john curl said:
...but it may look very simple when well executed... Bootstrapping of a source by a negative feedback by voltage reduces an effective gate-source capacitance from 30 pF to approximately 0.5 pF, so a Miller one can dominate if not to take care of a voltage swing on a drain, but when carefully implemented input capacitance and Miller capacitance variations may cancel each other.
But more complex solutions that reduce input capacitance by combination of negative and positive feedback do an opposite thing: reducing a constant part of an input capacitance they make a variable one dominant. In Russia it is called "Bear's Help"
OK, fellow designers, let us determine where the 'Bear' really sits!
The fundamental debate that started all this was the use of the 2SK170 as a microphone input stage, while working with default impedances of 1G ohm or more.
First of all, the 170 is a 'workhorse' device that is in many people's spare parts box, as a leftover from other projects. This is good and convenient, but IS the 2SK170 really the best device to use as an input device with condenser microphones?
In my opinion, there are other devices, with inherently lower input capacitance, that offer lower distortion performance, even with the same input circuit.
Still, these alternative devices may or may not be as easily available from our spare parts box, and they might be more noisy, and certainly have less Gm than the 170. Is it worth it?
Well, what are we working with? A large area or a small area microphone? A quality microphone capsule or something that we got from Radio Shack?
Can we really just use the mike capsule into a summing input and get a real advantage?
If we use a more conventional technique of hi Z loading, how about bootstrapping and cascoding to lower the effective capacitance of the 170? Even if the input capacitance is perfectly linear, what is the downside of having any significant input capacitance at all?
Is just 'linearizing' the input capacitance the same thing as 'lowering' the input capacitance?
So many questions, and I'm out of time. I hope that some here think about it.
The fundamental debate that started all this was the use of the 2SK170 as a microphone input stage, while working with default impedances of 1G ohm or more.
First of all, the 170 is a 'workhorse' device that is in many people's spare parts box, as a leftover from other projects. This is good and convenient, but IS the 2SK170 really the best device to use as an input device with condenser microphones?
In my opinion, there are other devices, with inherently lower input capacitance, that offer lower distortion performance, even with the same input circuit.
Still, these alternative devices may or may not be as easily available from our spare parts box, and they might be more noisy, and certainly have less Gm than the 170. Is it worth it?
Well, what are we working with? A large area or a small area microphone? A quality microphone capsule or something that we got from Radio Shack?
Can we really just use the mike capsule into a summing input and get a real advantage?
If we use a more conventional technique of hi Z loading, how about bootstrapping and cascoding to lower the effective capacitance of the 170? Even if the input capacitance is perfectly linear, what is the downside of having any significant input capacitance at all?
Is just 'linearizing' the input capacitance the same thing as 'lowering' the input capacitance?
So many questions, and I'm out of time. I hope that some here think about it.
dimitri said:
Of course, the world is not so perfect as in textbooks. I'm still waiting either for your apologies, or for technical explanation of "flaws" you were so rude talking about.
john curl said:
Nobody said it is the best device. Also, real impedance of a signal source is much lower than 1G (even half G because of 2 resistors in parallel on AC), it is almost capacitive and is determined by capacitance of M7 capsule.
The fundamental debate started with my statement about 2SK170 in your amp: if it works well in my condenser microphones, it would work much better in your arrangement. That's all. I'm out. Please don't call me anymore.
Now what did we do? Ask a few questions?
Dimitri, IF we could find some way of putting pages 5 and 6 on this website (just those two pages) of the AES preprint that you referenced me, it might clarify my questions and concerns. It would show how any real(even if perfect) capacitance causes distortion by interacting with the capsule capacitance (that is variable). I might be able to do it through SY, but you might be able to do it better and faster. In any case, all we want here is to put forth the best advice possible, because it might influence a design decision in future.
Dimitri, IF we could find some way of putting pages 5 and 6 on this website (just those two pages) of the AES preprint that you referenced me, it might clarify my questions and concerns. It would show how any real(even if perfect) capacitance causes distortion by interacting with the capsule capacitance (that is variable). I might be able to do it through SY, but you might be able to do it better and faster. In any case, all we want here is to put forth the best advice possible, because it might influence a design decision in future.
