Lowest noise amps

[Winfield Hill]

Ken Smith wrote...

Winfield Hill wrote:
[....]

The 2sk170 with its under-1nV spot voltage-noise level isn't best
employed in a follower circuit, because this low noise level would
force one to follow the JFET follower with an unusually-low-noise
BJT amplifier, such as one using Rohm's 0.55nV 2sd786 transistors.
That's not appealing, compared to a properly-designed common-source
JFET amplifier that provides all the advantages without the pain.

Yes, an amplifier where the feedback to the source is taken from
much later in the circuit is a better way to go. You usually want
to follow the cascoded pair of JFETs with a low noise PNP stage.

Actually that's a good place to switch to IC technology, such as a
fast opamp. Plus, a BJT often fits better than another JFET for the
input JFET's cascode stage.

 

[Robert Baer]

Ken Smith wrote...
[...]

Watch out in the capacitive micro-pre-amp. The gate capacitance of
the very low noise JFETs is enough to effect the linearity.

At 2mA, I think I'd look at the LSK170 for my low noise FET.

I just purchased 100 Toshiba 2sk170, and measured samples at 1 and
3mA: under 1nV. I like that their input capacitance is lower than
many others with similar or lower e_n. I purchased mine from MCM.

It's easy to devise circuits that eliminate any linearity effects
from the JFET's gate capacitance changing with drain voltage.

This is not the distortion I was refering to. In a capacitive microphone,
the sound moves one plate of a capacitor by the amount I'll call X. For
small spacings between plates, the capacitance of the capacitor varies as
1/X. If the charge on the capacitor is fixed, the voltage varies as 1/C.

You end up with 1/(1/X) or simply a voltage that varies with X.

If the gate capacitance is high, you end up with 1/(Cgs + 1/X) and hence
distortion.

Although in a 'follower' circuit, the effect of the gate-to-source
capacitance is considerably reduced. (Gate-to-ground capacitance is not
affected unless you bootstrap the surrounding metalwork)

Errr...
Methinks you are incorrect in saying that if the capacitor plate
moves by amount X, that the capacitor varies as 1/X.
That is not possible; moving the plate by 0.1 micrometer cannot give
a capacitance change a thousand t1mes more than a movement of 0.1
millimeter.
Your statement implies that a zero movement gives an infinite change
of capacitance (this naturally is in the limit).

So, start with the formula for *total* capaciance, where everything
is kept constant except the dielectric thickness X:
C=K/X where C is capacitance in some reasonable units like
picofarads, and K is a constant that includes fixed things like
dielectric constant, plate area, etc.
HERE AND ONLY HERE does the *total* plate seperation X change the
*total* capacitance by 1/X (aka "by definition".
*BUT*
The sound *changes* that seperation, and one wished to know how the
capacitance changes.
Hmmmm.....
Sounds like high math...maybe, hummmm, differential calculus, maybe????
So, d(1/X) = -(1/X*X)dX .
My, oh my!
Sounds like it is LINEAR - that is to say, the capacitance changes
the same way that X changes!
What a massive surprise!

 

[Robert Baer]

Ken Smith wrote...

Winfield Hill wrote:
[....]

The 2sk170 with its under-1nV spot voltage-noise level isn't best
employed in a follower circuit, because this low noise level would
force one to follow the JFET follower with an unusually-low-noise
BJT amplifier, such as one using Rohm's 0.55nV 2sd786 transistors.
That's not appealing, compared to a properly-designed common-source
JFET amplifier that provides all the advantages without the pain.

Yes, an amplifier where the feedback to the source is taken from
much later in the circuit is a better way to go. You usually want
to follow the cascoded pair of JFETs with a low noise PNP stage.

Actually that's a good place to switch to IC technology, such as a
fast opamp. Plus, a BJT often fits better than another JFET for the
input JFET's cascode stage.

You mean, as an example, the lower device on the totem is the
low-noise JFET, and the upper device is a bipolar transistor.

 

[John Woodgate]

I read in sci.electronics.design that Robert Baer

Methinks you are incorrect in saying that if the capacitor plate
moves by amount X, that the capacitor varies as 1/X.
That is not possible; moving the plate by 0.1 micrometer cannot give
a capacitance change a thousand t1mes more than a movement of 0.1
millimeter.
Your statement implies that a zero movement gives an infinite change
of capacitance (this naturally is in the limit).

