Cornish Buffer: why the 1n cap?

[Fancy Lime]

Hi there,

there was a post involving the Cornish Buffer today so I had a look at the schematic and found something I did not understand. There is a 1nF cap between B and E. What does that do? Assuming an ideal voltage follower, there should be zero AC signal across this cap, so I am assuming it has something to do with non-ideality. Other than that I'm blank. I am referring to C2 in this schematic:

http://www.next.gr/uploads/7/cornish_buffer.png

Thanks and cheers,
Andy

[r080]

It has to do with bootstrapping. I don't have any idea how to analyze it, but that might help you find something in literature about it.

[antonis]

IMHO, main bootstrap cap is C3 (4μF) for raising R3 effective value (high input impedance purpose..)

C2 looks more like RFI decoupling item..
(prevents B-E rectification..)

[PRR]

It is a quick hack to reduce radio reception.

You find it also in cheap old phono preamps.

[Fancy Lime]

It is a quick hack to reduce radio reception.

You find it also in cheap old phono preamps.

Thanks, that makes sense. However, I still don't get how or why it actually works. Is there a benefit of doing it "the Cornish way" vs. the more obvious 50pF to 100pF cap from base to ground?

Cheers,
Andy

[r080]

IMHO, main bootstrap cap is C3 (4μF) for raising R3 effective value (high input impedance purpose..)

C2 looks more like RFI decoupling item..
(prevents B-E rectification..)

That makes more sense. I knew one of them was for bootstrapping, anyway.

[antonis]

Using a more wide meaning of term, both are bootstrappers(*) - one for R3 and the other for itself..

(*) Stephen, is that terminology approved for Stomboxology Vocabulary..??

[duck_arse]

Using a more wide meaning of term, both are bootstrappers(*) - one for R3 and the other for itself..

(*) Stephen, is that terminology approved for Stomboxology Vocabulary..??

not for the Stomboxology Vocabulary, no. not espressive enough.

[PRR]

> cap from base to ground?

What is this "ground"? If you are getting radio, ground may be a vague concept.

The radio is not audible until it is rectified on a diode. The B-E junction is that diode. if you can damp RF there, problem solved.

In more complex circuits this hack may raise other problems. But in a pedal, bah.

[Fancy Lime]

> cap from base to ground?

What is this "ground"? If you are getting radio, ground may be a vague concept.

The radio is not audible until it is rectified on a diode. The B-E junction is that diode. if you can damp RF there, problem solved.

In more complex circuits this hack may raise other problems. But in a pedal, bah.

Thanks Paul!
I was about to respond to your answer yesterday and ask for clarification what you mean. I'm glad I did not have the time and instead had a night to sleep on it. I think I understand now. To you comment about ground at RF: I would normally assume "ground" in this case to be anything that has a low impedance path to earth. This may of course involve a rather lengthy chain of effects, cables, the amp and whatnot, so "low impedance" is indeed a vague and somewhat unpredictable parameter. Am I correct assuming that you are saying that I should not trust the impedance to be sufficiently low at RF?

Cheers,
Andy

[antonis]

Am I correct assuming that you are saying that I should not trust the impedance to be sufficiently low at RF?

Without speaking on behalf of Paul..

Rectification only needs one forward voltage drop of diode used for this procedure..
Considering B-E junction as a diode, 1nF cap simply shunts (by-pass) it..
So, voila.. Rectification prevented..!!

We don't wish to "completely" shunt RF, i.e. from signal path to circuit ground 'cause we should take into account signal source impedance (unknown!!) together with Q1 stage input impedance (known/estimated) for calculating C2 value for desired LPF cut-off frequency..

Another way of facing it could be considering 1nF impedance much less than Emitter intrinsic resistor one for medium to high RFs, resulting into a more "convenient" shunt path..

[r080]

Excellent explanations. Given that purpose, it seems there might be recommendations for a specific type there. Ceramic?

[antonis]

You're now opening the bag of Aeolus..!!   

[r080]

I see the Dirk Hendrik G2 schematic did find that Pete used ceramic there.

Bag of Aeolus? Is that a bad thing?

[antonis]

Bag of Aeolus? Is that a bad thing?

As bad as an ethernal discussion about capacitors could be..
(the best conclusion should be a perpetual discussion..)

[Fancy Lime]

...

Bag of Aeolus? Is that a bad thing?

It's a great pedal name, is what it is.

[Rob Strand]

Given the low value of the output resistor the 1n cap could be there to prevent oscillation under certain conditions.

See replies #30 and #31 from this old thread,
https://www.diystompboxes.com/smfforum/index.php?topic=119501.20

Point 5 shows how the larger output resistor helps, so we would expect a small output resistor to be worse.

There's a lot of hidden evil in buffers which need very close analysis.   You can read a lot of stuff from the links in reply #30.  The addition of the input and output resistor is a kind of recipe which helps but it might not work under all conditions unless you go through the details.

Another thing to factor in would be the input is driven by a guitar and cable.   So the input source impedance is complicated.   The guitar will be inductive with a high valued of resistance to ground from the pots and cable will present a capacitance to ground.