200W MOSFET CFA amp

[MagicBox]

But that was at 100W or 200W for powerful version.

Ah okay. I think it was my 50W design that did 1e-6% 1K full power. If I remember correctly, the 250W 8 Ohm was 2 or 3e-6%.

 

[dadod]

Ah okay. I think it was my 50W design that did 1e-6% 1K full power. If I remember correctly, the 250W 8 Ohm was 2 or 3e-6%.

My TT amp(VFA) simulates 3e-6% 1k at 100W 8 ohm.


edit: it looks like "my is smaller then yours"

 

[MagicBox]

Would you agree the outputstage and compensation used are the decisive factors in a negative feedback poweramp design?

 

[mcd99]

I would agree with that but adding in loop gain.

 

[wahab]

Yes, but don't forget distortion at 1 kHz .000001%, I never seen that in VFA amp simulation.

This means that if OLG THD1 is 0.1% then 100db NFB
are to be available to reach such a number....

A lateral fet OPS has about 0.2-0.5% THD depending
of the number of parralleled devices.

 

[dadod]

Would you agree the outputstage and compensation used are the decisive factors in a negative feedback poweramp design?

Technically I would and I like my TT VFA amp a lot. What is behind the magic of CFA amp I don't know as I never built one but I am intrigued.

 

[overall feedback]

My TT amp(VFA) simulates 3e-6% 1k at 100W 8 ohm.


edit: it looks like "my is smaller then yours"

Talking about values, can you attach some pictures?

 

[dadod]

Talking about values, can you attach some pictures?

If you want to see some picture of my TT VFA amp here is whole thread about it. http://www.diyaudio.com/forums/solid-state/182554-thermaltrak-tmc-amp.html

 

[FdW]

Some people are interested in more powerful CFA amp, and this is 200W//8ohm and 400W//4ohm....

Damir,

Can you add a noise sim to this (please ).

 

[dadod]

Damir,

Can you add a noise sim to this (please ).

There is no difference in the noise between 200W//8ohm and 400W//4ohm.

 

[MagicBox]

edit: it looks like "my is smaller then yours"

LOL yeah.. I may need to compensate my THD with brute heatsink size!

 

[mcd99]

LOL yeah.. I may need to compensate my THD with brute heatsink size!

That's the price we all pay.

 

[dadod]

How good is PSRR for this amp and how could be improved in quite simple way. I simulated tree different cases, no any aditional power line filtering, simple RC filter for input stage only, and simple C multiplier for input stage. All this was simulated on + power line only, - line PSRR should show similar results.
BR Damir

 

[sonidos]

After restoring my Adcom 535, building a chip amp, currently working on Richard Marsh's headphone amp, and using the LME49713, I am really hoping you guys come up with something beautiful here for my first big amp project. I'll just sit here and quietly lurk and learn.

 

[mcd99]

Dadod,

You are a mind reader. last night I was thinking about asking you about PSRR.

Would you be able to do similar sims for the simpler (lower LG) version. It would be interesting to see how/if LG affects PSRR.

Thanks

Paul

 

[Bonsai]

dadod, you need to use real world cap models to assess PSRR.

 

[dadod]

Dadod,

You are a mind reader. last night I was thinking about asking you about PSRR.

Would you be able to do similar sims for the simpler (lower LG) version. It would be interesting to see how/if LG affects PSRR.

Thanks

Paul

PSRR lower LG version.

 

[mcd99]

Thank you. It shows some minor benefits to the added complexity for the higher LG version. Although these are only at low frequency.

The cap multiplier looks like a easy win. The complexity is minimal and the extra voltage drop could be countered by a small dual secondary transformer added to the main PSU transformer.

 

[dadod]

dadod, you need to use real world cap models to assess PSRR.

I used LTspice capacitance models, changed ESR and ESL and no any significant differences showed. Tell me what real world cap model you suggest?

 

[dadod]

Thank you. It shows some minor benefits to the added complexity for the higher LG version. Although these are only at low frequency.

The cap multiplier looks like a easy win. The complexity is minimal and the extra voltage drop could be countered by a small dual secondary transformer added to the main PSU transformer.

JLH in his 80W mosfet amp used +-55V for OPS and +-50V for input stage not to saturate lateral mosfet when they go in very unlinear input capacitance(if I remember it correctly) so small voltage drop(1.2V) on the C multiplier is OK for me.