Nelson Pass said:
Do you mean it is only in combination with bootstrapping you
don't like the method, or do you want a CCS anyway? I am
asking since most or all designs I have seen that use lateral
MOSFETs use only a bias resistor without a CCS, but none of
them has been bootstrapped, I think.
Christer said:
probably becuase bootstrps went out of fashion in the 1970s,
.I have simulated quite a few circuits with bootstraps and they actually work very well in terms of performance, stability and distortion figures. I will re-run my version of citation 12 tonight which uses a bootstrap and the current variations though the bootstrap is like 0.01ma on ~40ma bias, if I recall correctly.
millwood said:
That's true, but to some extent for pedagogical reasons, I think,
and to make it possible to start by building a very simple amp.
I am not sure if he sells/used to sell PCBs for all of the designs
in the book, though. It is also true, that some of the points
PMA listed are not considered a problem by Slone except in
certain cases. It is mostly his more powerful amps that seem
to address all of those points.
I am also not sure how much he cares about selling PCBs and
kits nowadays. I think somebody mentioned his Optimos design
has gone into commercial production.
Hi there.....and I'm the fella who spends most of my time doing the mosamp repairs....the circuit configuration on all these amps hasn't really changed from the very day they came on the market......... I'm a severe critic of all the faults that time hasn't ironed out with busted mosfets/trannies, overheating with short frilly heatsinks and sloppy biassing techniques..the very stuff tubes grin over...
I've replaced so many complementary output pairs with other types that no-one even prof'ies could tell a difference in the sound...
Viva.... tubes anyday.
rich
I've replaced so many complementary output pairs with other types that no-one even prof'ies could tell a difference in the sound...
Viva.... tubes anyday.
rich
richwalters said:
Have you had a high failure rate also for amps with lateral
MOSFETs? Due to their negative temp. coefficient above some
low Id, they are basically supposed to protect themselves
which I guess is why usually only a resistor is used for biasing.
Since you seem to have seen a lot of blown amps, your
"statistics"on this would be interesting to hear.
Re: Reply
First pick Rod's goal. Then, has he reached it? Everyone here know that Rod's goal wasn't extreme high-end, more like easy to build (few parts) and decent sound quality.
Note also that Hugh's AKSA amp is similar and highly regarded and I haven't read a bad thing about his amp.
I (as Nelson Pass) think that the bias circuitry may make trouble beacuse of the somewhat unstable currents. It may not be a problem but I trusted too must to the negative temperature coeffcient in my QRO amp and the result is a somewhat unstable amp with respect to it's idle currents.
In Rod's case it may not be a problem.
hifiZen said:Bootstrapped VAS - yuk!Use of L-mosfets in source-follower is good though. Overall, it doesn't look like anything special; I can see plenty of potential enhancements.
amp_man_1 said:
These comments are similar which I have got about my things. None of you have actually measured nor heard the amp, so why so negative?PMA said:The project is unfortunately a combination of the worst of both worlds ..
First pick Rod's goal. Then, has he reached it? Everyone here know that Rod's goal wasn't extreme high-end, more like easy to build (few parts) and decent sound quality.
Note also that Hugh's AKSA amp is similar and highly regarded and I haven't read a bad thing about his amp.
I (as Nelson Pass) think that the bias circuitry may make trouble beacuse of the somewhat unstable currents. It may not be a problem but I trusted too must to the negative temperature coeffcient in my QRO amp and the result is a somewhat unstable amp with respect to it's idle currents.
In Rod's case it may not be a problem.
wrong again.........
"probably becuase bootstrps went out of fashion in the 1970s"
No..... The enclosed schematic section (R111,R112,C102) shows a bootstrap circuit by a pretty well regarded designer on a pretty recent design. A very popular kit amplifier which also sounds very good also uses a bootsrap circuit on the voltage gain stage.
The point that nobody seems to recognize is that bootstrap allows the circuit driving the output mosfet's gate to swing outside the supply rail and gives more output swing without seperate rails for the front end. It also provides a more resistive load at high frequencies for the voltage gain stage and most likely less high frequency parastics than a transistor current source.
I can't see why bootstraping and quasi- comp output stages are not as valid design approaches today as in the past.
Before everybody second guesses Mr. Elliot, I should point out that his circuit topology is very simular to some off the highly regarded B and K amplifiers.
Every one seems to forget a major advantage that the Mosfet follower has over the BJT follower in not loading the VAS stage with the load impedance divided by the Hfe of the output BJTs.
"probably becuase bootstrps went out of fashion in the 1970s"
No..... The enclosed schematic section (R111,R112,C102) shows a bootstrap circuit by a pretty well regarded designer on a pretty recent design. A very popular kit amplifier which also sounds very good also uses a bootsrap circuit on the voltage gain stage.
The point that nobody seems to recognize is that bootstrap allows the circuit driving the output mosfet's gate to swing outside the supply rail and gives more output swing without seperate rails for the front end. It also provides a more resistive load at high frequencies for the voltage gain stage and most likely less high frequency parastics than a transistor current source.