Hi John,
I would love to see those references!
The application questions you just raised in post # 9423 (!) are the things I was trying to point out. Also, there may be one perfect device for a situation, or a few. There are choices to be made and it really depends on the design goal as to what is (more) right and wrong (or less right really). Not unless someone grabs a completely undesirable part and uses that anyway.
Ken, the non-linear gate capacitance is more of a charge thing (no relation to the charge amplifier). This can be very non-linear or not so bad depending on device and operating points. So think in terms of charge and the current required to change that charge.
-Chris
I would love to see those references!
The application questions you just raised in post # 9423 (!) are the things I was trying to point out. Also, there may be one perfect device for a situation, or a few. There are choices to be made and it really depends on the design goal as to what is (more) right and wrong (or less right really). Not unless someone grabs a completely undesirable part and uses that anyway.
Ken, the non-linear gate capacitance is more of a charge thing (no relation to the charge amplifier). This can be very non-linear or not so bad depending on device and operating points. So think in terms of charge and the current required to change that charge.
-Chris
Anatech, at least I got you interested! That is a start. We will have to see about references. I have them, but I can't post them. Perhaps, I need your email address.
Equations:
C= K*Eo*A/D, where Eo= 8.854x10-12
where:
K is the dielectric constant of the material,
A is the overlapping surface area of the plates,
d is the distance between the plates, and
C is capacitance
Non linear, 1/d law...
V=sqrt(2W/C)
where:
W is the energy,
V is the voltage.
Again, non linear....
Where to find a perfect capsule for a perfect amp in the perfect world? 😀
C= K*Eo*A/D, where Eo= 8.854x10-12
where:
K is the dielectric constant of the material,
A is the overlapping surface area of the plates,
d is the distance between the plates, and
C is capacitance
Non linear, 1/d law...
V=sqrt(2W/C)
where:
W is the energy,
V is the voltage.
Again, non linear....
Where to find a perfect capsule for a perfect amp in the perfect world? 😀
Please, I will send you the references if you want. It is more complex than I first thought as well. Or maybe, I am worrying about nothing, BUT nobody has yet shown me that this is so.
I work with B&K capsules or better, so what they worry about is important to me.
I work with B&K capsules or better, so what they worry about is important to me.
I wouldn't repeat what JC told, here are obvious mistakes:
1) the JFET current is too low 200uA
2) the voltage across JFET is too high 25V
3) the preamp clips at 200mV peak input
4) the input capacitance is too high 5pF
enough for today
dimitri said:
1. Self noise level is way below ambient noises,
2. It sounds well with all kinds of phantom powers and input impedances in all available consoles I've found so far, and sounds perfect,
3. Show me clipping on any record from my site,
4. if something differs from your personal preferences that does not mean that it is wrong and don't give you the right for rude attacks on personalities of designers.
Enough for you forever.
Wavebourn said:
Wavebourn said:
Wavebourn said:
Wavebourn.
I think your research may be onto something good.
When can we have one Diy Project with at least one schematic.
So people can enjoy th perfpormance of your circuit idea.
Is it in less than 10 years from now?
Or maybe even in 2010. Two years from now?
Only want to know. Thamks.
----------------
I love this one, Wavebourn:
😎
If something differs from your personal preferences
that does not mean that it is wrong.
😎
It is one clever message, Wavebourn,
especially to all rigid old chaps in this Blowtorch topic.
----------------
Because this is not at all, how I have seen people in Blowtorch thinks. Not even you Wavebourn, really 😉
Your one-liner sends a message, but for no good ... to these supposedly clever ones.
Because when you ARE RIGID and not flexible in your mind,
then you ARE STUCK with your ways and ideas.
Hopelessly!
Until the day you die!
😀 😀
The only-my-way-is-good-enough-for-any-thing persons.
The percentage in this thread is
higher than any where else I have have been on the net.
Regards.
No offense, only some things to think about.
For each and every one, really.
Lineup audio blog - Thinking out-of-the-box!
- Status
- Not open for further replies.