3e Audio TPA3255 Amplifier Kit 480-1-29A Review

[daniboun]

Yes, but we will send our upcoming FB360 (and FB100) instead of D400M (we'll request to Amir around January). FB360 and FB100 have deeper PFFB (higher feedback factor) than the rest of the competition so the IMD and 32 tone tests will be better.

Just bought MORNSUN LOF550-20B48-C because it has an easy power on switch implementation. And it doesn't affect the FFT measurements which means it provides clean power.

Nice performance BTW for this 480-1-29A!, the performance is neck to neck with Topping PA7.
I mentioned before on the 3e thread that the mono version has a higher output power than the stereo version, which was proven correct with this mono and stereo review.

Also, I noticed that the stereo version is shutting down on 20Hz clipping which indicates that it has a marginal stability. Would be interested if one of its channels will be tested on a reactive load test, since 2 channels cannot be tested simultaneously. Dual mono is the way to go for getting the highest power and performance for the TPA3255 chip.

That is a good new ) looking forward to see that improved FB360 )
Even if the PA7 has comparable performance, we must not forget that it uses two TPA3255 chips in //

 

[Multicore]

One of TI's marketing claims is high efficiency. https://www.ti.com/product/TPA3251 "Class D efficiency" has been controversial here recently in one thread (although I was not quite sure if it was started as a joke).

@amirm, you occasionally provided efficiency measurements, which I especially value. Is there any chance you could add them here?

 

[Multicore]

Do the higher red, green and pale blue curves here

Correspond to the taller grass in the east (right) of this garden?

 

[HarmonicTHD]

Do the higher red, green and pale blue curves here


Correspond to the taller grass in the east (right) of this garden?

Their 5W values are somewhat related. Ballpark.

 

[AudioSceptic]

This is a review and detailed measurements of the 3e Audio 480-1-29A, TPA3255 based class D amplifier module. It was sent to me by the company in a reference design for ease of testing:
View attachment 332950
The module is the PCB to the right and costs US 89.99 from company Aliexpress store. This is the mono version (I reviewed the stereo version earlier). The differentiation is pulling the required class D filter into the amplifier feedback loop (PFFB) and with it, essentially eliminate the effect of speaker load on frequency response. Low cost implementations lack this which can make the tonality somewhat variable. Other optimizations are done to extract better performance out of the TI TPA3255 class D amplifier module.

When swapping the stereo module for mono above, I was impressed by the hefty machined aluminum heat coupler. It is quite substantial and spreads the heat nicely to large area of the case bottom.

The included power supply in this sample is MORNSUN LOF550-20B48-C. Specification is 48volts at 6.5 amps with air cooling and 11.5 amps with forced air. Latter allows more than 550 watts to be provided to the amplifier.

3e Audio 480-1-29A Amplifier Measurements
The initial set of measurements match the stereo version so I will quickly go over them:
View attachment 332953
View attachment 332954

View attachment 332955
View attachment 332956

View attachment 332957
View attachment 332958

View attachment 332959
View attachment 332960

We get more power:
View attachment 332961
View attachment 332962
We had 206 watts in stereo. Same improvement with 8 ohm load:
View attachment 332963
Stereo version produced 92 watts.

Interesting to see the transfer function change some in this version:
View attachment 332964
Cleaner than mono at first but then worse at higher frequencies. Here is the stereo version again:

I captured the switching frequency and was impressed by how low its level was, and how high the frequency was:
View attachment 332965

Amplifier Reactive Load Test
This is one robust amplifier, essentially not caring about the load down to impressive 2 ohms!
View attachment 332966

View attachment 332967

You get nearly 600 watts here! Note that the test only keeps the amp at these power levels for short period of time so this is not "continuous" amount of power in the pure sense of that term.

Conclusions
I don't know what to be more impressed by: the TI TPA3255 or 3e Audio's engineering skills. This is an incredibly well performing at this price level. Of course you have to account for adding a capable power supply and case to get these numbers but still, one can't help but be impressed.

I am happy to recommend the 3e Audio 480-1-29A, class D Amplifier.
------------
As always, questions, comments, recommendations, etc. are welcome.

Any donations are much appreciated using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/

This is so good! Does TI actually make the TPA3255 or just a reference design that can be made by anyone?

 

[delta76]

This is so good! Does TI actually make the TPA3255 or just a reference design that can be made by anyone?

They make the chip, and release a reference design of the board that uses the chip. Manufacturers then can make that board, a modification of it or design a new one (rarely) That is very common in semiconductor industry

 

[AdamG]

I dont ever remember seeing this many consecutive Golfing/Soccer panthers before.

Incredible!

Yeah that Golfing Panther is getting over exposed and probably close to demanding higher pay. But for us the readers this is a wonderful run of great product reviews just in time for Christmas!

