Symx said:
When I was young, I was much more serious, and my circuits
were more complicated. Now I am not so serious, and I take
pleasure in making good amplifiers which are simpler. I still
use cascoding, super-symmetry, Class A, variable bias and
so on, but I apply them to circuits with fewer gain stages. They
don't measure as well, but to my taste they sound better and
they serve as good tutorial examples.
Nelson Pass said:
Hmm. Age - that explains it. The older I get, the simpler I like things. When I start a new design, if it starts looking complicated or bandaided anywhere, I usually throw it out and start over. Some of the amp schematics here seem overly complicated to me. That said, the KISS method has limitations. It's good to go simple, but not too simple.
I was once told many years ago that the knowledge that goes into a design is inversely proportional to the number of potentiometers or select in test (S.I.T.) resistors in the circuit. This was not referring to volume, tone, etc., controls.
More on the term "cascode"
"First use by F.V. Hunt and R.W. Hickman in a 1939 article. They discuss circuits for application in what they call "low-voltage" (130V-600V) stabilizers. They describe two triodes in cascade, and think of it as a pentode replacement. Hence they thought of the name "cascode".
This name applies to a cascade of a common-cathode triode stage and a common-grid triode stage. This particular cascade configuration became very popular, and it still is. The term "cascode" also became very popular. Not only the circuit, but also its name evidently filled a need. Engineers liked the name, and still like it, because it distinguishes this popular stage from other cascade circuits. Later, in 1948, Wallman, MacNee, and Gadsen describe this as a stage which has the low noise factor of a triode and the gain of a pentode."
from: http://www.quadrivium.nl/language.html
JF
"First use by F.V. Hunt and R.W. Hickman in a 1939 article. They discuss circuits for application in what they call "low-voltage" (130V-600V) stabilizers. They describe two triodes in cascade, and think of it as a pentode replacement. Hence they thought of the name "cascode".
This name applies to a cascade of a common-cathode triode stage and a common-grid triode stage. This particular cascade configuration became very popular, and it still is. The term "cascode" also became very popular. Not only the circuit, but also its name evidently filled a need. Engineers liked the name, and still like it, because it distinguishes this popular stage from other cascade circuits. Later, in 1948, Wallman, MacNee, and Gadsen describe this as a stage which has the low noise factor of a triode and the gain of a pentode."
from: http://www.quadrivium.nl/language.html
JF
With due respect to Nelson, I find that the older I get, the better I was..... 
But my circuits definitely get simpler since I spend more time choosing circuit dimensions and components.
I find designing to get rid of (largely academic) problems seldom seems to give the desired results, and often introduces others, many of which damage the music.
But don't ask me which problems. I've forgotten........
Cheers,
Hugh
But my circuits definitely get simpler since I spend more time choosing circuit dimensions and components.
I find designing to get rid of (largely academic) problems seldom seems to give the desired results, and often introduces others, many of which damage the music.
But don't ask me which problems. I've forgotten........
Cheers,
Hugh
Nelson Pass said:
Hm, I guess that depends on how you define "serious". Isn't it
more serious to try designing better-sounding amps??
OTOH, I agree that the better one learns a subject and the more
experienced one gets in it, ones attitude towards the subject
becomes more relaxed (maybe less serious) but one usually
achieves better results because of this more relaxed attitude.
I started some listening tests midweek, based on suggests by Charles. Because the gears were in motion to build the Borbely Super Buffer, it was easiest for me to try different things with it as it was on the breadboard.
Now I think that listening tests are very difficult and best if the A/B'ing occurs instantaneously. Anyway the easiest A/B test was to have the amp configured with "global feedback" from the output, or to move the complementary output follower outside the feedback loop. (I guess this means that the load, in this case Sennheiser headphones, are not influencing what is fed back to the input.)
Well, it seemed to me that without the follower in the FB loop the music is softer, open, and lively. With "global FB", it seemed sharper, a little more congested, and more electronic... Again, I'll admit that listening tests are slippery. YMMV. However, the way I prefer to sit and just listen was with the follower outside the loop.
