MC12 confirmed
"submarine Tim" 's noise; <140nV/sqrt(Hz), 20µV(bw=20kHz) --> SNR exemplary 20*log_10(10V/20µV) --> 114dB !
SNR_1V ---> 94dB
SNR_0,1V ---> 74dB
The rescue submersible is not a world champion, but it is perfectly suitable as a working basis and, on top of that, better than the old hi-fi standard.
Disclaimer
"submarine Tim" 's noise; <140nV/sqrt(Hz), 20µV(bw=20kHz) --> SNR exemplary 20*log_10(10V/20µV) --> 114dB !
SNR_1V ---> 94dB
SNR_0,1V ---> 74dB
The rescue submersible is not a world champion, but it is perfectly suitable as a working basis and, on top of that, better than the old hi-fi standard.
Disclaimer
- My sketch is solely a working basis, all symbols represent nothing more than placeholders, their values or dimensions are only starting points.
- The existing layout, the circuit board, is also nothing more and nothing less than a breadboard - a starting point for later revisions, or for a preliminary final circuit board.
- A finished kit is completely different from a brainstorming session; there is no kit.
- All changes are reserved in principle.
- Incidentally, I (H. Bt.) am aware of all the weak points of the present circuit or the first step.
The distortion spectrum is monotonically decreasing and is dominated by K2, at normal to high room volume the THD_100 will practically not exceed 0.001% -> says MC12 with a real load of 8 ohms +/- j0.
MC12 is not my favorite spice tool, but this multitool is extremely powerful -> we just have to use it selectively and know what we are doing.
That settles the dispute for me. For all interested experimental archiologists I will gladly open a new thread, ... if desired in diary style.
HBt.
MC12 is not my favorite spice tool, but this multitool is extremely powerful -> we just have to use it selectively and know what we are doing.
That settles the dispute for me. For all interested experimental archiologists I will gladly open a new thread, ... if desired in diary style.
HBt.
The ac-input resistance is practically constant!
The criticized 5k1 resistor, in the range from 100Ohm to 100kOhm, serves as the running variable. The simulation not only confirms the theory, but also the "virtual zero point"!
I do not give up hope that the task and significance of this inconspicuous resistance will eventually become clear to even the most vehement critic.
MC is a great tool if you know how to use it, just like OrCAD or my old SPICE. It would be even better to be able to predict all this without a simulator, but that would require a sound knowledge of the basics - which is something that quite a few porpoises in the pond lack.
I leave it to the reader's imagination to draw conclusions from the last comments, especially with regard to the fact that ohmic resistors can never be directly responsible for non-linear distortions.
No hard feelings,
HBt.
Psst.
I would like to politely ask critic B once again to discuss (use only the PM option!) with me personally and, above all, not to engage in any further technical-physical misdirection.
thx
I sincerely apologize to the uninvolved reader who now stumbles across the unnecessary accumulation of strange claims, circuits and THD curves. I'm sorry about that; it also spoils my enjoyment of the subject of electrical engineering/electronics/communication transmission ...
The criticized 5k1 resistor, in the range from 100Ohm to 100kOhm, serves as the running variable. The simulation not only confirms the theory, but also the "virtual zero point"!
I do not give up hope that the task and significance of this inconspicuous resistance will eventually become clear to even the most vehement critic.
MC is a great tool if you know how to use it, just like OrCAD or my old SPICE. It would be even better to be able to predict all this without a simulator, but that would require a sound knowledge of the basics - which is something that quite a few porpoises in the pond lack.
I leave it to the reader's imagination to draw conclusions from the last comments, especially with regard to the fact that ohmic resistors can never be directly responsible for non-linear distortions.
No hard feelings,
HBt.
Psst.
I would like to politely ask critic B once again to discuss (use only the PM option!) with me personally and, above all, not to engage in any further technical-physical misdirection.
thx
I sincerely apologize to the uninvolved reader who now stumbles across the unnecessary accumulation of strange claims, circuits and THD curves. I'm sorry about that; it also spoils my enjoyment of the subject of electrical engineering/electronics/communication transmission ...
