New Dual TPA3255 Module + PFFB / Fully differential / OP amps DIP8 @ Great price

[Roland68]

In the end, what's the actual difference between the cheap and more expensive comp? Audible sound quality? Objective sound quality? Longevity?

In terms of both measured performance and power output, the inexpensive TPA3255-based boards are all inferior to affordable €50 amplifiers such as the AIYIMA A07 and Fosi V3 Stereo. Especially if you intend to house the circuit board in an enclosure anyway, you have to ask yourself whether it makes sense to purchase anything cheaper than a Fosi V3 Stereo or an AIYIMA A07 MAX (the latter of which is even bridgeable for use as a mono amp)—both of which are available for just €50.

In contrast to the 3E Audio A7/A7se, the sonic difference—specifically regarding transparency—is clearly distinguishable in a level-matched blind test using Elac DBR62 speakers, even for inexperienced listeners. Of course, the difference isn't night and day, but it amounts to more than just subtle nuances.
A comparison with the 3E Audio A5/A5se, Topping Mini 300, or PA5/II reveals an even more pronounced difference, as these well-engineered TPA3251-based units offer a touch more transparency.
Superior build quality is, of course, another key factor. Although I am generally considered somewhat pedantic when it comes to the quality of components, circuitry, and power supplies, it would make no sense for me to build something myself as long as I can purchase products like 3E Audio's A-series amplifiers or their new multi-channel models.

The DIY board discussed in this thread likely falls somewhere in between these examples; however, without actual measured data, this remains merely a conjecture.

 

[Tell]

In terms of both measured performance and power output, the inexpensive TPA3255-based boards are all inferior to affordable €50 amplifiers such as the AIYIMA A07 and Fosi V3 Stereo. Especially if you intend to house the circuit board in an enclosure anyway, you have to ask yourself whether it makes sense to purchase anything cheaper than a Fosi V3 Stereo or an AIYIMA A07 MAX (the latter of which is even bridgeable for use as a mono amp)—both of which are available for just €50.

In contrast to the 3E Audio A7/A7se, the sonic difference—specifically regarding transparency—is clearly distinguishable in a level-matched blind test using Elac DBR62 speakers, even for inexperienced listeners. Of course, the difference isn't night and day, but it amounts to more than just subtle nuances.
A comparison with the 3E Audio A5/A5se, Topping Mini 300, or PA5/II reveals an even more pronounced difference, as these well-engineered TPA3251-based units offer a touch more transparency.
Superior build quality is, of course, another key factor. Although I am generally considered somewhat pedantic when it comes to the quality of components, circuitry, and power supplies, it would make no sense for me to build something myself as long as I can purchase products like 3E Audio's A-series amplifiers or their new multi-channel models.

The DIY board discussed in this thread likely falls somewhere in between these examples; however, without actual measured data, this remains merely a conjecture.

You've measured any of these inexpensive TPA3255 boards? I've looked before but never really found anything so I really have no idea how they objectively perform. The only one I've found is a few simple measurements of the ~5 euro TPA3118 boards I've got and I think the SINAD of them was at around 65dB (distortion limited), and tbh I can't really hear much of a difference between those and my Topping Mini300 unless I turn the volume up quite high. Sure the frequency response probably deviates a bit but it's nothing substantial, and I do measure and EQ my speakers/room anyways so that difference is neglectable.

 

[Roland68]

You've measured any of these inexpensive TPA3255 boards? I've looked before but never really found anything so I really have no idea how they objectively perform. The only one I've found is a few simple measurements of the ~5 euro TPA3118 boards I've got and I think the SINAD of them was at around 65dB (distortion limited), and tbh I can't really hear much of a difference between those and my Topping Mini300 unless I turn the volume up quite high. Sure the frequency response probably deviates a bit but it's nothing substantial, and I do measure and EQ my speakers/room anyways so that difference is neglectable.

I actually already mentioned that in Post #39—the very one you quoted. Perhaps you should go back and read it again.
The dirt-cheap TPA3118 boards are usually fakes.
If you don't hear any differences in your setup, that's fine—that can certainly happen.

