Current conveyor as a voltage amplifier

[dadod]

Originally Posted by keantoken

I'll bet you could replace M1 and M2 with resistors without losing anything. Q17/18 will not improve PSRR, and bias loading here is relatively benign.

I suppose you mean CMRR not PSRR.
And you are right as usual, M1 and M2(CCS) and Q17/18 are not so important regarding CMRR, but without Q17/18 I've got three times higher distortion(it looks that main distortion source is output buffer now).
Now the best and quite simple(not as simple as Gaiwire with no CCS and no bootstrap) is to discard M1 and M2 CCS, but keep it in output buffer.
I think that this is now your circuit as much as mine Kean.
Damir

[keantoken]

Yes, I meant CMRR.

Dadod, here's something you can try for the output buffer. Put the zener in series with Q14/Q21 collectors. Then connect Q13/Q22 bases to Q14/Q21 collectors. This will get rid of the need for a bootstrap or CCS I think.

About current mirrors, I thought one of my recent posts was relevant:

diyAB Amp - The "Honey Badger"

[dadod]

Originally Posted by keantoken

Yes, I meant CMRR.

Dadod, here's something you can try for the output buffer. Put the zener in series with Q14/Q21 collectors. Then connect Q13/Q22 bases to Q14/Q21 collectors. This will get rid of the need for a bootstrap or CCS I think.

About current mirrors, I thought one of my recent posts was relevant:

diyAB Amp - The "Honey Badger"

Regarding current mirrors I am in accordance with you, look here Current conveyor as a voltage amplifier. Mismatch in CM emitter resitor is the main source of the current conveyor distortion.
And now simulation with your proposed output buffer changes. THD20k was dobled but that is still so low and I think this is the best and simplest WireGain circuit diagram. That distortion is with 300 ohm load, so perfect for Sennheiser HD 580 Precision headphones(I have this one).
Thanks Damir

[dadod]

This should be final version of this line preamplifier. It is not preamp only, but could drive a headphones, like Sennhaiser(300ohm) and Grado(32ohm). Whith the Sennhaiser there should be no problem and distortion is in ppm level. Grado with its 32 ohm load is a bit tougher load but still distortion is quite low and very benign, look attached FFT. It was simulated at 2mW and Grado defined 98dB at 1mW.
dado

[keantoken]

Collector of Q21 is broken.

[dadod]

Originally Posted by keantoken

Collector of Q21 is broken.

Thanks, and LEDs instead zeners(D3, D4).

[dadod]

There was an error in last schematic(output buffer) and here is correct one.
dado

[dadod]

Finally I did first listening. Power supply was a temporary one very simple with 2x3300uF giving +-27 V.
I am really satisfied with the result even with this simple power supply. My intention is to design a shunt regulator for +-32V and I expect to get even better result with it.
First listening was done with simple circuit without bootstrap in the current conveyor part, that I am going to do next.
I have Aikido tube lineamp and I like GainWare more (could be because I designed it).
The source I used was Denon DVD 3910, lineamp GainWire, poweramp my AudioCube described in solid state thread, and loudspeakers were my design with a ribbon from Newform Research and 8” woofer from ScanSpeak.
Sorry for bad photographs, I've got a camera in my cell phone only.
dado

[vac231]

Hi Dado,

Very happy to see your new little preamp ! Could you describe in a few words how is the sound and how it has improved ?

Best regards,

Nicola

[dadod]

Originally Posted by vac231

Hi Dado,

Very happy to see your new little preamp ! Could you describe in a few words how is the sound and how it has improved ?

Best regards,

Nicola

Nicola, it was very short listening session and with just a simple power supply, but it was very promising. I will report again with proper power supply.
In a few words, great dynamic, clear voices and hights.
BR Damir