Noisy powered speaker

I've had that on amplifiers and it was a noisy resistor! Also remake any soldered joints that look suspect.

Freezing spray might help to localise it.

It could be that digital noise is getting through into the audio stage - perhaps a failure of a filter component. Just because the caps are not visibly damaged does not mean they still hold there original capacity or have not gone high ESR. Given there are not many electrolytics, I would be tempted to swap them out anyway before getting too deep into further investigation.

I'd change the amplifier chips, TPA3118 - RS sell them for not much, probably cheaper elsewhere. They're class D amplifiers so are efficient and don't require heat sinking, but they're still going to get warm.

Oh, I'd assumed the ICs would be the least likely parts to degrade ...

Is that distortion that varies *with* the signal, or an *additional* noise? Does it happen on the analogue jack input?

You would in general be correct. They don't degrade, they DIE horribly and in pain. Pots and other mechanical devices including solder joints are no 1. Electrolytic and high voltage capacitors are No 2 and 'everything else' is no 3.

Burbly frying noises are often symptomatic of RF instability. which can be a bad capacitor.

Well like yourself, it does sound like inverter pick up. Many of these devices have the dc voltage internally higher than the supply voltage, and use an internal switch mode inverter to generate the voltages, so it really is probably a suppressor component on the inverter or dodgy cap as you deduce. Could be really just worth getting a circuit and seeing if there are any likely culprits and swapping them out.

Brian

I had a sub woofer and some gunge they had stuck around the components went slightly conductive due to moisture in the air. Sadly, it now has a problem where the protection relay makes a bad connection sometimes, turning it off with a bang. I do honestly wonder whether components are just not made so well any more. Brian

So would I those either work or don't usually its something else. I've never really understood class D. This was experimented with by Clive Sinclair many years ago, but they did not sound very good and chucked out a lot of interference. Brian

Doesn't vary with the signal, or the volume level, just a background mushy noise, the frequency of which wanders up and down, its most noticeable with an input connected but nothing playing.

Yes, as long as something is plugged in (even a 4" plug to plug lead connected to nothing) it makes the noise, so presumably the jack is one of those with a physical switch incorporated, so it "knows"; as soon as it's unplugged it mutes itself.

The whistling doesn't alter with the volume control (though a background hum does increase with volume) here's a recording

Interesting. sounds like filtered white noise. 'wind in the willows'.

Hmm. An educated guess is that its doing RF oscillations at some level. Without a scope it's hard to be sure.

check for something like a dry joint on a *small* capacitor.

It's semi-assembled with the the case hanging off, but other than making up five JST extension leads, no chance of getting my mits anywhere near it with scope probes, even then I wouldn't know what I'm looking for really.

I can disassemble again and try that

It's a Class D amp. It does PWM at 1.5MHz or less, then does a low pass filter of it, for regeneration of base band. By using output transistors in saturation (inside the TDA chip), the efficiency of the amp is very high.

Some of the filtering visible in the picture, is to prevent noise from going back on battery (and using any battery cable as an AM broadcast antenna).

If the PWM did not work at a constant frequency, perhaps modulation of the PWM operating frequency, would come through in the base band ?

Could there be a DSP hiding on there somewhere ?

How do the control buttons connect to the circuit ?

I feel like a portion of the circuit is missing. Another circuit board ?

And that glue to protect the electrolytics from vibration, that's nasty stuff, but it does not look like it has caused obvious damage in this case. On some devices with amps-in-sub-box, the copper tracks get eaten by the corrosive glue. The glue did not do that initially, but as it ages, and spends time at elevated temperature, it decomposes into "something else".

That glue idea, has destroyed a fortune in audio equipment.

One thing I don't know about Class D, is how much effect the ripple currents in the low-pass filter, affect capacitor life, and whether Class D eventually fails at the filter level.

The capacitor seams are intact, but that does not necessarily mean the capacitance value is still at the original level.

For people using Audacity at home, with your audio sample, snip the last little bit off the end of the sample, then select-all and Normalize to whatever level you want. That could make the sample easier to hear.

I was expecting to hear a bit more "burbling" (old cellphone noise), but the noise isn't exactly like that.

Paul

all well and good, but it is unlikely that the class D amp could produce that particular effect.

Again I cant see how it would

Dunno.

No. It sounds more like white noise and a voltage controlled filter :-)

I don't have any freezer spray (or equivalent) I doubt acetone evaporating would make much difference, tried tapping inductors and capacitors, probably pointless but nothing changed the whistling.

touching the input stages and the actual chip pins does introduce hum as you'd probably expect, but doesn't change the characteristic of the noise.

Here's the underside of the board, not very legible, it's not as horrific as I though when I first saw it, that's resin that spilled from where it's glued to the case, not battery juice.

Through that foil-wrapped flex/ribbon in picture1, the buttons are about

Yes, as i said there's a separate board with the buttons and a sub-module for bluetooth and NFC and mic, though I'm only using it as an analogue amp/speaker, I use the "smarts" of a chromecast audi to feed this thing, yes that's wifi but the whistling happens even with the chromecast unplugged and powered off, so it's not related, the chromecast is USB powered, and if I do power it from the USB outlet on the speaker it picks up horrible digital noises, makes it unlistenable.

For the two large ones, I could desolder one leg and measure them.

I've normalised and re-uploaded the file, in real life I just hear the warbly/whistly noise, the loud his is probably from normalising it to be too loud.

same URL

OK, the amp-pcb3 picture shows a power converter. That is changing battery voltage to something else. And with a 5 amp rating, it might be powering the amp chips.

VASKXX Bluetooth Speaker User Manual VAVA VOOM _20150911_V4.0-OUT-1 SUNVALLEYTEK INTERNATIONAL

The control board, one of the chips has a quartz crystal.

Paul

I'd suspect a bad electrolytic cap(s) as the cause of that noise. Press a new/good cap's leads onto each old one in turn while it's operating, if you hit a bad one it will clear up.

RF, easy to look for.