...
...........2015
Diamond Input and DC Servo
Vssa seems lot better with dumped Diamond input and DC servo.
Proof of concept:
-> www.esperado.fr - VSSA with Diamond input and DC servo
Vssa seems lot better with dumped Diamond input and DC servo.
Proof of concept:
-> www.esperado.fr - VSSA with Diamond input and DC servo
Hi Esperado;
great that you share your schematic
well, removing the 2 caps is really changing the cfa model ! your amps looks more like a great but usual CFA as per Marantz or Accuphase, anyway worth a try.
about the servo, it s always a pain to inject something in the signal or feedback loop as you also inject noise. What do you think of ajusting dc by changing the current in the front end, like done on the headset amplifier below by slightly changing the curent flowing down the led used for a ccs: A Pure Class A Dynamic Headphone Amplifier | HeadWize
great that you share your schematic
well, removing the 2 caps is really changing the cfa model ! your amps looks more like a great but usual CFA as per Marantz or Accuphase, anyway worth a try.
about the servo, it s always a pain to inject something in the signal or feedback loop as you also inject noise. What do you think of ajusting dc by changing the current in the front end, like done on the headset amplifier below by slightly changing the curent flowing down the led used for a ccs: A Pure Class A Dynamic Headphone Amplifier | HeadWize
Not really, on my opinion. Same stages in the feedback loop, same devices with same currents, same operation (same gain in the IPS and VAS).
Well, the purpose is to get the best results without adding complexity in the signal and feedback path (loosing audio details).
The noise ? Are-you scared by the noise generated by a TL072 at unity gain (20nV/rtHz ?) in the audio range, attenuated by... 56dB here (22/15022) ?
I think it is better than to mismatch the CSS symetry(distortion side).
Anyway, where is the difference if you modulate the current of a CSS by a noise, you will get this same noise out of it, not ?
Last edited:
Hi Esperando
Interesting schematics. It reminds me a lot like the Hiraga Super Class A but with lateral MOSFET output I did something similar to that but with 2SK1530/2SJ201 as outputs. However, I have not used DC servo. My friends and I found the MOSFET version was an improvement in sound compared to original Hiraga Super Class A.
Fab
Interesting schematics. It reminds me a lot like the Hiraga Super Class A but with lateral MOSFET output I did something similar to that but with 2SK1530/2SJ201 as outputs. However, I have not used DC servo. My friends and I found the MOSFET version was an improvement in sound compared to original Hiraga Super Class A.
Fab
Vssa seems lot better with dumped Diamond input and DC servo.
Proof of concept:
-> www.esperado.fr - VSSA with Diamond input and DC servo
Hi Christophe
Nice iteration, although I don't like diamond inputs, it will work OK. VSSA doesn't need DC servo since +/-5 mV DC offset is perfect result for power amp, also this DC coupled figures won't pass +/- 10 mV.
In extensive listening test over the summer period for new First One v1.3 I found out more important features (couldn't be found anywhere on DIY forum) to get perfect sound, so diamonds stays reserved for girls only hehe
L.C.
See first image in my modified article. I added a 40Hz distortion comparative measurement."DC Servo ?"...Reserved for the cars: breaks, steering, etc.
www.esperado.fr - VSSA with Diamond input and DC servo
Did the sim twice because i was shocked.
Nothing to add. I go back to my cave.
Last edited:
Christophe you did a great job, showed exact points, strenghts and weaknesses, I appreciate a lot, thanks.
After forthcoming exhibition VSSA v1.5 will be on my screen so most probably I'll consider DC coupled version in my listening tests and see if it's competitive.
Be sure report will follow but not before november.
Thanks again for your effort and good will.
I'm not even sure the distortion in the bass is not 'more musical', of course ;-)
Only listening will tell. But, at least on the paper, or on an objective and technical point of view, the benefit of DC coupled feedback is obvious... see what i mean.
About Servo, i never understood why they have bad reputation. Because they look ugly and complicated on the schematics ?
If you set their FC low enough (0.1Hz) they have no influence on phases, that we cannot achieve with caps in the feedback path. You can use little ceramic caps with no problem (LF). They don't add noise or distortion, because their output signal can be attenuated by some 50dB or more(the full gain of the OPA at DC) and they output no signal in the bandwidth.
Trying to use them to move the current of a CCS is stupid, on my point of view. Their noise (if any) will be present as well at the output of the CCS (so obvious). And one of the goals is precisely to avoid CCS mismatch.
I was never able to hear any difference in sound character, switching Servos in/out.
They take care of the offset whatever happens with temperature or transistor mismatch.
The only negative point is the boring +-12V to feed them. Or from outside, or from the main rails. But once done, it is for ever.
Just my two cents.
I'm moving next month to Portugal, I will lurk for your listening reports with great interest, once i will have ADSL there
Keep on the great work, Andrej.
Just for the fun (not tested in real), highly optimized DVSSA (1200V/µs, <4ppm H.D.) is here:
http://www.esperado.fr/vssa-diamond/dvssa.php
http://www.esperado.fr/vssa-diamond/dvssa.php
I thing so. But, if it is not requested by stability reasons, if it don't introduce too much phase shift at 20KHz, i think the benefit of higher inductance in output is to filter more RFI, damping those induced in the speakers wires before they enter in the feedback loop. (i use to shield my speakers wires as well).
As well as they filter a little more the very high order harmonics (crossover distortion) too (on measurements ;-)