Bloomy descriptions are nice but:
How about the actual schematic?
THD and THD+N graphs?
Last edited:
You could read a good book.
Now come the facts:
The orange trace is THD with 5,1k resistor and the blue trace is THD without 5,1k resistor.
In the schematic is 5k but I simulated with 5.1k, its irrelevant.
I don´t understand what you want to say
Yesterday we all agreed that the cause of excessive noise was the 100k base resistor that lacked a decoupling cap.
Fixing that brings everything to normal.
Now you post formulas without an actual schematic for reference.
For what?
Removing the 5,1k, fixing the bias network, adding a mirror, adding a pre-vas buffer, removing the pre-diff buffer brings a THD improvement of 60 dB.
In words: Sixty.
If you believe your schematic is good enough, yeah, why not...
You can try to talk to death everything.
By the way:
The schematic from the article does´t have the 5.1k resistor, it has the correct bias network and it has the decoupling cap on the 100k base resistor. !!!
So much for your achievements.
I quit this thread.
And not send PMs, I´m not interested..
In words: Sixty.
If you believe your schematic is good enough, yeah, why not...
You can try to talk to death everything.
By the way:
The schematic from the article does´t have the 5.1k resistor, it has the correct bias network and it has the decoupling cap on the 100k base resistor. !!!
So much for your achievements.
You kidding?
I quit this thread.
And not send PMs, I´m not interested..
Last edited:
The reference of the data I presented is the schematic from my posting #378.
I have lost all my (assumed childlike) joy in this little project. Why should I have to answer /accountability to the duo Bucks Bunny & Bernhard?
Back to the criticism, the wiring of the inputs and the associated noise, or the overall performance:
We can do something about the 20µV noise I mentioned above, we can wire both inputs in a variety of ways and or design and lay out the feedback loop ..!
It is not worth arguing about something essential (so trivial /basics) for all eternity or wanting to be right at all cost.
- At this stage of my approach or working method, the final values of the real components have not yet been determined.
- I am neither a pensioner nor a keyboard warrior, I earn my living with electronic circuits and the knowledge about them.
- This pastime is intended to be nothing more than a relaxing tinkering session.
I have lost all my (assumed childlike) joy in this little project. Why should I have to answer /accountability to the duo Bucks Bunny & Bernhard?
Back to the criticism, the wiring of the inputs and the associated noise, or the overall performance:
We can do something about the 20µV noise I mentioned above, we can wire both inputs in a variety of ways and or design and lay out the feedback loop ..!
It is not worth arguing about something essential (so trivial /basics) for all eternity or wanting to be right at all cost.
THD_1k Rload = 8Ohm:
So the approx -100dB posted (in #406, my circuit thought!) now becomes -160dB -> 1% / 10 /10 /10 /10 /10 /10 -> 1*10^-6% -> 0.000001%.
Nice
.#
In the meantime, I have successfully etched two circuit boards in true style. However, I will certainly never do that again, some of my material is 25 years old. The fact that the photo-coated base material was still functional is a miracle in itself. And improvising a chemistry lab in the bathroom takes time, exposing, developing, etching ... Getting everything back in order, oh dear; but I can't even get to my ancient etching machine in the cellar, what isn't there all around?! Well, it worked. I'll drill tomorrow. My God, I etched the last circuit board 15 years ago just for fun.
HBt.
The diagram in #404 expresses the following:
The signal source (far left) sees a quasi-constant load (the input resistance of the amplifier) of |10kOhms| (over practically the entire audio frequency range) -completely independent of the value of the demonized 5k1. In the idealized, simplified view, this 5k1 can also simply be thought as not really existing. If the signal source were to take note of this boy, then we would have to note 10k+5k1 = 15k1 on the input side.
Anyone who is thinking of a voltage divider here, is getting wrong. It is the resulting currents that flow to the node that play the music here, no voltage.
