A new concept to grasp.......
It is called "loop gain", boys and girls. Repeat after me........
The difference between open-loop gain and the closed-loop gain is called loop gain. In circuit with sufficiently high OLG, the CLG remains constant. Only the loop gain is lowered when OLG is reduced. (We aren't going into BW and -3 dB points.......)
In the case of this amp, the OLG is probably not high enough that lowering it will affect the CLG some.
All I see El-Phred-o doing is explaining it in terms that are in common usage, so the newbies don't go running off spouting nonsense next time they ask for help. It is hard to help people when they don't know how to ask the question in a manner that results in the answer that they need.
Jocko
It is called "loop gain", boys and girls. Repeat after me........
The difference between open-loop gain and the closed-loop gain is called loop gain. In circuit with sufficiently high OLG, the CLG remains constant. Only the loop gain is lowered when OLG is reduced. (We aren't going into BW and -3 dB points.......)
In the case of this amp, the OLG is probably not high enough that lowering it will affect the CLG some.
All I see El-Phred-o doing is explaining it in terms that are in common usage, so the newbies don't go running off spouting nonsense next time they ask for help. It is hard to help people when they don't know how to ask the question in a manner that results in the answer that they need.
Jocko
Hi, elders,
The discussion is getting more interesting. Thank you GRollins for answering my question. But there is one more question that is not been answered. What happens if I eliminate the lower resistor of the lower VAS (second stage) mosfet? The Source directly to -VCC, not via any resistors. I tried this in other amp (not aleph), and I think I hear better music details. Is it dangerous to try this at aleph design? Or will I get better music details?
For GRollins, I also have questions about tubes that I cannot answer. Please look at
http://www.diyaudio.com/forums/showthread.php?threadid=21349goto=newpost
about very low voltage tube preamp.
The more and more I study audio power amp, the question that raise in my mind becomes more and more silly. Some times it comes back on "how does 1 single transistor works?". I'm sure many of audio fanatics do not understand this either.
Without such knowledge (like Mr. Pass on mosfets), no one will be able to design ZEN like amp. Or aleph amp.
(I think thats why people gives Nobel Prices, the study of very basic knowledge is the key to built more complex things)
So Mr. Pass, how does 1 single mosfet works? If I put 1mA steady current or 1A steady current into a differential configuration, will the gate requirements be different? If I put the same preamp signal, will the output be different, or the music details be different? Since the mosfets aren't requiring current into its gate (like bipolar with its base with hfe figures), please tell me how it works.
The discussion is getting more interesting. Thank you GRollins for answering my question. But there is one more question that is not been answered. What happens if I eliminate the lower resistor of the lower VAS (second stage) mosfet? The Source directly to -VCC, not via any resistors. I tried this in other amp (not aleph), and I think I hear better music details. Is it dangerous to try this at aleph design? Or will I get better music details?
For GRollins, I also have questions about tubes that I cannot answer. Please look at
http://www.diyaudio.com/forums/showthread.php?threadid=21349goto=newpost
about very low voltage tube preamp.
The more and more I study audio power amp, the question that raise in my mind becomes more and more silly. Some times it comes back on "how does 1 single transistor works?". I'm sure many of audio fanatics do not understand this either.
Without such knowledge (like Mr. Pass on mosfets), no one will be able to design ZEN like amp. Or aleph amp.
(I think thats why people gives Nobel Prices, the study of very basic knowledge is the key to built more complex things)
So Mr. Pass, how does 1 single mosfet works? If I put 1mA steady current or 1A steady current into a differential configuration, will the gate requirements be different? If I put the same preamp signal, will the output be different, or the music details be different? Since the mosfets aren't requiring current into its gate (like bipolar with its base with hfe figures), please tell me how it works.
lumanauw said:
That is why I occasionally insert a tutorial on how Mosfets work
into the project articles. The "A75" and "Testing Mosfets"
articles contain such sections.
