diyAudio (
-   Class D (
-   -   TPA3122D2 based amp just clicking and hissing along (

cynric 5th July 2009 05:18 PM

TPA3122D2 based amp just clicking and hissing along
Hello everyone.
Yesterday I built -or tried to do so- a TI TPA3122 D2 based class D amp on a breadboard. I followed TIs appnote for the BTL config on page 15 of the datasheet, with the one exception that I used 10H/1.5F for use with 4Ohm speakers.
I also added a simple opamp-based buffer/inverter to get my SE inputs to balanced ones for the TPA. The power supply is buffered by 6000 FCs. All is powered by a 24V supply (not a good one though).
All caps are WIMA MKS film caps - except the stuff >=10, those are Panasonic FC low esr electrolytes.
I built two identical paths (L,R), of which neither work as they should.

At first I had the !shudown pin soldered directly to VCC and the mute to GND. This resulted in one amp just clicking an hissing, and the other one clicking and hissing and sometimes amplifying bits of the input inbetween. However, if I increased the input singal's volume the crackling time got more and the time the music came throug got less. Both amplifiers got pretty warm though.

Next I tried changing the Bypass-C to a 10 tantalum I had lying around with the result that the formerly partially working amp didn't work not either.

Now I fiddled some 10k Rs between !shutdown and VCC and between mute and GND, which seems to have reduced the heating of the amps and the hissing, but also killed any signal amplifiacation. The only thing coming out right now are a clinkg sound and barely recognizable pieces of music, both only right after plugging power in or out.

I do not have a scope or parts/low cap capacitors at home (hell, I even ran out of solder and desoldering wick during this project), but I will take it to work tomorrow, where there are an old scope and lots of parts.

Do you guys have any idea what might have gone wrong here? I checked the board several times for shorts or circuit mismatches, but everything looks as it is supposed to..

theAnonymous1 5th July 2009 05:56 PM

Hi cynric,

With my TPA3122D2 BTL boards I have SD connected directly to VCC and MUTE connected directly to GND without any problems.

How does your circuit compare to the measurements I took of my board below?

LIN (Pin 4) = 2.94V
RIN (Pin 5) = 2.94V
BYPASS (Pin 6) = 2.927V
VCLAMP (Pin 9) = 10.8V
LOUT (Pin 19) = 11.8V
ROUT (Pin 12) = 11.8V
BSR (Pin 13) = 21.5V
BSL (Pin 18) = 21.5V

BTW, BTL into 4R @ 24VCC is not a good idea with this amp; it will go into an over-current condition very quickly. It won't hurt anything though, you will just get limited output compared to an 8R load.

cynric 5th July 2009 06:14 PM

Hey theAnonymous1,
thanks for your reply.

My voltages - inputs and outputs disconnected:

LIN & RIN are at 3.0V
Bypass at 2.99V
VClapm about 11V
ROUT and LOUT cannot be measured with my mutlimeter as they fluctuate, seems to be just very few V, but I can't really tell.
BSL and BSR are varying between 10.5 and 12.0V

This is identical for both amps.

As for the speaker impedance, I'm not sure what speakers will be hooked to the amp, so I built it for 4R. I can still modify the filter if I have to. Or can a filter mismatch be the cause for my trouble?

theAnonymous1 5th July 2009 06:57 PM

The input side looks fine; there seems to be something wrong on the output side.

What type of caps are you using for BSL and BSR? Are you sure there isn't anything shorting one or both of the outputs to ground?

The filter values you chose shouldn't cause any problems with the stability of the amp, but there will be a slight peaking in the high frequency output when used with an 8R load.

cynric 5th July 2009 09:58 PM

The BS caps are WIMA MKS-02 220n polyester types. I'll see if i can swap those for ceramic types as suggested in the datasheet tomorrow.
No shorts to Gnd on the outputs, only the 4.7k resistor as per appnote (and the 10H high current coils and the 1.5F MKS2 caps).

Thanks again & good night (at least over here).

cynric 6th July 2009 05:27 PM

Well after lots of measuring and swapping parts I found the problem: The TPA didn't like my opamp stage. I used the main power for it, thus referencing its outputs to +12V -which i thought would not matter, because the amps inputs were AC coupled anyways. Well, turns out it did matter.
Now I put in a DC-DC converter to power the opamp, resulting in GND-referenced outputs and the one channel I finished is currently playing fine.

I'll finish the second channel sometime soon and report back then. Thanks for your support so far!

cynric 11th July 2009 02:22 PM

just to let you know: i got it working now. initially the second channel still crashed when at high input levels; i have kept my supply lines for both channels seperate except for one point in the center of the ICs close to the bypass caps, turns out that the amp with the slightly shorter path seems to have drained the other channels bypass cap too, causing it to stall.
Anyway it's fixed now and after fiddling with my too-small enclosure and some nasty buzzing i caught on my signal input wiring i now have a working stereo btl class D amp in a 17x11x3cm box capable of driving my speakers to uncomfortably loud levels :D

All times are GMT. The time now is 12:03 AM.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright 1999-2017 diyAudio

Content Relevant URLs by vBSEO 3.3.2