Help me understand - what do these do?

[antiuser]

I was looking at a schematic for a Zvex Box of Rock, and it looks like it's just a few SHOs chained together. A couple of things caught my eye though (highlighted in yellow). I would greatly appreciate it if someone explained to me what these sections do. I know the one by the output is an R-C filter of sorts, but why use parallel capacitors and why chain two of the same filters like this?

[antonis]

First on the left is for raisisng Q2 Gain at high frequencies..
(by shunting R6, hence lowering its effective value in denominator of closed-feedback loop..)
You may also consider it as treble booster..

The other on the right should be designed for -12dB/octave slope of LPF, if only there was a buffer between the 2 distinct configurations..(or R18 was 100k & C11/C22 100pF, at least..)
As it is, it needs transfer function analysis, 'cause 2^nd configuration severely loads 1^st one..
IMHO, it's just an attempt to use linear devices (like resistors & caps) in place of non-linear ones (like diodes) for clipping purpose..

[Marcos - Munky]

The one on the right are two low pass filters in series. I recall seeing this on some Runoffgroove's designs and on BSIAB2. In fact, this one is exactly as the output low pass filters from BSIAB2.

Why two caps imparallel? Well, two caps in parallel add the values together, so you have 2nF in each filter. Maybe Zachary was out of 2n2 caps, or maybe he had loads of 1nF caps (which was more likely).

[Mark Hammer]

Second-order, or 2-pole, lowpass filters have a steeper rolloff.  Why is that a good thing?
Easy.  If the rolloff is shallow, then you have to start the rolloff lower down in the spectrum in order to achieve the degree of treble cut you want in the range of frequencies desired.

As shown, a single pole, comprised of 10k and 2nf, starts its rolloff just under 8khz, yielding only 6db less treble content at 16khz.  So, given the frequency response of most guitar amps, would you even hear that sort of rolloff?  Not easily.  Two poles of lowpass at just under 8khz is no sort of "brick wall", but provides enough attenuation of any content above 8khz that you would easily hear the difference if you removed it.

[antonis]

Quite right Mark, but here we have 1st pole severe loading by 2nd one..
So, it's complicated a bit to calculate break frequency on parer..

[ElectricDruid]

+1 agree with both Mark and Antonis. It's a 2nd order/2-pole filter, but with passive filters like this, the second stage is going to load the first one.

The rule of thumb I picked up to keep this effect manageable is to have the resistors ten times bigger in the following stage. Like that, the loading isn't so significant. Obviously this requires scaling the caps down by the same factor if you're to match the cutoff frequencies.

So I'd have gone with 10K/2n2 followed by 100K/220pF.

[POTL]

The first filter imitates filtering in Marshall amplifiers, but the filter alone will not make this circuit sound even remotely reminiscent of Marshall.
The second filter cuts high frequencies, noise, hissing transistors, unpleasant sounds at very high frequencies.
All field effect transistors (MOS & J) require this filtering to make the sound enjoyable.