Advanced Solid State Amp

Status
Not open for further replies.
Jump to Latest
#42
Hienrich!

I would increase C6 to 2,2uF and C7 to 330uF, especially when the amp is playing as a subwoower amp.
For Q7 that big transistor is unnecessary, a small signal device will do the job, too.
And I would use complementary output devices.

Anyway nice design, congrats!

PS: I would increase the VAS current by R6.
 
#43
A lot of conflicting opinions here indeed. Perhaps if I offer mine it might confuse, err, I mean clarify the issue:

  • Keep the current mirror. Perhaps some people don't like it because the amp might have less distortion, including less of the 'nice' stuff, or perhaps it's a stability thing, since simply replacing resistor loading with a current mirror increases the gain which may decrease stability (good stablity is more important than many realize). That can be fixed with increased degeneration for the input transistors.

  • You should definitely degenerate the input transistors. Without degeneration, any mismatch between the transistors will have a bad effect and the bandwidth of the input stage will be low, necessitating a larger Miller compensation capacitor. Depending on how well matched the transistors are and how fast you want the amp to be, anywhere from 22 - 470 Ohms is good.

  • Similarly, you should degenerate the current mirror. You can get away without degeneration if you use a monolithic matched pair here, but even then the performance won't be as good without degeneration. I would suggest smaller resistors here, of perhaps 10-100 Ohms.

  • Consider buffering the Vas with an emitter-follower. The performance increase can be dramatic, although stability becomes a little more difficult to achieve. Removing the Vas degeneration resistor at the same time can also yield a small benefit.

  • Lose the diodes in the rails and use regulators instead, preferably a few volts higher than the output stage rails for maximum efficiency. It's not much more complicated than the diodes if you use IC regulators, but it works much better.

  • The earlier suggestion to wrap frequency compensation around two stages is good, but there is a significant risk of instability, so I wouldn't recomend trying something like that until you are very confident with amp design.

  • Bias the current source with a (non-blue) LED, 'cause they look cool😀
 
#44
Mr. Evil has a lot of good suggestions that I agree wholeheartedly. My only point is to use an RC network for the first stages vs. a regulated one that Mr. Evil suggested, for simplicity.

The first stages have a steady current drain, and the ltp has good psrr to begin with. a RC network (I usually use 470ohm+220uf) will be more than sufficient to decouple them from any rail droops.
 
#46
Yes....reinvented the AKSA

Or the Askaksa.

hehe


Carlos
 

Attachments

  • askaksa.jpg
    askaksa.jpg
    32.7 KB · Views: 557
#47
edl:

thanks for the compliments:angel:

well how did you know that this beast is driving my subs😀

ok... I'll try those changes :smash:

only one thing I've only got NPN output devices, well then still the amps sounds dynamic and tight, but yet hoping that I will find enough
time and money for there complimentaries....

rgds,

hienrich
 
#50
post 18

"it's better to use one floated resistor here" (resistor emitter driver)

My question:
For to substitute both resistors, how to calculate floated resistor?
For example, R (330 ohms) + R (330 ohms). Floated resistor is 660 ohms?.

Thanks.
Guillermo
 
Status
Not open for further replies.
Top Bottom