HARM-23, Jfet 100W 8ohm

[roender]

This is my new creation, an amplifier inspired by MikeB symasym but more powerful: 100w 8ohm.
What do you think about?

[roender]

FFT 20Khz

[lineup]

Originally posted by roender

This is my new creation, an amplifier inspired by MikeB symasym but more powerful: 100w 8ohm.
What do you think about?

Originally posted by roender

FFT 20Khz

The nice test results tells it all.
And SymaSym is a very good amplifier.

So, why is SymaSym so good?
Yes, because there has been a handful of the best members
HELPING EACH OTHER to make SymaSym better and better.
With ideas, discussion and good test equipement.

One Man alone can do pretty good, sometime.
Several minds together ARE ALWAYS more than ANY man.
We call it sometimes Brain Storming.

Several different people will see more than one.
Because will see each one from unique perspective - from each special side.
One person is limited - to one view - and may not see some side of thing.

United we stand .... or else we may fall.


Now, it is not as simple as always when people get together
will result in something great .. or even something good.

There has to be a friendly and accepting attitude
among most of those that are in lead of a project.
Or else it will block new ideas. Some of which may be the good and correct ideas.
All ideas are not good.
But blocking ideas, even if they are no good or even crazy or bad,
will stop even the good ones from coming.


congratulations, roender
you could not have made a better choice,
than start with SymaSym,
to make your own version, for your own need



Regards
to roender and MikeB
Quality of creation will tell also something about Quality of creator

lineup december 2006
http://lineup.awardspace.com/

[roender]

Hi all,
I need your critical opinion, specially at negative feedback calculation. I'm not so sure if it's stable
In my humble opinion, the gordian node of symasym is impedance unmatching between first and second differential amplifiers, which will make the first stage to be very unlinear.
The problem was solved by using p-jfets in second stage.

[viktor1986]

Well, with hope that this won't disappoint you (there is no reason why it
should, because your design is excellent!), I know for very similar design
done about 6 years ago by Dr. Jagodic (boraomega on this forum). In recent
years, several people built flat = LINE preamps based on this schematic
(with minor mods) and it is one of better line preamps I heard so far.
Those designs on his WEB site are mostly his medium class designs but his
HighEnd designs he never published because of commercial reasons. Just few
close friends have access to that collection. Here is the schematic. Sorry
no components values, it will be unfair to Bora if I put values.

[roender]

Originally posted by viktor1986
Well, with hope that this won't disappoint you (there is no reason why it
should, because your design is excellent!), I know for very similar design
done about 6 years ago by Dr. Jagodic (boraomega on this forum).

No problem
Nothing new on this world. If you look closely, you will see a lots of differences between those designs.

[lineup]

Mister viktor1986
did post that circuit here in wrong topic
... that one is a pre amplifier
... and Mister Viktor got it right the second time he posted,
in topic Super Preamp.

[MikeB]

Hi Roender, nice work!

A few comments, you shouldn't reference the cascode in 1st diffamp to plus-rail, use gnd instead. By cascoding the 2nd diffamp you already solved this impedance issue, you can skip the jfets here. Jfets here will make the bias in 2nd stage dependent from idss of these jfets.
C14 appears very large to me.
As long as r23 is 680ohms, the gain is only slightly decreased, the ol-gain is already greatly reduced by using jfets in 1st stage. I see no stability problems here.

You should check clipping behaviour, especially check reverse bias behaviour of q8. It might be necessary to use basestoppers to q8/9 and clamping diodes at output of 1st diffamp.

At 45v supply, it might get necessary to use 2 output pairs, especially with the weaker mjw0281/0302.

Also observe temperature of q12, you have increased current by 2.2, and increased voltage by 1.4. That gives factor 3 higher dissipation. A resistor in series with q12 collector might solve that.

Mike

[roender]

Thank you Mike, I will chewit slowly.