Vox/Matchless Cut control

[Turret]

Got a question regarding cut control in Vox and Matchless

Both Vox ac30 and Matchless dc30 uses 250k pot as a cut control. Matchless clubman schematic I have (googled) show 1M pot. As my understanding 1m pot will act more like aon of switch that way. Am I missing something here? Typo?

[lord preset]

Got a question regarding cut control in Vox and Matchless

Both Vox ac30 and Matchless dc30 uses 250k pot as a cut control. Matchless clubman schematic I have (googled) show 1M pot. As my understanding 1m pot will act more like aon of switch that way. Am I missing something here? Typo?

I recently did a Vox style cut control on a Hammond conversion project. I started with a 1m pot but the effective range of control was too narrow. I switched to a 250k pot but found it made little if any difference. Not sure why that was but in both cases the cut did function as a cut.

[matt h]

(deleted)

[pdf64]

The track taper type and wiring of the control probably has as much affect on its operation as the track value.
eg is full CW the most or least trebley setting?

[Darkbluemurder]

I have had good results with either 250k or 500k audio pots.

Cheers Stephan

[rock_mumbles]

The cap value depends on the grid reference resistor values ...
Vox uses 220k's and a 0.0047uf cap

in an 18 watt with 470k grid reference resistors you use a 0.0022uf cap

Most of my 18 watt builds have a Cut control ... as you turn the control up you cut the really high frequencies.

The Matchless Brilliance control is wired in reverse, turned down it removes the high end

[pdf64]

The cap value depends on the grid reference resistor values ...
Vox uses 220k's and a 0.0047uf cap

in an 18 watt with 470k grid reference resistors you use a 0.0022uf cap

I think that the grid leak value is less significant in determining the high end roll off break frequency than the plate resistance of the tube used in the LTP.

It's simplest to work out at the 'max cut' setting, in which the circuit forms a simple RC low pass filter, with the C term being the cut cap and the R being the circuit resistance, ie the 'source' resistance of the Norton equivalent for the circuit.

The circuit resistance will be (plate impedance // plate load resistor // grid leak resistor) x 2.
It 'x 2' due to the path to ground for the signal passed by the 'cut' cap being via the circuit resistance of the 'other' LTP plate.


So for Vox (62k5 // 100k // 220k) x 2 = 65k5, a 4n7F cap giving a break frequency of 517Hz
and for 18W (62k5 // 100k // 470k) x 2 = 71k1, a 4n7F cap giving a break frequency of 476Hz

So the same value of cap can be used in either amp without the break frequency being affected much.

As the 'cut' control is rotated away from the 'max cut' setting, resistance will be added in series with the cut cap, which will move the break frequency up and create a shelf or floor level in the filter's response.
The net result being that it become a frequency dependant potential divider, the R1 term being formed by the sum of LTP plate circuit resistances.

[M Fowler]

I use one value 250kL pot with .0047uf cap with great results.

[rooster]

Pdf64, your formula is interesting. Frankly it causes some confusion for me. Are you convinced that this 'break frequency' is what you say it is in either amp? And when you describe this as a 'break frequency', are you saying that the behavior of the control/circuit is like a low pass filter, the numbers you stated (517Hz and 476Hz) being the frequency that is not affected by the control?

In your equation you show the symbols '//' - what does this mean? I would like to do the math. But more confusion, when your equation delivers 65.5K, you inject the .0047uf cap and the resultant 517Hz break point. Where did this frequency come from? Is there math somewhere for this or a perhaps a chart? I would like to understand this. Thanks.

[JazzGuitarGimp]

// means "in parallel with". When paralleling more than two resistances, this formula works a trick:

1
----------------------
1 1 1
----- + ----- + -----
R1 R2 R3

You can solve for as many resistances as you like with this equation.

Once you have the total resistance (in this case 65.5K), and the capacitance (47nF), you can then solve for the break (or corner) frequency thusly:

1
Fc= ---------- where:
2piRC

Fc is expressed in hertz, R is expressed in ohms and C is expressed in farads.

1
So ------------------------------------- = 51.7 ohms
2 x pi x 65,500 x 0.000000047

Cheers,
Lou

[JazzGuitarGimp]

And I see the forum doesn't allow for multiple spaces between characters!

[pdf64]

See also http://en.wikipedia.org/wiki/Low-pass_filter
I used the term 'break frequency', whereas the wiki page uses 'cutoff frequency'; same thing.
There's many of useful online calculators at http://www.sengpielaudio.com/Calculations03.htm
eg http://www.sengpielaudio.com/calculator-RCpad.htm

[rooster]

OK, thanks for the links and the explanations.

Why did you figure for .047 (47nf), JGG? Typo?

[rooster]

For that matter, why do you write in your calculations:

'0.00000047' Where did this number come from and how is it relative to .0047uf, our target capacitance?

This confuses me still, sorry.

[pdf64]

I wouldn't get hung up about the particular values used in the example used by JGG.

My hunch is that the detail in my post may have been misread etc.

The key thing to take from it is probably the formulas for deriving f/R/C and parallel resistance, and the notation shorthand sometimes used.

Regarding the action of the cut control, to me the reality seems to match with the calculation; set to max cut, it certainly sounds like a low pass filter taking effect well below 1kHz.