Orange F.A.C knob frequency response

[pokus]

Hey there,
I was looking at the F.A.C knob by Orange and can't really figure out the frequencies there.
Does the 68n and the switchable capacitors in series form a high pass filter with the 100k? Or is there more going on?
I've got some problems recognizing what forms what when there are more than one or two caps and resistors next to each other.
Some also said that there is a gain of distortion, when switching through those presets. But how is that possible?
Thanks for help!

[antonis]

<Does the 68n and the switchable capacitors in series form a high pass filter with the 100k?>
In parallel with 220k grid bias resistor, also..

Strictly speaking, tube input impedance should also be involved in parallel combination but its very high value is severely dominated by 100k//220k so it can be safely ignored.. 

<I've got some problems recognizing what forms what when there are more than one or two caps and resistors next to each other.>
Generaly, when a cap is in series with signal path it can be considered as creating a High Pass Filter..
(there should be "somewhere" a respective shunt resistor..)
Same also stands for  series resistor and Low Pass Filter..
(even if there isn't an obvious shunt cap, various stray capacitances take care of it..)

<Some also said that there is a gain of distortion, when switching through those presets. But how is that possible?>
Although I can't get the "gain of distortion" term, distortion is depended on signal amplitude in conjunction with individual stage gain..
IF F.A.C. switch settings create distortion, that's due to frequency "selectivity" of Grid input signal voltage..
(Lows are more attenuated than Highs due to HPF, hence the seconds are more prone to distort - they income to next stage with higher amplitude..)

[pokus]

Alright, thanks a lot!
So I calculated that there is a lows cutoff starting from 33 Hz to 12 kHz. So I guess they've designed it to get good control of a wide range.
Just to be super sure, the grid bias resistor is the one going to ground?

With the term "gain of distortion" I meant that there is more distortion going on there. And I also think that the frequency shape let some assume that there's more gain or distortion like you've mentioned.

[antonis]

So I calculated that there is a lows cutoff starting from 33 Hz to 12 kHz.

Can't verify your calculations but, as long as you've considered as equivalent capacitance of 60nF cap in series with each one cap on selector the product of their individual values divided by the sum of them (caps in series should be treated like resistors in parallel) and as shunt resistor the parallel combination of 100k & 220K, you should be correct..

Just to be super sure, the grid bias resistor is the one going to ground?

Yeapp..

With the term "gain of distortion" I meant that there is more distortion going on there.

Actually, there is less signal going there..

P.S.1
The above are said considering signal coming from the left tube plate going to the right tube grid with one of jacks activated..
Also, didn't notice the 100k resistor between the two jacks input resistors..
(my bad..)

So:
In case of both jacks "open", RC should be calculated with R as series resistance of 100k + 220k ..
(the two 100k input resistors are not connected between signal path and AC GND..)

In case of one jack activated, RC should be calculated with R as (110k) // (100k + 220k)..
(no matter which jack 'cause resistor wiring results to same equivalent resistors setting..)

In case of both jacks activated (in case of such an option..), RC should be calculated with R as [100k] // [100k + (100k//220k)]

P.S.2
For a faithful gain calculation you have to also consider Plate resistor (220k) in parallel with any of following combination to find the Plate effective load.. 

Hint: Just follow all possible signal flow paths from a starting point of your taste till the end (which is right tube Grid..)..
Any item set between signal path and AC (or DC, of course..) GND at any specific node should be considered in parallel (shunt) - any item set before that specific node should be considered in series..

[pokus]

Oh maybe I should have added that, those are the send&return jacks of the amps. So is there a difference in tone shaping when they are connected or not connected? Or bypassed and connected through a pedal circuit?

[antonis]

those are the send&return jacks of the amps. So is there a difference in tone shaping when they are connected or not connected? Or bypassed and connected through a pedal circuit?

Can't say without amp Input & Output impedance info..

Considering a very high input (send) impedance and a very low (return), imagine send jack as open ciruit and return jack as geround short..

That guessed, left 100k resistor is effectively set out of circuitry while right 100k one is effectively grounded..

So, 68nF cap is set in series with any other series cap(s) selected, shunted by 168k75 resistor (100 + (100//220))

[PRR]

> can't really figure out the frequencies there.

Why do you care? I have only a rough idea what "Hertz" means to "tone".

It gives you 68,000pFd, 4,700pFd, down to 330pFd. A 200:1 spread. A 7+ octave range. More than the guitar can cover. Steps are only about an octave, which for a single-pole filter is "close ratio". The SP6T switch is cheap, none of the caps cost much. Build it and twiddle it.

Your image omits the point of the extra jacks. My insight is that the amp should play OK with nothing in the jacks. So the load on the caps is 100K+220K or 320K. The driving source is probably like 80K. (12AX7 with 100K load gives 60k rp and 38k effective source; so 220K would be about double.)

Distortion could change because the 3rd stage either has big bass, or doesn't. Big cap lets fat bass through to muddy-up the 3rd stage's overload; small cap feeds just the highs (typically weaker) for a different less-mud sound.

[pokus]

Alright thank you guys!
But why isn't it 100k//220k instead of 100k+220k, aren't they in parallel?
And how would the circuit look like when there were no send&return jacks? So that it acts the same.

[antonis]

But why isn't it 100k//220k instead of 100k+220k, aren't they in parallel?
And how would the circuit look like when there were no send&return jacks? So that it acts the same.

You already answered your query..!!

When send & return jacks are out of game, their 100k resistors are "floating" (no current pass) so the only path to GND is 100k in series with 220k..

[PRR]

> why isn't it 100k//220k instead of 100k+220k, aren't they in parallel?

[pokus]

Ok, got it now! Thanks for that!