The circuit sounds very good in either configuration. And I realize that just removing the output stage from the loop is not what ZFB is all about. However, I've got a configuration closer to Charles suggestions that I'm pleased with and am now ready to finish my project.
Thanks Charles.
JF
Now I think that listening tests are very difficult and best if the A/B'ing occurs instantaneously. Anyway the easiest A/B test was to have the amp configured with "global feedback" from the output, or to move the complementary output follower outside the feedback loop. (I guess this means that the load, in this case Sennheiser headphones, are not influencing what is fed back to the input.)
Well, it seemed to me that without the follower in the FB loop the music is softer, open, and lively. With "global FB", it seemed sharper, a little more congested, and more electronic... Again, I'll admit that listening tests are slippery. YMMV. However, the way I prefer to sit and just listen was with the follower outside the loop.
The circuit sounds very good in either configuration. And I realize that just removing the output stage from the loop is not what ZFB is all about. However, I've got a configuration closer to Charles suggestions that I'm pleased with and am now ready to finish my project.
Thanks Charles.
JF
Personally (and maybe it's just me) I get more results living
with a change for a while, like a week or two, and then going
back. The difference is often most apparent when you go back.
It's like drinking better wine - often the difference is really noticed
when you back to the cheap stuff.
with a change for a while, like a week or two, and then going
back. The difference is often most apparent when you go back.
It's like drinking better wine - often the difference is really noticed
when you back to the cheap stuff.
>I'm perfectly proud to have designed something before Hawksford, and less happy to have been anticipated by him on another.
Nelson, your circuit 3,995,228 and later variations from Pioneer 4,254,379, JVC 4,274,059 and Sansui 4,401,951 are different circuit realizations of Blomley idea (nonswitching), which was mentioned here several times. Hawksford idea (and at the same time Iwamatsu 4,476,442) is the error feedback with summing at the input of the output stage.
Nelson, your circuit 3,995,228 and later variations from Pioneer 4,254,379, JVC 4,274,059 and Sansui 4,401,951 are different circuit realizations of Blomley idea (nonswitching), which was mentioned here several times. Hawksford idea (and at the same time Iwamatsu 4,476,442) is the error feedback with summing at the input of the output stage.
Attachments
keep it simple KISS
lineup here
I am an old
so I have a liking for anything
that still in the year of 2006
in a civilization filled with complications
can be called simple
- close to simple
and yet do an excellent job
... even in many instances, cases
do it better!
than all them many complicated, sophisticated
and the numerous circuits
being overloaded with an overhead of components
that achieve not much more
that adds hardly anything to final quality
of the output, the resulting outcome
lineup who Dares to try: KEEP IT SIMPLE
and this
even in times where advanced technology is the Golden Calf they dance around
and complication beyond understanding is the norm
Just give the Class-D forum a visit ... and you see what I am talking about
================================================
PS. Anybody tried to repair their modern (model 2000-) car lately?
digitally controlled in so many the systems of the car
They do not make cars, like they used to do.
Not for do-it-yourself, that is for sure.
Symx said:
Nelson Pass said:
millwood said:
lineup here
I am an old
so I have a liking for anything
that still in the year of 2006
in a civilization filled with complications
can be called simple
- close to simple
and yet do an excellent job
... even in many instances, cases
do it better!
than all them many complicated, sophisticated
and the numerous circuits
being overloaded with an overhead of components
that achieve not much more
that adds hardly anything to final quality
of the output, the resulting outcome
lineup
who Dares to try: KEEP IT SIMPLE and this
even in times where advanced technology is the Golden Calf they dance around
and complication beyond understanding is the norm
Just give the Class-D forum a visit ... and you see what I am talking about
================================================
PS. Anybody tried to repair their modern (model 2000-) car lately?
digitally controlled in so many the systems of the car
They do not make cars, like they used to do.
Not for do-it-yourself, that is for sure.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.