Simple amp with directcoupled PP-follower

[revintage]

This one is really simple at the same time as it has double the effiency of the "normal" typical SE sourcefollower

Bias is through the gyrators fixed voltage. Temp-comp is achieved by the driver but it also affects midpoint voltage somewhat. No big deal here as we go single-rail.

The driver is very low distortion due to high local feedback and the path could be changed to include the output devices.

[revintage]

Forgot to mention that Class-A is a must for acceptable sonic result. 1,5A should do it.

[bobodioulasso]

How many volts are there across the Jfet when full positive swing?

[revintage]

Ca 40V.

If you are the worrying kind you could go for the 2SK223 (almost the same data as SK170) with Vds of 80V. Or you could go for lower voltages for both driver and follower.

[tiefbassuebertr]

Originally Posted by revintage

This one is really simple at the same time as it has double the effiency of the "normal" typical SE sourcefollower

Bias is through the gyrators fixed voltage. Temp-comp is achieved by the driver but it also affects midpoint voltage somewhat. No big deal here as we go single-rail.

The driver is very low distortion due to high local feedback and the path could be changed to include the output devices.

Because there is no global NFB I can consider this topology as two independed gain stages; first stage = voltage gain (in your case SRPP) and second stage = current gain (in your case unity gain push pull buffer).

Basicly there is a great similarity to the phono riaa head amp from X-pro (e. g. Creek OBH-8) go to post #5 about
How convert Creek OBH-8 to MC cartridges?

There are also two independed gain stages as above mentioned.

I would replace your push-pull output stage against a single ended version like that one from Andrea Ciuffoli - go to
http://www.audiodesignguide.com/Powe...wer/index.html

[revintage]

Hey tief....,

Read the first thread again!

I stated this an simple high efficiency alternative to the common SE follower. I also said the output stage could be included in the feedback path. Ciufollis is just one of hundreds presented.

How on earth can you think the driver is a SRPP? Just guessing? Your referance has nothing to do with this design. If you look carefully a SRPP takes its output signal from above upper devices source/emitter resistor.

This is a "normal" FET stage loaded by a gyrator instead of drainresistor. In combination with local feedback you get the characteristics of a high Gm, highly linear triode. Due to feedback you also get a Zout of in the ballpark of 500ohm.

We have discussed this solution at passlabs. Go search for "Schade".

[revintage]

bobod....,
You will have 45V at clipping so SK170 should be excluded as driver and Sanyo 2SK223 be used instead. If SK170 is a must, lowering voltages to 40/30V will be playing it safe.

[padamiecki]

Hej Lars!
If you use gnfb
you will lost sonic benefits of this design.

[revintage]

Originally Posted by padamiecki

Hej Lars!
If you use gnfb
you will lost sonic benefits of this design.

Hej Padamiecki,

Agree, this is my opinion too! As you noticed I said "could be changed". The greatest benefit with global would be to save the cost of one capacitor.

[padamiecki]

Originally Posted by revintage

Hej Padamiecki,

Agree, this is my opinion too! As you noticed I said "could be changed". The greatest benefit with global would be to save the cost of one capacitor.

Lars
gimme 5!