Good & interesting Thread.
PS. ( OT )
You can see on the PCB > EPICURUS >
EPICURUS and ZENO were two Greek philosophers (300 BC) that had opposing views of things.
as someone who has done many LM3886 amplifier projects, i must say that the sound of the TDA7293 we used in this design is very different from the LM3886. if you look at the FFT graphs, you will see that the secondary harmonics are higher than the tertiary harmonics. i can say that the sound of the amplifier is not the classic bjt sound.
I can't tell you good or bad, I can only tell you my own opinions. Modulus is a really nice amp, probably on paper. I made the Modulus and tried to listen to it. It was too dry for me and it was very tiring when I listened to it. After 3 songs, I would get a headache and feel sleepy. So I decided to make a more musical and warm chip amp and designed this. Putting aside the paper and measurements, the Epic amp is very enjoyable to listen to…
Very strongly agree with this. It is also very robust or can be built into a very robust amp. That on top of the crystal clear sweet detail means that I have been using the same amp very, very hard since 2010. I am not able to converse technically on this as electronics knowledge is very limited. Chimed in to say how robust and clean these are and inquire about the PSU implementation in the amp that I have. It uses one chip per channel for stereo, and the protected documents from the integrated systems supplier to product developers prohibit sharing. The amp is built by a company that supplies the amp stages to higher end euro and Asian vehicle manufacturers as well as this line of home systems
The documentation says that it uses a tracking power supply and the sound is much cleaner at higher volumes than another 'normal' 7293 amp that I have. Can this type of power supply make so much difference?
power supply selection and design depends on the place of use and load design. I prefer ultra low noise regulators in discrete circuits, standard regulators in circuits consisting of opamps, and linear psu's consisting of cap banks in amplifier feeds. There are reasons for all these preferences. Speed, load regulation, noise, etc.
Perhaps it refers to an arrangement like the one in the datasheet, page 13? I'm not sure whether the darlingtons trace or switch the higher rails, though. And I'm also not sure if this arrangement may introduce additional distortions in the transition between the low voltage and high voltage rails (but I'd be very interested to learn more about it).
Best regards!
I can't get to where the docs are stored, atm. I do recall a section about the class LH tracking PSU based design. There is a fair bit of talk about the tracking in the docs
Would be good to see someone here come up with a DIY version
Would be good to see someone here come up with a DIY version
I am not sure about the Class LH, the docs mention Class HD too. The only way I can describe the performance is that whatever component of music is peaking, whether vocals, effected instruments or dub echos of rim shots, just builds naturally and peaks without losing clarity when pushed hard. Sounds amazing dynamic range with good material for a very small and cheap amp. Whether on massed folk drums on A. R. Rahamen soundtracks or Greensleeves extended versions with the dub on the back half of the track
I wish there was a high-end Class HD diy project with the TDA7293 with the best in parts
I wish there was a high-end Class HD diy project with the TDA7293 with the best in parts
One is for the TDA8392B module and the other is for the TDA7293 one, maybe. I have both and the TDA7293 one sounds way better
Same, one amp that I retired too. In my case, it kept on heating and shutting down and the heat sinks and fan needed was too much. Didnt sound that great when pushed hard too
Sorry, didn't mean to hijack or derail your project. I think I have heard this chip singing well so far. Can it be even better? Your setup maybe?
I have been working on this amp non-stop. This amp has a class A native jfet buffer and the sound is really good. Our first improvement is to reduce the noise from 98uV to 30uV. It is cooler for +-15v and will work with all regulators with 78xx 79xx footprints. Rev1.2 will be presented to you with more improvements.
Design issue, I'd say. Heatsinking capability wasn't sufficient wrt. supply voltage and load impedance. Optionally you needed to parallel more chips.
Best regards!
Dear Randy,
It's a good sounded dirty girl.. I've listened it more than 2 months with different DACs.. However in my setup I could hear a deep (high freq) noise when I close to the tweeters about 10cms.. Then I decided to retire her after 2 months listenning.. Because I am a "noise obsessed" listener !!!🙁
But after weeks it occurred to me that the design needed to be improved. In fact we initially asked for help from Bob Cordell, who was the original designer, but he stated that he had no knowledge of LM317s and TDA series chipamps, so I started to examine the circuit. and I realized that the main source of noise was actually the ccs we were using. I replaced the lm317 based ccs with a current regulated diode and a jfet ccs, and had Sercan measure the results (since I did not have the necessary measurement equipment) and we were amazed to see that the background noise had dropped to 30uV (it was 100uV before). This is a very serious development, but I was disappointed that it did not improve the SNR. We are currently focusing on this issue and I believe that we will soon make progress in the SNR.