increasing the drive range of the drive pot in a Tube Screamer circuit

[mordechai]

I'd like to increase the available gain I can get on the drive pot for my TS project.  I currently have a 500K pot in place, and I'd like more gain, but I don't want to use a 1Meg pot (too much!). 

I tried putting a 220K resistor onto lug 3 of the 500K pot, but this actually seemed to limit how much drive I could get out of it.  I would have thought that the additional resistor would have "turned" it into a 720K pot -- was I incorrect about that?

Any thoughts on how I can avoid a 1Meg pot and coax more gain out of the 500K?

[Mark Hammer]

You can always stick a fixed resistor in parallel with a 1M pot to get a max resistance in between 500k and 1M.  Alternatively, consider using a toggle to stick something like a 220k-390k fixed resistor in series with your existing 500k pot.  That way, you can have two ranges of gain, with a little more dialability.  I might point out that one could easily use a DPDT stompswitch to go between the higher and lower gain settings.

As far as implementing goes, you have your 51k fixed resistor, the suggested additional fixed resistor, and the 500k pot.  The gain-range switch simply bridges/shunts the extra resistor (low-gain) or undoes the bridge (high-gain).

And since I can't resist putting more food on your plate than you can eat, note that the impact of a 1M pot can be easily offset by changing the value of the resistor on the ground leg.  Stock, the 4k7/47nf combo provides for a maximum gain of 118x and a low-end rolloff at 720hz.  If you make that resistor 6k8, you get a max gain of 156x and a low-end rolloff about 500hz.  So you get a less pronounced mid-hump and a usable amount of extra gain that lets you presevre what the TS does well.

[ashcat_lt]

Where was the other end of this 220K resistor? Increasing the 51K in the feedback path will slide the whole range upward, but not really give you a greater range of sweep.  In fact, since its kind of a logarithmic thing, you'll probably have less noticeable change between min and max.

Reducing the the 4K7 to ground from the feedback loop will probably get closer to what you want, but will also change the cutoff of the filters there, so you might need to scale up the caps to compensate.

Could you use a 1M linear pot with an appropriate resistor between the two normally used lugs to taper it and bring the value down closer to what you're looking for?

Doesn't do what you asked, but worth a shot - in my Rat I cut the track inside the 100K pot at the high end and then stuck a 180K resistor across the two lugs.  This way, the pot goes from 0 to almost 100K and then jumps to 180K.

All this said, my TS-type thing has a 500K resistor which I can switch in series with the gain pot so that it's kinda got a dual range: 0-500K and 500K-1M.  This might be an idea for you too, but I bring it up here because in my case the difference between the two extremes of the high range is surprisingly subtle.  Like I said above, its a logarithmic kind of thing, and while 1M - 500K = 500K - 0, 1M/500K is significantly less than 551K/51K.  Are you sure that 1M is that much more too much than 720K?

[mordechai]

Thanks for the quick reply Mark and Ashcat.  I had a small-scale epiphany a moment ago.   Increasing the gain of the pot does not necessarily increase the drive/distortion of the circuit, does it?  In fact, if my teency, reptilian brain is serving me correctly here, increasing the circuit's gain can actually decrease the amount of distortion the circuit can produce, n'est ce pas?  So perhaps I had less crunch coming out of the circuit BECAUSE I had upped the gain pot.    Would a lower value pot -- like, 250K -- yield more actual distortion??

I've already lowered the 4.7K/47nf combo to 2.4K/100nf, so that's given more more drive already.  But I thought the pot change might also help...

[ashcat_lt]

Well, Mark ninja'd me!  I ended up duplicating his suggestions, as well as one you've already tried.  I think my cut track/sudden jump approach is novel, though?

The way the TS does its thing is interesting and kind of complex, and definitely lends some credence to your question:  Does it clean up with more gain?

It's basically a form of crossover distortion, in that the signal swings closer to the zero crossings are affected while the louder peaks go through pretty much as normal.  The edges get steeper with more gain, but also more of the peaks get through at something like unity.  This would tend to lead to the idea that as you increase gain in the circuit the percent of time it spends in the distorted region actually decreases.  But...  The edges still keep getting steeper, and the breakover gets sharper and sharper, so you should get more of the higher harmonics even as it seems to get more transparent...or something...

[mordechai]

Don't feel bad, Ashcat, you both are ninja-ing the hell out of me. 

Anyway, here's what I finally came up with since last checking in:

4.7K/47nf changed to 2.4K/68nf -- good midrange, but not as "humpy"

clipping diodes: a Green LED against 2 x 1n4148 -- a more "open", bigger sound, crunchy but not sharp, and not as compressed as stock

one tant changed to a 47nf poly film cap, the other changed to an 18nf poly film cap, tone pot changed to a 10KB -- a more friendly tone control

input cap lowered from 47nf to 33nf -- focuses some more high range in just a bit for better treble response, works well with the new diode arrangement

500KB pot changed to 500KC -- better drive sweep throughout the range.

But the biggest change in terms of good drive characteristics was when I swapped out the op-amp to a stacked 4558DD/4559D combo.  That REALLY made a difference and I'm very pleased with how all of these interact to give the thing more muscle.  It still sounds like a TS, but doesn't have as nasal a quality and makes my little Fender Blues Jr. sound like a much more substantial machine.

[Gus]

Increasing the gain potentiometer or fixed feedback resistor value is not going to give more gain.

Remember this is a feedback diodes limited type clipping effect.
 
Once the diodes clamp the signal you don't get much more if any level change.

All increasing the feedback resistor(s) does is make the gain increase faster at the beginning of the gain control rotation.

If you want more level from the clipper, you need to increase the diode clipping level with more or different diodes

Leaving it stock give finer control of the gain setting.

[Mark Hammer]

We run into this blurring of the difference between "gain", "level", "distortion", etc., quite frequently, don't we?

My own sense from mordechai's OP was that he was looking for "more gain'" in the sense of being able to produce more distortion than stock (although I sensed a desire to get a little more level too).

Gus is absolutely correct that the clamping action of the diodes in the feedback path means you won't get much audible increase in level by simply increasing the feedback resistance more and more.  He is also absolutely correct in asserting that increasing the clipping threshold via changes to the type and/or number of diodes used will provide greater output level.  The cost of that will be confining clipping pretty much to the start of the note, with less during the sustain and decay portions of the note, but that can be fixed

One of the attributes that makes pedals like the Timmy, Zendrive, Lovepedal Eternity, the Xotic boosters, and so many other TS-derivatives appealing to players is their ability to produce hotter outputs without necessarily producing gobs of distortion.  That lets the user push the amp harder, and yield tones that are a nice combination of both pedal and amp breakup..

[mordechai]

I agree that the diode changes I implemented reduce the clipping for the duration of the note -- I saw that almost immediately with the LED/2x1N4148 combo I used.  But I think that's where the lowering of the 4.7K to 2.4K on the one hand and stacking the op-amps on the other hand made a difference.  And really, I was surprised at how the op-amp stacking helped here.  With the original 4558DD, it had a nice and strong tone, but when I added the 4559D to the mix, it REALLY beefed up.  So I think whatever distortion I lost with the diode change was really recovered with these other changes, or at least that's how it sounds to my ears.

[Rawk100]

I've already lowered the 4.7K/47nf combo to 2.4K/100nf, so that's given more more drive already.  But I thought the pot change might also help...

The cascade switch on the Jacques Overtube which gives more gain (is supposed to emulate 2 x TS stacked together)  has a 0.6k/0.1uf combination (I think).....