TPA3255 - all about DIY, Discussion, Design etc

Good evening amigos,

Here is the next step concerning the Drmodor TPA3255 module (then will come the Gan and the Purifi with Custom buffer). I decided to push the test of this amp with a well-known store in Paris and Lyon. I asked 3 experts from the store to test DrMordor's module for a week on a multitude of very high-end speakers. Since the Drmodor's module has not yet been measured, the exercise is to collect the opinion of professionals in the field to get an idea of ​​the quality of the module vs other amps in the shop.

It is a very famous shop which has very demanding customers and who have had a reputation for 45 years!

The founder will give me his expert opinion by next weekend.

I have already had the pleasure of listening to DrMordor's TPA3255 at the store plugged into wide band speakers with a deported "Pavilion", great ultra high ends speakers by the way ! It was impressive, and the customers were a little stunned.

Frankly I am confident and know these are very critical experts...

Some pictures + quick Video :

VID_20201205_162016

https://ibb.co/LvqBwkn
https://ibb.co/vw1zgFk
https://ibb.co/hmPw4c6
https://ibb.co/dpDzvdq
https://ibb.co/dWL5MX1
https://ibb.co/vzL9179

There are updates?:)
 
I think revolution is gonna be Gan Amplifiers ) It's promising but not mature yet
The way I see it, the TPA3255 is basically the holy grail for 99% of people out there. We already see that noise and THD performance can be under the point of audibility in most situations, and it also happens to be extremely cost-effective relative to its performance and with Class D power efficiency to boot. The only thing that's an unknown, to me, is the IMD performance with the latest designs, but that will be answered as soon as JLE/JLP sends a module to ASR.
 
The way I see it, the TPA3255 is basically the holy grail for 99% of people out there. We already see that noise and THD performance can be under the point of audibility in most situations, and it also happens to be extremely cost-effective relative to its performance and with Class D power efficiency to boot. The only thing that's an unknown, to me, is the IMD performance with the latest designs, but that will be answered as soon as JLE/JLP sends a module to ASR.

Amigos,

TPA3255 is the "today" amplifier. As I said, Gan will be the "tomorrow" amplifier. We must understand by revolution, a change or an evolution to what we are already producing )
 
Amigos,

TPA3255 is the "today" amplifier. As I said, Gan will be the "tomorrow" amplifier. We must understand by revolution, a change or an evolution to what we are already producing )
If GaN is going to be the next big thing, there has to be a compelling reason(s) for it to be considered a revolution... not just inching further on the scale of diminishing returns. Efficiency? Cost? Distortion? Size? Noise? Anything that we can hear, or are we just chasing smaller and smaller numbers at this point?

Hard to quantify what "revolution" means but I suggest that an improvement of minimum 3-5x on an important metric would be adequate.
 
Last edited:

Attachments

  • 32ED226A-F919-44AA-9017-E6769BDD0A94.jpeg
    32ED226A-F919-44AA-9017-E6769BDD0A94.jpeg
    292.8 KB · Views: 439
If GaN is going to be the next big thing, there has to be a compelling reason(s) for it to be considered a revolution... not just inching further on the scale of diminishing returns. Efficiency? Cost? Distortion? Size? Noise? Anything that we can hear, or are we just chasing smaller and smaller numbers at this point?

Hard to quantify what "revolution" means but I suggest that an improvement of minimum 3-5x on an important metric would be adequate.

For those who would not have understood :D:D:D

Theory of Audiophile Revolution

 
Last edited:
6.67Khz

hahaha...funny pic




Since the beginning of this design , I have wondered about the peak of 6.67Khz
True to our roots , we made hundreds of tests.
So those are my findings :6.67Khz is the last frequency that has both 2 and 3H (public knowledge). Its the highest frequency where you can see still h2/3 in 20Khz bandwidth.

However , distortion rise is seen staring at around 3Khz. After 6.67Khz is does not vanish and in fact it goes up to 20Khz.. this rising harmonics at 3 , 4 and 5Khz its right in the middle of most sensitive area of human hearing.
You can use feedback to attenuate partially . However one of my principles for the design is to correct each non linearity locally .. it is my firm belief that's the holy grail to SQ.

We found the solution .Using the right components , connected to the right (no noise) plane with the correct size loop . We see very small variation between THD at 1 and 6.67Khz .

One small step in a long journey.
 
Last edited:
Wouaaaaaaa ) Well done for this fine analysis!

Cdsgames : are you going to claim a nationality for the production of the modules? I guess you designed it but mass production will be Asian?

Audiophonics made me laugh by tagging their Hypex amp "made in France" ... rather to say" assembled "in France. It is as if they have appropriated the Hypex patent !

It may sound silly to you, but transparency says a lot about intentions)