BSIAB / JFET Theory

[seedlings]

If anyone has a little time, I'm trying to get a handle on tone shaping in BSIAB / JFET circuits:



Please let me know where I'm off-base:
1) C10 / R9 is RF input filter
2) Why is there both C3A and C3B bypass caps- did Mr. Guidry just need a 1.22uF cap to boos the proper range of frequencies?
3) C7 keeps the highs through the voltage divider R10/GAIN and C9 keeps them still through the GAIN pot
4) C8/R12 filters out more lows
5) Tone stack is like BMP
6) R4/C12, R5/C13 and R4+R5/C13 are all HP filters, or is this some sort of notch filter?
7) What is the purpose of C4 and C5?  It kind of looks like high frequency feedback to roll off the treble..?
8) C3, C6 and C8 are coupling caps

Thanks tons!
CHAD

Here's the full file:
http://gaussmarkov.net/layouts/bsiab2/bsiab2-project.pdf

[Davelectro]

These questions have been answered in several topics.

2) C3A + C3B +R15: this is a filter to boost frequencies from 200hz and up. There's no 1.22u cap, so...

6) R4 + C12 + R5 + C13: Second-order low-pass filter.

7) C4 and C5 do not roll off treble. They form a hi-pass with R2||R6 and R7||R3, respectively.

[seedlings]

These questions have been answered in several topics.

2) C3A + C3B +R15: this is a filter to boost frequencies from 200hz and up. There's no 1.22u cap, so...

6) R4 + C12 + R5 + C13: Second-order low-pass filter.

7) C4 and C5 do not roll off treble. They form a hi-pass with R2||R6 and R7||R3, respectively.

My apologies, but thank you for taking time to help.  6) is interesting as I thought an inductor was needed for passive second-order filtering.

CHAD

[WGTP]

Someone else posted and I have also simmed to verify that the 2nd order filter at the end is a myth.  It works as a 1st order, at least in LT Spice the Falstad applets.  The two 1st order filters are not isolated from each other, so the values combine into a single 1st order.  If one was before a gain stage and one after, I think it would work as a second order. 

[Davelectro]

OK. Let's call it dual first-order low-pass filter, then. 

[seedlings]

Someone else posted and I have also simmed to verify that the 2nd order filter at the end is a myth.  It works as a 1st order, at least in LT Spice the Falstad applets.  The two 1st order filters are not isolated from each other, so the values combine into a single 1st order.  If one was before a gain stage and one after, I think it would work as a second order. 

OK- that's what I remember from the old car audio days.

Another question about C4 and C5... it looks as if the signal enters the cap from the collector of Q4, then has a choice of a 1M resistance or Q1- and I would guess the transistor has the favored path.  So, is the lower FET amplifying a full-range signal and the upper FET amplifying a high-pass signal?  I can't quite get my head wrapped around this cascode deal.

Repeated thanks for your patience.
CHAD

[WGTP]

http://www.geofex.com/  Check out the 7.7.01 article about Mu/Srpp amps for some explanation.

I have a similar design on my breadboard, but it uses the Mu and then 2 SHO stages.   

To me the main Tone Tweaks on the BSIAB are:
Some similar circuits either don't use the final gain stage or change it to a buffer before the tonestack.  See the Spitfire for variations.
When a cap is in parallel with a resistor, it usually produces a shelving response rather than 1st order roll off
C4 high pass cap.  .22uf-.01uf
C3 high pass shelving cap  22uf to 1.22uf ends up being sort of a treble boost with low values
C7 high pass shelving cap can be used to tweak the high/midrange
.022uf caps are commonly used in between preamp stages in tube amps.  May be large enough that they don't reduce the bass much.
C9 high pass shelving with the gain control does nothing at max gain but can be tweaked to control treble at lower gain settings.
R11 & R13 could be replaced with 470ohm to 2.2k resistor and 22uf cap for more gain
If the simulations are correct, R5 & C13 can be eliminated and C12 doubled or .0022uf to .01uf for treble roll off.  It probably interacts with the BMP tonestack.
The BMP tonestack is probably the main determiner of tonal character.  Hope this helps. 

[seedlings]

Thank you WGTP!  The 7/7/01 post basically says caps C4 and C5 prevent current from being drawn from the gate of the top JFET, so perhaps the value is not very significant?
From the article:
"What happens with signal? If a small upward voltage is connected to the gate of the lower JFET, it turns the lower JFET on a bit, and the lower drain sucks in more current. The current can't come from the capacitor, as the capacitor is effectively in series with a 5M resistor. It can only come out of the source of the upper JFET."


