Hi, Mikeks,
I don't say it is a folded cascode, but it works like a folded cascode. It should be fast.
You are right, T8 is independent current source. It will always give 6mA. If T4 draws 3mA, then the 6V2zener will be passing 3mA. If T4 draws 4mA, then 6V2 zener will be passing only 2mA. Total will be always 6mA, because that is what T8 gives.
That is what I mean "like folded cascode".
I don't say it is a folded cascode, but it works like a folded cascode. It should be fast.
You are right, T8 is independent current source. It will always give 6mA. If T4 draws 3mA, then the 6V2zener will be passing 3mA. If T4 draws 4mA, then 6V2 zener will be passing only 2mA. Total will be always 6mA, because that is what T8 gives.
That is what I mean "like folded cascode".
mikeks said:Stocchino's approach to generating push-pull operation in the last stage of gain is ingenious.......at least conceptually
Nevertheless, true push-pull action in the last stage can only be guaranteed if said action also obtains in the second stage...
Current sink T6 is therefore essential...
http://www.diyaudio.com/forums/attachment.php?s=&postid=760513&stamp=1131303612
Note that the last stage of gain is bootstrapped...
viz:....the ANF current source T8 & sink T9 are intended to swing beyond their respective supplies...courtesy of the 220u capacitors and 1K54 resistors....
mikeks said:
Ok, now I follow the discussion when I see the "missing" link!
Maybe a minor thing, but I don't understand why the designer didn't use another MPSU57 instead of 1N4148 next to T5 as it is a current mirror, with now much worser matching!
What is the technical specification for this amplifier?
Cheers Michael
You're right, I just realized that the whole Rch(100),T2,T8 since the 70's is in a form of 1 resistor. CCS can be bootstrapped too
Mikeks,
Once I build something similiar, it turns out having limited bandwith (still above audio range, but cannot be very high). When I try to extend the power bandwith, it becomes not stable, oscilation. When I trace what happened, it turns out that positive swing and negative swing has a different transfer function. From input to T3/T4, everything is still symmetrical. But from T3/T4, the transfer function is different for positive half and negative half. In this schematic, the positive half is driven directly by T4.
BUT... the negative half cannot be driven directly by T3. It has to enter another transfer function, that is via T5 (current mirror with IN4148). Then it can drive negative half.
For comparison, symmetrical differential system can have power bandwith higher than this single differential+push-pull VAS (which will always have current mirror as T5+IN4148).
Am I right about different transfer function for positive and negative half?
jam said:
Not really...all one need do is insure the sink supplies enough current to drive the compensation network and subsequent stage....
Without T6, you'll not have push-pull action in the second stage, or indeed the third, as such action is comprehensively compromised by the minor loop about T4...
jam said:
Soon old chap...soon...
lumanauw said:
All amps. produce slightly different incremental foward path transfer functions....if only because each transistor's inter-electrode capacitances vary with voltage swing....
Which is why the response to square wave stimulus at rated output is recommended to confirm stability...
To answer your specific query....No....not to a significant degree...
LUMANAUW
"CCS can be bootstrapped too"
One of the first amplifiers using a constant current source had it bootstrapped, around 1970. I am absolutely unable to remember its name. Any reader knows ? Historically, it's interesting.
Bootstrapping is a very clever and conceptually fascinating technique, relatively unknown and too rarely used, which can be used in many circumstances with huge benefits : for example, you can bootstrap op-amp supplies, an emitter follower, the input resistor of an amplifier...
For those who hate negative feedback, we have to saus : it's positive feedback.
~~~~~~~ Forr
§§§
"CCS can be bootstrapped too"
One of the first amplifiers using a constant current source had it bootstrapped, around 1970. I am absolutely unable to remember its name. Any reader knows ? Historically, it's interesting.
Bootstrapping is a very clever and conceptually fascinating technique, relatively unknown and too rarely used, which can be used in many circumstances with huge benefits : for example, you can bootstrap op-amp supplies, an emitter follower, the input resistor of an amplifier...
For those who hate negative feedback, we have to saus : it's positive feedback.
~~~~~~~ Forr
§§§
forr said:
Indeed it is...only snag is, (with a purely passive TIS load), the positive feedback must be wrapped around a stage with exactly unity gain, +/-1 part per million, if the resulting impedance at the TIS output is to approach that of a simple ANF current source....
In practice, and with realistic loads, output stage gain is closer to 0.8 than unity, engendering a dispropotionately appalling loss in TIS gain.....
Bootstrapping a current source would appear to solve this problem.....
PREVIOUS POSTS
"I'm not clear. If the VAS stage is already CCS loaded, what's the point of bootstrapping it? What to achieve with bootstrapping the CCS (in VAS load position)?"
"With bootstrapping you can swing above/below the voltage supply rail helping to drive FET's fully, nothing else for CCS."
Bootsrapping consists in maintaining the voltage across a circuit (a simple resistor, a constant current source, an op-amp, etc...) almost constant. A cascode whose voltage reference of the command node (base, gate) is connected to the command node of the follower node (emitter, source) of the first device is a bootstrapped circuit.
A circuit with a constant voltage across it means no current to supply into some of its capacitors, the Cbc of the output transistor in a constant current source, for example.
~~~~~~~ Forr
§§§
"I'm not clear. If the VAS stage is already CCS loaded, what's the point of bootstrapping it? What to achieve with bootstrapping the CCS (in VAS load position)?"
"With bootstrapping you can swing above/below the voltage supply rail helping to drive FET's fully, nothing else for CCS."
Bootsrapping consists in maintaining the voltage across a circuit (a simple resistor, a constant current source, an op-amp, etc...) almost constant. A cascode whose voltage reference of the command node (base, gate) is connected to the command node of the follower node (emitter, source) of the first device is a bootstrapped circuit.
A circuit with a constant voltage across it means no current to supply into some of its capacitors, the Cbc of the output transistor in a constant current source, for example.
~~~~~~~ Forr
§§§
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.