Cop out, you can guess by inspection and I know it. Keep flailing behind the curtain.
Somewhere around 1500 to 1800 Ohms depending on the transconductance of the input after the bias trim. Oh yes it changes 10% or so with the 1Meg resistors. I used a gm of 5mS.
PMA (anybody)use your sim, stuff DC current into the high Z node and see. jcx did.
Last edited:
I realize that this may not make things more clear, but Thorsten has brought up something that has been addressed in the JC-3 to some extent. The type of compensation used in the JC-3, is called Miller Feedback compensation. One advantage of this kind of compensation is that it provides a lower drive impedance at high frequencies due to the feedback effect of the comp. cap(s) (highly linear only, please) that help to remove the problem of nonlinear caps in the output driver stage and also pads the intrinsic non-linear capacitance of Ccb in the voltage drivers, to make it essentially more linear. Everything is important.
I think there is much said here, without looking at the JC-3 for what it is, and why it was designed that way.
I think there is much said here, without looking at the JC-3 for what it is, and why it was designed that way.
OK let's walk through it, two equations is all you need. You try and stuff current into the high Z node. This current is supplied by the input stage (I = gm*Vin). But by inspection Vin = Vo*(R2/(R1+R2)) where R2 is the input resistor and R1 is the feedback resistor. The output resistance is (R1+R2)/gm*R2 or 1800 Ohms where R1 = 80k and R2 = 10k and gm = 5mS. There are no "real" missing terms that make a SIGNIFICANT difference.
EDIT - Add the 500k resistor you get 1580 Ohms. Night and day you decide.
EDIT - Add the 500k resistor you get 1580 Ohms. Night and day you decide.
Last edited:
There is a difference in the circuit initially used to make the line driver for the JC-2 and to convert it to the input stage for the JC-3. One of the changes was removal of the extra compensation caps that were originally in the JC-2. It just speeds up the amplifier a little bit, but I don't remember the details. This does make the compensation cap(s) that are inherent in the Ccb of each driver, not padded with a parallel linear cap, so they are slightly more non-linear. However they are summed together and are working in opposite voltage, so there should be first order cancellation of the voltage dependent total capacitance.
I am eagerly awaiting a Letter to the Editor
jan
Sorry I'm late to this "party". This is from 25 pages ago or so...
Gee, Jan, you sure didn't try to argue with Malcolm Hawksford about this exact same topic when you interviewed him for AudioXpress a couple of years ago. Malcolm said:
"The sharp switching
edges from the DAC output can only
be perfectly reproduced with an I/V op
amp that has infinite bandwidth and no
limit on slew-rate. Any practical circuit
will have nonlinearity and slew rate limits
such that a transient input signal can
slightly modulate the open-loop (OL)
transfer of the op amp.
Modulating the OL transfer function
means you modulate the circuit’s
closed-loop (CL) phase shift. What is
interesting is that it looks remarkably
similar to correlated jitter; they share
a family resemblance. It also is
similar to what people have been talking
about as dynamic-phase modulation in
amplifiers. Whenever an amplifier stage
needs to respond very quickly, it tends
to run closer to open loop and therefore
is more susceptible to open loop nonlinearity."
You didn't argue with him then. Why argue with Nelson now?
Charles are you seriously coming back after all this time to try to find out why I didn't do something that you think I should have done three years ago?
Wow! I'm flattered!
But, to answer the question, I don't see what I should have argued about. Malcolm made a factual statement that as far as I can see is correct (if he indeed did say what you quoted; I haven't checked it). What point should I have argued?
I also don't get the reference to Nelson; that discussion was about the 'sameness' or not, of degeneration and loop feedback, and related to Bruno Putzeys article in Linear Audio Vol 1. Letters to the Editor, commenting on articles, are almost always enlightening; you should send one!
jan
Wow! I'm flattered!
But, to answer the question, I don't see what I should have argued about. Malcolm made a factual statement that as far as I can see is correct (if he indeed did say what you quoted; I haven't checked it). What point should I have argued?
I also don't get the reference to Nelson; that discussion was about the 'sameness' or not, of degeneration and loop feedback, and related to Bruno Putzeys article in Linear Audio Vol 1. Letters to the Editor, commenting on articles, are almost always enlightening; you should send one!
jan
I still think the circuit is fine. The number I picked for gm was arbitrary the actual JC-3 gm might be much higher, the larger it is the lower the impedance anyway.
Last edited:
Why does the amplifier have to see the edges? There are ways to prevent that.
(All the way back on page 1580.)
You have got to be kidding me!
This is supposed to "prove" something? By making an artificial mathematical construct that is equal to the resistor (within the artificial mathematical construct that the circuit simulator represents)?
