Retaining treble in dirt pedals when you roll back on the gain control?

[Nocaster Cat]

Let's use a Distortion+ as an example. I like some of the lower gain settings but I find when I roll back on the gain control, the tone gets dull/a little muddy. Is there a way to implement something like a treble bleed mod that we use on guitars? Thanks.

[Gus]

Edit

You can make the high pass independent of the gain by making the capacitor in the gain control leg a greater value and adding a buffer before the distortion plus with a valuable high pass between the buffer and distortion section.  The gain setting will now not change the highpass AND you can control the highpass with another external control.

Something I posted in the past might help for ideas

[Groovenut]

You might try bypassing the gain pot with a cap from wiper (pin 2) to the junction of the 4k7 resistor (pin 3). a value of 470p to 1nF would be a good starting place.

[Tony Forestiere]

Good advice.

LPD: loved "Trial By Fire (Blackened Soles)". Great work. 

[FiveseveN]

Guys... the D+ is a very particular beast. As you turn up the gain control, the high-pass filter's rolloff frequency is also shifted higher, from 3 Hz to 720 Hz. Most other op-amp-based distortions keep this frequency fixed and vary the other branch of the feedback divider. In fact, if you want the same filtering independent of the gain pot's position that's probably the simplest mod you can implement.

[Mark Hammer]

Generally, it works the other way.  Let's take the TS-808 as an example.  The 51pf cap in the feedback loop of the clipping stage rolls off treble.  It interacts with the feedback resistance (51k + 500k pot) to shift that treble rolloff point downward as the gain is increased.  Of course since increasing the gain increases the treble, you essentially end up with high end that is comparable at higher gains to what is available at lower gains.

The gain adjustment of the Dist+ works in a somewhat opposite manner.  When the gain is increased via the ground leg, the treble remains unaffected, but the bass is rolled off as the gain is increased.  In that scenario, two things happen.  First, the clipping produced by more gain increases the treble content.  But as well, the loss of bass makes the treble all the more prominent.

One way to attempt to maintain a relatively even tonal balance as gain is increased is to use both paths simultaneously.  Instead of the variable resistance being ONLY on the ground leg, or ONLY in the feedback path, let's consider connecting the pot wiper to the inverting pin on the op-amp, with one side of the wiper being in the feedback path and the other side being on the ground leg.  We'll assume that there is also a fixed resistor in series with the feedback loop portion, and another resistor in series with the ground leg portion, along with a feedback cap, and a ground leg cap.

Why is this "better"?  Well, instead of doing all the gain chainging via the one path or the other, you change it in both manners at once, via the reciprocal action of the pot.  Gain is increased by higher feedback resistance, but it is also increased by dropping the ground-leg resistance.  Lucky for us, both those things happen at once.  It also works the other way.  As rotating the pot increases the ground leg resistance, it also drops the feedback resistance.

Let's do a gedanken experiment and use a 100k linear pot, with an added 1k resistor on the ground-leg end, and a 47k resistor on the feedback end.  We'll stick a 1uf cap on the ground leg side, and a 220pf cap in the feedback loop.

Okay, crank the gain so that all of that pot's resistance is on the feedback side.  Our feedback resistance is 147k, and our ground leg 1k, giving a gain of 148x; halfway between a TS-9 and a Dist+.  Our treble rolloff is at 4.9khz.  Our bass rolloff is at 160hz.  So, plenty of bite, and enough bass.  Rotate the pot the other way and you end up with 101k on the ground-leg side and 47k in the feedback, giving a gain of roughly 1.5x.  The treble rolloff is around 15.3khz, and the bass is at around 1.5hz.

Why is this better?  In the case of the TS-9, treble rollloff at the two extremes is 5.6khz (max gain) and 61khz (about 3.5octaves).  In the case of the Dist+, bass rolloff changes from 720hz (max gain) to 6.7hz (about 6.5octaves).  Ignore the specific frequencies, and focus on the amount of change at each end.  The change in treble rolloff is a littleover 1.5 octaves in our revised arrangement, although our bass rolloff changes by almost the same ratio as with the Dist+.

Lets skip the lowest gains by placing a 22k resistor in parallel with the ground leg, so that the ground leg resistance ranges between 1k and 23k instead of 1k-101k.  This makes our highest gain setting the same, but our lowest gain setting 3x, and our bass rolloff at 7hz at min gain, instead of 1.5hz.  You obviously won't hear that change, but it serves to illustrate a principle.: we've reduced the relative change in bass rolloff accompanying gain change.

To sum up, by making use of both sides of the wiper, in different roles, and being selective about cap values, and making clever use of parallel (and series) resistance, we can make an op-amp stage that has a pretty decent range of gain settings, without changing either bass or treble rolloff ponts all that much.

[dschwartz]

How about fixing the gain (and freq response) at 500x or so, and using a volume control before de opamp, like traditional preamps? Of course will need a buffer at the input to isolate the guitar..

[Groovenut]

You might try bypassing the gain pot with a cap from wiper (pin 2) to the junction of the 4k7 resistor (pin 3). a value of 470p to 1nF would be a good starting place.

This is kind of what I had in mind for a DS-1 type circuit


Freq plots for the circuit at differing gain pot levels


Adjust the resistor level for the amount of treble boost you want.

The plot at 100% gain is what all gain levels look like (without the increasing dBs) when the circuit is stock

[Groovenut]

You can try this for the Dist+

No free lunch of course, but it may do the trick. Mess with the series resistor to alter the amount of treble bleed.

[samhay]

>One way to attempt to maintain a relatively even tonal balance as gain is increased is to use both paths simultaneously...

I have done this a few times. The most recent and complicated is in the DiSCO (there are frequency plots on pp 3): http://www.diystompboxes.com/smfforum/index.php?topic=110346.0