Capacitance multiplier: which Darlington to choose?

[JohnRoberts]

Analog Circuit Design · Samuel Groner · Resources · Power Amplifiers

Page 31 for "stellar nonlinearities"...

QED

Sam is (was?) a well respected member here. I didn't find anything comprehensive on page 31 of the comments .

JR

 

[user 37518]

Sam is (was?) a well respected member here. I didn't find anything comprehensive on page 31 of the comments .

JR

That is because there is none, it is just a plot, it is a strawman to somehow try to prove a point not relevant to what I was saying. Which is that if you operate the MOSFET between subthreshold and saturation it will be non -linear.

 

[JohnRoberts]

That is because there is none, it is just a plot, it is a strawman to somehow try to prove a point not relevant to what I was saying. Which is that if you operate the MOSFET between subthreshold and saturation it will be non -linear.

there is no rule against mind reading but perhaps there should be...

give him time and space to make a credible response or correction.

JR

 

[user 37518]

there is no rule against mind reading but perhaps there should be...

give him time and space to make a credible response or correction.

JR

Well, if he just provided a plot with zero explanations from him (except QED), what do you expect me to do but guess?

 

[thor.zmt]

You are right. It is just that he constantly tries to boost his ego even when he is wrong.

Really!?

Sometimes a mirror is a useful tool.

Thor

 

[user 37518]

Really!?

Sometimes a mirror is a useful tool.

Thor

It is, and I should take it into consideration, I could use some modesty. Yet, I would advice you the same, since you seem to be adopting this attitude with everyone on the forum.

 

[Walter66]

That is because there is none, it is just a plot, it is a strawman to somehow try to prove a point not relevant to what I was saying. Which is that if you operate the MOSFET between subthreshold and saturation it will be non -linear.

Whats the consequence of that Statement? Will the IRF830 Not be ideal For using in a cap multiplier circuit of, lets say, 25mA Power consumption, because it works nonlinear?

 

[user 37518]

Whats the consequence of that Statement? Will the IRF830 Not be ideal For using in a cap multiplier circuit of, lets say, 25mA Power consumption, because it works Not linear?

I am sorry Walter, but I am afraid the thread has derailed. I apologize for this for which I've been a part of.

 

[JohnRoberts]

come on guys there should not be arguments about objective facts...

Play nice BOTH of you....

JR

 

[Walter66]

No Problem, just trying to understand. Was it a theoretical Dispute or does it have a practical relevance on using the IRF 830 on a Low current preamp?

 

[user 37518]

come on guys there should not be arguments about objective facts...

Play nice BOTH of you....

JR

You are right. I apologize for my behavior, with all (Thor included) Carry on with the original thread topic, I feel bad that Walter is left confused and no one is answering his questions. If Thor is ok with it, I suggest we leave it there, and maybe continue in other occasion, after both of us are more calm.

 

[JohnRoberts]

stop already

JR

 

[user 37518]

No Problem, just trying to understand. Was it a theoretical Dispute or does it have a practical relevance on using the IRF 830 on a Low current preamp?

From my perspective it was all theoretical, and not related to your query

 

[thor.zmt]

Will the Mosfet Work better with Higher current load?

No. It may not matter though.

Is there a difference in Low current Performance between bipolar and Mosfet?

Yes, a huge one.

For maximum simplification:

A bipolar transistor has increasing transconductance with increased current

A MOSFET has (near) constant transconductance up to a certain current, after that transconductance falls with rising current

In a more or less open loop follower circuit a Mosfet will have greater noise rejection and lower output impedance than the bipolar transistor at low currents, while at relatively high currents the bipolar can pull ahead.

Of course, if we use a 100A rated MOSFET with mOhm channel resistance and thus several 100A/V transconductance, than just 1A might be still far ahead of BJT's.

It comes back to selection of the appropriate parts for a specific job.

Thor

 

[user 37518]

stop already

JR

I already said I will, take it easy, John

 

[thor.zmt]

Which is that if you operate the MOSFET between subthreshold and saturation it will be non -linear.