CHAD

[Davelectro]

The 7/7/01 post basically says caps C4 and C5 prevent current from being drawn from the gate of the top JFET, so perhaps the value is not very significant?

It is significant in terms of tone since it's part of a frequency filter (although not an ordinary HP, as far as I remember).

If those two biasing resistors were smaller than 10M -let's say 1M- it would be easier for you to notice the progressive loss of bass as you decrease the value of that cap.

[amptramp]

This is not an SRPP / µ amp stage.  A proper SRPP stage returns the gate to the drain of the lower FET which is connected by a series resistor from the lower FET drain to the upper stage source with the output taken from the upper stage source.  This gives the upper stage some gate signal so it can reduce its current when the lower stage increases its current and vice versa.  The optimum resistor is 1/gm so if the upper FET has a transconductance of 5000 µmhos, the resistor would be 200 ohms.  The circuit shown gives you a current source as a load for the lower FET, which is a totally different design.  Note that an SRPP stage needs a relatively low load impedance since the current from the upper stage must add up to the current to the lower stage plus the current to the load.  With a high-impedance load, it would clip readily on low signals.

[seedlings]

My youngest son just turned 1, so I have a very good idea of what wobbly off-balance baby steps look like.  I'm trying to learn this stuff in wobbly off-balance baby steps.  Thank you again for making time to reply and clarify!

CHAD

[seedlings]

What is the proper term for the transistor arrangement in the first BSIAB gain stage?  I originally thought it was cascode...

CHAD

[Earthscum]

Mod your Mu-amp/Minibooster

Pretty much what Amptramp just summed up, written in a Keen way.

[seedlings]

Mod your Mu-amp/Minibooster

Pretty much what Amptramp just summed up, written in a Keen way.

I must have misunderstood when Amptramp said

This is not an SRPP / µ amp stage.

BSIAB gain stage is a SRPP / u amp stage. <-- true statement?

CHAD

*edit: I also read this: http://www.geofex.com/Article_Folders/foolwfets/foolwfets.htm

[amptramp]

What is the proper term for the transistor arrangement in the first BSIAB gain stage?  I originally thought it was cascode...

CHAD

Cascode is different.  Here is one of the least confusing explanations:

http://www.allaboutcircuits.com/vol_3/chpt_4/8.html

A cascode amp uses the fixed emitter/cathode/source voltage of the upper stage to hold the collector/plate/drain voltage of the lower stage constant, eliminating Miller capacitance changes and guaranteeing limited feedback so that oscillation does not occur.  Cascode was popular in RF stages of the better televisions and FM tuners.  This schematic shows a cascode RF stage in the upper left of this FM tuner schematic:

http://www.44bx.com/leak/Leak/TroughLine2_cct.gif

It shows a dual triode with the first stage connected as a grounded-grid amplifier and the second stage on top of it with fixed bias.  Since this holds the upper stage cathode to a small voltage variation, the lower stage plate is at the same voltage and neutralization is not necessary.

[amptramp]

Mod your Mu-amp/Minibooster

Pretty much what Amptramp just summed up, written in a Keen way.

I must have misunderstood when Amptramp said

This is not an SRPP / µ amp stage.

BSIAB gain stage is a SRPP / u amp stage. <-- true statement?

It is not a µ amp.  A µ amp uses a resistor from the lower drain to the upper source to provide a signal for the upper gate.  The BSIAB uses the upper transistor as a current source, which is a high impedance load.  I understand µ amp to be equivalent to SRPP whereas R G Keen does not.  Maybe I am wrong.

[Earthscum]

,

I understand µ amp to be equivalent to SRPP whereas R G Keen does not.  Maybe I am wrong.

If you're wrong then you are not the only one. I have come across this mention by several others. As I understand it has to do with the internal workings of a tube compared to a jfet, and the grid leakage. The jfet version doesn't mu until that resistor is added, which isolates the gate of the upper fet from the output. With the grid leakage in a triode the resistor isn't needed. But I am still learning and my knowledge is much less than I like to think it is, lol.,,,,

[amptramp]

I got my information here:

http://www.tubecad.com/may2000/

and here:

http://www.tubecad.com/june2000/

Warning: You can disappear into Tube CAD Journal and not come out until next week.

[seedlings]

I got my information here:

http://www.tubecad.com/may2000/

and here:

http://www.tubecad.com/june2000/

Warning: You can disappear into Tube CAD Journal and not come out until next week.

Funny, I was just reading those exact articles!  To clarify my brain: mu or µ essentially means gain? (a high-mu tube basically means high gain)  So the phrase "mu amp" or "µ amp" means... 'high gain' amp?

CHAD