TPA3116D2 Amp

Interference cancellation. Even shielded inductors emit RF noise especially at such short distances. By arranging them in pairs in close proximity and on right angles the inductors are most strongly magnetically connected and act like chokes at very high frequencies. It is important to note that they should not be magnetically connected at all at lower frequencies below about double the switching frequency. I can't imagine why this would cause them have lower overall loss, ie. get less warm, but I might have some ideas that I could test out.

Thank you for the explanation. I understand the magnetically coupled aspect, but still not certain how the RF cancels each other. The RF waves have to be 180 degree out of phase with each other to achieve that. How does one assure that?

Regards,



Regards,
 
Audiobah and Ydbz use 22uH inductors, so simple answer is 22uH aren't as likely to get warm🙂 But more powerfull 10uH's don't get warm either, which makes the Coilcraft specs suspicious maybe???

Well, do not mean to argue or anything, but you are simply stating the difference in observation and it is not necessarily the answer. To be fair, we are discussing about different inductors in different amps. So the cause of the heat might be different?

Regards,
 
Thank you for the explanation. I understand the magnetically coupled aspect, but still not certain how the RF cancels each other. The RF waves have to be 180 degree out of phase with each other to achieve that. How does one assure that?

Regards,

The two output inductors are sitting on the same plane, and as always in binary class D H-bridges as we're discussing here the outputs are differential, ie. exactly the same but with reverse polarity, so the magnetic field is also exactly (or at least very nearly) opposite.

Now, since we're talking about the TPA3116 it could also be used in ternary (tri-state) mode, and then they would not be exact opposites at low to medium outputs. Although the TPA3116 implementation of ternary mode could better be described as floating single ended at these lower output levels instead of true ternary mode but it's pretty much equal.
 
Last edited:
The two output inductors are sitting on the same plane, and as always in binary class D H-bridges as we're discussing here the outputs are differential, ie. exactly the same but with reverse polarity, so the magnetic field is also exactly (or at least very nearly) opposite ........

I have five Class D amps of different manufacture, various inductor placements, including a Feixiang 2.1 with six stock inductors in a row & almost touching, as well as an unboxed version of the one in the pic below.

No hum, EMI, heat or other issues, other than the ubiquitous on/off click on a couple of them

Lucky, I guess 🙂
 

Attachments

  • Clipboard01 small.jpg
    Clipboard01 small.jpg
    37.9 KB · Views: 545
I have five Class D amps of different manufacture, various inductor placements, including a Feixiang 2.1 with six stock inductors in a row & almost touching, as well as an unboxed version of the one in the pic below.

No hum, EMI, heat or other issues, other than the ubiquitous on/off click on a couple of them

Lucky, I guess 🙂

Why? That's exactly what I write is the optimal way...
 
I'm almost about to give that one a try, too. What do you think about it's additional sub crossover, does it work as supposed to?

Once I get my Feixiang 2.1 in, I will be testing the cross-over. Danzz, notified me he is going to try and take pictures of his Feixiang 2.1. We are hoping some board members can identify the various pathways on this board. It apparently includes a soft-start function as well, and has a mute board port.
 
On the Breeze 2.1 I found that the middle pot changes the slope of the low pass filter for the sub - with a corner freq of around 350Hz. Good for FAST but not for a regular sub. The freq doesn't shift, just how steep the fall off is.

i still don´t get over this.
how can the slope = steepness = order of a filter be manipulated by a single pot?
maybe the pot is for setting the q-factor of the filter?
but this would not make sense to me with a lpf (ok, not everything that is being built makes sense, so maybe...).
is there a circuit diagram?
 
Hi,
I would like to try a linear power supply (~21V) with a toroid transformer for my TPA3116.
Do you know where I can find shematics and documentation?
Probably it should be possible to buy one, but I don't know where I can buy one.

Any help?

Single voltage PSU can be a simple as a transformer, bridge rectifier and smoothing cap(s). to get 21v use a 15v transformer, 120va or larger. Caps should be 10,000uf or more, 35v or more. Filters and snubbers can be added but this is the basic and will work fine.

