Is PFFB in new tpa3255 class D amplifiers really a good idea, soundwise ?

[Musician]

The new chip based tpa 3255 amplifiers coming out now measure really good and uses post filter feedback technique , that is used global inside the amplifier. This way the filtering coils eventual sound-colour in the amplifer are cancelled, at least with static signals. Tests on audiosciencereview shows really good results. But how is the sound with music ?

Already in the -70 Matti Otala proved that to much global feedback can be bad for the sound. Is there a technological jump nowadays that makes his statement obsolete ?

Are the new class D tpa3255 amplifiers with PFFB really sounding better than the same tpa3255 amplifiers without PFFB? What are your opinion on this ?

 

[13th Duke of Wymbourne]

I don't think Otala proved anything new. His 'work' on transient overload in feedback systems had been well understood by control engineers for decades. NFB reduces non-linearities for sure but is often questioned as being bad for 'sound'. Putting aside transient overload I have never seen an analysis that shows that NFB does something wrong. I'd like to see less hand waving on this topic and more use of the hands to write down the equations to prove their points. As for the PFFB, that isn't new as Bruno Putzeys does it in his amps.

 

[Peter Stockwell]

Tests on audiosciencereview shows really good results. But how is the sound with music ?

I don't know, but I have plunged deep into some of the the threads over there about Post Filter FeedBack (PFFB) implementation. I've been running Class D amps for some time (Temple Audio Monoblocks 2017's), and I am inclined to believe that the designers of the amps with PFFB aren't just doing it because it's a flavour of the month.

PFFB has been around for a while, but, IIRC, it was covered by a Texas Instruments patent which has since lapsed. A number of new commercial applications are appearing (Fosi V3 monos, Eversolo Amp F-2, Starke Audio Fieras, to name a few).

 

[Musician]

I don't know, but I have plunged deep into some of the the threads over there about Post Filter FeedBack (PFFB) implementation. I've been running Class D amps for some time (Temple Audio Monoblocks 2017's), and I am inclined to believe that the designers of the amps with PFFB aren't just doing it because it's a flavour of the month.

PFFB has been around for a while, but, IIRC, it was covered by a Texas Instruments patent which has since lapsed. A number of new commercial applications are appearing (Fosi V3 monos, Eversolo Amp F-2, Starke Audio Fieras, to name a few).

With PFFB theres always more IC :s involved, I find it hard to believe that a Tpa3255 class D amplifier with PFFB and two ne5532 will sound better than the same amplifier with only one opa1612. Maybe Im wrong on this ? With correct PFFB implementation the measurements with static ( non musical ) signals will turn out to be better .