What's causing the sound difference between two class D amps.

[Oddball]

Don't give up just yet. In the second hand market you could get a deal of your life. If your market is not very developed in therms of used Hi Fi sales, then it could be that some of the more wealthy members will just dump their purchases to get rid of them. And they might be well worth the price.

 

[melomane13]

inductors were getting hot while the board was idle
I couldn't to hold my hand more than 10 seconds on them
...

Oscillation ?
What is the power supply?

 

[solderdude]

There's probably something wrong with your specific amp module if there's an obvious difference in sound.

It probably only needs a better op-amp ... fortunately it is in a socket so easy to swap.

I know... I know this isn't helping.

 

[kemmler3D]

Ty I'll try this but I don't have a line in port for sound it's only output
Maybe I'll just give up on trying to get better sound.
There is a saying ... You get what you paid for And I got what I paid for
Some things aren't meant for some people... Like sound stuff isn't my thing I guess

If the amp has load dependency it's possible to correct it via EQ for free on Windows or Linux. You would need some way to see the problem, though.

If you don't have any audio input, but you do have any devices with a microphone, you can at least try and run a sweep with REW to see the frequency response, then try EQ and see if it solves the problem. Ideally you use a measurement mic, but even a smartphone mic can reveal large deviations.

I also agree that if you end up looking for new gear, second-hand is the way to go. It's true whether you are spending $1 or $10,000, you can get more for your money that way. Most good gear will last many years so used is often a good option.

 

[LifeCapsule]

In your position, I would absolutely look for a good 2nd-hand receiver or Hi-Fi amplifier rather than buy very cheap and dubious products.

Class D and Class AB can sound equally great, but there are simply no real bargains as new products under $75+.

It might sound silly, but actually you would better off saving up rather than buying bad $10 product after bad $10 product all of which will disappoint. Much of the gear I have bought has lasted decades, making each piece cost the equivalent of much less than $5 a month ..

The problem is second hand stuff here in Iran have problems like they make a bad idle noise which can be heard from the speakers and it's not fun , I just needed something minimum for my pc speakers that sounded like garbage

 

[LifeCapsule]

Oscillation ?
What is the power supply?

Power supply is a laptop charger 20v 3.25A it belongs to my old laptop and it's fine and healthy! I tried 24 volts , same results, even 12v 5a switching board was the same

 

[LifeCapsule]

It probably only needs a better op-amp ... fortunately it is in a socket so easy to swap.
View attachment 540719

I know... I know this isn't helping.

You think problem lies there ? Why the inductors get hot without a load connected to them is this normal? Idle hotness is weird to me , the chips under heatsink are original .. I checked it , I guess other components have low quality or this specific module sounds like this ! Most people won't notice the difference if they aren't up close... The bass it produces is half baked, idk which part is responsible for this , I triple checked everything

 

[kemmler3D]

You think problem lies there ? Why the inductors get hot without a load connected to them is this normal? Idle hotness is weird to me , the chips under heatsink are original .. I checked it , I guess other components have low quality or this specific module sounds like this ! Most people won't notice the difference if they aren't up close... The bass it produces is half baked, idk which part is responsible for this , I triple checked everything

I think you are right, an idle Class D amp should probably not be getting very hot. This is considered one of the main benefits of Class D. Sometimes weird stuff like ultrasonic noise or DC offset can load the amp without producing sound, but that is only a guess, I'm not an engineer like some people here!

 

[LifeCapsule]

It probably only needs a better op-amp ... fortunately it is in a socket so easy to swap.
View attachment 540719

I know... I know this isn't helping.

These inductors get pretty hot without a load faster than other inductors, nothing is connected to them no load no input still makes heat with 24v or 20 v adaptors but not so much with 12v ...less hot on 12v but still sounds crappy

 

[NTK]

You think problem lies there ? Why the inductors get hot without a load connected to them is this normal? Idle hotness is weird to me , the chips under heatsink are original .. I checked it , I guess other components have low quality or this specific module sounds like this ! Most people won't notice the difference if they aren't up close... The bass it produces is half baked, idk which part is responsible for this , I triple checked everything

The popular TI class-D amplifier chips, such as the TPA3255 and 3251, operate using AD (2-level) modulation. Their output when idle is not zero (see below). Notice that when the sine wave (red curve) is close to zero, i.e. at zero output, the PWM waveform is a square wave with 50% duty cycle. The 50% duty cycle PWM square wave only has frequency components at the switch frequency and above, and are blocked by the output LC filter, and the output to the speaker is zero.

