3E Audio 260-2-29A : TPA3255 / HA3588 CoilCraft / PFFB Fully differential.

[S=klogW]

@S=klogW i don't have experience on the RC filter as i used to test with AES filter,someone diy maybe did this just google it should be help.

@daniboun the special verison is 12dB Gain,ther is nothing to discuss here so i suggest to maintain our topic on current version: 260-2-29A。
regarding inductor we probebly can't only refer to DC current chart as there mine not-linear chart in high frequency switching condiction.in our test result it did show HA3588 THDN earlier rise higher than 7G23B when output >100W/4R
but all in all,how is the different between 105dB and 108dB impact on sound quanlity,IMO it is really make not sense for most of people.

105dB / 108dB - No at my age I can't hear a difference. And yet here we all are on ASR regularly tuning in to read of the latest 0.1dB improvement in DAC SNR and who's topping the charts this week (Topping?), even though limits of audibility were breached about 10dB ago. The war is won and now it's about excellence in audio science; solid engineering; value for money; electrical safety and who's got the most convincing, big-screen LCD emulation of 1970's VU meters! Just because you can't hear a difference doesn't mean it's not important. At the Topping Board Meeting, when it was announced that the LA90 was ready and it has a 7dB edge on the Benchmark AHB-2's sinad, did anyone there greet this news with the statement: "What's the point in sending it to Amir ? That 7dB improvement over our competitors is likely inaudible anyway" OR, did they all freak out and party hard, long into the night and wake up next afternoon with crippling hangovers? It's a first world fetish, but it's our fetish and ASR recommendations cause much less pain in trying to feed the kids, compared with reading Sterilephial's Recommended Components 2023 Edition and feeling you're a worthless creature until you buy yourself the latest:

VACuous Statement 45point2 (lobotomized) IQ: $75,000 stereo; $150,000/pair as monoblocks.

 

[daniboun]

105dB / 108dB - No at my age I can't hear a difference. And yet here we all are on ASR regularly tuning in to read of the latest 0.1dB improvement in DAC SNR and who's topping the charts this week (Topping?), even though limits of audibility were breached about 10dB ago. The war is won and now it's about excellence in audio science; solid engineering; value for money; electrical safety and who's got the most convincing, big-screen LCD emulation of 1970's VU meters! Just because you can't hear a difference doesn't mean it's not important. At the Topping Board Meeting, when it was announced that the LA90 was ready and it has a 7dB edge on the Benchmark AHB-2's sinad, did anyone there greet this news with the statement: "What's the point in sending it to Amir ? That 7dB improvement over our competitors is likely inaudible anyway" OR, did they all freak out and party hard, long into the night and wake up next afternoon with crippling hangovers? It's a first world fetish, but it's our fetish and ASR recommendations cause much less pain in trying to feed the kids, compared with reading Sterilephial's Recommended Components 2023 Edition and feeling you're a worthless creature until you buy yourself the latest:

VACuous Statement 45point2 (lobotomized) IQ: $75,000 stereo; $150,000/pair as monoblocks.

Everyone knows that ordinary mortals need to be reassured) especially here.... the numbers reassure, the numbers soothe. For scientifico- skeptics the measurements make no sense and for good reason we manage to sell cables at $ 500. Marketing exists because there are prospects) I am the first to quibble over some DB that I wouldn't hear but I must admit that I like to play this game))) But it is a passion that keeps me busy after all )

I saw Wilson Benesh.170KE speakers... who is the stupidest? he who buys or he who sells? ))))

 

[S=klogW]

You absolutely right. My tired old brain had forgotten we were dealing with both channels driven and that I should have multiplied my suggested ampage for the psu by two, so 12amps not 6amps in the BTL configuration.