It may be that you are simply imagining that a transistor is
more complicated than it is. From the inside you might be
right, but from the outside it's just a thing with 3 pins.
lumanauw said:
I fully agree about the basic knowledge as the most important key. But, understanding the basic seems to be not always easy, as one question invites another question repeatedly. For instance, I am again confused with the loop gain of the op amp, which represents the voltage gain going all the way around the circuit, "from input to output and back to input". Eeeng...... And it is difficult for me to fix a reasonable approximation to make the things simple. Eeeng.....
Hi, Mr. Pass
I'm just wondering, how old are you? And how long have you been playing around with transistors?
You are a row model for electronic hobbyist. You are clever, can make good equipments, and most important can make money of it. Your product of passlab, aleph amp are famous up to here (Indonesia). They are highly appreciated by high-end enthusiast (most of them are very critical).
Some of the home audio that use your aleph amp said to me that the sound of it getting better and better as the amp is getting hot (untouchable).
Is it possible to make a good sounding amp without excessive heat? Cold as gainclones, but sounding like aleph.
What is your principle in designing amp? High current? Heat?
I'm just wondering, how old are you? And how long have you been playing around with transistors?
You are a row model for electronic hobbyist. You are clever, can make good equipments, and most important can make money of it. Your product of passlab, aleph amp are famous up to here (Indonesia). They are highly appreciated by high-end enthusiast (most of them are very critical).
Some of the home audio that use your aleph amp said to me that the sound of it getting better and better as the amp is getting hot (untouchable).
Is it possible to make a good sounding amp without excessive heat? Cold as gainclones, but sounding like aleph.
What is your principle in designing amp? High current? Heat?
lumanauw said:
I am 52
About 35 years
The Alephs get better as they warm up - all the adjustments
are made at 45-50 deg C on the sinks, and the transconductance
of the Mosfets increases as they warm up
Yes, you can build a good sounding amp without excessive
heat, but it depends on what you mean by "good" and
"excessive".
I see that the jackals have arrived...
I mean, really, you've got to love these guys. Self-appointed experts (just ask 'em!) able to design amps with both hands tied behind their backs--and blindfolded, too. They spend hours a day online, judging from the number of posts they make, yet cannot be bothered to answer a simple query.
Until...
I post.
Then they come crawling out of the woodwork, consumed by righteous anger, convinced that they need to 'save' unsuspecting newbies from the evil Grey. Oh, lawsy me...whatever would we do without them?
Note that in all their posts (and pictures), they did not once stoop to answer any of the questions asked, only to criticise my responses. Hours, even days, have passed, and still questions hang in the air, unanswered. They can't be bothered, I suppose.
I mean, jeez, even Nelson will talk to normal people as though they're...well, normal people. What a concept.
I reckon I'll take another whack at it. Then we'll wait for more nit-picking from those too lofty to speak to the commoners directly.
lumanauw,
The resistor under the output MOSFET (the bottom one) serves three purposes. One, it provides a current sense reading for the current-limiting circuit (the lower NPN and the voltage divider at its base). Two, it gives the output MOSFET an operating environment more similar to the upper MOSFET (the Aleph current source). By making the environment more similar, the two MOSFETs behave more similarly, reducing distortion by some small fraction. (For that matter, there's some local feedback involved, which will also reduce distortion.) And most importantly, three, it tends to damp out differences between devices operating in parallel.
There may be other things as well. I'll mull it over whilst I eat supper and try to get back in here if I come up with something.
Can you operate the MOSFET without it? In theory, yes. You'll need very well matched devices and you'd have to rework the front end so that the output MOSFET would bias properly, but it could be done.
As far as making amps that sound good, but run cool...that's been the subject of much research and head-scratching over the last hundred years or so, since Lee DeForest invented the triode. Progress has been made, but it's not likely that Nelson (or anyone else, for that matter) is sitting on some secret design that will run at room temperature, yet sound exactly like real music.
Grey
P.S.: Lest any newbies get intimidated, a quick perusal of the index in Diefenderfer, Horowitz & Hill, et. al. did not uncover any reference to the phrase "loop gain," separate and distinct from negative feedback. Open loop gain and closed loop gain, I saw, but no entry for "loop gain." Maybe it's some special kind of feedback they use up there on Mount Olympus. For us mere mortals, I think it's safe to stick with the terms "open loop gain," "closed loop gain," and "negative feedback." At least until the gods decide, in their righteous wrath, to hurl lightning bolts down to Earth to smite the unbelievers. After which point, such mundane concerns will no longer matter to us.