Does seem a bit unlikely.

So, start with the formula for *total* capaciance, where everything
is kept constant except the dielectric thickness X:
C=K/X where C is capacitance in some reasonable units like
picofarads, and K is a constant that includes fixed things like
dielectric constant, plate area, etc.
HERE AND ONLY HERE does the *total* plate seperation X change the
*total* capacitance by 1/X (aka "by definition".
*BUT*
The sound *changes* that seperation, and one wished to know how the
capacitance changes.
Hmmmm.....
Sounds like high math...maybe, hummmm, differential calculus, maybe????
So, d(1/X) = -(1/X*X)dX .
My, oh my!
Sounds like it is LINEAR - that is to say, the capacitance changes
the same way that X changes!
What a massive surprise!

Harrumph! dC/dX = d/dX(K/X) = (K/X^2). Inverse square law! However, the
excursion is minute, even compared with the small X in a miniature
electret capsule, so the even-harmonic distortion is negligible up to,
typically, 120 dB SPL, beyond which distortion is not the greatest of
your concerns.

 

[Robert Latest]

On 20 Aug 2005 06:06:20 -0700,

A cascode configuration reduces the Crss gate-drain capacitance, and
bootstrapping the cascode off the source

What does 'bootstrapping the cascode off the source' mean?

robert

 

[Winfield Hill]

Robert Latest wrote...

Winfield Hill wrote


What does 'bootstrapping the cascode off the source' mean?

Just that. Control the BJT cascode voltage from the JFET source.

 

[Ken Smith]

What does 'bootstrapping the cascode off the source' mean?

Vcc
!
/
\
/
\
! !
\! !
!--+---/\/\/-- GND
e/! !
! !
! ---
! ---
! !
!- !
---->! !
!-+----
s !


Asssuming there is lots of feedback brought into the source of the JFET,
all three terminals of the JFET have exactly the same AC signal on them so
the input impedance is infinite.

The highest input impedance I've ever been able to do at a few KHz was
200G. To do this, absolutely everything near the first stage has to have
the same signal on it.

 

[Ken Smith]

Errr...
Methinks you are incorrect in saying that if the capacitor plate
moves by amount X, that the capacitor varies as 1/X.
That is not possible; moving the plate by 0.1 micrometer cannot give
a capacitance change a thousand t1mes more than a movement of 0.1
millimeter.
Your statement implies that a zero movement gives an infinite change
of capacitance (this naturally is in the limit).

I think perhaps I was unclear in the statement. The sound changes X from
some initial value. At rest the two plates are at some distance.

This is how all of the high end microphones are in fact made.

 

[Robert Baer]

I read in sci.electronics.design that Robert Baer



Does seem a bit unlikely.



Harrumph! dC/dX = d/dX(K/X) = (K/X^2). Inverse square law! However, the
excursion is minute, even compared with the small X in a miniature
electret capsule, so the even-harmonic distortion is negligible up to,
typically, 120 dB SPL, beyond which distortion is not the greatest of
your concerns.

I suggest you go back to college.

 

[Robert Baer]

I think perhaps I was unclear in the statement. The sound changes X from
some initial value. At rest the two plates are at some distance.

This is how all of the high end microphones are in fact made.

Yes; what i have seen is a solid "front" with holes or slots, and a
metallized mylar diaphram is stretched over the "back" of that piece.
The plastic is the insulator, and a conductive ring makes the
electrical contact to the metallized surface *and* stretches the mylar
uniformly.
It is rare that the metallized mylar is placed between two solid (but
perforated) plates.

 

[John Woodgate]

I read in sci.electronics.design that Robert Baer

I suggest you go back to college.

I suggest you explain instead of just distributing your opinions.

 

[Glen Walpert]

Yes; what i have seen is a solid "front" with holes or slots, and a
metallized mylar diaphram is stretched over the "back" of that piece.
The plastic is the insulator, and a conductive ring makes the
electrical contact to the metallized surface *and* stretches the mylar
uniformly.
It is rare that the metallized mylar is placed between two solid (but
perforated) plates.

