a real constant-current LED driver: indicator LEDs can be driven directly from a digital output on a microcontroller, or switched with just a mosfet. Constant current is always going to be more components.

In AC-powered applications, making an LED not flicker at 100/120Hz requires capacitance for energy storage, which are bulky and unreliable.

Constant current LED drivers usually use PWM for brightness control. They still have a low value shunt resistor to set the fixed current limit. Making a system with variable current adds much more complexity.

True, some do, though some don't (and some have both). Even dedicated LED drivers using PWM can be better than simple PWM by using smarter modulation to get high resolution without reducing the frequency to something noticeable.

I've noticed a couple of LED bulbs (one with warnings that it's non-dimmable) where the PWM (presumably) means that even when I'm not looking for it, I'll move my hand quickly to grab something and it looks like I'm in a 20fps computer game, fun novelty but it wears off. Fortunately there are plenty where I don't see this.

I get this effect when using a dimmed led headlamp in the rain. I can see the rain drop fall in discrete steps.

I've definitely noticed that with my bike light. You don't notice the flickering normally, but rain makes it looks like you're playing a video game at 3fps. (I think I measured the flicker and the PWM frequency was like 400Hz, which is not quite as bad as 3fps.)

The most jarring thing is seeing the flicker when you move your eyes. If you scan your vision past a flickering light, you'll have multiple discrete images of that light left on your retina.

Strobing can also be dangerous around rotating machinery, like a lathe, because at certain RPM the machine will appear to not be spinning.

I believe the solution is to use a buck converter with enough inductance to ensure that the current to the LED is not interrupted.

Is there no way to manufacture more persistent phosphor for white LEDs?

The white ones are actually 'blue' - but all that is not needed, the LEDs are current driven devices and even if they get a bit more relative output flux with higher currents, it's close to negligible. Overall phosphorus is 'bad' for the LEDs, reducing luminosity, increasing temperature.

Also the PWM can be in the high 30KHz w/o much of switching sacrifice loses, not even dogs would react on it.

The other way is to have a constant current source, this gives you much better control versus the voltage source.

The simple answer - (awfully) bad LED drivers

I hate the 50-60hz flicker, but most people don't seem to notice it or care. Manufacturers don't want to spend the extra 10 cents per unit on a decoupling capacitor.

The capacitor means lower power factor. The flickering ones are a sign of a terrible driver, just avoid them.

Depends what LEDs you’re talking about. Christmas lights are really bad. They’re worse at 30hz because it’s cheaper not to rectify the AC current so they only chooch half the time.

> why is flickering LEDs so common?

Maybe you have a "leading edge" dimmer, whereas most LEDs needs trailing edge.

You can buy good LED drivers with up to around 20kHZ PWM, but even 800 hz is enough to not see the flicker anymore, though it will show up on a video camera as stripes. It's the cheap 250 hz drivers that are painful to the eyes.

Here's an example for a controller going up to 16 kHZ. http://www.iskydance.com/index.php?c=product_show&a=index&id... (no affiliation). Would need a DMX fader to control it though.

Can I buy wifi-enabled bulbs that use these?

I use TP-link Kasa color-changing bulbs (although I'd be happy with any that can change colour temperature between daylight and candlelight), and the flicker sometimes bothers me.

Thank you for asking this. Flicker drives me absolutely mad, and it seems like no one else I've met experiences it (and trust me, I've asked a lot of people).

  • Cheap lights on a bike? Flicker.   
  • Lights on an expensive SUV? Flicker.   
  • Lights at Starbucks or at my library? Flicker.  
  • My phone (Pixel 5)? Flickers (I use a PWM reducing app that works by applying an overlay).  
  • Nearly 30% of all lighting at homes/commercial spaces and nearly 100% of those LED Edison bulbs? Flicker.

And if it didn't give me a headache often I wouldn't mind that much, but unfortunately it does.

God I wish I knew how to fix them. I can hear all of those, it's a nightmare, I can't sleep if there is one within 10m and no walls in between. It triggers my tinnitus, which then makes sleeping even harder

Two google queries: low frequency transformer whine AND low frequency electronic whine will get you much closer to an answer.

Thank you for these!

We recently had some new plug sockets put in and I swear every plug in the house has started making a noise since then. My search history was getting very extravagant trying to find answers! Hopefully this will help!

I bought an Intel engineering sample CPU to play with it. It's basically a beta chip they send to motherboard makers. When I enable the E cores it chirps and whistles like a bird, not like the annoying coil whine from a GPU. I think this must be caused by a different phenomenon, but I'm not sure.

I really doubt it's coming from the CPU. Most likely from the motherboard, near the CPU area. Many motherboards have coil whine nowadays, especially when power saving features (like C-states) are enabled.

You may be right. It's hard to isolate, but it's definitely near the CPU. It's not the E cores, like I thought as it does it with them disabled. I just haven't heard a similar coil whine noise before.

Is it too costly?

Very likely.

It's not just the cost of making the the coils quieter, you also have to put them through Quality Control procedures. That's where the expense is.

Then there is also the old adage: "If it works, don't fix it."

Coil whine is merely the modern equivalent of transformer noise.

Most people (especially older folks) can't hear it or don't even notice.

Cost.

Would need an additional 'manual' step during manufacturing, for a small payoff

It's a problem that goes away as you get older, for better or worse.

This is just because You using cheap electronics, for which cost of production is too important.

I have experience in past, with very expensive products, and they where perfectly quiet.

Unfortunately, for some reason, looks like it is global progress slowdown, such compromise-less production near extinct, and each year it is harder to find perfect products.

Dell's XPS series is notorious for being both expensive and exhibiting coil whine. It's not about the price of electronics, it's about what the consumer cares about for the given price - size/weight/other fancy features/coil whine/etc.

You don't even imagine, what is REALLY expensive.

Example is something like IBM System z, when it was new product, priced from 300k in simplest configuration.

Other examples where SGI stations priced around 50k, when top PC costs ~5k.

So even macbooks pro are cheap.

Reply on news.ycombinator.com