I mean, really, you've got to love these guys. Self-appointed experts (just ask 'em!) able to design amps with both hands tied behind their backs--and blindfolded, too. They spend hours a day online, judging from the number of posts they make, yet cannot be bothered to answer a simple query.
Until...
I post.
Then they come crawling out of the woodwork, consumed by righteous anger, convinced that they need to 'save' unsuspecting newbies from the evil Grey. Oh, lawsy me...whatever would we do without them?
Note that in all their posts (and pictures), they did not once stoop to answer any of the questions asked, only to criticise my responses. Hours, even days, have passed, and still questions hang in the air, unanswered. They can't be bothered, I suppose.
I mean, jeez, even Nelson will talk to normal people as though they're...well, normal people. What a concept.
I reckon I'll take another whack at it. Then we'll wait for more nit-picking from those too lofty to speak to the commoners directly.
lumanauw,
The resistor under the output MOSFET (the bottom one) serves three purposes. One, it provides a current sense reading for the current-limiting circuit (the lower NPN and the voltage divider at its base). Two, it gives the output MOSFET an operating environment more similar to the upper MOSFET (the Aleph current source). By making the environment more similar, the two MOSFETs behave more similarly, reducing distortion by some small fraction. (For that matter, there's some local feedback involved, which will also reduce distortion.) And most importantly, three, it tends to damp out differences between devices operating in parallel.
There may be other things as well. I'll mull it over whilst I eat supper and try to get back in here if I come up with something.
Can you operate the MOSFET without it? In theory, yes. You'll need very well matched devices and you'd have to rework the front end so that the output MOSFET would bias properly, but it could be done.
As far as making amps that sound good, but run cool...that's been the subject of much research and head-scratching over the last hundred years or so, since Lee DeForest invented the triode. Progress has been made, but it's not likely that Nelson (or anyone else, for that matter) is sitting on some secret design that will run at room temperature, yet sound exactly like real music.
Grey
P.S.: Lest any newbies get intimidated, a quick perusal of the index in Diefenderfer, Horowitz & Hill, et. al. did not uncover any reference to the phrase "loop gain," separate and distinct from negative feedback. Open loop gain and closed loop gain, I saw, but no entry for "loop gain." Maybe it's some special kind of feedback they use up there on Mount Olympus. For us mere mortals, I think it's safe to stick with the terms "open loop gain," "closed loop gain," and "negative feedback." At least until the gods decide, in their righteous wrath, to hurl lightning bolts down to Earth to smite the unbelievers. After which point, such mundane concerns will no longer matter to us.
Try getting a better book
http://focus.ti.com/lit/an/slod006b/slod006b.pdf
"TI Opamps For Everyone: page 81
The quantity AB is so important that it has been given a special name: loop gain. In Figure 5-7, when the voltage inputs are grounded (current inputs are opened) and the loop is broken, the calculated gain is the loop gain, AB."
Jocko
http://focus.ti.com/lit/an/slod006b/slod006b.pdf
"TI Opamps For Everyone: page 81
The quantity AB is so important that it has been given a special name: loop gain. In Figure 5-7, when the voltage inputs are grounded (current inputs are opened) and the loop is broken, the calculated gain is the loop gain, AB."
Jocko
Loop gain
It is most surprising if loop gain is not defined in Horowitz &
Hill, since that seems to be the book everobody recommends.
Anyway, it is commonly used in app notes etc. as Jocko
pointed out, and it is defined and extensilvely used in classic
textbooks like my old Schilling & Belove from 1968. Which reminds
me of what a famous professor at MIT use to say: "Read the
literature, even if it is more than five years old".
It is most surprising if loop gain is not defined in Horowitz &
Hill, since that seems to be the book everobody recommends.
Anyway, it is commonly used in app notes etc. as Jocko
pointed out, and it is defined and extensilvely used in classic
textbooks like my old Schilling & Belove from 1968. Which reminds
me of what a famous professor at MIT use to say: "Read the
literature, even if it is more than five years old".
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.