 

[LSPhil]

In terms of both measured performance and power output, the inexpensive TPA3255-based boards are all inferior to affordable €50 amplifiers such as the AIYIMA A07 and Fosi V3 Stereo. Especially if you intend to house the circuit board in an enclosure anyway, you have to ask yourself whether it makes sense to purchase anything cheaper than a Fosi V3 Stereo or an AIYIMA A07 MAX (the latter of which is even bridgeable for use as a mono amp), both of which are available for just €50.

In contrast to the 3E Audio A7/A7se, the sonic difference, specifically regarding transparency, is clearly distinguishable in a level-matched blind test using Elac DBR62 speakers, even for inexperienced listeners. Of course, the difference isn't night and day, but it amounts to more than just subtle nuances.
A comparison with the 3E Audio A5/A5se, Topping Mini 300, or PA5/II reveals an even more pronounced difference, as these well-engineered TPA3251-based units offer a touch more transparency.
Superior build quality is, of course, another key factor. Although I am generally considered somewhat pedantic when it comes to the quality of components, circuitry, and power supplies, it would make no sense for me to build something myself as long as I can purchase products like 3E Audio's A-series amplifiers or their new multi-channel models.

The DIY board discussed in this thread likely falls somewhere in between these examples; however, without actual measured data, this remains merely a conjecture.

You are making several very confident claims here without presenting any actual measurements of the board discussed in this thread.

You criticize the inductors, PCB traces, and overall build quality, yet you clearly never saw a datasheet for the coils used on my board, nor do you know the copper thickness or current capability of the PCB. Simply looking at photos and assuming “cheap board = poor engineering” is not technical analysis.

Also, visually similar layouts do not automatically mean identical quality but the opposite is also true. A €25 board is not automatically “bad” simply because a branded product costs €150. Nobody economically manufactures fake high-current inductors for these boards; tooling and production costs alone would exceed the possible profit margin.

You also mention “clearly distinguishable transparency differences” in blind tests. That is a very strong statement for TPA325x amplifiers operating within their linear range. Without published level-matched measurements, distortion spectra, noise floor data, or ABX methodology, this remains subjective opinion, not established fact.

Regarding the original question whether a 500 W PSU is sufficient the answer is still yes. Real music is not a continuous sine-wave stress test. In none of the music material I measured did the actual power demand exceed roughly 170 W average consumption. Peak capability and continuous RMS test conditions are two very different things.

I responded with concrete numbers regarding load impedance and power delivery. You responded mostly with assumptions and generalizations about “cheap DIY boards.” Those are not the same thing.

If you have actual measurements of this exact board, post them. Otherwise, please present your statements as personal impressions instead of objective facts.

 

[Ruediger]

Regarding the original question whether a 500 W PSU is sufficient the answer is still yes. Real music is not a continuous sine-wave stress test. In none of the music material I measured did the actual power demand exceed roughly 170 W average consumption. Peak capability and continuous RMS test conditions are two very different things.

But it has to withstand testing the published specs, doesn't it?

Can't just pre-determine what kind of signals people have to send through it to be OK.

 

[LSPhil]

But it has to withstand testing the published specs, doesn't it?

Can't just pre-determine what kind of signals people have to send through it to be OK.

Absolutely, and I never argued against verifying published specifications under controlled test conditions.

The TI datasheet clearly defines operating conditions, output power, thermal behavior, and measurement methodology. I have no reason to doubt those numbers. But we also have to distinguish between laboratory stress testing and real-world music playback.

Continuous sine-wave testing is essentially a worst-case thermal and power-supply stress scenario. Yes, under those conditions the TPA3255 absolutely requires proper cooling, especially when driven hard for extended periods. I have personally managed to bring a 320 W PSU to its knees using continuous sine signals.

With actual music material, however, the situation is completely different. Music has crest factor, dynamics, pauses, and varying spectral content. Even the most demanding track I mentioned never produced anything close to continuous sine-wave load conditions in my testing.

