Yes. Actually, a triode always has intrinsic local feedback because of the internal capacitance between A and G, even without the external capacitor.
Jan
I'm not concerned with semantics so I probably shouldn't have mentioned local vs global thing.
Whatever you choose to call it, the cap connected to the OT secondary works well and there is no audible instability (I don't have a scope or other sophisticated test gear) and there is absolutely no hum/noise.
I'm mostly interested in the possible installation of the switch I described above.
Whatever you choose to call it, the cap connected to the OT secondary works well and there is no audible instability (I don't have a scope or other sophisticated test gear) and there is absolutely no hum/noise.
I'm mostly interested in the possible installation of the switch I described above.
No, in the position of unbypassed cathode resistor the bypass capacitor is floating.If reconnected the charge current will give a plop.
Mona
So the switch just shouldn't be thrown while the amp is powered up.
Would the placement of the switch (on the cathode side (+) of the cap vs. the OT side (-) matter? Any advantage of one over the other?
Last edited:
In principle no, as long as the capacitor is switched out of the circuit.
But if you get a short to ground with the switch on the transformer side no harm done.On the cathode side it's harmfull for the tube.
Mona
Thanks Mona. I'm not sure if I'll implement the switch or not. I'll have to scrounge around in my junkbox to see if I even have one that would work.
The amp sounds really good with the Local NFB.
but the OPT is a passive component, we are counting active device is it not? as long as one active device is concerned, that is local feedback...
TTT149 Amplifier Local & Global Feedback Pt1 - YouTube
Watch this video see it changed you view.
Watch this video see it changed you view.
Nice video. And nobody is counting the number of stages, who came up with that nonsense? Global is, well, global, it is a simple English word, should be easy even for non-native speakers.
If it isn't global, if it isn't from all the way output back to input, it is local.
In this case, taking the feedback from the xformer secondary is global.
Think about it, if you wrongly call that local, how do you then call it from tube anode to grid? Superlocal?
I am not splitting hairs, it is not about semantics. It is about accurately describing something so that anyone anywhere in the world understands what you mean. Modifying definitions and making them internally inconsistent as well because you have this urge to add something personal to our heritage collapses the community to just you and your buddies. A huge step backward.
If you want to add to our heritage, come up with a smart circuit, but don't mess up our common understanding.
Jan
If it isn't global, if it isn't from all the way output back to input, it is local.
In this case, taking the feedback from the xformer secondary is global.
Think about it, if you wrongly call that local, how do you then call it from tube anode to grid? Superlocal?
I am not splitting hairs, it is not about semantics. It is about accurately describing something so that anyone anywhere in the world understands what you mean. Modifying definitions and making them internally inconsistent as well because you have this urge to add something personal to our heritage collapses the community to just you and your buddies. A huge step backward.
If you want to add to our heritage, come up with a smart circuit, but don't mess up our common understanding.
Jan
jan.didden,
The "intrinsic local feedback" of a triode that you talked about in post # 41, normally only occurs above audio frequencies.
As you already know, the plate to grid capacitance, times the gain, is the Miller Effect capacitance.
But most designers drive the grid from a low enough impedance, so that the Miller Effect capacitance does not cause a high frequency rolloff at audio frequencies. Otherwise, the amplifier will not be flat to 20kHz or beyond.
That means, the 'intrinsic local feedback' of the triode is effectively and essentially removed.
Some newbies might get the wrong idea about that intrinsic local feedback of all triodes, so I had to comment.
The "intrinsic local feedback" of a triode that you talked about in post # 41, normally only occurs above audio frequencies.
As you already know, the plate to grid capacitance, times the gain, is the Miller Effect capacitance.
But most designers drive the grid from a low enough impedance, so that the Miller Effect capacitance does not cause a high frequency rolloff at audio frequencies. Otherwise, the amplifier will not be flat to 20kHz or beyond.
That means, the 'intrinsic local feedback' of the triode is effectively and essentially removed.
Some newbies might get the wrong idea about that intrinsic local feedback of all triodes, so I had to comment.
Last edited:
For a small change of a triode plate voltage, the rp is relatively constant.
plate voltage change/rp = plate current change
We did not change the grid voltage, and yet the plate current changed.
Simple, it is not negative feedback.
The triode in common cathode mode inverts input to output phase (from grid to plate).
Changing phase is not negative feedback.
A transformer changes phase 180 degrees from primary to secondary (at very high frequencies).
That is not negative feedback.
plate voltage change/rp = plate current change
We did not change the grid voltage, and yet the plate current changed.
Simple, it is not negative feedback.
The triode in common cathode mode inverts input to output phase (from grid to plate).
Changing phase is not negative feedback.
A transformer changes phase 180 degrees from primary to secondary (at very high frequencies).
That is not negative feedback.
Last edited:
I have to disagree. Internal anode-to-grid feedback in a triode is always present; it is determined by the physical construction.
If you are aware of it you can indeed design your circuit such that its influence for audio frequencies can be neglected, but that doesn't remove it.
Jan
Yes, but a changed plate voltage will be transmitted to the grid through the plate-to-grid capacitance and that is negative feedback.
Jan
Exept grounded grid stage
Question global versus local feedback, somewhat confusing.
Take a two stage amp, feedback output to cathode first stage ; global.
Feedback output to cathode output tube ; local.
Now the TS removes the firststage. Without modifing the output stage the feedback suddenly changes from local to global but is still the same.
Mona
I use the term Global, to mean in the control theory sense of the word.
From DUT output to input, so that includes OPT, zobel on secondary (if used) all within the global loop.
Feedback over two stages, such as plate to cathode of a driver stage preceding it, Is a Local FB loop.
If used at the same time as Global (Output to Input) FB then that local is really a nested local FB loop within the global FB loop.
I'm not sure why there is so much debate over something that has been defined for almost a century.
It may be confusing that some FB is "near local" and others are "far local", but they're all local.
From DUT output to input, so that includes OPT, zobel on secondary (if used) all within the global loop.
Feedback over two stages, such as plate to cathode of a driver stage preceding it, Is a Local FB loop.
If used at the same time as Global (Output to Input) FB then that local is really a nested local FB loop within the global FB loop.
I'm not sure why there is so much debate over something that has been defined for almost a century.
It may be confusing that some FB is "near local" and others are "far local", but they're all local.
Last edited:
Exept grounded grid stage
Question global versus local feedback, somewhat confusing.
Take a two stage amp, feedback output to cathode first stage ; global.
Feedback output to cathode output tube ; local.
Now the TS removes the firststage. Without modifing the output stage the feedback suddenly changes from local to global but is still the same.
Mona
Well. You didn't change the actual feedback wiring but you did change the circuit big time. I find this change from local to global entirely logical and consistent.
Jan
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.