After my own personal Ohm's Law drop-off yesterday (well maybe not Ohm's Law, more I forgot we had stereo, two channels to address, not one). I awoke this morning with the clear and firm conviction that there is a further Ohmic miscalculation that we all have been making. Take the TPA3255's stated power output spec from the TI datasheet:

7.6 Audio Characteristics (BTL)
Audio frequency = 1 kHz, PVDD_X = 51 V,
GVDD_X = 12 V, RL = 4 Ω, fS = 450 kHz, ROC = 22 kΩ, TC = 75°C, Output Filter: LDEM = 10 μH, CDEM = 1 μF, mode = 00,
AES17 + AUX-0025 measurement filters,unless otherwise noted.

>>> RL = 4 Ω, 1% THD+N 255W output per channel <<<

So 255 watts times by 2 for the two channels =510 watts divide by 4 for the four Ohm load and square root to get 11.29 amps required if we are going to run full power tests with a contant 1kHz sinewave, both channels driven. Hence, 3eAudio have chosen the Morsun LOF550-20B48 because for the short period of time it takes to make the full power test it can provide 48Volts and 11.46amps without the need to provide the 25cubic feet/minute forced air cooling. Morsun also specs "Over-current Protection ≥105%Io, hiccup, self-recover"; so it should be possible to sink up to 12.00amps before the supply starts to hiccup. We need 11.29 amps, we've got the Morsun giving 12.00 amps - Great! We've got it covered - Simple! WRONG!!!

Answers on a postcard, please to :

Georg Simon Ohm
Boffin's Shed
Le Jardin des Délices

 

[antcollinet]

And yet here we all are on ASR regularly tuning in to read of the latest 0.1dB improvement in DAC SNR and who's topping the charts this week

Are we?

Personally I use the Sinad chart as a broad brush filter. Anything in blue and green is audibly perfect - choose between them based on other criteria.

After my own personal Ohm's Law drop-off yesterday (well maybe not Ohm's Law, more I forgot we had stereo, two channels to address, not one). I awoke this morning with the clear and firm conviction that there is a further Ohmic miscalculation that we all have been making. Take the TPA3255's stated power output spec from the TI datasheet:

7.6 Audio Characteristics (BTL)
Audio frequency = 1 kHz, PVDD_X = 51 V,
GVDD_X = 12 V, RL = 4 Ω, fS = 450 kHz, ROC = 22 kΩ, TC = 75°C, Output Filter: LDEM = 10 μH, CDEM = 1 μF, mode = 00,
AES17 + AUX-0025 measurement filters,unless otherwise noted.

>>> RL = 4 Ω, 1% THD+N 255W output per channel <<<

So 255 watts times by 2 for the two channels =510 watts divide by 4 for the four Ohm load and square root to get 11.29 amps required if we are going to run full power tests with a contant 1kHz sinewave, both channels driven. Hence, 3eAudio have chosen the Morsun LOF550-20B48 because for the short period of time it takes to make the full power test it can provide 48Volts and 11.46amps without the need to provide the 25cubic feet/minute forced air cooling. Morsun also specs "Over-current Protection ≥105%Io, hiccup, self-recover"; so it should be possible to sink up to 12.00amps before the supply starts to hiccup. We need 11.29 amps, we've got the Morsun giving 12.00 amps - Great! We've got it covered - Simple! WRONG!!!

Answers on a postcard, please to :

Georg Simon Ohm
Boffin's Shed
Le Jardin des Délices
View attachment 310528


View attachment 310527

You effectively have the two speakers in parallel as far as the PSU is concerned - so a 2 ohm load.

Put it another way - that 12(say)A is split between the two channels if both are driven equally - 6A each - 36x4 = only 144W/channel.

(Same as 12A into 2 ohm = 288W)

This was indicated in my post above where I pointed out that a 6A PSU could only deliver 3A per channel, and thereby 36W only.

With real music, unlikely to be a big problem as both channels are unlikely to be driven to full power at the same time - and even then such peaks will be very transient unless you are driving the amp into massive distortion.

 

[S=klogW]

Are we?

Personally I use the Sinad chart as a broad brush filter. Anything in blue and green is audibly perfect - choose between them based on other criteria.

You effectively have the two speakers in parallel as far as the PSU is concerned - so a 2 ohm load.