The way you wrote in your original reply I read it as "this always happens and there is poor linearity" maybe I misread?

My experience is that practical Mosfet's one might sensibly use are very linear, moreso than many possible BJT's.

Thor

 

[JohnRoberts]

take it easier....
====
@Walter66 Be careful asking questions on internet forums. People will tell you everything they think they know even if it does not address your question. There are lots of very smart people around here so good luck sifting through all that shared knoweledge.


JR

 

[thor.zmt]

From my perspective it was all theoretical

From my view it was all practical, supported by theory.

''theory without practice is empty; practice without theory is blind''

Charles Mordechai

Thor

 

[user 37518]

From my view it was all practical, supported by theory.

''theory without practice is empty; practice without theory is blind''

Charles Mordechai

Thor

Come on Thor, I am trying to put this to an end. I already apologized to you and the rest, can we leave it there?

 

[thor.zmt]

take it easier....

I am being pretty easy.

I already bit my proverbial tongue repeatedly.

I do remember a little bonmot we had in East Germany... It translates poorly as there are multiple layers of irony here that pretty much only work in German.

"Jenigen die denken alles zu wissen gehen denen die es wissen auf die Nerven!"

"Those who believe they know it all really annoy off those of us who do know!"

Now I don't pretend to much, but I like finding answers to questions and it is my experience that, unless we attach the "extreme simplification" warning Tag, most answers are complex.

In fact I find that really complex questions have really complex answers, while conversely really simple questions have enen more complex answers.

Like:

What is the transconductance of a MOSFET?

Clearly is NOT:

gm = 2*ID/(Vgs -Vth), Vth being the threshold voltage.

It is governed by a much more complex relationship, as we all have established here.

So offering this equation is incorrect, even though the equation is correct for PART of the question.

We can either accept an answer that partially correct and partially incorrect as correct, or we can call it incorrect (or partially correct).

It is about simple facts. Ego's should not enter here.

Thor

 

[thor.zmt]

Come on Thor, I am trying to put this to an end. I already apologized to you and the rest, can we leave it there?

Ok, apologies for riding my hobby horse (truth) too hard.

Let's bury our hatchets.

We both try to help Walter to progress his project, let's be practical and helpful.

Thor

 

[Walter66]

take it easier....
====
@Walter66 Be careful asking questions on internet forums. People will tell you everything they think they know even if it does not address your question. There are lots of very smart people around here so good luck sifting through all that shared knoweledge.


JR

Thank you John. Will try to bring theory to life by Testing the circuits. There we're given many good hints in this thread, many thanks for those.

 

[abbey road d enfer]

Whats the consequence of that Statement? Will the IRF830 Not be ideal For using in a cap multiplier circuit of, lets say, 25mA Power consumption, because it works nonlinear?

Non linearity is a non issue in this case, because there is significant NFB.
As to the relevance of Gm as the dominating factor of performance in this application, hmmm...

 

[Walter66]

I cant write about theoretical Performance and even about subjective listening experience or measurements due to Holidays. If t measures and sounds good I'll go for it. But what I try to understand is, why did Shindo San Not having choosen a Mosfet over BJT If the superiority is obvious from a technical Point of view. He was a guy who constantly studied the sound of every single part in his amps, a Samurai using a solder Iron instead of a sword. Wasn't he aware about the existence of high voltage Mosfets? Hardly imaginable.
But I'll give it a try anyway. I think what we have seen is the superiority of the circuit compared to older solutions like tube regulated PSU.

 

[thor.zmt]

But I'll give it a try anyway. I think what we have seen is the superiority of the circuit compared to older solutions like tube regulated PSU.

Again a reminder. When doing listening comparisons make sure the comparisons are "apples to apples" as well as level matched and unbiased.

You can of course compare two different circuits with two different electrical responses and decide "I like X better than Y", but it has no reflection on anything more.

It's like me hating bananas and loving oranges. It's a valid preference, but not an indication that oranges are better in any way, except to me as fruit to eat. Now crazy enough, I do like bananas when fried, though I'd probably still prefer a nice juicy Jaffa Orange.