 

[MediumRare]

Can someone explain to me why this graph doesn’t say that, for high frequencies where distortion is audible (cymbals, for example) that -62/-72 dB at normal listening levels is not especially good? The Topping PA5 II at 20W is 10 dB better.

 

[staticV3]

Can someone explain to me why this graph doesn’t say that, for high frequencies where distortion is audible (cymbals, for example) that -62/-72 dB at normal listening levels is not especially good? The Topping PA5 II at 20W is 10 dB better.
View attachment 333136

See post #36

 

[MediumRare]

See post #36

I saw that post, still doesn’t explain why 10 dB worse is a golfing panther.

 

[staticV3]

I saw that post, still doesn’t explain why 10 dB worse is a golfing panther.

Because if you were to weight this measurement to reflect how people will actually use the Amp (speech or music instead of test tones), then this discrepancy would largely disappear.

Just like THD+N vs Frequency with 90kHz BW, the measurements are interesting from an academic standpoint, but can quickly become misleading if one does not realize their synthetic nature.

Pushing distortion up in frequency and power where regular content rarely goes, is perfectly acceptable in my book. Just like noise shaping.

 

[AudioSceptic]

On the way... but I can't tell more for the moment )

Audiophonics?

 

[Billy Budapest]

Kind of interesting the top of one of the caps has “ASR” written on it. I hope this doesn’t mean it is a “golden sample.”

 

[Billy Budapest]

Pushing distortion up in frequency and power where no regular content ever goes, is perfectly acceptable in my book. Just like noise shaping.

Agreed. Kind of like people complaining about the switching frequency of Class-D amps or FM carrier leakage.

 

[MediumRare]

Because if you were to weight this measurement to reflect how people will actually use the Amp (speech or music instead of test tones), then this discrepancy would largely disappear.

Just like THD+N vs Frequency with 90kHz BW, the measurements are interesting from an academic standpoint, but can quickly become misleading if one does not realize their synthetic nature.

Pushing distortion up in frequency and power where no regular content ever goes, is perfectly acceptable in my book. Just like noise shaping.

I suppose the way to tell if it’s relevant to an audio chain would be to compare distortion v frequency of a very low distortion speaker at various SPLs. One could use white noise, for example. Would there be a measureable difference between amps? Or re-run the 32 tone signal at 20w and 50w.

 

[IAtaman]

I suppose the way to tell if it’s relevant to an audio chain would be to compare distortion v frequency of a very low distortion speaker at various SPLs. One could use white noise, for example. Would there be a measureable difference between amps? Or re-run the 32 tone signal at 20w and 50w.

Best Revels tested here has ca 45dB harmonic distortion at 5KHz at 96dB SPL. Even if you double the power at 5KHz to 16W (which happens to be what I calculated from the music spectra graph staticV3 shared) 80dB distortion of the amp would still be lost in the the distortion of the speaker most likely. If you need cheap amplification, this is excellent. Having said that, personally I would not use this as my main amp. I think two of these amps in bridged mode would be great for a powered sub-woofer solution.

 

[OldTimer]

It's just the same balanced input only, no single ended.

 

[MediumRare]

Best Revels tested here has ca 45dB harmonic distortion at 5KHz at 96dB SPL. Even if you double the power at 5KHz to 16W (which happens to be what I calculated from the music spectra graph staticV3 shared) 80dB distortion of the amp would still be lost in the the distortion of the speaker most likely. If you need cheap amplification, this is excellent. Having said that, personally I would not use this as my main amp. I think two of these amps in bridged mode would be great for a powered sub-woofer solution.

I agree with your conclusion, tho at 10k and 96 spl the relevant figures are closer: 42 dB distortion for Revel beryllium tweeters and -73 dB for the amp. At 15k the amp is 10 dB worse. Not sure if SPL and distortion dB can be compared directly like that. Also not sure how to evaluate audibility. So if we are left looking at things from a pure engineering standpoint, -63 dB distortion at 15kHz is not "great".

 

[IAtaman]

I agree with your conclusion, tho at 10k and 96 spl the relevant figures are closer: 42 dB distortion for Revel beryllium tweeters and -73 dB for the amp. At 15k the amp is 10 dB worse. Not sure if SPL and distortion dB can be compared directly like that. Also not sure how to evaluate audibility. So if we are left looking at things from a pure engineering standpoint, -63 dB distortion at 15kHz is not "great".

30dB between 42dB and 73dB is still 0.1% of the power, and 73dB is still probably better than most (if any) speakers. 15KHz's second harmonic is at 30KHz so although it matters from an engineering perspective, it does not matter perceptually.

 

[MediumRare]

30dB between 42dB and 73dB is still 0.1% of the power, and 73dB is still probably better than most (if any) speakers. 15KHz's second harmonic is at 30KHz so although it matters from an engineering perspective, it does not matter perceptually.

IMD goes down as well as up. We can see in the 32 tone even at 5 watts the rising floor. I’m not going to continue this. The fact remains it’s not "great".