3e Audio TPA3251 TPA3255 Finished Amplifier with PFFB is coming!

[SMen]

Are there not two different measurements in discussion here? SINAD vs THD up to 5 watts where most of the power is relevant anyway?
It's a question of whether the THD is audible, or thresholds of THD audibility surely? I'm sure power matters though, just how much vs speaker sensitivity.

What is interesting about the 3251 model is the 'over engineering' aspect with PBTL - it gives more current possibly for those lower impedance loads, while not delivering extra watts, which may or may not be necessary.

 

[MCH]

Worth resurfacing this 3251 vs 3255 comparison

• xx51 has slightly lower THD than xx55 up to around 5 watts
• 5 watts to 100 the chips perform similarly
• above 100 watts the xx51 starts to break up, the xx55 goes to over 200W before also breaking up

if you don't see yourself pushing serious power (either for high volume or insensitive speakers), I would choose xx51 for providing slightly better performance for most typical listening ranges. I concede the point that both amplifiers are well in to inaudible THD range, so the real world difference is negligible. Still, the engineer within me would be happier knowing I have the more optimal chip for my use case.

That graph is mislabeled. The difference you see at <4 w is noise, that could be relevant for some use cases, maybe...

 

[harkpabst]

Are there not two different measurements in discussion here? SINAD vs THD up to 5 watts where most of the power is relevant anyway?
It's a question of whether the THD is audible, or thresholds of THD audibility surely? I'm sure power matters though, just how much vs speaker sensitivity.

What is interesting about the 3251 model is the 'over engineering' aspect with PBTL - it gives more current possibly for those lower impedance loads, while not delivering extra watts, which may or may not be necessary.

Very true.

And let's not forget that 200 W vs. 100 W allows just 3 dB higher SPL. This can be relevant at times, but in many situations it's just not. More sensitive speakers can so the same and potentially benefit from lower noise. As so often it's not a quzof what is better but what is better suited.

3e Audio already made their point why they offer the TPA3251 variants at all. What's the point in asking them again? Who would expect them to change their mind right now?

@Tofffsy, there wil be no difference in sound stage. Channel separation is best with monoblocks, in theory. In practice I predict it's already beyond any perceivable threshold with the stereo amps.

The number of opamps will be identical between A5 and A7.

 

[yo!]

Are there not two different measurements in discussion here? SINAD vs THD up to 5 watts where most of the power is relevant anyway?
It's a question of whether the THD is audible, or thresholds of THD audibility surely? I'm sure power matters though, just how much vs speaker sensitivity.

What is interesting about the 3251 model is the 'over engineering' aspect with PBTL - it gives more current possibly for those lower impedance loads, while not delivering extra watts, which may or may not be necessary.

Yes, PBTL is the most tempting method, PostFilter PBTL is without a doubt the best choice, but even with the simpler PreFilter PBTL that most designs are (2 coils vs. 4 coils for output filters per chip/channel, higher THD but still pretty good) they're superior in stability and with the fitting SMPS it doubles the power too, but this is not the point. You know it does the job without any doubt and the chip stays cool. It's not this expensive when you think about the versatility of such an amplifier and long term usage.

Comparison for the 3255 PBTL post and pre filter, not much to worry @ 0.05 THD / 50 W when stable coils are used.

 

[harkpabst]

Yes, PBTL is the most tempting method, PostFilter PBTL is without a doubt the best choice, but even with the simpler PreFilter PBTL that most designs are (2 coils vs. 4 coils for output filters per chip/channel, higher THD but still pretty good) they're superior in stability and with the fitting SMPS it doubles the power too, but this is not the point. You know it does the job without any doubt and the chip stays cool. It's not this expensive when you think about the versatility of such an amplifier and long term usage.

Comparison for the 3255 PBTL post and pre filter, not much to worry @ 0.05 THD / 50 W when stable coils are used.

I noticed that you edited your post, but it looks like you got it wrong now where it's been correct before: Pre-filter PBTL offers better performcance that Post-filter PBTL.,

However, I cannot remember anyone even discussing this over the course of the last one or two years. You see two coils per channel on the 3e Audio DIY boards and on the boards for the new A5/A7 series of finished amplifiers. It's Post-filter Feedback that set them apart from the crowed as they were the first ones to really get it going in a commercial design (as far as I remember).

 

[yo!]

This was a sideline only, not talking about feedback but filtering before or after bridging and I hope we see some nice 3e bridged amps. Pre/Post Filter PBTL and PFFB combined is another story.

 

[harkpabst]

This was a sideline only, not talking about feedback but filtering before or after bridging and I hope we see some nice 3e bridged amps.

I see.

But still, this ...

Yes, PBTL is the most tempting method, PostFilter PBTL is without a doubt the best choice, but even with the simpler PreFilter PBTL that most designs are (2 coils vs. 4 coils for output filters per chip/channel, higher THD but still pretty good) ...

... makes no sense. Pre-filter PBTL offers both advantages, fewer parts and better performance. It's just not supported by all TI Chips. The TPA325x family does support pre-filter PBTL.

Pre/Post Filter PBTL and PFFB combined is another story.

It's exactly the story told by 3e Audio.

