Charles Hansen said:
Apparently you didn't read John's original statement.
He said:
I also use Global feedback in ALL of my power amp designs and lots of it...
The emphasis on "ALL" is John's.
Well, John himself says he's uses (note he says "use" not "used") LOTS of global feedback in ALL of his power amp designs and that he's not promoting feedback-free operation.
...so I how can I be promoting feedback free operation?
se
John,
We must be the same age..... That's exactly how I feel!
My own preference is always for local feedback. At present I'm working on a power amp which uses only feedback around the driver and output stage; voltage amplification is done without any feedback except local.
Folks,
The longer you spend in audio, the more you realize no one circuit can do it all. If it's great on trumpets, deep bass might suffer; image depth, sound stage height might be affected.
This is the reason I have all but given up on the complex, math arguments; you really need to build it, listen to it, and 'voice' it with good component choices. Tragically this is almost all empirical, the more so since we must use commonly available parts.
Even if we agree on whether it is feedback or emitter degeneration, or how much, or why, or what definition precisely we should be using, will this advance the pursuit of audio? Will it be other than academically interesting? I would remind the many mathematicians present that 2 + 2 = 5 for large values of 2.
Cheers,
Hugh
We must be the same age..... That's exactly how I feel!
My own preference is always for local feedback. At present I'm working on a power amp which uses only feedback around the driver and output stage; voltage amplification is done without any feedback except local.
Folks,
The longer you spend in audio, the more you realize no one circuit can do it all. If it's great on trumpets, deep bass might suffer; image depth, sound stage height might be affected.
This is the reason I have all but given up on the complex, math arguments; you really need to build it, listen to it, and 'voice' it with good component choices. Tragically this is almost all empirical, the more so since we must use commonly available parts.
Even if we agree on whether it is feedback or emitter degeneration, or how much, or why, or what definition precisely we should be using, will this advance the pursuit of audio? Will it be other than academically interesting? I would remind the many mathematicians present that 2 + 2 = 5 for large values of 2.
Cheers,
Hugh
gm & hfe, Re, nfb
Actually Ic is stabilized against beta variations by Re. Without Re, as in an unbypassed common emitter stage, any change in beta, or hFE, will change Ic, and thus the small signal transconductance, gm, will change as well. Adding Re stabilizes Ic against beta variations. Re is providing SERIES negative feedback. After all, gm is merely small-signal collector current, ic, divided by small-signal collector to emitter voltage, vce, which can be computed as (nkT/q)Ic. Your model and equations are valid as far as predicting circuit behavior goes. They, however, do not explicitly demonstrate the series nfb present here. The EF circuit's tendency to neutralize beta AND hie (input resistance) variations is due to the presence of Re, forcing a constant Ic, which results in stable gm. The collector current Ic is hFE times base current Ib, given by Vin divided by the Thevenin input resistance which consists of Rb, plus hIE, plus (hFE + 1)*Re. In order to stabilize the circuit against hfe variation, Re must be large compared to hIE, or in the denominator, (hFE + 1)*Re >> hIE. In the numerator we have (Vin*hFE ). Since (hFE + 1)*Re >> hIE, we can neglect hIE, leaving (hFE + 1)*Re. The collector current Ic, becomes (Vin*hFE)/((hFE + 1)*Re), or approx. Vin/Re. Better accuracy is obtained if we subtract 0.65 V from Vin to account for the base-emitter forward drop. As long as hFE is reasonably large, hFE/(hFE + 1) approaches one. This is why EF circuits are stable with regard to varying beta. The transconductance gm, is only as stable as Ic, which is only as stable as (hFE + 1)*Re being greater than hIE.
Also worth noting, if an EF has no feedback, why can they exhibit local oscillations? The hfe at high enough frequencies, becomes complex, having a phase lagging component. Instead of pure real, hfe may be 29 - j40 at high enough frequency. If the load is complex, resistive and capacitive, 50 - j60 for example, the complex impedance is reflected back to the base side can have a negative real part, i.e. -900 - j3800. This is essentially a negative resistance. The result is local instability, since the degenerative or negative feedback, becomes regenerative or positive feedback producing local oscillation. To stabilize, we must cancel the negative resistance by inserting some positive resistance in the base side. Any circuit designer who has used emitter followers (I've used dozens) knows this. A no-feedback or open loop system is incapable of oscillating. It's difficult to explain without feedback, isn't it?
I can't see how anyone can deny that an EF posesses feedback. Every argument to the contrary can be refuted upon closer scrutiny. Best regards to all.
