Current feedback high speed headphone amp with diamond output buffer

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Jump to Latest
C

Christer

Member
Joined 2002
Please guys,

the schematic was just posted for reference to the simulation figures
I posted earlier. It wasn't proposed as anything new. It is the
standard version straight from Jungs article, but with other
transistors and maybe some minor details differing.

I am sure there are zillions of possible ways to alter it, but that
was not the reason I posted it. Not that I am not interested in
suggestions for modiified version, though.
 
C

Christer

Member
Joined 2002
PMA said:
Christer,

I understand, but the schematics is worth paying attention, as it is one of the great circuit ideas ever published.
Click to expand...

Agree, I just wanted to point out that my reason for posting it
was not to start a discussion on the circuit per se, since it is
"the standard version" of it. BTW, it is biased in class A in this
case. The output transistors I use are perhaps not the best
choice either in a low-voltage application, but I took what models
I had. I made the distorsion simulations mainly with ideal BJT
models to experiment with parameter mismatching, but the PSRR
simulations were done with the BJTs in the schematic to give
(hopefully) more realistic figures.

CFP outputs sounds novel, to me at least, but will we get the
same Vbe cancellation effects then? Then I guess it gives a
risk of oscillation and increases the output impedance? Still
worth considering I suppose.
 
C

Christer

Member
Joined 2002
jam said:
I ment input and output coupling caps.

Regards,
Jam
Click to expand...

Yes, I had a feeling that might be the case, but I don't have any
in the schematic and since you directed the comment to me, I
couldn't quite see any other interpretation.

If you are concerned about DC offset, yes I suppose that may
perhaps be a problem if using the buffer stand alone. Within
the loop of an op amp I guess not. Input coupling caps seem
always wise to have, I think, but had no purpose for the particular
simulations I did.
 
PMA

PMA

Member
Joined 2002
pmacura.cz/audiopage.html
Paid Member
Christer said:


CFP outputs sounds novel, to me at least, but will we get the
same Vbe cancellation effects then? Then I guess it gives a
risk of oscillation and increases the output impedance? Still
worth considering I suppose.
Click to expand...

CFP output stage (please see push-pull follower on my page) does not change Vbe cancellation, as the B-E junction of output devices would be connected across small resistor inserted to the collectors of your Q2 and Q4. Oscillations, if occurred, can be easily damped by Ccb capacitors. Class A operation would probably not bring oscillation problem (no turn-ons anf turn-offs). Output impedance is highly reduced, not increased. It is reduced as an effect of additional local NFB, created by CFP's. The concept is proven and works, brings about 10x less distortion compared to original Jung's circuit.
 
C

Christer

Member
Joined 2002
PMA said:


CFP output stage (please see push-pull follower on my page) does not change Vbe cancellation, as the B-E junction of output devices would be connected across small resistor inserted to the collectors of your Q2 and Q4. Oscillations, if occurred, can be easily damped by Ccb capacitors. Class A operation would probably not bring oscillation problem (no turn-ons anf turn-offs). Output impedance is highly reduced, not increased. It is reduced as an effect of additional local NFB, created by CFP's. The concept is proven and works, brings about 10x less distortion compared to original Jung's circuit.
Click to expand...

OK, yes I didn't really try to analyze it, just speculated a bit I
admit. Since I know you use discrete diamond buffers a lot,
I presume you have built and tested the CFP version too? Does
it have any sound improvement too, or just measure better?
 
PMA

PMA

Member
Joined 2002
pmacura.cz/audiopage.html
Paid Member
Christer,

I have proven the concept at the class A output stage of the power amplifier. The very low THD combined with NFB brought sound improvement that can be described as better resolution, clarity and smoothness of the sound, this compared to DEF (double emitter-follower) output stage.

The buffer circuit of this kind is just prepared by a colleague of mine and supposed to be tested soon.

Pavel
 
C

Christer

Member
Joined 2002
PMA said:
Christer,

I agree. CFP output stages may bring problems in class B or AB (narrower crossover region results in higher THD harmonics), but not in class A.

The simulation results we have for the Jung's buffer are very promissing (THD lowered about 10x).

Pavel
Click to expand...

Ah, that is interesting. Basically all of Slones amps are class AB (or
class B depending on ones definitions) so that might explain why
he has reverted to using followers again, perhaps.

The case with the buffer is interesting, especially if one wants to
use it as a plain buffer and not within a feedback loop of an op amp.
 
P

pjacobi

Member
Joined 2004
linearaudio.de
jam said:
[...]
P.S. I have built a similar circuit and found the offset drift too high for direct coupled preamp applications. I suppose a higher beta device or darlingtons for the input pair might help. I had to use a servo (nasty) to direct couple it. [...]
Click to expand...

Hi Jam,

The input offset current can be made small with diamond of darlingtons, e.g. 0.3uV in this test circuit:
http://www.linearaudio.de/scratch/dmnd-test-2a.pdf
Then drift wouldn't be a problem, I suppose?

Or are you referring to the drift due to Vbe drift? I did assume, that this can be made small by using equal currents for first and second stage and thermal coupling.

Best Regards,
Peter Jacobi
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Top Bottom