Since no real inductor is ideal, there will be losses and the losses will turn into heat. That's the reason the inductors of the LC filters get hot. See this section of the TI application guide.

 

[solderdude]

You think problem lies there ?

Nope it certainly isn't hence all the emoticons.
It was an inside joke from the many 'this op-amp sounds better in this (or that) TPA3255 amp thread.

Why the inductors get hot without a load connected to them is this normal?

They used cheap inductors intended for mains filter to cut costs. Good inductors are more expensive. That's why they get hot. Even without a load the caps at the output are a substantial load. These caps have to be there in order to filter out the huge amount of RF crap that would otherwise be at the output.
Better inductors would have become less hot and look like this:

Idle hotness is weird to me , the chips under heatsink are original .. I checked it , I guess other components have low quality or this specific module sounds like this !

I won't make any comments about the 'sound' as I don't have one nor am I using the same load.

Most people won't notice the difference if they aren't up close... The bass it produces is half baked, idk which part is responsible for this , I triple checked everything

If bass is weird I would flip polarity of one of the speakers and listen if bass improves. Maybe the labelling on the amp or PCB layout is done wrong (just taking a stab at it because of the weird bass sound).

 

[JSmith]

This module is around $3 on AliExpress... wouldn't be expecting much from that.


JSmith

 

[RandomEar]

Honestly: As cheap as it is, I would still not expect this to sound really bad unless pushed. Most of the magic is in the amp chip, so even the cheapest board manufacturer really can't f*ck up that much.

Low power? Sure. Load dependent? You bet. Fails quickly? Maybe. But at normal listening volumes at home it should still sound OK at first glance. Maybe this specific board is broken or there is some egregious design flaw, but I would not bet on it for now.

@LifeCapsule I think the first thing you need - as mentioned by other members already - is some objective test. Use a phone mic + app or something else you've got on hand or can borrow and measure the frequency response. Do the same for your old class AB amp. Compare and report back here.

I hate to bring this up again and again, but right now, all you have is a sighted test. Likely without matched levels and with a long wait time in-between for switching the amps. This form of testing is notoriously unreliable and people report huge differences all the time which then immediately vanish when tested again with the right controls in place (blind, level-matched, quick switching between amps). That's why actual measurements are so valuable.

 

[Tell]

I have an amplifier here based on two really cheap TPA3118 boards and they sound more than fine. Sure they are load dependent (which can be fixed with EQ) and don't have tons of power, but I've been using them for years for various applications and I've had no problems with them. For around 5 euro each the bang for buck ratio is really just up the roof!

 

[iGmole]

Ty I'll try this but I don't have a line in port for sound it's only output
Maybe I'll just give up on trying to get better sound.
There is a saying ... You get what you paid for And I got what I paid for
Some things aren't meant for some people... Like sound stuff isn't my thing I guess

Do not fall into despair. Maybe consider doing IEMs if you can stand them? There is epic sound available at the 10-20 dollar range.

 

[LifeCapsule]

Nope it certainly isn't hence all the emoticons.
It was an inside joke from the many 'this op-amp sounds better in this (or that) TPA3255 amp thread.

They used cheap inductors intended for mains filter to cut costs. Good inductors are more expensive. That's why they get hot. Even without a load the caps at the output are a substantial load. These caps have to be there in order to filter out the huge amount of RF crap that would otherwise be at the output.
Better inductors would have become less hot and look like this:
View attachment 540774

I won't make any comments about the 'sound' as I don't have one nor am I using the same load.

If bass is weird I would flip polarity of one of the speakers and listen if bass improves. Maybe the labelling on the amp or PCB layout is done wrong (just taking a stab at it because of the weird bass sound).