Actually Ic is stabilized against beta variations by Re. Without Re, as in an unbypassed common emitter stage, any change in beta, or hFE, will change Ic, and thus the small signal transconductance, gm, will change as well. Adding Re stabilizes Ic against beta variations. Re is providing SERIES negative feedback. After all, gm is merely small-signal collector current, ic, divided by small-signal collector to emitter voltage, vce, which can be computed as (nkT/q)Ic. Your model and equations are valid as far as predicting circuit behavior goes. They, however, do not explicitly demonstrate the series nfb present here. The EF circuit's tendency to neutralize beta AND hie (input resistance) variations is due to the presence of Re, forcing a constant Ic, which results in stable gm. The collector current Ic is hFE times base current Ib, given by Vin divided by the Thevenin input resistance which consists of Rb, plus hIE, plus (hFE + 1)*Re. In order to stabilize the circuit against hfe variation, Re must be large compared to hIE, or in the denominator, (hFE + 1)*Re >> hIE. In the numerator we have (Vin*hFE ). Since (hFE + 1)*Re >> hIE, we can neglect hIE, leaving (hFE + 1)*Re. The collector current Ic, becomes (Vin*hFE)/((hFE + 1)*Re), or approx. Vin/Re. Better accuracy is obtained if we subtract 0.65 V from Vin to account for the base-emitter forward drop. As long as hFE is reasonably large, hFE/(hFE + 1) approaches one. This is why EF circuits are stable with regard to varying beta. The transconductance gm, is only as stable as Ic, which is only as stable as (hFE + 1)*Re being greater than hIE.
Also worth noting, if an EF has no feedback, why can they exhibit local oscillations? The hfe at high enough frequencies, becomes complex, having a phase lagging component. Instead of pure real, hfe may be 29 - j40 at high enough frequency. If the load is complex, resistive and capacitive, 50 - j60 for example, the complex impedance is reflected back to the base side can have a negative real part, i.e. -900 - j3800. This is essentially a negative resistance. The result is local instability, since the degenerative or negative feedback, becomes regenerative or positive feedback producing local oscillation. To stabilize, we must cancel the negative resistance by inserting some positive resistance in the base side. Any circuit designer who has used emitter followers (I've used dozens) knows this. A no-feedback or open loop system is incapable of oscillating. It's difficult to explain without feedback, isn't it?
I can't see how anyone can deny that an EF posesses feedback. Every argument to the contrary can be refuted upon closer scrutiny. Best regards to all.
Hi Folks! Actually, I would prefer not to use feedback, BUT I can make an acceptable audio product with feedback, if it is necessary to meet mid-fi specs as well. Charles Hansen knows this and has stated it here.
Now who is MY competition? Well, Charles for one, Nelson Pass for another. YES, we really compete with each other in listening contests like the CES, for audio reviews, and in the audio marketplace. I think that we are pretty good sports about it as well. We will still talk to each other and do not publicly 'badmouth' each other, even if we don't necessarily agree on everything. This is called professionalism, folks, or something like it!
Now who is MY competition? Well, Charles for one, Nelson Pass for another. YES, we really compete with each other in listening contests like the CES, for audio reviews, and in the audio marketplace. I think that we are pretty good sports about it as well. We will still talk to each other and do not publicly 'badmouth' each other, even if we don't necessarily agree on everything. This is called professionalism, folks, or something like it!
john curl said:
Well, I know what winning is like and I know what losing is like. I prefer winning too.
Okay, I'm getting the drift here. Again, I'm just interested in building one amplifier for my self, but of course there are commercial interests at stake here. And hehehe... I'm not interested in doing my own listening tests with this project. It would be too much like having two clocks and then not knowing what time it is. It's taking long enough time to select a topology, procure and test components, and put it into a nice enclosure. I think I'm going to have to wait to try a "no feedback" design.
Thanks guys for taking the time to contribute : ).
JF
Re: Re: It's not clear...
Hello all,
Just a quick post to say that I am appalled at the general
conduct in this thread, in particular, of Steve Eddy.
It could have been very interesting, and I have a lot of
experience with design, implementation and sonic comparisons
of 0 FB circuitry, but decline to participate.
I have put in a formal complaint about SE's general conduct
here and feel pretty strongly about it. It is the type of behaviour
that I don't expect from my kids. There is a high level of personal
attack and it is directed in particular at people who have decades
of real world design and manufacturing experience. In my world
these people desrve respect however things appear somewhat
different around here.
I stand by this post and won't be caught into any further
arguments over it, however I will be happy to talk to moderators
about any issues.
Terry Demol
Hello all,
Just a quick post to say that I am appalled at the general
conduct in this thread, in particular, of Steve Eddy.
It could have been very interesting, and I have a lot of
experience with design, implementation and sonic comparisons
of 0 FB circuitry, but decline to participate.
