TPA3116D2 Amp

Did you also change the return input(neg) decoupling?

try tack soldering balanced inputs to bottom legs of decoupling cap all 4 nearest chip. this will show if input circuit is responsible. if it persists its input caps or traces or input connection. if it stops then after input.

i found my same issue was input caps.

Hi,
Is "return input" the small cap by the big one? If so, then yes, I replaced them also.
Sory for a noob question, maybe, but what exactly is tack coldering? Does it mean "shortening" the input path to go directly from terminal to the chip? What if I just remove all input caps? Because I have noise even with no source connected.
 
Also, I was having soldering session yesterday to check for all kinds of combinations, and my "oscons" seem to be a problem for sound quality. The amp sounds great for like half a minute, and then sound quality just deteriorates rapidly. When I switch back to stock power caps, this doesn't happen.
 
Hi,
Is "return input" the small cap by the big one? If so, then yes, I replaced them also.
Sory for a noob question, maybe, but what exactly is tack coldering? Does it mean "shortening" the input path to go directly from terminal to the chip? What if I just remove all input caps? Because I have noise even with no source connected.

When I say " tack solder" I mean just a temporary joint, like tack welding.

Did anyone answer your question about the oscons before? I had a quick look for data sheets, but have real trouble finding oscon data sheets. If they are adding to noise its probably best to remove them and put something else in while you track down where rest of noise is coming from.

Are all your inductor outer cases in A1 condition. Damage to these does happen!
 
When I say " tack solder" I mean just a temporary joint, like tack welding.

Did anyone answer your question about the oscons before? I had a quick look for data sheets, but have real trouble finding oscon data sheets. If they are adding to noise its probably best to remove them and put something else in while you track down where rest of noise is coming from.

Are all your inductor outer cases in A1 condition. Damage to these does happen!

All inductors seem ok. The oscons aren't really adding the noise, but make the amp crash all the time when playing a richer music like rock. Like 330uF is not enough to power the chip or something.. The stock nichicon is 1000uF.

Will try to remove all 4 input caps and see if the TV still disturbs the amp so loudly.
Because of the lodness of these sound artefacts i am almost sure that the problem is on the input side.
 
Last edited:
All inductors seem ok. The oscons aren't really adding the noise, but make the amp crash all the time when playing a richer music like rock. Like 330uF is not enough to power the chip or something.. The stock nichicon is 1000uF.

Will try to remove all 4 input caps and see if the TV still disturbs the amp so loudly.
Because of the lodness of these sound artefacts i am almost sure that the problem is on the input side.

330uF crashing amp? As in shutting it down? If this the case then other more experienced and technologically savvy folk on here may be able to identify problem. I am pretty sure the data sheet says down to 100uF is usable, depending on supply etc. I swapped out the 1000 for 470uF and it improved no end. I know others reported 220uF improved too.

As I said, tack a balanced input connection to input pins nearest chip underneath, that'll bypass board input and show for definite if input side of board is issue. I just cut end off an old headphone lead and connected, using Mobile device as source. I just used some old spare electrolytic from draw to imitate the input caps. Signal was well under 1V so the electrolytic was fine. You could just use the caps you removed from stock, as the problem still there it couldn't have been them.
 
Got it working! With wimas and oscons and everything.
Don't know what exactly was wrong but I basicaly resoldered everything. Even damaged the pcb during this solder-rage. I used 0.22mmq soft copper wire instead of my usual cat5.
The amp is now reasonably quiet and sounds very good.
Highs are not as strong as with yjred and thus listening is not fatigueing at all, bass isnt that strong which I would actualy like more, but overall, a nice amp. Now I need to figure out a case for it. Or go with legos again🙂
 
With most PSU voltages used, less than 750mv give maximum output you mean.

Looked at Sure website, 21V PSU, 0.74V input, 25/27 dB, they print today. Thermally limited 50W 4 ohm, 10% THD.

With forced cooling tpachip, higher psu voltage might give more output and need some more input.
 
Voltage gain can easily be calculated..

26db is a gain of 19.95. With 19V supply voltage and BTL configuration, there is 19V*0.707=13.43Vp_rms. So, 13.43/19.95=0.683Vp_rms at the input is the clipping limit.

You should consider some more loss at the output due to switching -19V-x
 
Last edited:
Just to brag a little: listening to a Duke Ellington's record I' ve listened many times, with this modded amp, it feels like owning at least a new record if not a turntable much better than the rp1 I have... level of detail is unprecedented (at least in my living room🙂)
 
I have a simple question. What speaker impedance would be best to match up with one of the pre made amps such as Breeze 2.1 or SMSL SA-60? I have one of each and am looking for new speakers. Probably will go use one of these to power up a pair of bookshelf speaks and use with a powered subwoofer. Thanks
 
I have a simple question. What speaker impedance would be best to match up with one of the pre made amps such as Breeze 2.1 or SMSL SA-60? I have one of each and am looking for new speakers. Probably will go use one of these to power up a pair of bookshelf speaks and use with a powered subwoofer. Thanks

it is difficult to predict. It depends on the impedance characteristics (vs frequency) of your speakers. Without that knowledge, typically one has to assume the nominal impedance quoted for your speakers (4 ohm, 8 ohm etc.) applies to the whole frequency range (20-20KHz) and choose the output filter accordingly. However, as xrk pointed out, the differences in sound from average speakers usually do not vary greatly. One has to listen for themselves to decide whether the filter needed to be optimised for a particular speaker.

Regards,
 
Thanks for the feedback. I guess the advice to listen for yourself is the best advice anyway. For whatever it may be worth two rather different speakers I'm considering for these are the Fusion Alchemy 8's and a front ported bookshelf speaker with the Alpair 10.3. They will be used in an office space and 90% at modest to medium levels, with the occasional crank it up session at the end of the day. Also virtually no options on room placement since the will have to sit on a table top no more than a foot from the back wall which is a double pane laminated glass wall. The office is shared by 4 people and we all like music and one of us is a classically trained pianist, and we all love music but there is simply zero floor space for speakers. In any case I have hopes for the little TPA3116 in the office.