637 Amplifiers

[quanghao]

Originally Posted by Jay

It is usually not much different in noise that can be seen in simulation of different CCS topologies. Worse, lower noise doesn't always mean better sounding. I'm still not sure if I want to go the trouble building such a complex current source. What is your experience. Is it a big improvement soundwise? I mean, are you just relying on simulation numbers?

ADD: Note that the amp already uses regulated front end supply

??????

my experience is to practice many times,
and tested on actual measurements, tests by many people listen and compare!

current source I have is simple and I noticed not like, I want a smaller noise, so I need to change! I Want amplifier with parameters better technically.

[quanghao]

Originally Posted by Jay

It is usually not much different in noise that can be seen in simulation of different CCS topologies. Worse, lower noise doesn't always mean better sounding. I'm still not sure if I want to go the trouble building such a complex current source. What is your experience. Is it a big improvement soundwise? I mean, are you just relying on simulation numbers?

ADD: Note that the amp would probably be using regulated front end supply

I do not understand this sentence, you just tell me!

my experience is to practice many times,
and tested on actual measurements, tests by many people listen and compare!

current source I have is simple and I noticed not like, I want a smaller noise, so I need to change! I Want amplifier with parameters better technically.

[Jay]

I think removing the current mirror is not in line with the spirit of using such a complex current source for the LTP...? Any reason for removing those transistors?

[quanghao]

Originally Posted by Jay

I think removing the current mirror is not in line with the spirit of using such a complex current source for the LTP...? Any reason for removing those transistors?

you say about current mirror for different pre??

or other?

New CCS its is no good for it??

[Jay]

Originally Posted by quanghao

[/B]
I do not understand this sentence, you just tell me!

I mean, by looking at perfectionism in you, you might use regulated supply for the front end LTP. This regulated supply will help lower the noise so complex ccs will be less critical than if using no regulated supply.

Originally Posted by quanghao

[/B]
my experience is to practice many times,
and tested on actual measurements, tests by many people listen and compare!

What I want to know is if you have taken the trouble to "voice" the current source. I mean, have you listened and compared several constant current sources. If so, then good, I might try it out. I myself doubt if I can hear improvement because the power supply for the opamp itself will probably have negative effect to the amp.

[Jay]

Originally Posted by quanghao

you say about current mirror for different pre??

or other?

New CCS its is no good for it??

Why have you removed these 2 transistors?

[quanghao]

Originally Posted by Jay

Why have you removed these 2 transistors?

me you want to menu simple it

[quanghao]

Originally Posted by Jay

Why have you removed these 2 transistors?

this

[quanghao]

Originally Posted by Bonsai

You have removed the LTP current mirror. Although I expec this design will work without it, loop gain will be lower and current balance in the LTP will not be optimum. Both of these will lead to higher distortion.

yes i see it!

Thanks

[keantoken]

Quanghao, the input stage to the OPA6x7 is not a conventional LTP. Notice the phase is actually reversed; the Jfet sources actually point up; they are P-channel. This could give nice advantages to the way compensation is applied.

There are ways to perform this phase reversal with an N-channel or NPN LTP (something like cross-comp) but I think these may cause stability problems.

It would be interesting to see if you can make it work with the same input topology as the OP6x7. Otherwise, the circuit is actually pretty conventional, if performing well.