TPA3116D2 Amp

[Saturnus]

Originally Posted by KJA 2013

They seemed like a good choice to me too, but there are so many 10uH inductors to choose from. Have you used these before?

They are used in the amp10 from 41hz. The only drawback is that they're through-hole which makes it a little annoying when designing an amp meant for having components only on one side and the reverse full ground plane thermal glued to a heatsink.

[KJA 2013]

Originally Posted by Saturnus

They are used in the amp10 from 41hz. The only drawback is that they're through-hole which makes it a little annoying when designing an amp meant for having components only on one side and the reverse full ground plane thermal glued to a heatsink.

Luckily the physical size of the inductor is such that by bending a small foot on each lead enables the inductor to fit perfectly on this board. Wasn't sure they would fit, but they fit quite nicely.



.

[Saturnus]

Originally Posted by nesha

Small TPA3110 runs filterless and sounds good, and I did not notice any problems even with my 10m loudspeaker cable. Wushuliu reported good performance and Chakija (on diyaudio.rs) tried and liked it with his Pi clone speakers. I'm no expert,maybe 3110 uses somethig else for filtering...

TPA3110 is a much smaller amp and specifically designed for "filterless" output due to a EMI suppressing modulation scheme. Filterless in this respect means ferrite beads and/or very short leads. Running completely filterless with long leads means you're sending out loads of RFI noise, and I strongly advise you not to do that.

You can also run TPA3116 with ferrite bead output filter only but it's highly recommendable to use a full LC output filter due to the increased efficiency and sound quality you achieve with a properly designed LC output filter.

You should under no circumstances run a TPA3116 completely filterless!

[mondogenerator]

Originally Posted by xrk971

Regarding the LC filter: it is 10 uH compared to the almost 1 mH inductance of speaker voice coils I think the effect on overall reactance is minimal. If you were to drive multiways with huge coils in the XO, even less important of an effect.

Agreed. 10uH is significant when using a low Le full ranger though

It the fact its an LC, not LR or LC which bothers me. Flat impedance is actually quite important with these amps id guess. Its a downside,not necessarily a dealbreaker

[wushuliu]

Originally Posted by FantasicoGrande

Ah yes, as I look closer at you board I see it is the same as mine.
What caps do you have in there? Looking at the first photo it looks like you have replaced them all?

They look like Wima caps.

[KJA 2013]

Originally Posted by wushuliu

They look like Wima caps.

Yes, the red caps are .68 Wimas that are about to be replaced with some .68 ERO caps that I prefer. The grey ones are ERO and the .01 are polystyrenes.

[radiosmuck]

Originally Posted by KJA 2013

The top set in this first photo, and they are also shown installed in this second photo. The lower set in the first photo are the inductors from Ice Components.

Is there a "real" sonic advantage to be had replacing the coils because the board now looks pretty ugly?

[wushuliu]

Originally Posted by KJA 2013

Yes, the red caps are .68 Wimas that are about to be replaced with some .68 ERO caps that I prefer. The grey ones are ERO and the .01 are polystyrenes.

Are the EROs MKP or MKT?

[KJA 2013]

Originally Posted by wushuliu

Are the EROs MKP or MKT?

I tried both polypropylene and polyester and ended up prefering the MKTs that I had on hand.

[wushuliu]

Originally Posted by KJA 2013

I tried both polypropylene and polyester and ended up prefering the MKTs that I had on hand.

Cool. Thanks.