And that's a pretty nice memory, isn't it?
Perhaps once again as a reminder:
I'm only talking about an attempt to rescue the sunken ship Stellema (or shall we call the boat Tim) - not about a Leach, Blameless or anything else. If we wanted to design an ultimate 25W amplifier from scratch, then of course I would proceed differently.
I turned around my circuit to inverting and great wonders happened. THD 20dB down.
Stability is not perfect and perhaps I changed a detail unknowingly, need check.
ClassA version
Stability is not perfect and perhaps I changed a detail unknowingly, need check.
ClassA version
Ok, virtual ground - inverting affects 2nd harmonic only, by 20dB. 3rd remains exactly the same.
If that circuit was downgraded - no mirror, no cascodes, the impact on the higher THD would be neglible imho.
So what we have here? Second down to -184 from -164, all else equal.
One could say that the normal version has a nicer falling harmonic spectrum
Green trace = inverting classA, red trace = normal classA
5th harmonic and up depend on hf compensation, nothing to see here.
If that circuit was downgraded - no mirror, no cascodes, the impact on the higher THD would be neglible imho.
So what we have here? Second down to -184 from -164, all else equal.
One could say that the normal version has a nicer falling harmonic spectrum
Green trace = inverting classA, red trace = normal classA
5th harmonic and up depend on hf compensation, nothing to see here.
The input is 0V, the bases are not. Also the emitters of the inner diff transistors are on -1,4V, so that is very shallow water...
The original schematic had unequal resistors in that divider, so the vas input was at about 4V.
The base of the diff input transistor is on 0V and has nearly no ac voltage swing.
Have you already read the entire article? I see you're starting to think your way into the original circuit 😉.
Believe me, I've chewed the circuit up and down - number 250 is the least ("das kleinste Übel") of it.
Believe me, I've chewed the circuit up and down - number 250 is the least ("das kleinste Übel") of it.
Yes, I read it and can not find anything extraordinary, except the inverting thing which is not a game changer obviously.
He talks about linear input impedances, uses buffers, I use buffers too.
I was a bit confused by the many different schematics you uploaded, one time without coupling caps, throwing the gain away, Another time in post #250 those caps are a bit small, resulting in -3dB at 10Hz.
The original is slightly better, -1,3dB at 10Hz.
Anyway: coupling caps... tech from the last millenium.
IIrc, I asked you before, what do you think is great about that original schematic? Please enlighten me.
He talks about linear input impedances, uses buffers, I use buffers too.
I was a bit confused by the many different schematics you uploaded, one time without coupling caps, throwing the gain away, Another time in post #250 those caps are a bit small, resulting in -3dB at 10Hz.
The original is slightly better, -1,3dB at 10Hz.
Anyway: coupling caps... tech from the last millenium.
IIrc, I asked you before, what do you think is great about that original schematic? Please enlighten me.
That's completely unnecessary.
Please let's clear up the side skirmish via PM, wrangling over competence does not belong in the public domain.
Thank you,
HBt.
A suggestion for goodness sake (and explanation of how I work):
Think of the designated symbol of the 1µF capacitor as a placeholder that conveys something special to me personally, like a yellow Post-It.
And you don't need to harp on about the doctrine that "capacitors in the signal path are bad", because I haven't liked that for a long time.
In any case, ac couplings are not a sign of a bad design or even an exclusion criterion!
Simply start a simulation of all relevant parameters, characteristic values ... as a function of these two placeholders.
With MicroCap12 this is no problem and can be done quickly.
It should also be noted that the current source or sink is also only a placeholder 😉.
Please dear Bernard,
don't say now, as you have done so often, that you are too "lazy" [quote Bernard] to do it. Or that this is my job. If you are going to criticize, it should be well-founded and watertight.
No offense intended.
In any case, I am happy to receive any constructive input.
#
I'll get to work on the layout at the weekend - after all, DIY is all about the joy of crafting.
If wahab doesn't pull the emergency brake or I don't notice anything else that speaks against implementation, then this plan will remain in place.
regards,
HBt.
Think of the designated symbol of the 1µF capacitor as a placeholder that conveys something special to me personally, like a yellow Post-It.
And you don't need to harp on about the doctrine that "capacitors in the signal path are bad", because I haven't liked that for a long time.
In any case, ac couplings are not a sign of a bad design or even an exclusion criterion!
Simply start a simulation of all relevant parameters, characteristic values ... as a function of these two placeholders.
With MicroCap12 this is no problem and can be done quickly.
