Hello Charles,
thank you for your reply. I completely agree with you that it would be necessary to compare both amps in the same audio chain. I was exhibiting my preamp with Wilson's, but I was not able to take the AYRE to the same audio chain. I really appreciate your frankly words about possible non-feedback circuit specific properties. On the other hand I was greatly surprised by Ayre's ability to play jazz music (to "bring the feeling") etc. Nice experience.
Best regards,
Pavel
thank you for your reply. I completely agree with you that it would be necessary to compare both amps in the same audio chain. I was exhibiting my preamp with Wilson's, but I was not able to take the AYRE to the same audio chain. I really appreciate your frankly words about possible non-feedback circuit specific properties. On the other hand I was greatly surprised by Ayre's ability to play jazz music (to "bring the feeling") etc. Nice experience.
Best regards,
Pavel
It's also been my experience that "open loop" solid-state amplifiers tend to sound nicer (slightly sweeter, richer, more relaxed) than feedback types, but sometimes less "precise" with complex music in terms of imaging and detail. I forgave this shortcoming after realizing that, whether because of or in spite of it, these amps IMO sound more like the real thing.PMA said:
Younger readers may not be aware of this, but the "open loop" design concept has been around and in commercial use for over 20 years in successful solid-state products from Jeff Rowland and Threshold as well as in the more recent offerings from Ayre, BAT, and Pass among others. (Not to be US-centric, but that's mostly what I've been exposed to).
So, as far as I'm concerned, the concept is well proven. This isn't to say that solid-state amps using global feedback can't sound good, but my feeling is that if you start off with an intrinsically linear topology with a substantial bias on the output stage, you may well be better off going "open loop." A lot depends on how well the circuit tolerates global NFB. If you can't use a hot-running output stage, first make the circuit as simple and fast as possible, then apply global NFB.
I've learned a lot from DIY experience, and think that the best advice for the strongly motivated is to build an amp that can tolerate different kinds of NFB and try them for yourself.
PMA said:I will add few details. It seems that it is quite useless to load speakers with frequencies and rates of voltage rise that they are unable to radiate. Following this rule you will obtain non-aggressive, smooth, pleasant and natural sound. Specs looks worse, but sound is much better.
What is the high frequency limit that you use?
What is the slew rate?
Typical THD?
Where do you apply this high frequency limit in the amp (feedback loop, miller caps, increase open loop gain, ...)?
Leach paper describes splitting the feedback into two paths: Output of amp and driver stage. See http://users.ece.gatech.edu/~mleach/lowtim/fdbk.html. Do you use a smililar approach?
Sorry for all these questons but I am curious...
Thanks
response to Ultima Thule question
The difference is plenty, almost 2 to 1. With a diode, the current drawn from the input source equals the current in the load resistor. The input source is loaded by an impedance equal to that of the diode plus the resistor. With a transistor having hfe = 1, the input source only supplies half of the current in the emitter resistor load, with the collector current supplying the other half. In this case the input source is loaded by the diode impedance plus twice the emitter resistance. Since hfe = 1, the collector current equals the base current, but the emitter current equals the base current multiplied by (1 + hfe).
Pretty easy question. What was your point in asking it, if you don't mind? Best regards.
The difference is plenty, almost 2 to 1. With a diode, the current drawn from the input source equals the current in the load resistor. The input source is loaded by an impedance equal to that of the diode plus the resistor. With a transistor having hfe = 1, the input source only supplies half of the current in the emitter resistor load, with the collector current supplying the other half. In this case the input source is loaded by the diode impedance plus twice the emitter resistance. Since hfe = 1, the collector current equals the base current, but the emitter current equals the base current multiplied by (1 + hfe).
Pretty easy question. What was your point in asking it, if you don't mind? Best regards.
Claude,
that was over 2 month ago and I wrote in the context of feedback when people started to argue wethere there are feedback in a Emitter follower, but you are quite right about your point.
I realize now, whe reading your post, that I wasn't that accurate with my point when stating hfe = 1.
I could and should have said hfe = 0,0001 (or so, a value that small so it could be said it's of insignificant value in the context of a transistors significance as an EF, but still a transistor...)
Or, I could have asked: What is the difference between a transistor with B & C tied together and a diode if viewed from an exteranal point of view in the context of feedback.
I tried to rise peoples attention and give a new view about "feedback" in a single component in a time when some claimed that an amplifier with ANY emitter follower is not a "pure non-feedback" amplifier, but I was surpised nobody payed any attention to it....or?
Put the transistor with tied B & C in a black box with two legs and the diode in a black box and it's hard if not imposible to distinguish them apart from each other.
So my point was... IF there's feedback in a emitter follower then there's as much feedback in a diode....but I may be wrong... I just wanted to show the small and ridicoulus significance of arguing a single component having claimed feedback and thereby disqualifie an amplifier as an non-feedback such.
Someone said just for some time ago that there's also feedback in a CE transistor stage, maybe we can't then build non-feedback amplifiers at all with semiconductors.... tube dudes(at least one!) would say they can....I don't know enough about tubes so I can't answer that question....
Well, it was an interesting thread, but who cares..... and it was then....or...
Cheers
that was over 2 month ago and I wrote in the context of feedback when people started to argue wethere there are feedback in a Emitter follower, but you are quite right about your point.
I realize now, whe reading your post, that I wasn't that accurate with my point when stating hfe = 1.
I could and should have said hfe = 0,0001 (or so, a value that small so it could be said it's of insignificant value in the context of a transistors significance as an EF, but still a transistor...)
Or, I could have asked: What is the difference between a transistor with B & C tied together and a diode if viewed from an exteranal point of view in the context of feedback.