 

[SMen]

yo! said:

Pre/Post Filter PBTL and PFFB combined is another story.

It's exactly the story told by 3e Audio.

Just to make sure I understand this correctly - 3e Audio are combining pre filter PBTL with post filter PFFB?

 

[harkpabst]

Just to make sure I understand this correctly - 3e Audio are combining pre filter PBTL with post filter PFFB?

Well, I don't know their schematics nor board layout, I'm just counting components.

With every form of bridging we need fully symmetrical filters (there's no common ground reference). BTL uses two outputs to feed one speaker, PBTL four amplifier outputs This image is from the same application report SLAA822 as the previously posted distortion figures for pre-filter vs. post-filter PBTL:



A stereo power amp employing PBTL needs two TI TPA325x chips and at least four inductors in the case of pre-filter PTBL. Using post-filter PBTL one would need eight inductors, which are obviously not there.

 

[yo!]

Too many PBTLPFFBBTLPFFB letters
Not sure about the e3 mono architecture. They claim 20 dB gain for the module, isn't it true for any kind of PBTL in combination with PFFB, that the gain factor drops to around 15 dB only?
Is it BTL+PFFB then?

 

[SMen]

Well, I don't now their schematics nor board layout, I'm just counting components.

With every form of bridging we need fully symmetrical filters (there's no common ground reference). BTL uses two outputs to feed one speaker, PBTL four amplifier outputs This image is from the same application report SLAA822 as the previously posted distortion figures for pre-filter vs. post-filter PBTL:

View attachment 410538

A stereo power amp employing PBTL needs two TI TPA325x chips and at least four inductors in the case of pre-filter PTBL. Using post-filter PBTL one would need eight inductors, which are obviously not there.

View attachment 410539

Excellent, Mr. Holmes! So it looks like pre-filter PBTL

 

[harkpabst]

Too many PBTLPFFBBTLPFFB letters
Not sure about the e3 mono architecture. They claim 20 dB gain for the module, isn't it true for any kind of PBTL in combination with PFFB, that the gain factor drops to around 15 dB only?
Is it BTL+PFFB then?

PBTL or BTL shouldn't make any difference regarding gain. PBTL just increases current output capabilities over BTL, voltage and thus gain remain the same.

The "open loop gain" of the TPA325x chips isn't all that high and to the best of my knowledge non-configurable (theses are not multi purpose op-amps, but highly specialized audio amplifiers after all), so the effectiveness of overall feedback is limited. The application report suggests a setup resulting in ~6.5 dB lower gain.

@3eaudio have already confirmed the PBTL design of the A5, A7 and A7 mono (vs. BTL for the A5se and A7se). It's right on their web page.

A5x A7x Series Next-Gen PFFB Amplifier | 3e Audio

www.3e-audio.com

 

[yo!]

OK, then we should expect a less gain with PFFB implemented too, hopefully this will be covered when tested.

Wonder if PFFB will do some magic regarding signal overshoot, a way bigger thread than all these THD focused comparisons. Some say the higher the frequency, the less it's effective.
Wish to see step response measurements for comparison in the future.

 

[harkpabst]

Gain figures of the individual products have already been provided.by 3e Audio (for balanced and unbalanced) connection:

3e Audio TPA3251 TPA3255 Finished Amplifier with PFFB is coming!

@3eaudio , will any of the amplifiers make it for the 11.11 sale? aggressive but early of Dec should be ok.

www.audiosciencereview.com

The TI chip can be complemented with another gain stage, it just has to be low noise, low distortion.

Few testing facilities publish amplifier step response measurements nowadays. They are easy to perform, easy to read ... and incredibly hard to interpret in any meaningful way.

 

[yo!]

Oh, clean preamps are included, this makes things easier with lower output signal gear.
And the steppies ... yes, hard to say if those nanosecond overshoots are problematic or just adding a little 'freshness' to the sound or maybe nothing tremendous at all. Fact is, a 'perfect' amplifier will not have any significant overshoot, as many well built AB or A amps can do. It's wishfully simply the next step for class D evolution.

 

[wkamil]

Today is December 1st, do you think it will actually be possible to order A5/A7 today as 3ea Audio was tempting?

 

[Graham849]

Archimago is reviewing the upcoming models

Part I: 3e Audio - A5 Stereo and A7 Mono Class D + PFFB, TI TPA3251/3255 Amplifiers - Features and initial measurements.

A blog for audiophiles about more objective topics. Measurements of audio gear. Reasonable, realistic, no snakeoil assessment of sound, and equipment.

archimago.blogspot.com

 

[3eaudio]

Today is December 1st, do you think it will actually be possible to order A5/A7 today as 3ea Audio was tempting?

just share some progress here, release postpone to 10/Dec due to factory over capacity,sorry for this delay.

 

[Inertiaman]

just share some progress here, release postpone to 10/Dec due to factory over capacity,sorry for this delay.

Appreciate the timely update. Transparency is much better than silence.

Even if order entry isn't yet available, can you share specific pricing of each model?
Will all models me available simultaneously on ~ Dec 10, or is there a sequence of availability?

 

[yo!]

What is the tester talking about? DAC overshoot? Come on, this overshooting is typical TPA behavior ... you can see it on Ti specsheets.