I can't see why bootstraping and quasi- comp output stages are not as valid design approaches today as in the past.
Before everybody second guesses Mr. Elliot, I should point out that his circuit topology is very simular to some off the highly regarded B and K amplifiers.
Every one seems to forget a major advantage that the Mosfet follower has over the BJT follower in not loading the VAS stage with the load impedance divided by the Hfe of the output BJTs.
Attachments
I haven't tried to really understand bootstrapping since I got
the impression it is an old technique not a par with the more
modern solutions. Maybe I should reconsider and try to get
a better understanding of it. One thing that has bothered me
about it though is that it seems to depend on choosing a
very good capacitor, basically a lytic with very good HF
performance. Is that a correct observation?
Fred,
with my limited experience and intuition I have similarly
dismissed quasi-complementary outputs as old-fashioned,
and probably inferior technology due to the follower/CE
assymetry. To me it seems that even rather ill-matched
complementary devices should perform better.
However, I saw you post today that maybe they
should gain popularity again using MOSFETs. I admit to not
having done any serious thingking about it yet, but intuitively
I see that we may perhaps get a better symmetry here due
to the MOSFETs being more symmetrical devices regarding
drain and source than BJTs are regarding C and E. Is that
what you meant, or should I go back to the drawing board
and think deeper before I ask?
the impression it is an old technique not a par with the more
modern solutions. Maybe I should reconsider and try to get
a better understanding of it. One thing that has bothered me
about it though is that it seems to depend on choosing a
very good capacitor, basically a lytic with very good HF
performance. Is that a correct observation?
Fred,
with my limited experience and intuition I have similarly
dismissed quasi-complementary outputs as old-fashioned,
and probably inferior technology due to the follower/CE
assymetry. To me it seems that even rather ill-matched
complementary devices should perform better.
However, I saw you post today that maybe they
should gain popularity again using MOSFETs. I admit to not
having done any serious thingking about it yet, but intuitively
I see that we may perhaps get a better symmetry here due
to the MOSFETs being more symmetrical devices regarding
drain and source than BJTs are regarding C and E. Is that
what you meant, or should I go back to the drawing board
and think deeper before I ask?
no.......
"I think it is stable once the rail voltages are fixed"
The current through the bias resistor for the mosfet gates is defined by the base to emitter voltage for the common emitter voltage gain stage and the resulting collector current for the value of the base to emitter voltage. This base is driven by a current mirror so the base emitter voltage is not as well defined as with a load resistor at this location. Even with a fixed load resistor for the diff pair, changes in the VAS transistor quiescent current with temperature will more the MOSFET bias with a just resistor to set the gate to source voltage.
"I think it is stable once the rail voltages are fixed"
The current through the bias resistor for the mosfet gates is defined by the base to emitter voltage for the common emitter voltage gain stage and the resulting collector current for the value of the base to emitter voltage. This base is driven by a current mirror so the base emitter voltage is not as well defined as with a load resistor at this location. Even with a fixed load resistor for the diff pair, changes in the VAS transistor quiescent current with temperature will more the MOSFET bias with a just resistor to set the gate to source voltage.
reach for the sky .......
Fred: "I saw a guy dancing in a pair of bell bottom jeans at a disco club, like Travolta in Grease. Bell bottom jeans must be in."
Careful........ I am old enough to have worn bell bottoms in my misspent adolescence. I do miss the girls reaching up to put their books in the top of their lockers while wearing miniskirts...... walked into a wall the first time that happen.
Fred: "I saw a guy dancing in a pair of bell bottom jeans at a disco club, like Travolta in Grease. Bell bottom jeans must be in."
Careful........ I am old enough to have worn bell bottoms in my misspent adolescence. I do miss the girls reaching up to put their books in the top of their lockers while wearing miniskirts...... walked into a wall the first time that happen.
Re: no.......
I am not sure how you came to your conclusion but if I were to calculate the bias, it would be quite simple. The lower mosfet's gate is about -4v (look at its Vgs), give or take a little.
so at idle, the current going through R7/R8 got to be, per Ohm's law, (Ve-4v)/(R7+R8). Rod seems to be setting his bias at about 5ma so it works out to be (Ve-4v)10ma, very close to the formula he gives on his site.
you can then calculate Vr1 to be 8v/5ma = 1.3kohm. I tend to double that so a ~2.7kohm pot will do the job.
is that too hard to express?
Fred Dieckmann said:
I am not sure how you came to your conclusion but if I were to calculate the bias, it would be quite simple. The lower mosfet's gate is about -4v (look at its Vgs), give or take a little.
so at idle, the current going through R7/R8 got to be, per Ohm's law, (Ve-4v)/(R7+R8). Rod seems to be setting his bias at about 5ma so it works out to be (Ve-4v)10ma, very close to the formula he gives on his site.
you can then calculate Vr1 to be 8v/5ma = 1.3kohm. I tend to double that so a ~2.7kohm pot will do the job.
is that too hard to express?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.