Good Differential J113 JFET Headphone Amplifier

This is a no BS headphone amplifier.
Fully discrete with Class A output stage.

Differential J113 Amp_05.jpg
 
Is there any downside to adding a current mirror (U5, U7) compared to a plain LTP?

Looking at the U1 - U2 - U5 - U7 signal loop, normally, I'd expect that an increase in U1 current results in a complementary decrease in U7 current, but is there any frequency where this undergoes an inversion?
 
Might also look at LSK170B/C and the LSK389 family for the input fets. Recommend resistor on emitter of U4 to supply rail or remove the mirror, as designed there is no voltage margin for the current mirror. From experience this design should work well. (I have several generally similar discrete op-amps and power amps that use this topology.)
 
Have you actually built this? The reason I ask is that the voltage across your mirror and resistors is just the Vbe drop across U4 which isn't optimal.

My note was just some suggestions of things to consider, your blunt response seems to indicate you took offense at the fact that I made those suggestions, which was not my intent.
 
@lineup Going forward you need to clearly state that these designs are simulations only and have not been tested in the real world. It is great to share ideas, but fundamentally this is a form of misrepresentation as many of us (myself included) assumed that you were actually building and testing at least some of your ideas.

If this is a mental design exercise that's fine, you just have to be upfront about it. I currently design consumer electronics for a living (my specialties are analog, mixed signal, and power) and am not exactly clueless.
 
  • Like
Reactions: tubesguy