It should also be noted that the current source or sink is also only a placeholder 😉.
Please dear Bernard,
don't say now, as you have done so often, that you are too "lazy" [quote Bernard] to do it. Or that this is my job. If you are going to criticize, it should be well-founded and watertight.
No offense intended.
In any case, I am happy to receive any constructive input.
#
I'll get to work on the layout at the weekend - after all, DIY is all about the joy of crafting.
If wahab doesn't pull the emergency brake or I don't notice anything else that speaks against implementation, then this plan will remain in place.
regards,
HBt.
I still have one request to make: You use my name abbreviation as a prefix in your files. This is quite misleading for a casual reader of this thread, because your circuits and diagrams are not mine (they have nothing to do with me or my thoughts, suggestions).
Maybe you can find a better solution for your file management in the future, (better) replace the prefix 'hbt' with something non-personal. As I said, it's not my circuits you are simulating and the results you are posting.
HBt.
Maybe you can find a better solution for your file management in the future, (better) replace the prefix 'hbt' with something non-personal. As I said, it's not my circuits you are simulating and the results you are posting.
HBt.
Lets look at it like this: Technically a coupling cap will not improve sound. Subjectively it could produce effects that some people find pleasant.
Especially with values over 10µF when film caps start to be to large, ESR and D are introduced with use of lytics. Small lytics tend to dry out over time.
There are whole threads about capacitor sound, everybody has his opinion, I have mine.
The original schematic has three ac coupled stages, tiny errors add up.
Simulating your circuit or the original does not make sense for me because the outcome is very predictable as neither uses the usual measures to lower thd.
One sentence:
The diff pair is the royal couple, king and queen. They need to match and everything has to be done to make them comfortable.
Especially with values over 10µF when film caps start to be to large, ESR and D are introduced with use of lytics. Small lytics tend to dry out over time.
There are whole threads about capacitor sound, everybody has his opinion, I have mine.
The original schematic has three ac coupled stages, tiny errors add up.
Simulating your circuit or the original does not make sense for me because the outcome is very predictable as neither uses the usual measures to lower thd.
One sentence:
The diff pair is the royal couple, king and queen. They need to match and everything has to be done to make them comfortable.
And yet they are so incredibly practical 🙂.
#
Are we still concentrating on L.Stellema's throw, or #250? Otherwise I'll have to get off at this bus stop.
@Bernhard,
Once again you leave me speechless in this thread, either I play ping pong with you or put you on my personell ignore list for a longer time, not just a few days.
This statement is nonsense from any perspective.
You are vehemently trying to discredit a legitimate technical solution - to a (well known) problem.
But you would really have to explain this point in detail and explain it correct and plausibly in a technical, physical and physiological context.
And here you have landed at the completely wrong address - with real components (and physics)! In your list, you also omitted dielectric absorption ...
Are you confusing the terms opinions and views with the word believe?
Infinitely small errors add up to an infinitely small error - which our control loop hopefully erases.
What would be the usual measure to reduce the distortion factor? I already regret my question. You don't have to answer, please. I know all the measures - and so do countless other forum members.
Bye,
HBt.
(I am putting you on my ignore list for a long time now, communication via PM is open (and up) to you, the channel is free - use it!)
Last edited:
I invite you to skip all coupling caps in a real-world amp. And I bet your wife will hear from the kitchen the difference in sound - to the worse because this amp will not work anymore
D stands for dielectric absorption
My Genrad 1688 LC bridge measures D down to 1ppm. So I have very accurate information about every cap I use or not use.
The rest, forget...
Input DC blocking cap - not needed.
Series cap for the feedback shunt resistor - not needed.
Cap across a bias diode - that diodes have a low impedance, maybe a few hundred ohms, I would call that a de-coupling cap. Won´t hurt.
What else?
My good old M-25
Why did they use two high quality film caps on the input + a smaller one, if caps don´t matter? Why not just a 10 µF bipolar lytic?
Why did they use two high quality film caps on the input + a smaller one, if caps don´t matter? Why not just a 10 µF bipolar lytic?
Sorry:
DA does not only stand for the dielectric absorption option in Germany (alone) - ancient measuring bridges do not determine this property of a capacitor!
D is also synonymous with the dissipation factor tan_delta in Germany and Bavaria.
And this is what your bridge shows you.
I assume that you are not familiar with the DA determination method.
And now please put an end to the /that nonsense; I would like to avoid the intervention of the 'moderation team' if it's possible!