It's as big of a waste of time as those parlor tricks where you ask someone to take a number, multiply it by something, subtract something, divide by something, multiply it again, and end up with the original number. Not impressed...
(Again, back from page 1580.)
Two roads diverged in a wood and I - I took the one less traveled by, and that has made all the difference.
Robert Frost
The AD844 (and a couple of long discontinued Burr-Brown parts designed by their long-disbanded German team, that had +/-5 volt rail limits) are the only CFB op-amps that had a pin connected at what Dr. Gilbert referred to as the "transconductance node".
If only more companies would spend the $0.001 per part to add another wirebond lead and connect it to an unused pin on the leadframe, these parts would be ever so more versatile.
For many application-specific parts, absolutely yes.
But when National made their series of "audio-specific" parts and left out the connection, it was a crime.
I believe the argument at that point was that degeneration isn't equivalent to negative feedback - with the "evidence" being that Nelson's 2 different example circuits, applying feedback differently didn't have the "the same" distortion spectrum
so I tried to show that Nelson's degenerated common source amplifier example fits in the "Output Current sensing, Input Voltage feedback" scheme of negative feedback loop classification as shown in EE textbooks
by using "spice tricks" I sensed the output current proxy (the V across the Drain R, closely equal to the source current "sensed" by the degeneration R), feed it back as a V that subtracts from the input V
when you do that the spicy circuit shows the same input Z, output and distortion spectrum to within the limits of the approximations - certainly good enough for audio frequencies
clearly when "loop gain" falls at high frequency there are differences between the spicy "thought experiment" and degeneration
but where "loop gain" is large I showed you can in principle get Blackman's Z transformations, Baxandal “harmonic multiplication” - with the same spectrum to a very fine level of detail
I, and EE textbooks claim this shows that degeneration is a classic form of negative feedback
clearly some here design at a very high level without knowing these formalisms - but I don't see how that gives them a "free pass" to make specific claims about what feedback in electronic amplifiers is, does in ignorance of and contradicting professional EE understanding, definitions as taught for over 1/2 a century now in University programs as “textbook” knowledge at the undergrad level
on the DAC I/V - I think the call for high GBW in op amp I/V can be found in Iout DAC hybrid/discrete module app notes - as well as the "insight " that we generally don't want to reproduce the switching edges but rather a properly anti-imaging lowpass filtered output
by using a lead C in the feedback the "problem" then changes to needing a high current slew rate - and low Z Class A BJT Q buffers inside the op amp loop are also 30-40 years old DAC I/V techniques
so I tried to show that Nelson's degenerated common source amplifier example fits in the "Output Current sensing, Input Voltage feedback" scheme of negative feedback loop classification as shown in EE textbooks
by using "spice tricks" I sensed the output current proxy (the V across the Drain R, closely equal to the source current "sensed" by the degeneration R), feed it back as a V that subtracts from the input V
when you do that the spicy circuit shows the same input Z, output and distortion spectrum to within the limits of the approximations - certainly good enough for audio frequencies
clearly when "loop gain" falls at high frequency there are differences between the spicy "thought experiment" and degeneration
but where "loop gain" is large I showed you can in principle get Blackman's Z transformations, Baxandal “harmonic multiplication” - with the same spectrum to a very fine level of detail
I, and EE textbooks claim this shows that degeneration is a classic form of negative feedback
clearly some here design at a very high level without knowing these formalisms - but I don't see how that gives them a "free pass" to make specific claims about what feedback in electronic amplifiers is, does in ignorance of and contradicting professional EE understanding, definitions as taught for over 1/2 a century now in University programs as “textbook” knowledge at the undergrad level
on the DAC I/V - I think the call for high GBW in op amp I/V can be found in Iout DAC hybrid/discrete module app notes - as well as the "insight " that we generally don't want to reproduce the switching edges but rather a properly anti-imaging lowpass filtered output
by using a lead C in the feedback the "problem" then changes to needing a high current slew rate - and low Z Class A BJT Q buffers inside the op amp loop are also 30-40 years old DAC I/V techniques
In order:
1) Agreed, in the sense that in the cases of our better linear circuits the
additional harmonics are down a lot farther than the Baxandall example.
The graph trotted out for that shows the 3rd as about 11 or 12 dB down
from the 2nd, but in real examples, 30 dB is more like it. Hard to get
excited about.
2) OK, and back to 1)
3) OK
4) Nobody is claiming that degeneration is not a form of feedback, I
just think that for clarity in these discussions it is appropriate to identify the
type of feedback applied.
(From page 1595.)
“Those who don't know history are destined to repeat it.” -- Edmund Burke
It's been over 30 years since Erno Borbely gave a tutorial on how to design a 100 watt DC power amplifier in The Audio Amateur. It just requires attention to detail. Something that some people on this forum seem to lack.
- Status
- Not open for further replies.