Tone pot in Brian Wampler's modded MXR Distortion +

[fishfude]

Ah ok, that makes sense, thanks.

So how does changing C3 allow for the ''Same frequency response''?

Here's the article in case anyone is interested: http://www.premierguitar.com/articles/mxr-distortion-plus-mods-1

Cheers,

ff

[Mark Hammer]

In a non-inverting op-amp configuration (and remember, any op-amp can be configured as inverting OR non-inverting), the gain is set by the ratio of the feedback resistance to the ground-leg resistance (i.e., the resistance from pin 2 to whatever is being used for ground).

The feedback and ground-leg resistance act like a potentiometer, and "divide" the feedback from the output back to the input.  Since the feedback is negative, having more feedback reduces gain, and bleeding off more feedback increases gain.  So, when the path from pin 6 to pin 2 is a straight wire, and pin 2 to ground is open, there is no gain at all, since all negative feedback is preserved.  Conversely, when the feedback resistance is much greater than the ground-leg resistance, it's like a volume pot that has been turned way down, and the amount of feedback available at the junction ("wiper") of those two resistances, at pin 2, is pretty small, hence less negative feedback, and more gain.

The trick to thinking about such non-inverting stages is that treble rolloff is dictated by caps in the feedback loop, and bass rolloff is dictated by caps in the ground leg (C3 in your drawing).  Where the rolloff occurs is given by the formula F = 1 / (2 * pi * R * C).  Still thinking about that how-much-negative-feedback-is-retained/lost aspect, note that a cap in parallel with the feedback resistance will provide an expressway for high-freq negative feedback, while a cap in series with the ground leg will dictate how much negative feedback is retained or lost above/below a given corner frequency.

Leaving both R4 and C3 constant, as you make R2+R3 smaller, the gain is increased (more negative feedback is bled off), but the feedback bleedoff tend to be more confined to mids and highs.  If one holds C3 and R2+R3 constant, and varies the value of R4, the gain will change, but there will be no impact on low end, as there is when you vary R2+R3.

How does this play out?  On a Dist+, you lose bass as you turn up gain, because it uses the ground leg resistance to adjust gain.  On a Rat or TS, you lose treble as you turn up gain, because they hold the ground leg constant and vary gain with the feedback resistance.  Neither is better or worse, but each can be used strategically to achieve the sound you want.

Another way to tackle it is to connect the wiper of the gain pot to pin 2, such that resistance is removed from the ground leg as it is added to the feedback path.  That can often be used to vary gain over a respectably wide range without affecting the bass and treble rolloffs too too much.

[fishfude]

Wow, thanks Mark for taking the time to explain that. It clears so much up for me! I really appreciate it. I think I'm finally ready to finalise this beast!

[slacker]

So how does changing C3 allow for the ''Same frequency response''?

It doesn't, if you just changed C3 the frequency response would change. The frequency response is set by C3, R2 and the setting of R3, at maximum gain R3 is 0 so the response is set by C3 and R2, like the article say, this boosts frequencies above about 720 Hz, the gain these frequencies get is roughly R4/R2 1M/4.7k = 212.

If you change C3 to 0.22uF and R2 to 1k, then that combination allows the same frequencies to be boosted only now the gain is about 1M/1K = 1000 so you have more gain.

EDIT: looks like Mark beat me to it, but hey I've typed it so I might as well post it Try sticking the different values into the equation Mark gave and you'll see .22uF/1k is the same frequency as 0.047uF/4.7k.

[fishfude]

Thanks Slacker. I threw those digits in there and, YUP, same result .

Cheers

ff

[Mark Hammer]

EDIT: looks like Mark beat me to it, but hey I've typed it so I might as well post it

Been there, done that, bought the t-shirt, worn it out, and used it to polish the car. 

Fishfude, Just note that since the bass-rolloff is a function of both the resistance and C3 cap value, it is possible to maintain the same low-end rolloff by increasing the cap value as you reduce the resistance.  So, holding the feedback resistor constant at 1M, a 10k resistance and .047uf cap on the ground leg would give us a gain of (1M+10k)/10k = 101x, with a bass rolloff of 338hz.  A .01uf cap and 47k resistance to ground would also give us a rolloff at 338hz, but a gain of (1M+47k)/47k = 22x.

I have my suspicions that, historically, since the Dist+ was released before the advent of hum-rejecting single-coil-style pickups (dual rails, stacked humbuckers, etc.), the loss of bass response as the gain was cranked had the side-effect of making any 60hz hum not quite as objectionable.  That may not have been the deliberate strategy, but simply one of those things where someone went "Hey, that doesn't sound nearly as hummy as I thought it would".  Additionally, the Dist+ does not have gobs of gain (max is 213x), so shaving off the bass at highest gain settings makes the added treble seem that much more prominent.

[fishfude]

Got it. So what I was thinking was going Bat crap crazy and having a 470R resistor for a gain of ~x2128.659574468085 

I've been messing with the cap to ground and I kinda like 3x 0.1uF caps in parallel (0.3uF). So that gives me a bass rolloff at 1/2*pi*R*C = ~442Hz.

Mark, I had decided to go with your original, Stupidly Wonderful Tone Control, with R1 1K, R2 A500k and C1 0.01uF. My R3 is A100k. I though this would have given me a corner frequency of 15923.6 when turned completely down. But I couldn't help but get lots and lots of treble attenuation from the start, which just got more extreme as I turned the pot (until interestingly the last 10% where some came back - at a guess this is because of R3?). My R1  (1K) is replacing R6 in the original schematic (originally 10K).

I'm tempted to just leave out the tone control or maybe give the SWTC2 a blast.

[Mark Hammer]

1) the drawing in the Premier Guitar article is sort of correct, but mostly wrong.  Initially, MXR used a 10k pot but lost too much volume with that, so higher pot values are recommended, such as 50k or 100k (log, of course).

2) The original SWTC should usually employ a tone pot of a lower value than the volume pot that follows it.  So if the volume pot as 100k, you probably don't want to use a tone pot any higher than 50k, and would probably do better with 10k-20k.

3) The choice of cap to ground also has to factor in the 10k leading up to the clipping diodes.  So, if one was to use a 20k tone pot in conjunction with a 1k resistor in series with it, and a .01uf cap from the wiper to ground, the minimum rolloff is happening at  1 / (2 * pi * .01uf * .011Meg) = 1446hz, and the max rolloff will start around 513hz, because the 10k resistor is included in the calculation.  The diodes to ground are just a distraction when making this calculation.

So here is what I'd suggest.

• drop the 1k min resistance because you have the 10k in there anyway
• use a 50k tone pot and 100k volume pot (using a clipping diode complement with a higher clipping threshold than the original pair of 1N270s will yield a high-enough output level that you can afford to lose a little via the tone control)
• use a 3300pf cap to ground; at min resistance (the cap is essentially placed in parallel with the clipping diodes), the rolloff will start around 4.8khz, and at max resistance (10k + 50k) the rolloff will start around 803hz

That should get you a reasonable range of tone settings, without too much signal loss.  Keep in mind that's a 6db/oct rolloff, so there will continue to be some grind in there, no matter what setting you have it at.

[Gus]

Something I posted before