sorry to seem so dramatic but this is probably the 6th time I post on a forum (instructables, adafruit makezine, element14, etc) and havnt found any answers. My TPA6021A4 amplifier I made has alot of distortion that goes away when i touch the positive terminal of the audio input with my finger. the weird thing is, every single (5) audio amplifier ive tried to make so far has had the same problem. the reason I am using the TPA6021A4 chip is because I emailed Ben Heck and that is what he suggested I use, but I cant get it to work. I am making them on breadboards mabey thats the problem? please help.

sort by: active | newest | oldest
fossilshark (author) 1 year ago

OK so heres what i have now:

i am using 47 nanofarad caps since thats all i have. i have the 10uf cap and the 47nf cap connected to the ground (pin one) and +5v (pin 3) just like the schematic. i have the 330uf cap connected to the +out signal (pin 20) with the 1k resistor going to ground. i have a 47nf capacitor connecting audio in+ to the right input (pin 4). i have nothing connected to the left input and output pins. the 10k, and 100k pot no longer changes the volume (pin 18) but it does not turn on and off. the music just has alot of distortion. could this be because of the breadboard? or is there another problem.

fossilshark (author)  fossilshark1 year ago

and just because i hooked up the left out and left in pins and wired the out in parralel with the right out and helped ever so slightly with the distortion

fossilshark (author) 1 year ago

sorry about that forgot to add the pics

iceng1 year ago

Does it play well when someone else touches it ?

Ok, let's try to get it all together...
You followed the standard schematics for the chip.
You have distorted audio unless you touch the input.
You use a wallwart as the power supply.

Is it a standrad type wallwart or electronic switchmode?
The later often have troubles supplying a stable voltage free of interference, a filter and big cap can help to keep the unwanted signals out of the amp.
Best option would f course be to actuall check the voltage output with an oscilloscope to make sure it is a clean DC.
The standard setup according to the data sheet is stereo, to get the big speaker out of the pic try with some headphones or a tiny 0.5W speaker.
You can not use a massive speaker on a tiny amp!
And again: what is your input?
Without a proper input the amp will not work.
As suggested grounded inputs should turn the amp silent.

I suspect two things:
a) The power supply is crap.
b) The amp is constantly over the limit because the load is far bigger than anything the amp could supply.

The input biasing on this amp shows ALL the inputs connected via 0.47uF caps. I suspect he's upset the biasing

fossilshark (author)  steveastrouk1 year ago

how do i...err...un-upset the biasing?

1.) Use the correct values for the caps ?

2.) Use good decoupling on the power supply ?

3.) Measure the supply volts with an oscilloscope and see if the supply is dropping out.

4.) Reduce the input amplitude to the acceptable range for the amplifier, confirmed with a scope.

5.) Don't randomly connect pins together to see what happens ?

fossilshark (author)  Downunder35m1 year ago

i was using a 5 volt regulator connected to a 12 volt dc supply but when i changed over to the wall wart nothing happened. the input is a audio jack connected to my phone. 3w speaker is the smallest i have, i do not have access to an oscilloscope. The problem isnt distortion anymore. it will shut off every 2 seconds and only stay on for 2 seconds when i have music playing.

fossilshark (author)  fossilshark1 year ago

scratch that, on and off problem is fixed, distortion is back.

...only via a 470nF cap, for this amplifier.

You have 5 and 7 to AGND via 470nf caps ? Do you have 19 tied to AGND too ?Do you have /Shutdown tied high and /fade tied low ?

You are substituting a 10 X smaller cap than specified in the datasheet, that won't help a bit.

What's your source ? Can you turn the source volume down ? Does it distort less when you do ?

iceng1 year ago

I don't see the very necessary 10uF electrolytic and the o.47uF ceramic capacitors across the power input.

fossilshark (author)  iceng1 year ago

when i put them (electrolytic but it shouldnt matter) it stopped working all together

It matters a lot ! The .47 must be a ceramic (though a 0.1uF will probably work) The circuit is probably suffering from noise injected from the supply brick.


Not uncommon.

You generate a current, your heart beat, radio signals, and even muscle movement.

So when you touch the positive you add a circuit to your amp.

You need a input filter.

Ether a cap and inductor or a resistor and inductor calibrated to the frequency range you are amplifying.

fossilshark (author)  Josehf Murchison1 year ago

im trying to amplify the overall signal im not trying for just highs or lows

That's not what he's saying - you shape the frequency response of the circuit to the audio band.

Post your schematic.

Since it is an audio amp an audio range filter on the input.

I can see the 3 watt speaker shutting down the IC you need something smaller or a power amp using the IC as a preamp.

What power supply are you using ?

fossilshark (author)  steveastrouk1 year ago

5 volts coming off a wall wart directly to the breadboard. was using a 5v regulator but it got too hot for a 12v supply so ill throw that on the final design

seandogue1 year ago

When an amplifier isn't connected to a valid load, it will act as an antenna.

When not connected, try shorting the amplifier's input leads. your noise should go away.

fossilshark (author)  seandogue1 year ago

im not worried about the noise when its not connected just when my music is playing.

Schematics, images, details????????
And I would not even bother building an amp on a breadboard, there is a reason people use circuit boards and keep connection between parts short.

And there are reasons people use breadboards to learn how to build circuits...

Yes, to see if it is working but for audio amps?
I would expect to get a lot of unwanted iterference from the wires, not to mention possible connection problems.
What works fine to light a LED not alwys works as good for an audio signal.

If I am notmistaken the chip would need a heatsink for 2W output already, the speaker looks slightly bigger than 2W to me.
What is the actual input? The chip only uses 0.3V +-
Do you use a coupling capacitor for the input?

fossilshark (author)  Downunder35m1 year ago

i put in all the necessary caps from the schematic in the datasheet but they didnt help at all. and the chip doesnt heat up one bit.

The point was, this guy is clearly on the early part of the learning curve, and may want or need to use the same components in several circuits as he goes - not easy if you are using PCBs (and may still lack in soldering skills).

TL;DR - cut the guy some slack!

fossilshark (author)  Kiteman1 year ago

i solder fine lol im just trying to get this breadboarded so i can make sure it works

Just saying... :(
Maybe it would be a good idea to try with some headphone instead of the massve speaker ?

fossilshark (author) 1 year ago

UPDATE: those pictures were my old setup but you get the idea. i am now using a 3w speaker and the chip does not get hot at all. it works but it will stay on for a few seconds and shut off for a few seconds, i suspect the chip cant put out a full 2 watts mabey? becaue it seems like its shutting down for overload protection im not sure. its connected to an audio jack into my phone.


If the audio input is not connected to a mic or some other device, it should be shorted to ground