Transforming TPA3255 into ABD modulation.

[Hayk]

It works.
I thought the piano will be the best to test, so I listened this

The piano is on purpose intended to be in the background, so the adjustment,(I replaced the 330k resistor with 470k adjust) should bring forward or take backwards. Nothing like that , very disappointing.
I wanted to hear how the texture of the sound changes, so I switched to this.

Very clear variation of mildness to expressive. The double C is the most clear soft to firstly plucked. All this at 1w average I estimate, where the simulator calculates 0.01% THD.
Although I ended with flu, still, by several days of lack of sleep, I found the mild position so relieving that for sure I love this circuit.
I will proceed with the high impedance version and started the PCB to be installed replacing the opamp. It will be for both types to be populated accordingly. By this for $4 one can get 5 PCBs, a pair populated for high impedance and another pair for variable texture.
With high impedance, the simulator gives a THD of 0.0006% for 1w@8ohms and 0.006% for 100w@8ohms. The major advantage is the reduction of the 3255 noise by 20db, which has already been reduced by ABD mod.

 

[KL....]

Thank you, @Hayk, are there no opamps that do this type of Difference amp, as effectively? Although/perhaps not with High input impedance? Your suggestion is that R7 controls/trims Phase, is that correct?

• Are you connecting input via RCA or XLR or both?
• Are you connecting to the 3255 via Single or Balanced output/input?

 

[Hayk]

Thank you for reminding me about balanced input. I will modify the preliminary PCB so that when wired for high impedance, potentiometer connector becomes the negative input by the grounding capacitor. The width of the PCB is 1.5 times the 8 DIP, an adapter that comes with the PCB holds it vertically while inserted in the socket.

 

[Hayk]

This is a trial PCB with 8 dip adaptor to separate.
I will start wiring a prototype today but I am not so hopeful to hear anything better than tpa3110 can do. It will be only good for measurments particularly with differential input as the interface has 0.00004% THD at 0.4vp and 0.0003% at 2vp.
The 0 ohm resistors are to short according to the selected circuit.

 

[Hayk]

I listened for about 2 hours with the potentiometer at full expressive position. The speaker I use is the lowest efficient I have, 89db 1w/m. Very impressive bass, punchy, high presence that sounded my modest speaker, full spectrum.
I switched to my very high efficiency one, 106db, 9x12in 100db with giant horn 230Hz cut, nothing significant effect. To know that the higher is the input signal, the more it generates distortion, this why low frequencies which are higher in amplitude, the more they get treatment. In other words, the lower is the efficiency, the wider is the effect mild-expressive.

 

[KL....]

I listened for about 2 hours with the potentiometer at full expressive position. The speaker I use is the lowest efficient I have, 89db 1w/m. Very impressive bass, punchy, high presence that sounded my modest speaker, full spectrum.
I switched to my very high efficiency one, 106db, 9x12in 100db with giant horn 230Hz cut, nothing significant effect. To know that the higher is the input signal, the more it generates distortion, this why low frequencies which are higher in amplitude, the more they get treatment. In other words, the lower is the efficiency, the wider is the effect mild-expressive.

Thank you, @Hayk, some intrigueing Speakers that you have :=) The Horn looks a lot like an Avantgarde Horn although the surface is very Textured :=)

 

[KL....]

This is a trial PCB with 8 dip adaptor to separate.
I will start wiring a prototype today but I am not so hopeful to hear anything better than tpa3110 can do. It will be only good for measurements particularly with differential input as the interface has 0.00004% THD at 0.4vp and 0.0003% at 2vp.
The 0 ohm resistors are to short according to the selected circuit.

Thank you, @Hayk, very nice :=)
Re 0 ohm resistors.... that is one way to insulate a wire jumper, with a tiny/negligable amount of Resistance, isn't it?

To ask, do you happen to know of NCD/NewClassD/Lars Clausen?

 

[Hayk]

Do you have a circuit descrption of NCD?

 

[KL....]

Do you have a circuit descrption of NCD?

Thank you, @Hayk, this is an early/2006 example of the NCD Control loop, several/many changes since then, usually comes with a Buffer as you are suggesting/indicateing....

Singularity 3 High Speed

 

[Hayk]

Thanks for info.
I finished the high impedance interface and passed on bench test. Installed to be compared with the low impedance one.

 

[Hayk]

The price of 2x300w boards are in falling trend as the 2x600w is now less than $20 with free shipment.
I just found this on AliExpress:
US $63.27 | DC18-50V TPA3255 Audiophile HIFI Digital Audio Amplifier Board Power Supply 600W+600W 10A HiFi Amp Dual Channel Stereo Module

https://a.aliexpress.com/_c3kFzjE5

 

[KL....]