Some people are intentionally pushing amplifiers far beyond realistic usage by modulating signals down to 10 Hz or using heavily compressed bass torture tracks. That is perfectly valid if the goal is to explore the absolute limits of an amplifier design, but it does not represent normal listening conditions anymore.

So yes: published specs should survive proper testing. But we should also be honest about the difference between synthetic stress tests and real music playback.

 

[Tell]

I actually already mentioned that in Post #39—the very one you quoted. Perhaps you should go back and read it again.
The dirt-cheap TPA3118 boards are usually fakes.
If you don't hear any differences in your setup, that's fine—that can certainly happen.

You mentioned what? I've read your post a few times and I see no answer to my question about any measurements or anything else.
In what way is the TPA3118 boards fake? They make sounds just as they are intended so I don't see a problem? And yeah tbh I don't think you'd hear much actual difference either, if the frequency response due to load dependancy is corrected of course.

 

[Ruediger]

Continuous sine-wave testing is essentially a worst-case thermal and power-supply stress scenario. Yes, under those conditions the TPA3255 absolutely requires proper cooling, especially when driven hard for extended periods. I have personally managed to bring a 320 W PSU to its knees using continuous sine signals.

With actual music material, however, the situation is completely different. Music has crest factor, dynamics, pauses, and varying spectral content. Even the most demanding track I mentioned never produced anything close to continuous sine-wave load conditions in my testing.

Some people are intentionally pushing amplifiers far beyond realistic usage by modulating signals down to 10 Hz or using heavily compressed bass torture tracks. That is perfectly valid if the goal is to explore the absolute limits of an amplifier design, but it does not represent normal listening conditions anymore.

One of my first thoughts when seeing these DIY options is always subwoofers. Be it as a poweramp or a DIY plateamp. I guess that exceeds the normal music application, although the variations, pauses etc. also apply.

The model discussed here might not be very appropriate for such a speaker (at least as a plateamp). But I think people do use TPA amps for passive subwoofers and at times also challenge power and cooling a little bit.

 

[frabor]

@LSPhil "sine-wave testing is essentially a worst-case thermal and power-supply stress scenario"

I was under the impression that square pulses were worse case, not challenging you, just asking .

I have a diy subamp with two 150W power supplies connected two 500W Tamp. The unit has a 1 Amp resetable fuse 120 V ACv( needs to be upgraded). I have not been able to ever trip it with music but once I play tones at about 100 dB at listening positions, it always blows. Point is the insane difference in requirements (DC power and cooling) when playing tones vs playing music.

 

[LSPhil]

@LSPhil "sine-wave testing is essentially a worst-case thermal and power-supply stress scenario"

I was under the impression that square pulses were worse case, not challenging you, just asking .

I have a diy subamp with two 150W power supplies connected two 500W Tamp. The unit has a 1 Amp resetable fuse 120 V ACv( needs to be upgraded). I have not been able to ever trip it with music but once I play tones at about 100 dB at listening positions, it always blows. Point is the insane difference in requirements (DC power and cooling) when playing tones vs playing music.

A 1A fuse at 120V AC gives you 120W max input power. After losses in the power supply (say 80–85% efficiency) and amplifier (maybe 60–70% for Class T/Tripath), you’re realistically getting ~50–60W total output power for both channels combined. That’s only ~25–30W per channel into speakers.

With a typical 83 dB/W/m speaker, 30W (17dB) gives you about 83+17=100 dB at 1 meter, exactly where you’re blowing the fuse with test tones.

Your real bottleneck isn’t the amp it’s the AC input fuse and likely the power supply capacity.

Disclaimer:
The information provided in my posts is intended for general informational and educational purposes only. While I make every effort to provide accurate technical information, I make no warranties or guarantees regarding completeness, accuracy, or suitability for any particular purpose.
Any actions taken based on information from my posts are strictly at your own risk. I assume no liability for damages, injuries, losses, or other consequences resulting from the use or misuse of this information.
I am not a licensed electrician, certified engineer, or legal professional. Always consult a qualified professional and follow applicable safety standards, laws, and manufacturer guidelines before modifying electrical equipment, power supplies, amplifiers, or fuses.