Put it another way - that 12(say)A is split between the two channels if both are driven equally - 6A each - 36x4 = only 144W/channel.

(Same as 12A into 2 ohm = 288W)

This was indicated in my post above where I pointed out that a 6A PSU could only deliver 3A per channel, and thereby 36W only.

With real music, unlikely to be a big problem as both channels are unlikely to be driven to full power at the same time - and even then such peaks will be very transient unless you are driving the amp into massive distortion.

Obviously you got the right answer: it's a 2ohm load, but you are missing my point; this is not about the real music case, it's about test and measurement, a.k.a. mensuration:
"The act, process, or art of measuring"; choosing the correct psu so you generate correctly executed measurements for the spec sheet. 3EAudio chose a 11.5 amp power supply with a <5% current overload capability so as to generate their spec sheet for their full power 1%, 4ohms both channels driven, and we on ASR didn't immediately question this choice. It took me a day, but then I posted, because I do question this choice. To run this test properly, even if they were only getting 220 watts per channel they should use a sqrt(440/2) capable psu and that's a 14.85 ampere unit; at TI's stated 255 watts you need 16 amps.

This is probably why the graph stops its vertical trajectory just above 1% distortion; it's current starved.
Just saying that if we are going to be objective, we need to have standards (keep 'em at the BIPM, St. Cloud, Paris ).

IMO, 3eAudio fell into the 4ohm/2ohm trap which is why they chose a psu that just covered the erroneously calculated 11.29amps.

Thirsty little buggers these tinychip ampeaters!

 

[S=klogW]

Obviously you got the right answer: it's a 2ohm load, but you are missing my point; this is not about the real music case, it's about test and measurement, a.k.a. mensuration:
"The act, process, or art of measuring"; choosing the correct psu so you generate correctly executed measurements for the spec sheet. 3EAudio chose a 11.5 amp power supply with a <5% current overload capability so as to generate their spec sheet for their full power 1%, 4ohms both channels driven, and we on ASR didn't immediately question this choice. It took me a day, but then I posted, because I do question this choice. To run this test properly, even if they were only getting 220 watts per channel they should use a sqrt(440/2) capable psu and that's a 14.85 ampere unit; at TI's stated 255 watts you need 16 amps.

View attachment 310536
This is probably why the graph stops its vertical trajectory just above 1% distortion; it's current starved.
Just saying that if we are going to be objective, we need to have standards (keep 'em at the BIPM, St. Cloud, Paris ).

IMO, 3eAudio fell into the 4ohm/2ohm trap which is why they chose a psu that just covered the erroneously calculated 11.29amps.

Thirsty little buggers these tinychip ampeaters!

Looking more closely at the 3eAudio graph, the near-vertical trajectory reduces in slope twice. The first time at around 0.07% THD+N that's around 230 or 240 watts, so the current starvation, leading to measurement inaccuracy starts there.

 

[antcollinet]

Looking more closely at the 3eAudio graph, the near-vertical trajectory reduces in slope twice. The first time at around 0.07% THD+N that's around 230 or 240 watts, so the current starvation, leading to measurement inaccuracy starts there.

Those changes in slope are not a result of current limiting (I don't think). The slope gets shallower - less (distortion increase)/watt. If current or voltage limit was getting worse you'd expect the slope to get more vertical.

The real question is: Is @3eaudio measuring with both channels driven to full power at the same time. If so how are they getting simultaneous 200W+ into 4 ohm with only 12A from the PSU?

 

[S=klogW]

Those changes in slope are not a result of current limiting (I don't think). The slope gets shallower - less (distortion increase)/watt. If current or voltage limit was getting worse you'd expect the slope to get more vertical.

The real question is: Is @3eaudio measuring with both channels driven to full power at the same time. If so how are they getting simultaneous 200W+ into 4 ohm with only 12A from the PSU?

Yep, that's the real question. I guess I took the Norwegian coastal road to get to my point! : for full confidence: 16 amps.

