CFB
I haven't read it.
Hi Stein,
You're right about the definition of "current-feedback" in old school text books. But nowadays it makes little sense to stubbornly adhere to the old definitions. The paper below differentiates between four basic FB topologies:
1. Series-shunt
2. Shunt-series
3. Series-series
4. Shunt-shunt
I think this nomenclature is less confusing.
Cheers,
E.
I haven't read it.
Hi Stein,
You're right about the definition of "current-feedback" in old school text books. But nowadays it makes little sense to stubbornly adhere to the old definitions. The paper below differentiates between four basic FB topologies:
1. Series-shunt
2. Shunt-series
3. Series-series
4. Shunt-shunt
I think this nomenclature is less confusing.
Cheers,
E.
How to lower distortion -
Ciruits which present devices with virtual ground low impedance nodes are now generally becoming labelled as current-mode circuits.
How to lower distortion ---
Fundamentaly, volts and currents are associated to electric charge. Driving charge into high Z nodes will result in large voltage excursion. While charge driven into low Z nodes will result in small voltage excursions.
Thx-RNMarsh
Ciruits which present devices with virtual ground low impedance nodes are now generally becoming labelled as current-mode circuits.
How to lower distortion ---
Fundamentaly, volts and currents are associated to electric charge. Driving charge into high Z nodes will result in large voltage excursion. While charge driven into low Z nodes will result in small voltage excursions.
Thx-RNMarsh
Last edited:
The impedance of a node to which the feedback is returned does not determine the type of feedback applied.
It's just extraordinary how difficult it is to make simple points of fact on this forum.
So untrue, VSSA 1 mV DC stable without any servo, PSSR 100 dB.
Afcourse with the better components these days and thermal coupling, matching yes it can be done,.
PSST 100 dB, it is simple these days also, components are of high quality, but if not done thermal coupling offset go run, however I have play a current feedback amp without servo some years ago, it did walk between 2 and 90 mv so that was oke, some audio junkies want 1 uV or so before she consider there amp high end.
A servo is a nice tool get amp balanced, I have now a double mu stage hybride and this have current feedback to the first tube cathode, and yes this need a servo who act also as a tube age corrector.
Current feedback make the amp bandwith go in the megaherz sometimes so a taming is needed otherwise we have a receiver sometimes..
I am not as a calculating man, I make combinations with amps and a simulator.
Nice schematics from marsh, I did use this as a amp in a amplifier a la Alexander with the discrete opamp like marse but come not further the simulating.
For current feedback, it because of the low impedance node, make voltage a current, and if node is high impedance it get voltage, if I read internet it seems that
much people don,t understand current feedback, honestly I do also not now all, only that it is wat the word say, currentfeedback and it makes the amp independent
of bandwidth of open loop, therefore you see this current feedback a lot in video opamps and such, even opamps for medical purposis like hartbeat amplifiers.
regards
Last edited:
Just being practicle now --
See first sentence of line #564 above, again -->
Unfortunately, right or wrong doesnt matter any more, others have decided what to call the topology and all IC makers use that meaning of the term. Its locked in now. In this case majority rules. So I call it by what many others now call it... current-mode feedback. Its just the explanation of how it works that you can aurgue but even that has been concluded and accepted by designers, engineers and manufacturers, alike. World wide.
Thx-RNMarsh
The impedance of a node to which the feedback is returned does not determine the type of feedback applied.
It's just extraordinary how difficult it is to make simple points of fact on this forum.
See first sentence of line #564 above, again -->
Unfortunately, right or wrong doesnt matter any more, others have decided what to call the topology and all IC makers use that meaning of the term. Its locked in now. In this case majority rules. So I call it by what many others now call it... current-mode feedback. Its just the explanation of how it works that you can aurgue but even that has been concluded and accepted by designers, engineers and manufacturers, alike. World wide.
Thx-RNMarsh
Last edited:
The text on page 14 says: "Feedback from the output to the invering input acts to force the inverting input current to zero, hence the term current feedback"
This is completely untrue. The input stage is a push-pull class AB complimentary arrangement that will either source current to or sink current from the feedback network regardless of the feedback and dependent on the polarity of the input voltage.
In other words, the feedback doesn't act to " force the inverting input current to zero..."
Rather, completely consistent with orthodox feedback control theory, the voltage divider that is the feedback network, connected in parallel with the output, drops a fraction of the output voltage across the inverting input. This is then subtracted from the input voltage to produce the error signal which drives the amplifier's forward path.
Last edited:
Not true necessarily: the voltage developed depends on the current supplied to the node, regardless of its impedance.
There is no such thing as "voltage sampling current feedback".
The nearest thing to what you've described is shunt (voltage) derived-shunt (current) applied negative feedback or transimpedance feedback.
However, I am positive that what you're really refering to is shunt (voltage) derived-series (voltage) applied negative feedback or voltage feedback. This should never be described as current feedback.
The truth is sacrosanct regardless of whether others choose to believe a lie.
Last edited:
The same questions are asked of app engineers such as Erik Barnes of Analog Devices in a column called Ask the Applications Engineer -#22.
I would refer interested parties to read his answere and comment to him.
www.analog.com/library/analogDialog/Anniversary/22.html
Thx-RNMarsh
I would refer interested parties to read his answere and comment to him.
www.analog.com/library/analogDialog/Anniversary/22.html
Thx-RNMarsh
What if --- a voltage is 'sampled' and applied to a low/zero Z Ohm port?
If the port Z is virtual or not, the port Z behaves as if it was zero. Only current flows. Thus the name. You are making this WAY more complicated than it is. Who cares? ... shunt/series/voltage/current. The formulas for gain et al are different for the applications engineer.
The first paragraph of the web site refered to above says it about as clearly as it is going to get. The rest is important in describing the effective differences between VFB and CFB and the resulting differences.
Thx-RNMarsh
Last edited:
So untrue, VSSA 1 mV DC stable without any servo, PSSR 100 dB.
You never support your claim with any measurement or at least simulation.
Here is my simulation of the VSSA PSSR.
BR Damir
Attachments
dadod and his simulations pics
Tell me in db what's the ratio betwween 1 Vrms and 10 uVrms. FLUKE doesn't lie, your sims worth nothing, total BS.
P.S. Output DC offset is stable at +1 mV
dadod and his simulations pics
Tell me in db what's the ratio betwween 1 Vrms and 10 uVrms. FLUKE doesn't lie, your sims worth nothing, total BS.
P.S. Output DC offset is stable at +1 mV
As I expected, no any argument accepted.
P.S. I don't say that this amp is not good, but you behave as a salesman.
Well, I don't accept your cartoons, you don't accept reality measurements from FLUKE. Who's right?
You could be more careful with words, for you cartoons, for others technical arguments.
By the way could you explain in more details how did you measure the PSSR with "FLUKE"?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.