 

[frabor]

A 1A fuse at 120V AC gives you 120W max input power. After losses in the power supply (say 80–85% efficiency) and amplifier (maybe 60–70% for Class T/Tripath), you’re realistically getting ~50–60W total output power for both channels combined. That’s only ~25–30W per channel into speakers.

With a typical 83 dB/W/m speaker, 30W (17dB) gives you about 83+17=100 dB at 1 meter, exactly where you’re blowing the fuse with test tones.

Your real bottleneck isn’t the amp it’s the AC input fuse and likely the power supply capacity.

Disclaimer:
The information provided in my posts is intended for general informational and educational purposes only. While I make every effort to provide accurate technical information, I make no warranties or guarantees regarding completeness, accuracy, or suitability for any particular purpose.
Any actions taken based on information from my posts are strictly at your own risk. I assume no liability for damages, injuries, losses, or other consequences resulting from the use or misuse of this information.
I am not a licensed electrician, certified engineer, or legal professional. Always consult a qualified professional and follow applicable safety standards, laws, and manufacturer guidelines before modifying electrical equipment, power supplies, amplifiers, or fuses.

TKS for the analysis.Agree, the fuse was placed temporarily when I had only one channel and no buffer. The Tamp require 1.8 V for Max output but it was feed by a minidsp with max output .9V, so it was not a problem before. Now, with a buffer and two amps, the fuse needs to be increased to 3 A. Also , I need to fuse the toroidal that runs the preamp/ buffer before I changed the main fuse.

 

[Aimarketingbot]

90% Efficient Class-D Operation (4 Ω), according to the datasheet, so at 150 Wpc, about 30W of heat. Laptops with a similar power load require a cooling fan. Seems possible but borderline.

Dual Tpa3255 in pbtl, with proper cooling can do close 250 Wpc with a 48V/ 10 A

Only if you listen to test tones.
I have the BRZ PAD-30, also dual 3255, you need to really blast it (200W 8ohm towers) to get it to go over room temp. Would be easy to add additional heatsink or aluminium base underneath.

 

[frabor]

I do listen to test tones from time to time, lol. I find relaxing to blast the units and measure how much A they are drawing.

The BRZ PAD-30 seems like a really nice unit. Considering the case makes more sense than the dual TPA in this discussion.

What power supply do you use and how do you like the tks

 

[Aimarketingbot]

I do listen to test tones from time to time, lol. I find relaxing to blast the units and measure how much A they are drawing.

The BRZ PAD-30 seems like a really nice unit. Considering the case makes more sense than the dual TPA in this discussion.

What loser supply do you use and how do you like the tks

I understand test tones and measuring enjoyment! For me it's more about tuning the sub/speaker positioning, crossovers, room via physical optimisation. Then EQ it.

And yes, same conclusion and I put my money where my mouth is. By the time you put it in a case((s) especially for mono 3255s), terminals, wiring, volume knob etc, you can get that decent dual 3255 for less, also takes up less room lol.

I do LEDs so have lots of 48-54V PSUs laying around. One I have spare is LRS-350-48, it is 7.3A and adjustable voltage. It drives the towers to X-mech on the bass cones (4x89db/W 6" total), more than enough for musical peaks. Test tones are different, then you want more power, and you will run into heating with any compact 3255 especially dual 3255 with extra current on tap. But the base is very flat, you could easily thread tap or even C clamp (lol) a heatsink on with some thermal paste and it'd make a big difference.

Absolutely impressed by the amp. Beautifully quiet, excellent noise floor, detail, everything you'd expect from an amp 10-20x + the price, except volume control feel (cheap) and minor QC blemish on case/lower quality screen printing etc. I didn't expect this (took an educated gamble) but it is superior to thousands of dollars of British BI-amp I am replacing.

 

[frabor]

Talking about cooling , the dutch and Dutch cooling for the pascal amp.

 

[Ruediger]

Looks like a Noctua with the rubber feet kit.