Do you mean no overall phase inversion? If so, most SS amps already do that.
If you mean no phase inversion in any stage, then that is possible but why would you want it? Why complicate things? Why restrict yourself to followers and common base/gate circuits (and their combinations e.g. LTP)?
If you mean no phase inversion in any stage, then that is possible but why would you want it? Why complicate things? Why restrict yourself to followers and common base/gate circuits (and their combinations e.g. LTP)?
No. It could be worse, as a common base/gate stage has a higher output impedance so is more affected by stray capacitance than common emitter/source.
Yes, I mean no phase inversion in any stage.
I would be interested because I have noticed that the primary determination of sound quality is the topology.
But if the CB output sees the same impedance, viz a CCS, as the CE then the parasitics would give almost the same phase shifts anyway, c'est pas?
Any other thoughts?
Hugh
I would be interested because I have noticed that the primary determination of sound quality is the topology.
But if the CB output sees the same impedance, viz a CCS, as the CE then the parasitics would give almost the same phase shifts anyway, c'est pas?
Any other thoughts?
Hugh
Member
Joined 2009
Paid Member
An amplifier with absolutely no signal inverting stages anywhere in the signal path, cannot be operated as a negative feedback control system. You can't subtract the (scaled) output from the input for global NFB. So your amplifier must needs be a member of the "zero dBs of global feedback" club.
It seems like a good way to obtain voltage gain without signal inversion might be a (follower + common base + current source load) stage, reminiscent of the input quad of the uA741 opamp IC. Since voltage headroom in power amps is extremely expensive, you probably want a combination that doesn't sacrifice any more headroom than a standard emitter follower. That probably means using depletion-mode devices (JFETs and/or depletion MOSFETs) for either the follower, or the "common base" (common gate) device, or both.
Then for output drive, a cascade of 3 emitter followers in series, with 2nd emitter pointing to the rail instead of to the output, will save another 2*VBE of headroom.
It seems like a good way to obtain voltage gain without signal inversion might be a (follower + common base + current source load) stage, reminiscent of the input quad of the uA741 opamp IC. Since voltage headroom in power amps is extremely expensive, you probably want a combination that doesn't sacrifice any more headroom than a standard emitter follower. That probably means using depletion-mode devices (JFETs and/or depletion MOSFETs) for either the follower, or the "common base" (common gate) device, or both.
Then for output drive, a cascade of 3 emitter followers in series, with 2nd emitter pointing to the rail instead of to the output, will save another 2*VBE of headroom.
An amplifier with absolutely no signal inverting stages... cannot [use] negative feedback control
Mmmm... not really. If it has an output transformer, simply using the anti-clockwise output pin of the OPT delivers inverted output for negative feedback purposes. Of course, this point of view hinges on accepting that it is a symmetric and entirely passive component. If you agree to that, then ANY transformer at any stage may act to invert phase for NFB utility. Indeed, one could ground all secondary center taps, use only the in-phase winding for an exclusively non-inverting amplification path, yet use the otherwise unused inverting secondary winding for NFB at every stage
Moreover, one could even go so far in the design to have no amplification of signal amplitude at every stage (via cathode-follower / emitter-follower / source-follower topology choice), and with interstage transformers, get all the voltage gain one might desire almost trivially. Expensively, for sure, but trivially.
GoatGuy
Mmmm... not really. If it has an output transformer, simply using the anti-clockwise output pin of the OPT delivers inverted output for negative feedback purposes. Of course, this point of view hinges on accepting that it is a symmetric and entirely passive component. If you agree to that, then ANY transformer at any stage may act to invert phase for NFB utility. Indeed, one could ground all secondary center taps, use only the in-phase winding for an exclusively non-inverting amplification path, yet use the otherwise unused inverting secondary winding for NFB at every stage
Moreover, one could even go so far in the design to have no amplification of signal amplitude at every stage (via cathode-follower / emitter-follower / source-follower topology choice), and with interstage transformers, get all the voltage gain one might desire almost trivially. Expensively, for sure, but trivially.
GoatGuy
Yes! Thank You! Many years of IC design, in which fifty more transistors costs "zero" and one transformer costs "infinity," clouded my reasoning. Back in the days of firebottles and Germanium discretes, transformers were ubiquitous.
Progress is only possible if an unreasonable man proposes an unreasonable statement.
You may be right, but you may be wrong too. Not a good idea to make categorical statements.
Transformers are probably not the answers; they have too much phase shift cf. the negative global fb of the Williamson amp of 1947 which was limited to 15dB with a high quality Partridge xformer.
There is a circuit block that does not invert yet accommodates a nfb node. A Rush cascode.
Fascinating responses, perhaps in years to come this will be possible.
Hugh
Not true. Most progress comes from reasonable people making creative leaps from what is known, not unreasonable people who may (in some cases) be ignorant of what is known.AKSA said:
True statements can be made with confidence. It happens to be true that the excess phase shift created by a stage is not directly related to whether it inverts or not, as both inverters and non-inverters have other sources of phase shift. Hence your basic premise is flawed.AKSA said:
What will become possible? A rewriting of circuit theory, so your hypothesis suddenly changes from false to true?
@DF 96....
Hugh has been giving plenty in the community farther more for a person of his magnitude he has been humble enough to allow plenty of us to bother him with many questions and often some of them stupid and always he has been giving the answers with a smile in his face ...
Allow him to be even wrong ... you will see that eventually something good might come out of it .
Just my opinion .... It sounds interesting though ...i will keep a eye
Friendly regards
Sakis
Hugh has been giving plenty in the community farther more for a person of his magnitude he has been humble enough to allow plenty of us to bother him with many questions and often some of them stupid and always he has been giving the answers with a smile in his face ...
Allow him to be even wrong ... you will see that eventually something good might come out of it .
Just my opinion .... It sounds interesting though ...i will keep a eye
Friendly regards
Sakis
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.