I just made also a test with other sonys with 70ohm (MDR-XD200) and also with some small iphone earplugs, but the situation does not change: still it lowers the volume level automatically after a few secs..
I'm not so sure the opamp is having difficulties in driving all my headphones, and additionally looking at the specs they seem to me really flexible opamps.. probably I'm not running into output-overcurrent imho..
I also did the snubber/decoupling for the noise and the situation now is really ok with the linear 12V, still no better with the 24V switcher.. the only thing i noticed is that the basses sound less punchy now.. could this be?
I'll now going to doublecheck the connections again..
tent:wq
I'm not so sure the opamp is having difficulties in driving all my headphones, and additionally looking at the specs they seem to me really flexible opamps.. probably I'm not running into output-overcurrent imho..
I also did the snubber/decoupling for the noise and the situation now is really ok with the linear 12V, still no better with the 24V switcher.. the only thing i noticed is that the basses sound less punchy now.. could this be?
I'll now going to doublecheck the connections again..
tent:wq
Hi again,
ok ok sorry: I now tried again with some other headphones.. something like those skypephones with microphone included.. no idea of the impedance there (old sony md-200) but the fact is that using those I do not get any "protection" problem so far.. even if I use the 12V linear trafo.. and no background noise.. (!) wow..
so it is definitely the LM opamp to be so "nasty" in driving low impedanced phones?? is there no "patch" I can do in my circuit to overcome this?
so what is the protection that is kicking in? the output-overcurrent indeed?
tent:wq
ok ok sorry: I now tried again with some other headphones.. something like those skypephones with microphone included.. no idea of the impedance there (old sony md-200) but the fact is that using those I do not get any "protection" problem so far.. even if I use the 12V linear trafo.. and no background noise.. (!) wow..
so it is definitely the LM opamp to be so "nasty" in driving low impedanced phones?? is there no "patch" I can do in my circuit to overcome this?
so what is the protection that is kicking in? the output-overcurrent indeed?
tent:wq
Ok, thanks, I can adopt a buffer, but this would require redesign the whole thing or I understood that Appendix 1 of original Cmoy article is the solution for me:
HeadWize - Project: A Pocket Headphone Amplifier by Chu Moy
There I can decide to put more resistance on R5 (and also less noise) loosing some volume control, or put a cap, better if switched, so I can decide when I'll use hi-Z or low-Z phones.. I do not know if another option would be to have an adjustable gain too.. anyway the root of the problem basically is virtual ground inbalance, so another solution is also what tangent was suggesting by using a TLE2426 but also this would require me a complete redesign.. so for this project I'll use solution b.. next time I'll redo it I'll probably redesign a better virtual ground and also think about a buffer when using LM chips.
Thanks,
tent:wq
HeadWize - Project: A Pocket Headphone Amplifier by Chu Moy
There I can decide to put more resistance on R5 (and also less noise) loosing some volume control, or put a cap, better if switched, so I can decide when I'll use hi-Z or low-Z phones.. I do not know if another option would be to have an adjustable gain too.. anyway the root of the problem basically is virtual ground inbalance, so another solution is also what tangent was suggesting by using a TLE2426 but also this would require me a complete redesign.. so for this project I'll use solution b.. next time I'll redo it I'll probably redesign a better virtual ground and also think about a buffer when using LM chips.
Thanks,
tent:wq
I don't think there's any unusual behavior - princple is very old, see Pass Labs forum current discussions of J2/"SRPP"/Aleph CS mod - also some similarity to the "Allison" output stage bias
as I said the core circuit is the 2 Q bjt regulated ccs used everywhere - just turned upside-down, the RC bootstrap bias adds a little resistive load but no high frequency oddness
I might put series RC across the regulating Q EC to damp any really high frequency behavior
the circuit I show above likely has more issues with the Sziklai/cfp than the modulated cs but I haven't "wrung out" the cs circuit in the lab with Cload, steps, injected signal, ect
of course any time you see series R in an output you need to verify that Cload won't add too much phase shift and destabilize the amplifier negative feedback loop - for ~1 nF headphone cable load and <10 MHz loop gain intercept I don't think there's a problem in the circuit as shown
as I said the core circuit is the 2 Q bjt regulated ccs used everywhere - just turned upside-down, the RC bootstrap bias adds a little resistive load but no high frequency oddness
I might put series RC across the regulating Q EC to damp any really high frequency behavior
the circuit I show above likely has more issues with the Sziklai/cfp than the modulated cs but I haven't "wrung out" the cs circuit in the lab with Cload, steps, injected signal, ect
of course any time you see series R in an output you need to verify that Cload won't add too much phase shift and destabilize the amplifier negative feedback loop - for ~1 nF headphone cable load and <10 MHz loop gain intercept I don't think there's a problem in the circuit as shown
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.