I have been working on this amp design in simulation, off and on, in various incarnations, for about 12 years now. I've picked it up, played with it, and set it aside, several times. It's never been so hopeless as to abandon completely, but has always had lingering problems. I have always been intrigued by the idea of an opamp at the input of a power amp. Such topologies can be notoriously nasty to stabilize, and up until recently I never thought I had anything worth actually building.
The amplifier shown below will produce rather impressive simulated THD results. Please have a look. On +/- 54V rails:
----------150 W------1 W
THD20kHz 0.000881 0.000286
THD10kHz 0.000331 0.000140
THD1kHz 0.000078 0.000068
The amplifier has just over 210 dB of loop gain, so it's not for those who shy away from negative feedback. I'm not sure if this amplifier will sound well enough to be worth the effort, but I suppose I'll have to build one. I've always looked on it as sort of an academic exercise up until now. Before I do that, I thought I'd ask for opinions, suggestions, and other feedback.
If I've done it right, attached is a .jpg picture of the schematic, and a .zip file with the LTSpice .asc file and a .lib file w/ models.
Your comments are appreciated,
Mike
The amplifier shown below will produce rather impressive simulated THD results. Please have a look. On +/- 54V rails:
----------150 W------1 W
THD20kHz 0.000881 0.000286
THD10kHz 0.000331 0.000140
THD1kHz 0.000078 0.000068
The amplifier has just over 210 dB of loop gain, so it's not for those who shy away from negative feedback. I'm not sure if this amplifier will sound well enough to be worth the effort, but I suppose I'll have to build one. I've always looked on it as sort of an academic exercise up until now. Before I do that, I thought I'd ask for opinions, suggestions, and other feedback.
If I've done it right, attached is a .jpg picture of the schematic, and a .zip file with the LTSpice .asc file and a .lib file w/ models.
Your comments are appreciated,
Mike
Attachments
Last edited:
Let me say right up front that I'm very aware of how critical the models are in all of this. There's a fairly good group of models in the .lib file, and the results vary significantly, so yes, there's a little bit of gamesmanship involved here because this is the best combination of device types and models I've found. Still, the results were low enough to be interesting. Even if the real thing is an order of magnitude worse, that's still pretty good.
I am very interested in this amplifier.
From LTspice simulation with the attached files, I determined that there is a x175 voltage gain from the circuit that follows the Op amp output.
The overall voltage gain of this amp is x25 = ( 20K + 820 )/820.
I am puzzled to see, the amp is stable. The OPA is unity gain stable, how come there is no stability trouble with this x175 gain wich is well over x25
From LTspice simulation with the attached files, I determined that there is a x175 voltage gain from the circuit that follows the Op amp output.
The overall voltage gain of this amp is x25 = ( 20K + 820 )/820.
I am puzzled to see, the amp is stable. The OPA is unity gain stable, how come there is no stability trouble with this x175 gain wich is well over x25
Last edited:
The push pull Vas in this amp is very interesing.
It is made of two folded cascodes
Q16 ( current mirror Q11 Q10 ) Q3
Q17 ( current mirror Q14 Q12 ) Q2
This topology is nice.
High bandwith. ( Gain Band 30MHz )
Good PSRR
Good linearity ( 0.01% THD @100KHz )
Very flexible ( easy to change for gains from 10dB to 50dB )
Are they other examples on DIYaudio using this "folded cascode" topology ?
It is made of two folded cascodes
Q16 ( current mirror Q11 Q10 ) Q3
Q17 ( current mirror Q14 Q12 ) Q2
This topology is nice.
High bandwith. ( Gain Band 30MHz )
Good PSRR
Good linearity ( 0.01% THD @100KHz )
Very flexible ( easy to change for gains from 10dB to 50dB )
Are they other examples on DIYaudio using this "folded cascode" topology ?
- Status
- Not open for further replies.