Thor

 

[thor.zmt]

But what I try to understand is, why did Shindo San Not having choosen a Mosfet over BJT If the superiority is obvious from a technical Point of view.

You would need to ask. Seeing everything is hardwired, a BJT may just have needed fewer components. Or one day while browsing around the Akihabara component shops he found a huge box of the transistors for 1,000 Yen per kg blowout. Who knows.

Seeing that 30H chokes and additional filtering follow the electronic circuit, I'd almost wager that as long as any differences in actual DC voltage are compensated, the choke's DCR & Inductance and the filtering of the RC will swamp out any differences between BJT, Mosfet and even 317 3-Pin regulator chip.

But I'll give it a try anyway. I think what we have seen is the superiority of the circuit compared to older solutions like tube regulated PSU.

That totally depends on the tube regulates supply. Combine a pentode with the right load to maximise AC gain and then use beefy low impedance regulator tubes the results can be very good, objectively & subjectively. Like so:


All MKP capacitors. It was used to power a stereo 300B SE Amplifier (EL84 and later C3m driver) without further filtering for the output stage. The design is mostly by J.C. Verdier from Sound Practices.

In part this PSU was to be able to adjust the voltage to use many different output Valves, like here with a Pair of #45:



Glorious 1,600 milliwatt!

Thor

 

[Walter66]

Don't know why, but I Just like the sound Shindo creates, thats why I'm after it for understanding and then improving on the circuits. To understand something and Not only to be a customer is part of the fun here. Understanding, trying variations and deciding what sounds best, subjective analog brainfuck. Love this Game.

Unfortunately, my EL156 pentode Regulator produces more than the regular noise. So first trying to bring this down.
Need this Low noise voltage for the phonostage Experiment. It does now hum Like hell.
It has EF804S in the gain Stage and very Low Output impedance (Just an extra Choke in Front of Screen grid, otherwise only one Cap and the rectification).

Compared with the EL34 triode Regulator of the Fairchild 670, which one will be better suited for an audio PSU, pentode or triode regulator?

I thought a triode regulator has lower Output impedance, but might be wrong on this.
Both regulated PSU needs extra 200V, means the Transformers have to supply this extra voltage. If I would use my Siemens Tube E235L regulated PSU, it would only need 100V extra voltage in triode mode. Much more efficient.

Would that be beefy enough, it can deliver 100mA like the EL156 PSU. Don't know why the Folks of Wandel& Goltermann didn't use special regulator tubes but EL156 instead.

P. S. Nice amp! My F2A uses EL84 driver, too. Thats quite the Sakuma principle, beefy driver section for good sound.

 

[Walter66]

---

You would need to ask. Seeing everything is hardwired, a BJT may just have needed fewer components. Or one day while browsing around the Akihabara component shops he found a huge box of the transistors for 1,000 Yen per kg blowout. Who knows.

He liked to use cheap TV or Radio set tubes, so why Not cheap B+ regulators, too? LOL

Seeing that 30H chokes and additional filtering follow the electronic circuit, I'd almost wager that as long as any differences in actual DC voltage are compensated, the choke's DCR & Inductance and the filtering of the RC will swamp out any differences between BJT, Mosfet and even 317 3-Pin regulator chip.

You think he destroyed the advantages of achieving a Low impedance B+ Output with heavy filtering after the IC? So any Choke Filters should be placed prior to the active component? I really think he put passive filtering behind it, but its no safe bet.. Prior to His IC are just Caps and a resistor, means the active Filter Starts very near after rectification.

He could have applied it the wrong way, Just because to His ears it meant sounding better. He gave nothing about those technical Terms of circuits If it doesnt Sound the best way. I've seen many examples for this Kind of Viewing technical Things by Ken San. Very unconventional thinking indeed.

That totally depends on the tube regulates supply. Combine a pentode with the right load to maximise AC gain and then use beefy low impedance regulator tubes the results can be very good, objectively & subjectively.

Will this, technically speaking, be the best serial Tube regulator? A pentode? Maybe I got this wrong? Please explain. Thanks.