using the Stupidly Wonderful Tone Control (SWTC) version 2 on the ROG Peppermill

[mordechai]

I think I'd like to try implementing this modification to the Peppermill, but I have a few questions.

1. on the AMZ website, the pot with lug 2 wired to the cap is titled "R1".  But in Mark Hammer's original SWTC, the pot is R2 which follows a separate R1, so I am not sure if the AMZ version has "R1" of the SWTC2 replacing "R1" of the original SWTC.  Here's the AMZ site with both:

http://www.muzique.com/lab/swtc.htm

Applied to the Peppermill, should I keep the 8.2K toward the tail end of the circuit and place the SWTC2 after it, or should I have the "R1" pot of the SWTC2 r-- say, a 10KB pot -- eplace the 8.2K resistor?

2.  Should the cap in the SWTC2 mod replace the .0022uF cap-to-ground?  Or should I keep that cap in place right before the 100KA volume pot and use a different value cap coming off lug 2 of the SWTC2 pot?

3.  Depending on the answers to points 1 and 2 above, should the cap in the SWTC2 mod be the same as the .0022uF cap in the original Peppermill?

[Mark Hammer]

The basic principle is that all consecutive (uninterrupted) resistances in series with the output level pot are treated as one big resistance.  If there was 50k leading up to the 50k volume pot's input lug, then the whole thing would behave like a 100k pot that is never turned above halfway.

At the same time, all consecutive series resistance leading up to the cap to ground forms a single-pole lowpass filter, whose rolloff is given by 1 / [2*pi*(Ra+Rb)*C].  Here, the "R" is divided into whatever portion of the total resistance that first bit of the tone pot makes up, and any additional series resistance leading up to it.

The Peppermill has, in a sense, already done what the SWTC does.  There is an 8k2 resistance in series with a 100k pot, comprising a 108.2k pot that can never be turned up beyond 92% of the way.  The 8k2 and .0022uf cap form a 1-pole lowpass filter with a rolloff starting around 8.8khz.

How to implement a variable rolloff?

Let's start with the 8k2 resistor being replaced by 1k in series with a 10k pot.  That turns our 108k2 "pot" into a 111k "pot" that will never be turned up beyond 90%.  So, we won't lose too much signal here.

When the wiper of the tone pot is rotated all the way over to the 1k resistance, we will have our highest rolloff point, and when the wiper is rotated the other way, so that there is now 11k of resistance leading up to the wiper, we will have our lowest rolloff point.  That will help us to find the nearest common cap value that gives a desirable range of tone control.

Using the stock 2200pf, we get a range of rolloffs from 72khz (i.e., no audible rolloff) to just under 6.6khz.  So, a little bit more treble trimming that what 8k2/2200pf gives but not enough to hear or be useful.  Clearly we have to think about using a bigger value cap.  The control will have an approximate 11:1 range, so we can afford to use a much bigger cap.  A value of .01uf will give a control range of 1.4khz up to 16khz.  A big chunk of that (from about 10khz upward) will be wasted, so we can afford to go even bigger.  A cap value of .015 (15nf) will give a range of control from 10.6khz down to 960hz, which still leaves a bit of meaningless twiddle room at the top but it useful across most of its range.  A .018uf cap gives is a range of 8.8khz down to around 800hz; i.e., our original rolloff point to something noticeably lower.

While 800hz may seem awfully low, the shallowness of the rolloff still leaves enough mids and top end to not sound muffled, just..."tamer".

[mordechai]

Thanks Mark, that's very helpful and very clearly explained.  If I understand clearly, then...the original .0022uF cap to ground is entirely replaced by the cap coming of the wiper of the variable resistor, and that cap should be rather higher (.018uF) before the signal goes to the volume pot?


I'll post a build report as I continue to noodle with the circuit...

[Mark Hammer]

Yes.

I am assuming, perhaps naively, that the intent is to tame the treble moreso than it currently is.

[mordechai]

Yep.  I just want to have the circuit be able to be a bit twangy and alternately a little darker.  Nothing too dramatic, though, beyond the inherent characteristic tonal character of the effect.

[mordechai]

Using the stock 2200pf, we get a range of rolloffs from 72khz (i.e., no audible rolloff) to just under 6.6khz.  So, a little bit more treble trimming that what 8k2/2200pf gives but not enough to hear or be useful.  Clearly we have to think about using a bigger value cap.  The control will have an approximate 11:1 range, so we can afford to use a much bigger cap.  A value of .01uf will give a control range of 1.4khz up to 16khz.  A big chunk of that (from about 10khz upward) will be wasted, so we can afford to go even bigger.  A cap value of .015 (15nf) will give a range of control from 10.6khz down to 960hz, which still leaves a bit of meaningless twiddle room at the top but it useful across most of its range.  A .018uf cap gives is a range of 8.8khz down to around 800hz; i.e., our original rolloff point to something noticeably lower.

While 800hz may seem awfully low, the shallowness of the rolloff still leaves enough mids and top end to not sound muffled, just..."tamer".

Just a bit more clarification here Mark -- are these directions for the SWTC version 2 as found on the AMZ website (where the cap comes off the wiper of the pot at one end and is tied to the signal path -- NOT to ground -- on the other) or are these directions for your original SWTC design where the cap goes right to ground?

[Mark Hammer]

for mine. But you may like jack's too

[mordechai]

Would the directions/modifications you suggested above work efor Jack's adaptation?  I.e., would I still have a smaller fixed resistor (~1K) before the pot (~100K), and would the increased-value cap going from the pot's wiper to the signal path before the volume pot thus replace the need to the cap-to-ground scenario you designed in the original SWTC?

[Mark Hammer]

To be able to do the calculations for the lowpass function, the tone cap to ground has to have some kind of minimum series resistance.

You will note that, of the three drawing Jack shows, #1 (the "classic SWTC) has an "R1" as the minimum series resistance, and so does drawing #3.  Both of those incorporate a lowpass function: #1 being only lowpass, and #3 having a combination lowpass/highpass function.  #2 is only highpass, so it does not really require the minimum series resistancve in this particular application.