Here is a sample one I grabbed off the web. C2 isn't needed as bypass caps are on the 3116 board. Try to find a good quality cap with low ESR, using multiple caps to get the required capacitance lowers ESR as well.
 

Attachments

  • PSU.jpg
    PSU.jpg
    29 KB · Views: 611
Last edited:
Single voltage PSU can be a simple as a transformer, bridge rectifier and smoothing cap(s). to get 21v use a 15v transformer, 120va or larger. Caps should be 10,000uf or more, 35v or more. Filters and snubbers can be added but this is the basic and will work fine.

Here is a sample one I grabbed off the web. C2 isn't needed as bypass caps are on the 3116 board. Try to find a good quality cap with low ESR, using multiple caps to get the required capacitance lowers ESR as well.

post #5685

That darn web.

It put the bridge rectifier in wrong.

Rotate it 90 degrees clockwise and it will be OK.

🙂
 
Thanks irribeo for explaining Sure temps. They don't sound so bad after all. When my Sure arrives I will also be doing some tests.

Finally I would like to make a dual PBTL system using Sure boards and 24V SMPS, but I will first see how it goes with only one board.
 
Finally made bootstrap snubber mod to YJ b/b. I didn't notice any big differences, but maybe the sound became a bit lighter (which can be considered as a good thing).

Kemet 330 pF ±5 % C0G are the caps I used.

Previous mods are listed here.

Good job, nice and tidy, easy isn't it!

I recall it was after doing that mod my YJ amp really came alive, in particular clarity in the tops. These changes are not always easy to hear straight off but longer term listening becomes more enjoyable.
 
Feixiang 2.1

eqsOne;4106792[I said:
]I'm almost about to give that one a try, too. What do you think about it's additional sub crossover, does it work as supposed to?[/I]

Overall, the YJ B/B board is far superior to the Feixiang 2.1IMHO,but it's not 2.1

The Feixiang xover control isn't the best .... again, IMHO. The opamps are socket mounted so you can play around with those easily, but there isn't any improvement with any of the popular opamps. The best result I got was putting a simple subwoofer LC xover between the amp's subwoofer speaker output & the subwoofer speaker, but I had the xover lying around doing nothing. By the time the xover is costed in, something like a 100W Yung subwoofer plate would be far superior & with plenty of control, for not a lot more $s

However, if your normal speakers don't have much bottom end, your expectations are not too high, you have a passive 4 Ohm subwoofer & all you want is more bottom end, then it's a very cheap & cheerful solution that works 🙂
 
Overall, the YJ B/B board is far superior to the Feixiang 2.1IMHO,but it's not 2.1

The Feixiang xover control isn't the best .... again, IMHO. The opamps are socket mounted so you can play around with those easily, but there isn't any improvement with any of the popular opamps. The best result I got was putting a simple subwoofer LC xover between the amp's subwoofer speaker output & the subwoofer speaker, but I had the xover lying around doing nothing. By the time the xover is costed in, something like a 100W Yung subwoofer plate would be far superior & with plenty of control, for not a lot more $s

However, if your normal speakers don't have much bottom end, your expectations are not too high, you have a passive 4 Ohm subwoofer & all you want is more bottom end, then it's a very cheap & cheerful solution that works 🙂

Despite that I got this amp long ago, I havent had the chance to try it yet, I make my diy audio things in an old house that belonged to my father, but some burglars went in and robbed me a few things, speakers, a couple of subs, tools... anyway enough for not letting me finish a 2.1 project boombox I had in hands for a present for a friend...

Anyway back to the business, I believe that with a few mods this board can sound much better, of course it wont be a high end thing, but Im sure that it can sound really good for the money it costs... The mods I would do are a few, first of, due that the op-amps are in sockets I would upgrade them by lm4562 (I have a couple in my components box, but any suggestion its welcome), easy upgrade that can be a really surprise, at least in my lepai it was... second mod, changing caps, upgrade input caps before the main pot by wima mks2, as they fit nicely, then upgrade dc coupling caps, as suggested in this thread, oscons or nichicons, upgrade main power caps by panasonic FC or nichicon FG ( also have a couple of them around), and a mod to study if its possible it would be boostrap snnuber, there is not much gap for it, I would do only this mod in both satellite outputs.