jazz-amplifier

[anli]

Idea, first picture: input (I3), NFB (via R11), and output-zero-adjusting (I2) are summing at Q5 emiter. It's SE in class A (I1).

Implementation (soldered and under listening now) - second picture.

Advantages:

- it's DC amplifier,
- high PSRR (not-regulated PSU is in use, output Vrms is about 0.55mV),
- high current (and, as a result, good frequency characteristics) in the bottom mosfet gate chain,
- linear input ("ALPS-ready") - no preamp is needed,
- no need for input jfets to be stongly complemental,
- efficient (near max theoretical efficiency for SE topology).

[anli]

Originally Posted by anli

...are summing at Q5 emiter...

- must be read as "at Q4 emiter", of course, in accordance with the first picture.

[Nico Ras]

At first glance this is a very elegant design. Well done Anli

Kind regards

Nico

[anli]

Originally Posted by Nico Ras

At first glance this is a very elegant design. Well done Anli

Nico, thanks! I'm listening to the amp fourth day, and still am very excited

[juma]

Since i tried Mini-Aleph for the first time I noticed that HEXFETs sound worse when supplied with less than 20V Vds. I'm aware that you lowered PS voltage to decrease the total Pd of this amp, but it would be interesanting to hear it with at least +/-22V PS.
Otherwise, lovely amp, congratulations !

P.S.
When it comes to SE amps I always lust for a pair of good 16 Ohms speakers

[AKSA]

Andrew,

I agree with Nico. This is brilliant, I'm still figuring out some of the subtleties.....

I've noted your designs from the past. Some are in my schemats folder, you are 1 smart guy, perhaps you should keep your best ideas to yourself?

Cheers,

Hugh

[anli]

Originally Posted by juma

Since i tried Mini-Aleph for the first time I noticed that HEXFETs sound worse when supplied with less than 20V Vds. I'm aware that you lowered PS voltage to decrease the total Pd of this amp, but it would be interesanting to hear it with at least +/-22V PS.

I don't know a concrete schematics you are saying about, but, at any case, I think, the reason of such behaviour is simple - it is sufficient to look at datasheet plot showing input capacitance vs Vds. To cure the problem gate chain current must be increased, and the design permits such increasing (at my case the current > 30mA, but I just don't use mini-heatsinks for bjts and 9610). And, say, R10 just "takes" some heat from CCS.

Originally Posted by juma

Otherwise, lovely amp, congratulations !

Thanks!

Originally Posted by AKSA

I agree with Nico. This is brilliant, I'm still figuring out some of the subtleties.....

Thanks!

Originally Posted by AKSA

...perhaps you should keep your best ideas to yourself?

Not being English native speaker I'm not sure I understand "keeping" meaning. Do you mean some kind of, say, amps kit producing? If it's the case - well, I'm far from business, and the only way I have ATM is to share own thinkings with music lovers.