Back in the 60's when I worked at the University of Rochester
Accoustics Lab part of my job was replacing the aluminized Mylar
diaphragms in a custom microphone designed for -3 dB @ 2 MHz (for
shock wave measurements). This was the only Mylar diaphragm
microphone in the lab, and it was a real pain because of the pathetic
stability of Mylar; the diaphragms needed to be changed once a month
or so to maintain reasonable performance. Real measurement
microphones (Bruel & Kjaer, etc.) designed for reproducible response
over a long term always use solid metal foil diaphragms. The slotted
front cover is always removed for any precision measurements as it
severely degrades the performance of the microphone; measuement
microphone are always provided with response curves with and without
the cover, the curve without being the good one, the curve with being
for harsh field applications where the cover must be left on to
protect the diaphragm.

Regardless of diaphragm material there is always an air gap between
the diaphragm and the rear electrode which changes thickness with
diaphragm movement, exactly as JW and KS have described. Perhaps some
low end mics use the Mylar as the insulator between the capacitor
plates (at the edge support), but I have not seen this design; usually
the stationary electrode is insulated from the mic body and the
aluninized Mylar or metal diaphragm (moving electrode) is connected to
the mic body. The Mylar serves only to support the aluminum moving
electrode. Parallel plate capacitor theory is a decent 1st
approximation even though the diaphragm deflects into a shape closer
to spherical (with a very large radius) than planar (the edges do not
move).

I will attempt to post a description of condenser microphones with
construction diagrams from Acoustical Engineering by Harry F. Olson
(1957) to ABSE. Construction details have changed a bit since then
but the principle of operation has not.

 

[Robert Baer]

I read in sci.electronics.design that Robert Baer



I suggest you explain instead of just distributing your opinions.

Not an opinion.
No explaination needed for anyone that can read and knows a little math.

 

[Winfield Hill]

Robert Baer wrote...

Not an opinion. No explaination needed for anyone that can
read and knows a little math.

How about getting off your horse and joining the discussion?

 

[Glen Walpert]

Robert Baer wrote...

How about getting off your horse and joining the discussion?

Methinks he is moaning about a missing minus. Perhaps picky in a
discussion of amplitude, where use of absolute value could reasonably
be presumed.

Did my PDF ever show up on ABSE?

 

[Pooh Bear]

Robert Baer wrote...

How about getting off your horse and joining the discussion?

My sentiments entirely !

Graham

 

[John Woodgate]

I read in sci.electronics.design that Glen Walpert <[email protected]>

Methinks he is moaning about a missing minus. Perhaps picky in a
discussion of amplitude, where use of absolute value could reasonably
be presumed.

Oh, my goodness. I shall have to leave s.e.d immediately, in disgrace!
Oh, the SHAME. A missing MINUS!

 

[Robert Baer]

Methinks he is moaning about a missing minus. Perhaps picky in a
discussion of amplitude, where use of absolute value could reasonably
be presumed.

Did my PDF ever show up on ABSE?

It is a simple differential of the equation for capacitance as a
finction of the plate seperation.
The capacitance change has a simple linear relation to the change in
seperation (leaving out constant multiplier).
And someone is tying to imply that it is 1/(X*X)...

 

[Glen Walpert]

I read in sci.electronics.design that Glen Walpert <[email protected]>


Oh, my goodness. I shall have to leave s.e.d immediately, in disgrace!
Oh, the SHAME. A missing MINUS!

No need for such sarcasm, John. I'm sure we can all think of many
much better reasons why you should leave SED in disgrace!

 

[Glen Walpert]

It is a simple differential of the equation for capacitance as a
finction of the plate seperation.
The capacitance change has a simple linear relation to the change in
seperation (leaving out constant multiplier).
And someone is tying to imply that it is 1/(X*X)...

We all find ourselves needing a bit of a refresher now and then, it is
quite easy to look at a problem backward or upside down without
realizing it. An inverse relationship is not linear. Capacitance is
proportional to 1/x or x^(-1), where x is the distance between
parallel plates. You should be able to verify that, as well as the
solution of d/dx{x^(-1)], in your dusty old college textbooks. In the
meantime we will award you 2 points towards a peppermint sneaker award
.