I have put in a formal complaint about SE's general conduct
here and feel pretty strongly about it. It is the type of behaviour
that I don't expect from my kids. There is a high level of personal
attack and it is directed in particular at people who have decades
of real world design and manufacturing experience. In my world
these people desrve respect however things appear somewhat
different around here.
I stand by this post and won't be caught into any further
arguments over it, however I will be happy to talk to moderators
about any issues.
Terry Demol
Believe me, Terry, it's twice as painful for me- I'm friends with both those guys.
But Steve's technical point still stands, as far as I can see- it doesn't seem possible to make a zero feedback amp, only amps that eliminate particular feedback topologies. And John basically agrees with that, too. Too bad this got lost in the noise of the food fight.
I hope you can shake your head, chuckle, then contribute your thoughts.
But Steve's technical point still stands, as far as I can see- it doesn't seem possible to make a zero feedback amp, only amps that eliminate particular feedback topologies. And John basically agrees with that, too. Too bad this got lost in the noise of the food fight.
I hope you can shake your head, chuckle, then contribute your thoughts.
SY said:
I'm very sure that the starter of the original thread meant this.
By the way, welcome to the childrens garden.
This is like cooking without salt.
What if there is already salt in the meat ?
There is even salt in mineral water.
Everybody will agree upon that cooking without salt means cooking without throwing any salt into the pot.
But if somebody wanna split hairs there will never be cooking without salt.
Back to the usual tricks
From you know who.
""No, John, "global" isn't a third group of feedback with respect to series or parallel."
There are two types of feedback. Degenerative and regenerative.
That's it.
After that it's simply a matter of implementation.
It can be implemented as a series element, a parallel element or a combination.
It can be implemented locally in a single stage, less locally across two or more stages, or globally around the entire circuit.
But it's all still feedback."
Lecturing John Curl on feedback
Simply a matter of implementation. I don't think any one who designs amplifiers would call it simple........... and in fact is the what makes amplifier design challenging.
Don't you have anything to do but persecute Mr. Curl
I would rather him devote his time to useful advice on the forum, and not to seeing what you are slandering him over today. It is really pathological and you haven't damaged his credibility or reputation. I don't have a clue what the point is, do you?
Trying to increase sales for the company you work for, or become famous in the audio industry? I don't think it's working..........
From you know who.
""No, John, "global" isn't a third group of feedback with respect to series or parallel."
There are two types of feedback. Degenerative and regenerative.
That's it.
After that it's simply a matter of implementation.
It can be implemented as a series element, a parallel element or a combination.
It can be implemented locally in a single stage, less locally across two or more stages, or globally around the entire circuit.
But it's all still feedback."
Lecturing John Curl on feedback
Simply a matter of implementation. I don't think any one who designs amplifiers would call it simple........... and in fact is the what makes amplifier design challenging.
Don't you have anything to do but persecute Mr. Curl
I would rather him devote his time to useful advice on the forum, and not to seeing what you are slandering him over today. It is really pathological and you haven't damaged his credibility or reputation. I don't have a clue what the point is, do you?
Trying to increase sales for the company you work for, or become famous in the audio industry? I don't think it's working..........
In brief...
...too much effort. I don't think my ears are up to the possible subtleties involved--not that my ears are bad that I know of...
Well, I think it is possible to make a better sounding amp by applying hints and techniques from others. Global feedback remains an aspect of my project that I still wonder about.
(Now off to make some enchiladas : ).
JF
Steve Eddy said:
...too much effort. I don't think my ears are up to the possible subtleties involved--not that my ears are bad that I know of...
Steve Eddy said:
Well, I think it is possible to make a better sounding amp by applying hints and techniques from others. Global feedback remains an aspect of my project that I still wonder about.
(Now off to make some enchiladas : ).
JF
Though perhaps a minority opinion...
In Steve's defense--not that he needs it from me--I think he does a pretty good job of making the issues technical. Also, when attacked personally, he seems to take it much less personal than the average Joe IMHO. On the other hand, his questions are rather pointed... Perhaps a valuable aspect of a public forum...
JF
In Steve's defense--not that he needs it from me--I think he does a pretty good job of making the issues technical. Also, when attacked personally, he seems to take it much less personal than the average Joe IMHO. On the other hand, his questions are rather pointed... Perhaps a valuable aspect of a public forum...
JF
I'll say! Some of us (I include myself here) are not great at being diplomatic or sparing feelings when gnawing on the bone of technological understanding.
Coincidently, I'm a moderator-type on an advanced cooking forum. It ain't as simple as that. Low-sodium diets, a necessity for many with hypertension, involve awareness of the inherent sodium content of the ingredients. There's an analogy here....
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.