β is complex, involves a total feedback. Overall feedback is frequency dependent but also voltage-dependent. It is important to know when it can be considered constant.
Mathematical notation has a simple rule that clear and everyone understands it.
Vdif = U1*sin(ωt) - U2*β*sin(ωt+α)
IMHO correct
Vdif = U1*sin(ωt) - U2*β*sin(ωt+α)
IMHO correct
You are slower in deleting your incorrect posts than me in quotting them.
Have you read '1984' by George Orwell?
Have you read '1984' by George Orwell?
Some are trying to divert the thread from the problem as you can. Dependence Vdif delays in the amplifier is quite obvious. Limit Vdif for each topology as well known.
It is really difficult, John. You are a fair man. If we disagree, we stand our points of view.
With him, any fair discussion is impossible. If you followed this discussion, you would see it.
You know there is a measure of success. You are successful as an amplifier designer. If he was successful in math, he would have some professional support. He has not, not a single person.
Federmann,
You first raised a thread in which you claimed that you had designed an amplifier which the more experienced members here were confident was dangerous...you defended your design against all criticism from well regarded and experienced designers and engineers. That thread was closed by the moderators before you produced any evidence of the existance of one build of that amplifier in completed mode.
HAVE you now actually BUILT that amp? Or are you simply a theory guy who is flaming to get at certain members for your own satisfaction?
When we read the link which PMA has posted we see that you are now appearing to simply using DIYaudio forum to continue a battle which you have ben running on YOUR OWN forum and in your native language.
There is thankfully little bitterness across this forum; we need less, not more.
If any is interested I can provide detailed easy to follow instructions on where to find the "unsubscribe" button... or is this thread just to entertaining 🙂
Merry Christmas all.........
Merry Christmas all.........
For many years I worked in the development of integrated circuits. It was a real world. It was not possible to take one thousand transistors and a processor or other circuit.
Had to modeling and modeling and create a chip. Then check how the model corresponds to reality. If the models were no good, and no experienced designers, would never have failed to produce anything.
Right you are, Mooly. I often feel like this. But then, after reading several posts ....... I assume you would understand.
The only check would be a working amplifier, measurements and comparative listening tests.
Here, at diyaudio, you may find plenty of posts that prove how unreliable are free, public transistor models.
I have already written, if you are not satisfied, write a better model. It's not a big problem.
I founded this thread, to show that the relationship is the phase shift the output voltage amplifier and distortion, its sound. I founded this thread, to show that the THD measurement is not meaningful for audio amplifiers.
Some detractors, however, thread and make just fun to deflect from the thread.
Federmann....
Please answer this... how does your model take into account all the real variables that will be introduced when you take real components and real wiring, stray capacitance etc etc.
Please answer that 🙂
Please answer this... how does your model take into account all the real variables that will be introduced when you take real components and real wiring, stray capacitance etc etc.
Please answer that 🙂
BJT models are quite fine. Both JFET and MOSFET public models suffer from poor implementation of non-linear capacitances. Of course that manufacturers have much better models, just ask Scott Wurcer. But they are know-how, not public.
Mooly is very right about parasitic inductances at first, and capacitances as well. But the one who does not build amplifiers does not care, no reason to do so.
Mooly is very right about parasitic inductances at first, and capacitances as well. But the one who does not build amplifiers does not care, no reason to do so.
It's a matter of experience and practice.
1st Surround the design must be such that not added parasitic elements that would alter the characteristics.
2nd If you can not avoid the parasitic elements, the spatial design is eliminated.
3rd If the parasitic elements can be avoided, nor is eliminated, it must be added to the simulation and count them.
Never, repeat never, to me does not make that model did not fit the facts, must I know what I choose and in the production account.
1st Same element, another element of origin is different ...
2nd The difference is the size of the feedback capacity ...
3rd I have to take into account the voltage MOS transistors in UGS ...
4th Etc.
Stray parasitic elements are always present... even in the simplest designs... I often tweak compensation components even making the choice between say 5 or 7 pf when the circuit is built. I don't accept you can simulate like that. And if someone uses a different PCB layout the values change again.
I am at a loss as to what to say 🙂
I am at a loss as to what to say 🙂
- Status
- Not open for further replies.