 

[daniboun]

Hi amigos

On the way. )

I will start to work this week-end

Obviously you got the right answer: it's a 2ohm load, but you are missing my point; this is not about the real music case, it's about test and measurement, a.k.a. mensuration:
"The act, process, or art of measuring"; choosing the correct psu so you generate correctly executed measurements for the spec sheet. 3EAudio chose a 11.5 amp power supply with a <5% current overload capability so as to generate their spec sheet for their full power 1%, 4ohms both channels driven, and we on ASR didn't immediately question this choice. It took me a day, but then I posted, because I do question this choice. To run this test properly, even if they were only getting 220 watts per channel they should use a sqrt(440/2) capable psu and that's a 14.85 ampere unit; at TI's stated 255 watts you need 16 amps.


This is probably why the graph stops its vertical trajectory just above 1% distortion; it's current starved.
Just saying that if we are going to be objective, we need to have standards (keep 'em at the BIPM, St. Cloud, Paris ).

IMO, 3eAudio fell into the 4ohm/2ohm trap which is why they chose a psu that just covered the erroneously calculated 11.29amps.

Thirsty little buggers these tinychip ampeaters!

In the end, even if the 3E Audio graphics remain quite mysterious, your measurements show a good 105.8dB at 5watts into 4ohms with a 36V/9A PSU which is close to the measurements announced by 3E Audio.
I think we can all agree that the module seems to measure well =) And this is what we are trying to demonstrate in this thread for less sophisticated users. ))

 

[daniboun]

Apart from that... work in progress, I received everything and I'm getting to work this weekend)
I inspected the module, the finishes seem to be of good quality. No complaints !

As far as I'm concerned, Module Rev A; July 22th, Inductors Coilcraft HA3588-BL

 

[daniboun]

On the way....

Shots of the intended design (nothing is fixed yet) :

 

[Nabussan]

On the way....

Shots of the intended design (nothing is fixed yet) :

On the way....

Shots of the intended design (nothing is fixed yet) :

What components are you using?

 

[daniboun]

What components are you using?

Look on page 1, you have the complete description

 

[MCH]

Hey Daniboun, two questions.
What is that DC filter board? I cannot find it mentioned in the thread. Do you have a link for it?
Also, why the amp module so far from the connectors? (Not saying it is wrong, I have no idea, it is just a question)

 

[Nabussan]

Hey Daniboun, two questions.
What is that DC filter board? I cannot find it mentioned in the thread. Do you have a link for it?
Also, why the amp module so far from the connectors? (Not saying it is wrong, I have no idea, it is just a question)

If the 3rd component is a filter board, do you know why it might be required?

 

[Guddu]

@daniboun
You are using xlr, is 3e board balanced implementation by design?

 

[daniboun]

Hey Daniboun, two questions.
What is that DC filter board? I cannot find it mentioned in the thread. Do you have a link for it?
Also, why the amp module so far from the connectors? (Not saying it is wrong, I have no idea, it is just a question)

Hi,
It is a simply EMI Power Filter with a rectifier bridge in series to eliminate noise. It's purely optional. I had one available so I added it
For the second question, I tried several positions in the enclosure and it was the most consistent in my opinion

 

[Guddu]

@3eaudio
You have finished amplifier product based on previous modules, are you planning to release finished product with new modules as well?

 

[daniboun]

@daniboun
You are using xlr, is 3e board balanced implementation by design?

I don't know why most people are confusing the DS_EAUMT-0260-2-A_Rev1.0 module and the very latest 3E Audio module 3E Audio 260-2-29A (fully differential with Coilcraft Inductors / PFFB on). The latest 3E Audio module is completely full balanced....no SE inputs possible by default (unless you modify it by adding new OP amps.)
See page 1 as well )

 

[Guddu]

I don't know why most people are confusing the DS_EAUMT-0260-2-A_Rev1.0 module and the very latest 3E Audio module 3E Audio 260-2-29A (fully differential with Coilcraft Inductors / PFFB on). The latest 3E Audio module is completely full balanced....no SE inputs possible by default (unless you modify it by adding new OP amps.)
See page 1 as well )

Thanks