I tried to rise peoples attention and give a new view about "feedback" in a single component in a time when some claimed that an amplifier with ANY emitter follower is not a "pure non-feedback" amplifier, but I was surpised nobody payed any attention to it....or?
Put the transistor with tied B & C in a black box with two legs and the diode in a black box and it's hard if not imposible to distinguish them apart from each other.
So my point was... IF there's feedback in a emitter follower then there's as much feedback in a diode....but I may be wrong... I just wanted to show the small and ridicoulus significance of arguing a single component having claimed feedback and thereby disqualifie an amplifier as an non-feedback such.
Someone said just for some time ago that there's also feedback in a CE transistor stage, maybe we can't then build non-feedback amplifiers at all with semiconductors.... tube dudes(at least one!) would say they can....I don't know enough about tubes so I can't answer that question....
Well, it was an interesting thread, but who cares..... and it was then....or...
Cheers
Well I for one argued exactly the same thing. Except I used a resistor for the comparison. If one claims a transistor is a feedback device then one also has to claim a resistor is a feedback device and therefore everything is a feedback device!
Dmitri - your argument applies to a resistor: the output current depends upon voltage difference across the terminals!
The fact that an output current depends upon a voltage difference DOES NOT IMPLY FEEDBACK!
So we are back to the discussion if an emitter follower is a circuit with feedback or not 
I don't think this was the aim of the thread starter, and from my point of view it would be much more interesting if a discussion about circuits would start, as Charles Hansen pointed out.
Couldn’t it be the following agreement:
The circuit has feedback, if the feedback loop can be opened on the hardware and the open loop gain as well as the closed loop gain can be measured at the same circuit.
And if this cannot be done on a circuit, it would go through as a non feedback design (i.e. an emitter follower).
Obviously, integrated circuits with integrated feedback loop or unknown internal circuit are prohibited
Sorry for my bad English
, I hope my point is nevertheless understood.
Regards
Tino
I don't think this was the aim of the thread starter, and from my point of view it would be much more interesting if a discussion about circuits would start, as Charles Hansen pointed out.
Couldn’t it be the following agreement:
The circuit has feedback, if the feedback loop can be opened on the hardware and the open loop gain as well as the closed loop gain can be measured at the same circuit.
And if this cannot be done on a circuit, it would go through as a non feedback design (i.e. an emitter follower).
Obviously, integrated circuits with integrated feedback loop or unknown internal circuit are prohibited
Sorry for my bad English
, I hope my point is nevertheless understood.Regards
Tino
I should think it's clear by now that armchair pendants would rather argue than build circuits.
Some people talk about the fish that got away, others get up off their butts and go catch 'em.
For my part, I'm content to treat followers et. al. as black boxes and design circuits around them. My question at that point devolves to: Are there explicit feedback connections between the black boxes? It's a simple yes or no answer. If the answer is no, I consider it to be a no-feedback circuit.
Charles Hansen was kind enough to offer a circuit earlier. Why don't you use that as a starting point and go on?
Grey
Some people talk about the fish that got away, others get up off their butts and go catch 'em.
For my part, I'm content to treat followers et. al. as black boxes and design circuits around them. My question at that point devolves to: Are there explicit feedback connections between the black boxes? It's a simple yes or no answer. If the answer is no, I consider it to be a no-feedback circuit.
Charles Hansen was kind enough to offer a circuit earlier. Why don't you use that as a starting point and go on?
Grey
Well I would like to thank dimitri for his post.
He is the first person to explain it in a way that sheds some light on the issue for me. I will need to think some more on it but I can now intuitively see how an emitter follower could oscillate.
However I still think is a very different animal than multistage o/p to i/p FB.
mike
He is the first person to explain it in a way that sheds some light on the issue for me. I will need to think some more on it but I can now intuitively see how an emitter follower could oscillate.
However I still think is a very different animal than multistage o/p to i/p FB.
mike
Moderator hat on 
Hi,
IMHO splitting the feedback issue into circuits that has a feedback topology intentionally and devices with no intentional feedback components around is a non-issue. All active devices have internal feedback mechanisms, even a single triode has anode-grid feedback. The difference is that feedback internal to a device has usually much higher loop bandwidth but is still feedback, as the oscillating emitter follower (and source follower/cathode follower) proves.
It makes more sense IMHO to split into overall feedback and local feedback schemes.
Cheers
IMHO splitting the feedback issue into circuits that has a feedback topology intentionally and devices with no intentional feedback components around is a non-issue. All active devices have internal feedback mechanisms, even a single triode has anode-grid feedback. The difference is that feedback internal to a device has usually much higher loop bandwidth but is still feedback, as the oscillating emitter follower (and source follower/cathode follower) proves.
It makes more sense IMHO to split into overall feedback and local feedback schemes.
Cheers
Pjotr said:Hi,
IMHO splitting the feedback issue into circuits that has a feedback topology intentionally and devices with no intentional feedback components around is a non-issue. All active devices have internal feedback mechanisms, even a single triode has anode-grid feedback. The difference is that feedback internal to a device has usually much higher loop bandwidth but is still feedback, as the oscillating emitter follower (and source follower/cathode follower) proves.
It makes more sense IMHO to split into overall feedback and local feedback schemes.
Cheers
Pjotr,
/philosophy mode on
What continues to amaze me is that people try to break everything in pieces and put it in different boxes. I always think that you can understand things better from a wide perspective. When you realise that feedback can be internal AND external, intended AND unintended, it increases your understanding.
If you realise that amplifiers, car driving, even breathing, can only happen because of negative feedback, doesn't that increase your understanding of the PRINCIPLE? And doesn't better understanding of the principle not allow you better understanding in specific cases?
/philosophy mode off
Jan Didden
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.