Charles & SY, this may help. It doesn't matter that a bipolar has three terminals - since one is a current source. Both a potential divider of two resistors and an EF with resistive load have the same equation. Now if Andy is saying that the EF is a feedback system because of the form of the equation then he must also be saying that a potential divider is a feedback system. My assertion is that neither are.
Attachments
Charles
I was essentially talking about darlington "packages" but if memory serves me correct the schematic "C" just didn't do it for me musically when I was inquisitive enough to take the lid off a case and have a peek inside, to why a supposed "brilliant" amplifier was lacklustre or a inexpensive amp worked so well and way beyond my expectations. This was 20 or so years ago when I worked for Billy Vee Sound Systems in Lewisham, S.E. London UK.
I'm sure I have heard all of the configurations at one time or another but that is what I was trying to get across, the packages rather than the output configuration. Yep I have seen some seemingly wonderfully built and designed amplifiers sound just so so and some modestly built amps perform brilliantly.
one that sticks in my mind as being excellent musically was the Onyx OA21 and at a bargain price of 180 UKP and at the other extreme was something like the Musical P270 & P370 which were soundly beaten by their smaller stablemate the P170....and excellent amp if a little spiteful at the top end if badly partnered components were used. My all time favorite power amplifier was the Electrocompaniet 2 chanel power amplifier (yes the original one that used to self destruct with monotonous regularity).I honestly can't remember how they were designed but the hype and fuss around certain products always made me chuckle when we eventually sat down to listen to them at length.
I'm not attempting to fuel anything or antagonize people, just an observation.
regards
Kevin
and keep it up guys, I'm actually learning something
I was essentially talking about darlington "packages" but if memory serves me correct the schematic "C" just didn't do it for me musically when I was inquisitive enough to take the lid off a case and have a peek inside, to why a supposed "brilliant" amplifier was lacklustre or a inexpensive amp worked so well and way beyond my expectations. This was 20 or so years ago when I worked for Billy Vee Sound Systems in Lewisham, S.E. London UK.
I'm sure I have heard all of the configurations at one time or another but that is what I was trying to get across, the packages rather than the output configuration. Yep I have seen some seemingly wonderfully built and designed amplifiers sound just so so and some modestly built amps perform brilliantly.
one that sticks in my mind as being excellent musically was the Onyx OA21 and at a bargain price of 180 UKP and at the other extreme was something like the Musical P270 & P370 which were soundly beaten by their smaller stablemate the P170....and excellent amp if a little spiteful at the top end if badly partnered components were used. My all time favorite power amplifier was the Electrocompaniet 2 chanel power amplifier (yes the original one that used to self destruct with monotonous regularity).I honestly can't remember how they were designed but the hype and fuss around certain products always made me chuckle when we eventually sat down to listen to them at length.
I'm not attempting to fuel anything or antagonize people, just an observation.
regards
Kevin
and keep it up guys, I'm actually learning something
Christer
I was making a case against darlington "packages" and on memory the CFP configuration........you have surprised me with one amplifier using darlington packages and being well regarded all round, I hadn't heard of one before
.... but I digress, I'll use anything if it gets my mojo going, that is my only criteria for use all other parameters being suitable (power, reliability etc.). Anyhow I'll be listening to this thread in interest.
Kevin
I was making a case against darlington "packages" and on memory the CFP configuration........you have surprised me with one amplifier using darlington packages and being well regarded all round, I hadn't heard of one before
.... but I digress, I'll use anything if it gets my mojo going, that is my only criteria for use all other parameters being suitable (power, reliability etc.). Anyhow I'll be listening to this thread in interest.Kevin
PrimeCase said:
Hello Kevin,
Sorry if it seemed like I was picking on you, but I wanted to make a point that was only tangentially related to your post.
I figured that you were referring to Darlington "packages", because they don't perform nearly as well as two discrete devices configured in a similar way. The newer ones have gotten better, but the older ones were especially slow. Still I would never use in one of my designs, because they would create an artificial limitation.
It's really too bad, because Sanken makes some otherwise wonderful output devices that contain the bias sensing diodes on the substrate of the transistor. It's a shame that they are Darlington "packages".
But my point was there there are all kinds of circuits that can properly be called "Darlingtons". Some of them offer much higher performance for audio amplifiers than others. And so to me (at least) it seems useful to develop a vocabulary that distinguishes them. Douglas Self started on that job when he coined the terms "Type I" and "Type II". Unfortunately, they don't mean anything to someone who has read his book. It's nicer when you can get an idea of what the person is talking about because the term is self-descriptive.
Kind of like calling an amp with only local degeneration at each stage "zero feedback" -- most people understand what is meant quite readily. Others seem to prefer to be obtuse....
Cheers,
Charles Hansen
Just to clarify my thoughts regarding Darlingtons, the circuit I use in the Ayre V-5x is the one shown in my second post with pictures of Darlingtons. So I clearly think that type of Darlington has a lot to offer, but think there are a lot of problems with Darlington "packages".
(The image was taken from Marshall Leach's web site for the Leach amp, and the circuit itself was developed by Bart Locanthi, one of the great audio designers of all time. When Locanthi introduced it, it was called the "Twin-T" circuit, presumably because the marketing folks at JBL thought it sounded cool.)
