There are a LOT of variables at play! Most of which are invisible until you’re actually serious about mass-production.
For one-off, low-power? Sure! Just buy a cheap chip that runs in the low MHz range and move on.
For mass production, significant power? Price breaks aren’t always your friend there, and so it really can become cheaper to roll your own discrete thing. Then you have ALL the variables at play, and the frequency/efficiency/cost tradeoff really comes to light.
Also, the design skills aren’t necessarily what you think they are, even for large corporations. Those are based on people, not budgets, and they may not necessarily end up attracting the right people for a variety of reasons.
It doesn’t have to. Any ground loop anywhere, with analog signals involved anywhere, even if they’re not related to each other, can get “ground loop noise” into that analog signal. I’m not surprised at all.
And yes, the way to kill the noise could very well be to isolate the digital signal. Again, not surprised at all.
Until this paragraph I felt you knew a lot about SMPS but sorry now you’ve lost a lot of credibility. The Chinese make dollarcent switchmode power supplies and you find them in everything from USB chargers to battery powered equipment to LED lamps… And NO SMPS in the world operates on sub-kHz frequencies. Sorry, that just doesn’t exist.
That said, good luck on your quest Andrew, and please let us know how you fare!
You might spend an extra penny to do it right, but the corporate bean-counters won’t unless there’s a VERY good pure-business reason. If everyone can hear it but very few people think to complain, then it doesn’t constitute a problem yet. (and they advertise it as the best thing ever, which it absolutely is NOT!)
At any rate, I still think it’s a ground loop causing all of this, and fixing that will make this go away too. It’s technically still switching at an audible frequency, but as long as it stays out of the analog signal path, it’s fine.
Disagree.
I have it on good authority that a ground loop present, that carries no current, will cause zero problems. It is only when current travels through it, that it becomes an issue. This why you can just run two parallel long XLR cables with connected shields between two devices and still have no problems. You CAN get problems, but it’s not a given. Regardless how many dimmer packs and power transformers those cables run past.
Ground loops are evil sorcery, that much we can agree on.
And a lot of them are deathtraps with fake manufacturer and compliance logos:
for just one example among many.
Oh yes they do! At least in the high-audible range, which is indeed not “sub-kHz” to be pedantic with your post. I’ve heard them myself. In laptops, battery chargers, quite a few things.
The external power brick might be inaudible, but that’s not the only power supply. There are others inside, that take the one incoming DC voltage and convert it to a whole zoo of different voltages and current capacities for different purposes. It was one of those, that caused the audible high-pitched warble when I moved the mouse on a laptop, for one example, and constant high-pitched drone when I left it alone.
Unplugged the laptop from the wall, or the power brick from the laptop, and the whine/warble stopped. Isolated the USB connection to an otherwise analog mixing console (token 2-ch USB sound card permanently wired inside), and that also fixed it, with the laptop’s external power still plugged in and charging.
The high-frequency part of the posted noise sample could be. It’s hard to get people to agree on the same term for the same sound, even for the same person across time.
The fact many are not up to safety standards isn’t the point. I never said they are safe. But they all have cheap chips that have no problems to switch quite fast enough. And that was what you claimed. That cheap transistors couldn’t cope. Which is simply false. I know zero transistors that have trouble switching at more than 30kHz
And when I said sub-kHz I of course meant sub-1000Hz. I’m fully aware there are those that work on 10 kHz or thereabouts, and sure you can hear those. But in the OP’s case we see a 60Hz base frequency with harmonics, which I called, for better or worse, a “60Hz square wave”. I see no way that can stem from a healthy power supply, of any type. Way too many harmonics in this sound. But what do I know.
Ok folks, while the discussion of ground loops etc. is not 100% off topic, lets try to keep in mind the goal here is to help with OP with the initial problem, lets not get to far into the weeds that it is hard to see the answers directly related to helping the OP. Thanks!
I guess I should have been clearer. Cheap large or discrete transistors, 'cause a lot of switcher designs are still discrete and not an all-in-one IC. And even if they can switch way above audible, there’s still some pressure to keep the frequency as absolutely low as possible, just to minimize the switching losses.
There very well could be a failing reservoir capacitor in a linear or cheap switching supply somewhere, which causes a bigger current spike than usual at every AC peak. That’s certainly possible, and would make all the harmonics stronger, so some that you wouldn’t have detected normally, start to show up too.
But I still hear multiple things in the same recording. It’s not all harmonics of 50/60Hz. I didn’t actually measure anything like some did above, but I do hear a “low-frequency” switching supply in there too (as mentioned before and argued to death about the mere possibility), which would be towards the top end of the audible range, and easily picked up by a ground loop as well. Not related to the raw AC at all.
And it’s also possible that the “mid-high hiss” is also not harmonically related, but comes from yet another “low-frequency” switcher, but this one is “spread spectrum”. Essentially randomizes its switching times with no particular frequency. Also receivable by a ground loop.
So I still stick to my original advice of focusing on a ground loop, somewhere. Fix that first, ignore everything else until the AC buzz is gone, and see what else goes away with it. Could be everything, as argued to death already. If not, we’ll go from there.
The details are simply a response to a common notion that a physical complete loop of ground connections absolutely must sound like 50/60Hz with strong harmonics. It does not. It could pick up and sound like anything, and I gave some detailed examples of some of the other things it could be.
Multiple sources mix together just like channels on a console, but fixing the loop itself is like muting the master. Done!
I’ve heard them myself. In laptops, battery chargers, quite a few things.