Practically nothing flows through the 5k1 terminal /resistor,
practically nothing flows into the base electrode,
so the current through the 10k input resistor must determine everything else.
This current now also flows through the 100k negative feedback resistor with a 180° phase relationship. The sum in the node will hopefully be zero. There is a virtual zero point!
Of course, things do not look quite so ideal in reality, in practice. Wasn't there something about thermal noise?
What is it about #404?
The only point is' that it is completely unnecessary to open a big barrel because of this diabolical resistance and to legitimize the terror that has been organized on this basis.
Hopefully the THD(+N)-terrorism of the Tripple B Gang has now come to an end.
HBt.
The signal source (far left) sees a quasi-constant load (the input resistance of the amplifier) of |10kOhms| (over practically the entire audio frequency range) -completely independent of the value of the demonized 5k1. In the idealized, simplified view, this 5k1 can also simply be thought as not really existing. If the signal source were to take note of this boy, then we would have to note 10k+5k1 = 15k1 on the input side.
Anyone who is thinking of a voltage divider here, is getting wrong. It is the resulting currents that flow to the node that play the music here, no voltage.
Practically nothing flows through the 5k1 terminal /resistor,
practically nothing flows into the base electrode,
so the current through the 10k input resistor must determine everything else.
This current now also flows through the 100k negative feedback resistor with a 180° phase relationship. The sum in the node will hopefully be zero. There is a virtual zero point!
Of course, things do not look quite so ideal in reality, in practice. Wasn't there something about thermal noise?
What is it about #404?
The only point is' that it is completely unnecessary to open a big barrel because of this diabolical resistance and to legitimize the terror that has been organized on this basis.
Hopefully the THD(+N)-terrorism of the Tripple B Gang has now come to an end.
HBt.
Everything back to the beginning
Points of criticism were:
- THD depending on the amount of negative feedback
- Need for regulated PSU
- V_CE_0 far too low
- inverting
- crazy design
'wahab' has detected instability, inadequate frequency compensation. I have subsequently determined that the topology is fundamentally difficult with regard to the stability criterion.
No more and no less.
Attachments
I assume you are asking if I am building the actual power amplifier. Yes, I am. The so-called Diamond buffer, BB insists (also azubinski did, first) that the voltage follower is a diamond and is named after its creator, Mr. Diamond. That's what it's all about, the practical test of this configuration, actually completely independent of the IP-VA stage. And if I like this configuration, a KRILL (from Steve Dunlap) will eventually follow. All for just one reason: curiosity and because I have never heard or built such a configuration before.
The prefix (the superlative) "the best" should already indicate that there are technically more sensible solutions, in the sense of more economical or more stable - and that I have /had already recognized the weak points, so a wink with the eye.
After drilling the circuit boards, I am now curious about the behavior of this power amp circuit, as a follower (i.e. an appendage of the IP-VA stage) or as a member within the global feedback loop.
In addition, I have two more and further attempts in mind, for which I need some exotic circuit -> that's another reason why the clumsy attempt at terrorism really annoyed me.
Are you interested in this configuration? Search for the "Krill"
.kind regards,
HBt.
(thx to Franz for the tip)
Can we hope for a detailed explanation?
All right, so where does he belong?
Which trump card have I still not played - and why do I drink tea with milk and sugar?
Let's take the so-called frequency response (amplitude and phase, the angle) as a means of describing the complex transfer function of this quadripole and look at the graphical representation of the same.
Happy to read you again,
HBt.
(reference is "submarine Tim", without Ck 10pF keramisch)
Attachments
If you can't let go - and try to explain the number /case in a well-founded (factual, technically correct) way or offer a new construction site (or pull out the THD, THD+N foam club again), then make sure you don't fall into the obvious trap.
With this in mind, best regards to Munich,
HBt.
PS
I don't hold grudges, I'm a peaceful fellow
With this in mind, best regards to Munich,
HBt.
PS
I don't hold grudges, I'm a peaceful fellow