(The image was taken from Marshall Leach's web site for the Leach amp, and the circuit itself was developed by Bart Locanthi, one of the great audio designers of all time. When Locanthi introduced it, it was called the "Twin-T" circuit, presumably because the marketing folks at JBL thought it sounded cool.)
You see, if people are going to condem feedback as an undesirable charateristic and others claim their circuits have no feedback, then it must be worth defining feedback.
If we define what is essentially Ohm's Law as a feedback mechanism then we're surely lost as we cannot avoid this in any circuit.
So what is feedback exactly? Or less stringently, how do we identify the type of feedback that causes sonic degradation (allegedly)?
If we define what is essentially Ohm's Law as a feedback mechanism then we're surely lost as we cannot avoid this in any circuit.
So what is feedback exactly? Or less stringently, how do we identify the type of feedback that causes sonic degradation (allegedly)?
semantics, feedback
if something is called feedback, its something that must be feed into some device > back <, what means it comes out of the devices, and is feed back in.
this means >it<, the feedback subject, must have some path outside the device the feedback is applied to. As a transistor is a single device, and not an IC, the feedback must have at least one part outside the transistor.
not scientific but semantics.
if something is called feedback, its something that must be feed into some device > back <, what means it comes out of the devices, and is feed back in.
this means >it<, the feedback subject, must have some path outside the device the feedback is applied to. As a transistor is a single device, and not an IC, the feedback must have at least one part outside the transistor.
not scientific but semantics.
till said:
This feed-back path outside the device is the load
The base current flows across the load and the drive voltage is between base and load ground, this is feedback
Obviously, no-feedback requires the drive current to be flowing in a separate circuit not connected to the load [ie: common emitter]
emitter followers - feedback?
It's been a long held belief of mine that an emitter follower, or "EF", is indeed a negative feedback circuit. Of course, "long held beliefs", mine or yours, may or may not be correct. Maybe we're arguing over semantics.
Often, we tend to think of feedback in a "parallel" sense. A feedback circuit, which may be as simple as a single resistor, is connected at one end to the output, and at the other end to the input of the amplifier stage. This is PARALLEL feedback. An emitter follower circuit has no parallel feedback. It does however, posess a feedback mechanism which I understand to be SERIES feedback. If a voltage source is the signal inputted to the base resistor of the EF, it sees an input impedance which consists of the base resistance, Rb, plus the emitter Thevenin resistance (emitter resistor in parallel with load), Re, multiplied by (hfe + 1). The total resistance determines the current in the base. If hfe increases, the tendancy would be to drive more current into the load (output), and the emitter voltage wrt ground would increase. As a result, the base to emitter voltage, Vbe, would decrease, which would decrease the output current, and voltage. Resistance placed in the emitter tends to cancel out beta, or hfe variations. This is series type negative feedback, since the input is in series with the output. Also, devices with differing hie, or input resistance, would be stabilized by the series negative feedback inherent in an EF.
Many EE's I've worked with in 26 years, with discrete circuit design, consider an EF as posesssing negative feedback. Series, not parallel. Of course, I am not an audio circuit designer, but rather an EE generalist. Maybe the audio community has different definitions of the term. So be it. Let's not engage in a flame war over this. Best regards.
It's been a long held belief of mine that an emitter follower, or "EF", is indeed a negative feedback circuit. Of course, "long held beliefs", mine or yours, may or may not be correct. Maybe we're arguing over semantics.
Often, we tend to think of feedback in a "parallel" sense. A feedback circuit, which may be as simple as a single resistor, is connected at one end to the output, and at the other end to the input of the amplifier stage. This is PARALLEL feedback. An emitter follower circuit has no parallel feedback. It does however, posess a feedback mechanism which I understand to be SERIES feedback. If a voltage source is the signal inputted to the base resistor of the EF, it sees an input impedance which consists of the base resistance, Rb, plus the emitter Thevenin resistance (emitter resistor in parallel with load), Re, multiplied by (hfe + 1). The total resistance determines the current in the base. If hfe increases, the tendancy would be to drive more current into the load (output), and the emitter voltage wrt ground would increase. As a result, the base to emitter voltage, Vbe, would decrease, which would decrease the output current, and voltage. Resistance placed in the emitter tends to cancel out beta, or hfe variations. This is series type negative feedback, since the input is in series with the output. Also, devices with differing hie, or input resistance, would be stabilized by the series negative feedback inherent in an EF.
Many EE's I've worked with in 26 years, with discrete circuit design, consider an EF as posesssing negative feedback. Series, not parallel. Of course, I am not an audio circuit designer, but rather an EE generalist. Maybe the audio community has different definitions of the term. So be it. Let's not engage in a flame war over this. Best regards.
Claude,
I don't think "series" feedback as you put it is anything other than Ohm's Law in action. It's no different from two resistors in series (as I showed earlier).
I suspect, if this thread ever defines feedback, that the "parallel" form will be more akin to it.
I initially thought this debate was worthless, but I'm now starting to realise there may be a much more significant implication of the fact that differentiating feedback from basic network theory seems so elusive.
I don't think "series" feedback as you put it is anything other than Ohm's Law in action. It's no different from two resistors in series (as I showed earlier).
I suspect, if this thread ever defines feedback, that the "parallel" form will be more akin to it.
I initially thought this debate was worthless, but I'm now starting to realise there may be a much more significant implication of the fact that differentiating feedback from basic network theory seems so elusive.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.