vox-style treble/tone cut

[stormbringer]

Hi all,

I have got a rough idea how the vox-style treble cut (across the driver outputs) works, but I don't know how to pick values of the cap or pot or to estimate the LPF cutoff frequency. I'm hoping someone here can point me in the right direction.

Thanks in advance,
John

[jazbo8]

Just f=1/(2*Pi*R*C).

[stormbringer]

Just f=1/(2*Pi*R*C).

Thanks - is the R the series VR?

[Enzo]

Here, download this little bit of software, it's free. You can fiddle with a number of tone stack configuation and see the changes in real time.

TSC

[kleuck]

Thanks - is the R the series VR?

R is the output impedance of the tube. Adam's Amplifiers: Triode Calculator

[jazbo8]

That's right, the pot varies the amount of the cut, Vox's Top Cut works backwards, i.e., 0 - max cut, 10 - min cut.

Typical FR blow, C of 1nF, ignore the response below ~1k, it's from something else...
Vox Top Cut.pdf

Jaz

[tubeswell]

R is the output impedance of the tube. Adam's Amplifiers: Triode Calculator

He didn't post a schematic, but I think he's talking about the pot+resistor tone control after the LTP in an AC30, so R would be the pot (in Ohms) and C would be the cap (in Farads). Of course, exactly what you cut depends on what you have to begin with (which is where the output impedance and coupling cap from the driver stage comes in).

[stormbringer]

He didn't post a schematic, but I think he's talking about the pot+resistor tone control after the LTP in an AC30, so R would be the pot (in Ohms) and C would be the cap (in Farads). Of course, exactly what you cut depends on what you have to begin with (which is where the output impedance and coupling cap from the driver stage comes in).

Thanks - yes, I did mean the post PI cut. This cut circuit (inter-line pot, cap) is also in parallel with the el84 grid leaks, so I am not sure that just plugging in the cut circuit R and C into the formula posted above would capture the frequency response of the cut control.

I am asking not because of any inherent interest in the Vox circuit, but rather because I'd like to tame some high end fizz on a high-gain project amp (posted here: http://music-electronics-forum.com/t31084/). I don't want to add a pot until I've tested out different R and C values and get a good understanding of the control. I didn't want to start blindly using the 4n7 cap that is in the Vox, but I have yet to figure out a better option...

update:
here's what I now know (thanks to all who have helped me out, especially Kevin O'Connor): the top cut circuit is indeed in parallel with the el84 grid-leaks, and so a first approximation to the LPF cutoff would be f = 1/(2pi*R*C), where R is the total grid-leaks (standard 220k+220k = 440k). However, the circuit and grid-leaks are in parallel with the driver plate loads (standard 82k+100k), and so the net R in the equation is 182k || 440k = 128k. So assuming a C of 4n7 (typical vox value), the first and second approximations are surprisingly low: 77Hz and 263Hz. Lowering the value of the cut cap will increase the cutoff, and placing a R or VR in series with the top-cut cap will reduce the depth of the cut. I also note that the VR in series with the cut cap is like a cross-drive-line MV, except that it attenuates only those frequencies above the cutoff.

[stormbringer]

That's right, the pot varies the amount of the cut, Vox's Top Cut works backwards, i.e., 0 - max cut, 10 - min cut.

Typical FR blow, C of 1nF, ignore the response below ~1k, it's from something else...
[ATTACH]21355[/ATTACH]

Jaz

Thanks Jaz. Could you please clarify what the different functions represent? Are they different value caps or different settings of the pot?

[jazbo8]

The curves show moving the pot in 10% increment from 0-90%.

Jaz

[Dave H]

Thanks - is the R the series VR?

I wouldn’t use the VR value. I’d use 60k in the equation. 60k will give the approximate -3dB frequency with the VR set to zero resistance. I think it works well if it’s wired with the VR at zero resistance when it’s fully CCW (not backwards like the VOX). It has a smooth taper then with a 250k or 500k log pot.

[J M Fahey]

The Vox "Cut" tone control is similar to all other treble cut controls, but the doubt (not properly answered yet) is what's the "R" value to be considered .
It's neither the plate resistors nor the pentode biasing ones, but a combination, which must also include the tube plate resistance.
Grid bias: 220K+220K=440K
Plate resistors: 82K+100K=182K
Triode (internal) plate resistance: varies but a typical value is around 68K , so 2 plates = 176K
All 3 are in parallel, so final R amounts to around 75K
That with a cap of 4n7 gives a -3dB cutoff frequency around 450 Hz and then drops 6 dB/octave at higher frequencies..
The series Pot value changes the amount of cut above that, but not the frequency.

[stormbringer]

The Vox "Cut" tone control is similar to all other treble cut controls, but the doubt (not properly answered yet) is what's the "R" value to be considered .
It's neither the plate resistors nor the pentode biasing ones, but a combination, which must also include the tube plate resistance.
Grid bias: 220K+220K=440K
Plate resistors: 82K+100K=182K
Triode (internal) plate resistance: varies but a typical value is around 68K , so 2 plates = 176K
All 3 are in parallel, so final R amounts to around 75K
That with a cap of 4n7 gives a -3dB cutoff frequency around 450 Hz and then drops 6 dB/octave at higher frequencies..
The series Pot value changes the amount of cut above that, but not the frequency.

Thank you kindly, JM! It's posts like these that help me learn what the heck I'm doing!
One minor edit: the net resistance would be around 66k because the sum of the plates would be 136k.
One addendum: my amp has equal 100k plate resistors on the driver, so for me the net R would be 68k. That brings the -3dB cutoff to about 495Hz. And if I were to change to 2n2, the -3dB cutoff would be around 1058 Hz.

Now I know what numbers to try and generally what to expect. Thanks again.
-- John

[J M Fahey]

One minor edit: the net resistance would be around 66k because the sum of the plates would be 136k.

Yes, you are right.
It was around 1:30 AM and I was already quite sleepy .
Thanks for catching it

[Dave H]

The Vox "Cut" tone control is similar to all other treble cut controls, but the doubt (not properly answered yet) is what's the "R" value to be considered .

I suggested 60k because that's the value I worked out from my simulation with a 2n2 capacitor (see pdf below).

VOX cut 2n2.pdf