The price of 2x300w boards are in falling trend as the 2x600w is now less than $20 with free shipment.
I just found this on AliExpress:
US $63.27 | DC18-50V TPA3255 Audiophile HIFI Digital Audio Amplifier Board Power Supply 600W+600W 10A HiFi Amp Dual Channel Stereo Module

https://a.aliexpress.com/_c3kFzjE5

Thank you, @Hayk, $35.74, complete with w/2 channel (600W/600W, cannot bridge channels, no need?), w/solid caps (low ESR, where appropriate), w/encapsulated (low EMI, audio/linear?) inductors, w/ PFFB?, w/@Hayk Modulation/Phase Control? :=)

• Did you notice a 48VDC, >10A, 12VDC (for fan, if needed) SMPS that would be effective :=)
• Do you think that this can be made more Phase Accurate?

Their Wordage....

• Maximum 1200W high-power output, supporting 2-8 ohm speakers, suitable for various audio needs.
• It is recommended to use a DC power supply with a voltage of 36-48V and a current of 10A or above.
• High end HIFI sound quality, signal-to-noise ratio of 90dB, distortion less than 0.05%, and pure sound.
• Air cooled cooling design, silent and long-lasting fan, ensuring continuous full power operation.
• High power digital amplifier dedicated inductor, low magnetic loss, high current, excellent sound quality.
• Thick brushed heat sink, made of abundant materials, with exquisite appearance and excellent heat dissipation performance.

 

[Hayk]

This board is probably 2x tpa3255 in PBTL. It is not useful for ABD modulation, but it's falling price will oblige the 2x300w ones to decrease further theirs.
If you want to get something cheaper on Ali, put the item in the cart and wait.

 

[KL....]

This board is probably 2x tpa3255 in PBTL. It is not useful for ABD modulation, but it's falling price will oblige the 2x300w ones to decrease further theirs.
If you want to get something cheaper on Ali, put the item in the cart and wait.

Thank you, @Hayk, perhaps it is BTL (not indicated))? ABD modulation works with BTL & SE, doesn't it, or is it only SE?

 

[StormBreaker]

That's some serious modding! Changing AD to ABD on a TPA3255 is no joke. Cool idea, I'm curious how the filter turns out. Good job!

 

[Hayk]

If advertised 2x600w it can only be 2x PBTL. For ABD modulation, the 3255 can only work with 4xSE, to have channel A bridged with C and Channel B bridged with D.
I repeat again, this chip has a flaw in its design, in BTL and PBTL, the sound is not as good as in 4xSE bridged. In SE mode, it doesn't have DC speakers protection.

 

[KL....]

If advertised 2x600w it can only be 2x PBTL. For ABD modulation, the 3255 can only work with 4xSE, to have channel A bridged with C and Channel B bridged with D.
I repeat again, this chip has a flaw in its design, in BTL and PBTL, the sound is not as good as in 4xSE bridged. In SE mode, it doesn't have DC speakers protection.

Thank you, @Hayk, Ah, a couple of Ah/Ah there, A/C BTL & B/D BTL :=(
4xSE bridged is A/Master with BCD BTL/Slaves, like/as with the TDA7293, exempt of crossover distortion, is that correct?

Do you think that the 2x300w TPA3255 is Channel A bridged with C & Channel B bridged with D, or be Configured in this Configuration, and still have DC Speaker protection?

 

[Hayk]

With BTL, M1=0,M2=0, A is bridged with B, C with D. With PBTL,M1=0,M2=1, A//C bridged with B//D. There is no master slave story here. With 4xSE, you get 4 independent inverting amplifiers but the carrier phases are the same for A paired with C and B with D.

 

[Hayk]

This is the schematic of the filter, and the PCB layout. I'll posted the gerber file when all finished.

 

[KL....]

With BTL, M1=0,M2=0, A is bridged with B, C with D. With PBTL,M1=0,M2=1, A//C bridged with B//D. There is no master slave story here. With 4xSE, you get 4 independent inverting amplifiers but the carrier phases are the same for A paired with C and B with D.

Thank you, @Hayk, for clarifying. With your diagram, I will need to see how it is connected (in place), just not seeing the complete picture, as yet, so will need to keep reading (at least, for the moment). Should I, again, read from post 1? :=)

@Hayk, to read, are you able to provide a link to this PDF? Noteing that TI refer to a Master/Slave Relationship although do not indicate which Channels are Bridged and 1x TPA3255/2x BTL is 2 Channel....