TDA7293+JFE2140(LSK489)composite feedback amplifier

Again, I appreciate what you're doing and not trying to be negative, just sharing my thoughts.
Of course I don't know how the amp sounds BUT aren't THD, THD pattern, and noise widely accepted as an objective criteria reflecting the ability of an amp to correctly reproduce the input signal?
Introducing an input stage limits the output-input thermal coupling and reduces to some extent the distortion at low output power but as the open loop gain has not been increased the overall performance can't get much better.
The idea behind the composite amplifiers is to significantly increase the open loop gain and improve the performance at a given close loop gain proportionally to the open loop gain increase.
Take for example other composite amps using TDA7293 (https://www.kaltecs.com/tda7293-composite-amplifier/), where the open loop gain is increased by 40-50dB and the distortion is reduced by a similar degree.
 
I must admit that we could not achieve the desired THD improvement in this project. So if you ask what we achieved with all this mess 🙂 ;
  • We proved that TDA7293 chip works stable at 20dB gain..
  • First of all, we provided a very high impedance input buffer. This allowed our amplifier to be easily connected to any source, potentiometer, preamp, etc. without any loss in performance.
  • Thanks to the high-performance JFETs we used at the input, we achieved a significant improvement in noise and therefore SNR (103dB @1W SNR and 18uV noise while inputs shorted are good figures in my opinion)
  • Also, if we look at the FFT graph, we will see that the 2nd harmonic is dominant. This proves that the sound is more tube-like. Because I also got a similar effect in my various listenings.

Conclusion; I am in favor of giving it a chance since it is a different application that does not have a very serious cost. Especially the fact that the design, which originally belonged to Bob Cordell, has not been tried before in this form may be attractive to some of us. (I must also say that it has a really beautiful sound.)
On behalf of Sercan and myself, we would like to thank you all for your interest in the project.
 
Again, I appreciate what you're doing and not trying to be negative, just sharing my thoughts.
Of course I don't know how the amp sounds BUT aren't THD, THD pattern, and noise widely accepted as an objective criteria reflecting the ability of an amp to correctly reproduce the input signal?
Introducing an input stage limits the output-input thermal coupling and reduces to some extent the distortion at low output power but as the open loop gain has not been increased the overall performance can't get much better.
The idea behind the composite amplifiers is to significantly increase the open loop gain and improve the performance at a given close loop gain proportionally to the open loop gain increase.
Take for example other composite amps using TDA7293 (https://www.kaltecs.com/tda7293-composite-amplifier/), where the open loop gain is increased by 40-50dB and the distortion is reduced by a similar degree.
I see it completely differently!
 
  • Like
Reactions: sercan85
I've been tinkering with the TDA7293 for 30 years.
I usually rely on others when I'm not building a breadboard layout.
I got the boards from Sercan; they're very good quality and free of layout errors. I also got the boards from Kaltec (fully assembled). The layout is extremely weak, the construction is sloppy, and the component quality is poor, and all for €180 + €43 customs fees. Honestly, most Chinese boards are better. Additionally, I ordered boards with the TDA7293, but received boards with the TDA7294.
I'm looking forward to building Sercan's amplifier. I'm not sure yet whether I'll use the Kaltec boards, as I consider the stereo boards from China by Cirmech to be much better.
"...Now you can talk about the last 0.01% of distortion again!"
1000031957.jpg
1000031956.jpg
1000030414.jpg
1000031757.jpg
1000031758.jpg
1000030706.jpg
1000024576.jpg
 
If you wanto use single ended (rca) input, Just connect the - input to gnd. I have already listen to rca.

by the way we worked this amplifier with 20db gain. if you are thinking of making pcb for tda7293 and other chips yourself, there is a risk of oscillation in faulty pcb designs at low gain. that is why st recommends high gain. the most optimum design for tda7293 is our pcb.
 
Last edited:
The analytical sound of the LM3386 bothers me. That's why I'm interested in tda7293s. But I'll definitely try the LM1875. By the way, the parallel connected TDA7293 is not bad at all.

 
  • Like
Reactions: Franz Shottky
Hello,
What do you think it is possible to power the first stage of the TDA7293 (legs 7 and 8 which are IPS and VAS?) from a separate power supply?
I have a very good experience with other amplifiers when I powered the first stages from separate transformer and power supply.
If so are you sure you should get the same voltage as the output stage?Screenshot_20250428_012347_Word.jpg
 
Last edited:
Hello,
What do you think it is possible to power the first stage of the TDA7293 (legs 7 and 8 which are IPS and VAS?) from a separate power supply?
I have a very good experience with other amplifiers when I powered the first stages from separate transformer and power supply.
If so are you sure you should get the same voltage as the output stage?View attachment 1453889
It would be a good experience.. If you consider that the input stage is more sensitive to power supply ripples and transitions then I would add linear regulators to the input stages power input.. But of course the voltage level must be equal to the outputs power supply..
 
  • Like
Reactions: sercan85
and why I so enjoy the LM1875 chip amps. Listened all day today to one of the three LM1875s amps I own, and barely a hint of listener fatigue (and then only on less well recorded tracks).
Almost - or perhaps as good - at minimizing listener fatigue as my PASS and Musical Fidelity Class A amps...
The IC I’ve complained about most for causing listener fatigue is the LM3886. But I’ve heard from many that the LM1875 is truly excellent. I’d like to try my JFET-balanced, buffered CFB amplifier with the LM1875—definitely using a JFET 2140.
 
  • Like
Reactions: kenzo3
Hello,
What do you think it is possible to power the first stage of the TDA7293 (legs 7 and 8 which are IPS and VAS?) from a separate power supply?
I have a very good experience with other amplifiers when I powered the first stages from separate transformer and power supply.
If so are you sure you should get the same voltage as the output stage?View attachment 1453889
Of course it is possible, BUT with EXTRA care!!!
If the positive buffer driver supplying voltage (pin 7) is lower than the voltage on pin 13 you'll enjoy nice real explosion - the chip literally explodes. You can try this for fun.
If the buffer driver supply voltages are ~7.0V higher than those on the output stage you can omit the bootstrap capacitor. This will lower the thd by 7-8%.
If the buffer driver supply voltages are well filtered (low ripple and noise) than the thd can be additionally improved.
I've tried all these years ago and you can really have some sound improvement, BUT no point to do it if you can use a well designed composite amplifier. This is completely different story.
 
  • Like
Reactions: sercan85
If you wanto use single ended (rca) input, Just connect the - input to gnd. I have already listen to rca.

by the way we worked this amplifier with 20db gain. if you are thinking of making pcb for tda7293 and other chips yourself, there is a risk of oscillation in faulty pcb designs at low gain. that is why st recommends high gain. the most optimum design for tda7293 is our pcb.
I have no intention of manufacturing the TDA7293 amplifier. I'm just interested in the JFET input.
 
  • Like
Reactions: sercan85