Large electrolytics that need attention regarding quality, cost money, and will detoriate over time.
Great achievement.
👍right!
We are in 2024. You can built amplifiers that have no capacitors in the signal path, except the ones for stability.
Sometimes, it is more fun to polish a turd than a Blameless one.
Thank goodness the MF-A1 is a real audio amplifier!
A slightly higher quiescent current, a modern BJT on the output side (as selected in posting #1 by "Zen Mod"), a little more GNFB ... and the existing A1 is already significantly improved (?).
In other words: it gets a personal, very individual touch.
The initial idea of using a MOS-FET as an interface (or bridge) between the two stages is also not a stupid idea, but a charming one.
Since the usual cries for THD values below the 0.0001% mark under full load could be heard again, there is no other choice but to drastically increase the OLG - and in the present case (the version M0rten is kindly working on) this is exorbitant. Which compensation scheme is the only valid and truthful one, I deliberately leave to the individual expert.
Perhaps the user Bernhard could please explain his understanding of the "signal path" or define the signal path in a feedback system.
Personally,
I would like to know which companies Bernhard works for as a consultant designer. A confidential PM would suffice for me.
In future, I would welcome it if the usual destructive act were no longer reeled off.
The last three (or five) pages of this thread were not constructive.
Dear Bernhard,
do good - and explain in your own words the well-known but obviously not understood stability criterion.
Or hold back. Be constructive (by the way, you only suspect instability; and you can always bring that up again, again and again ... 😉 I didn't expect anything else. )
Best wishes to Bavaria,
HBt.
Psst
And now please continue with the next step, a first PCB.
I think, jxdking and myself are happy to be in the boat and will continue to participate constructively in this task.
A slightly higher quiescent current, a modern BJT on the output side (as selected in posting #1 by "Zen Mod"), a little more GNFB ... and the existing A1 is already significantly improved (?).
In other words: it gets a personal, very individual touch.
The initial idea of using a MOS-FET as an interface (or bridge) between the two stages is also not a stupid idea, but a charming one.
Since the usual cries for THD values below the 0.0001% mark under full load could be heard again, there is no other choice but to drastically increase the OLG - and in the present case (the version M0rten is kindly working on) this is exorbitant. Which compensation scheme is the only valid and truthful one, I deliberately leave to the individual expert.
Perhaps the user Bernhard could please explain his understanding of the "signal path" or define the signal path in a feedback system.
Personally,
I would like to know which companies Bernhard works for as a consultant designer. A confidential PM would suffice for me.
In future, I would welcome it if the usual destructive act were no longer reeled off.
The last three (or five) pages of this thread were not constructive.
Dear Bernhard,
do good - and explain in your own words the well-known but obviously not understood stability criterion.
Or hold back. Be constructive (by the way, you only suspect instability; and you can always bring that up again, again and again ... 😉 I didn't expect anything else. )
Best wishes to Bavaria,
HBt.
Psst
And now please continue with the next step, a first PCB.
I think, jxdking and myself are happy to be in the boat and will continue to participate constructively in this task.
Last edited:
Like I said: Input to output. I don't count the negative input of the ips, it is a fixed constant in a usual voltage feedback amp.
While we're at it, what do you think where the delay / phase shift comes from, that causes instability at some high frequency point ?
Perhaps from those transistors ?
Last edited:
Nice try. I know your intentions.
Look at my plot and you see it. Or buy yourself a book.
Remember: You are the one who posted the schematic of the obviously unstable amplifier, not me.
Thanks, I don't need you to tell me what to do.
Simulation says unstable. Mentioning flaws is not constructive ? I don't want others to make layouts for oscillators.
Still waiting for your original gain phase plot for cross check.
Last edited:
But that is very strange now, because you are constantly proclaiming that this proposal will oscillate!
I have just checked “my simulation” again and nothing oscillates.
However, I am happy to admit that the present circuit is susceptible and therefore I welcome the unrestrained commitment of every attentive neighbor.
"Lieber Bernhard",
I wish you every success with your test setup, a test setup that pays unconditional attention to the unconditional symmetry and the insane OLG of the two circuit halves (with their complementary components).
Do you have all the necessary components to hand?
Caution,
you must observe the absolute symmetry and the safe level /swing limit. (Don't forget that a THD of 0.0005% must not be exceeded, and you may be able to avoid any harmful coupling capacitors in the signal path.)
Perhaps we could turn the tables - you could send me your *.cir file (plus models used) and your book recommendation for me.
thx,
HBt.
If M0rton
has not yet lost interest in this crazy task, he is definitely waiting for the "ok" from Bavaria.
Everyone else is already bored by the destructive dispute between the German state capitals or is relaxing and listening to music.
I'm curious to see how the opposing team will stabilize the current draft while retaining the ludicrous THD - I'm really looking forward to the Bavarian solution.
Respectful greetings,
Team leader of the Northmen
HBt.
has not yet lost interest in this crazy task, he is definitely waiting for the "ok" from Bavaria.
Everyone else is already bored by the destructive dispute between the German state capitals or is relaxing and listening to music.
I'm curious to see how the opposing team will stabilize the current draft while retaining the ludicrous THD - I'm really looking forward to the Bavarian solution.
Respectful greetings,
Team leader of the Northmen
HBt.
Joking aside:
together we will solve any problems that arise with the prototype, that's no problem at all.
together we will solve any problems that arise with the prototype, that's no problem at all.
You really keep me on my toes.
I would like to refer to your
posting #214 and posting #232; the snot doesn't oscillate with the models I use!
C88 in your schematic must always be a tiny bit bigger than the other two bangers, you have to keep that in mind as well as the unconditional symmetry of the circuit halves.
I hope you'll take another look at Tim's original from time to time -> 2k2 || 680pF -> and the result on the phase rotations.
I would like to refer to your
posting #214 and posting #232; the snot doesn't oscillate with the models I use!
C88 in your schematic must always be a tiny bit bigger than the other two bangers, you have to keep that in mind as well as the unconditional symmetry of the circuit halves.
I hope you'll take another look at Tim's original from time to time -> 2k2 || 680pF -> and the result on the phase rotations.
Interlude
The voltage gain of one half of the circuit (upper or lower branch) without gnfb is already more than 100000 ! Load corresponds to 5.7Ohm.
The voltage gain of one half of the circuit (upper or lower branch) without gnfb is already more than 100000 ! Load corresponds to 5.7Ohm.