I flipped polarities before I saw no difference ... Like you and many other people said it's because the whole thing is crap !

 

[LifeCapsule]

Do not fall into despair. Maybe consider doing IEMs if you can stand them? There is epic sound available at the 10-20 dollar range

Do not fall into despair. Maybe consider doing IEMs if you can stand them? There is epic sound available at the 10-20 dollar range.

I like speakers idk why lol I have wired apple earpods jack 3.5mm and they sound excellent, warm , clear and with good bass... Excellent for gaming , movies , music for long periods of time but I crave good speaker sound , humans never get enough..... Even when they have everything they look for another thing ، this is the human curse :/

 

[LifeCapsule]

Honestly: As cheap as it is, I would still not expect this to sound really bad unless pushed. Most of the magic is in the amp chip, so even the cheapest board manufacturer really can't f*ck up that much.

Low power? Sure. Load dependent? You bet. Fails quickly? Maybe. But at normal listening volumes at home it should still sound OK at first glance. Maybe this specific board is broken or there is some egregious design flaw, but I would not bet on it for now.

@LifeCapsule I think the first thing you need - as mentioned by other members already - is some objective test. Use a phone mic + app or something else you've got on hand or can borrow and measure the frequency response. Do the same for your old class AB amp. Compare and report back here.

I hate to bring this up again and again, but right now, all you have is a sighted test. Likely without matched levels and with a long wait time in-between for switching the amps. This form of testing is notoriously unreliable and people report huge differences all the time which then immediately vanish when tested again with the right controls in place (blind, level-matched, quick switching between amps). That's why actual measurements are so valuable.

Sadly I can't test it since my old amp which was ab class is broken I tried to fix it fix the buzz ended up destroying it completely ... And now I seriously have to look for amps or just give up since my desktop USB speakers work just fine , what you said to test is too much hassle to me , I just decide with my ears, whichever tone is nicer that's my go-to amp , all I want is a warm sound with a clear soft bass that melts you , in short I like warm enjoyable music not cold and precise.... It's like warm music is making love to your ears ... Specially during meditation
Since in my country there is zero Chance of getting a good amp imma give up and just buy a board like my old ab class amp if I can. Find any ...

 

[GGroch]

This inexpensive amplifier probably does not have post-filter feedback (PFFB). There is a lot on this forum about that, but in Class D without PFFB you can get peaks at high frequencies with some speakers that might cause the tinny sound you describe.

I read the OP differently, PFFB is not likely the culprit.

He describes "shallow sounds in the mids" and later that his tone controls do not impact it because he has no midrange tone control. PFFB, even in worst cases, impacts frequencies from 10kHz (just barely) to 15kHz (perhaps 1dB) to 20kHz (quite a bit, potentially). All of these frequencies are well above not just midrange, but any music fundamentals at all.

If you have great young ears it may show up as a slight increase or decrease in "air". Frankly, I have yet to learn of any blind test that proves you can hear it at all. If course, of you read about how awesome PFFB is, you will hear it. I think something else is going on here.

 

[pjug]

I read the OP differently, PFFB is not likely the culprit.

He describes "shallow sounds in the mids" and later that his tone controls do not impact it because he has no midrange tone control. PFFB, even in worst cases, impacts frequencies from 10kHz (just barely) to 15kHz (perhaps 1dB) to 20kHz (quite a bit, potentially). All of these frequencies are well above not just midrange, but any music fundamentals at all.

If you have great young ears it may show up as a slight increase or decrease in "air". Frankly, I have yet to learn of any blind test that proves you can hear it at all. If course, of you read about how awesome PFFB is, you will hear it. I think something else is going on here.

You might be right, but without PFFB frequencies below 15KHz can be affected as well. Usually measurements we see just show effects of different pure resistance loads. Peaking can be much worse with reactive loads of some speakers.

edit: also I did not realize the OP was noticing this not with music but with gunfire sounds. I have a hard time grasping how such sounds would be affected by possible problems with the amp but it is easier to hear frequency response changes with pink noise than with music so maybe this is similar in that way.