Gain control in Bluesbreaker based circuit adds volume and treble. Why?

[skiraly017]

I've seached for this issue but have not found it.

The project started off as a GGG Bluesbreaker with a few of the more commmon mods. I liked it but it wasn't quite "there" so I continued experimenting and adjusting. I now have the ciruit where I want it tonally but I've run into a problem that has me stuck for a solution. The gain control not only increases the gain, but also increases the overall volume and seems to add a little treble as well. If I sub the 100k gain pot for a 10k (I worked down to the 10k through trial and error) I can almost eliminate the volume/treble increase issue but then it seems that the full sweep range of gain is reduced or is more subtle. Is any of this making sense? Any input would be greatly appreciated as I'm sure I'm stumbling over something very simple. Thanks.

[Mark Hammer]

You will note that in the Bluesbreaker, the wiper of the Gain pot is tied to the output of that first op-amp.  What that means is that when you move the wiper, it simultaneously adjusts both the feedback resistance of the first stage and the input resistance of the second stage.

In the first stage, reducing the feedback resistance decreases gain.  because the second stage is an inverting op-amp stage, changes to its input resistance also alter the gain; smaller input resistance = more gain.  The way the stock circuit works, moving the wiper clockwise increases the stage-1 feedback resistance and decreases the stage-2 input resistance at the same time, increasing their gains simultaneously.

Now, it also happens that in an inverting op-amp, factoring out all the other factors that can influence it, the input signal to that stage is also a function of the joint action of the input resistor and capacitor.  For any given input resistance value, making the cap value smaller raises the point where the low-end is being rolled off, however it does not affect the gain per se.  In contrast, making the input resistance value smaller, while holding the cap value constant, results in both the gain changing (increasing), at the same time that the low end rolloff starts to move upward.

So, ignore all the complications and let's assume, hypothetically, that the input to IC1b consisted only of a 4k7 fixed resistor and the .22uf cap.  With a 220k feedback resistor, that gives a gain of x46.8 (that may not seem like much, but multiplied by the gain of stage 1 it gets big fast).  Where is the low end rolloff?  Around 154hz.  So, not too much bass cut.  Okay, now imagine someone wanted to goose the gain a bit, and decided to do it via reducing the input resistance from 4k7 to, say, 3k9.  At the same time, the 220k feedback resistance is held constant.  The stage gain goes from just under 47 to just over 56, and the low-end rolloff moves up to 185hz.  So, you can see that one starts to lose bottom end as the gain is increased in this manner.

Does that explain what's happening to your pedal?

[skiraly017]

Nope, low end stays pretty constant. I really like the ways it sounds right now and the reduced value for the gain pot is more than liveable. My question was more for personal knowledge gain and a fix if possible more than "Why is this not working!". I suspect an adjustment can be made somewhere to fix the issue but I just haven't messed with the right part(s) yet.

[d95err]

The gain control not only increases the gain, but also increases the overall volume and seems to add a little treble as well.

This happens in every single distortion/overdrive/fuzz I've ever seen. Gain is essentially the same as adding volume, until you get massive amounts of clipping. With distortion comes added overtones which sounds like more treble. So, from your description it sounds completely normal.

[skiraly017]

The gain control not only increases the gain, but also increases the overall volume and seems to add a little treble as well.

This happens in every single distortion/overdrive/fuzz I've ever seen. Gain is essentially the same as adding volume, until you get massive amounts of clipping. With distortion comes added overtones which sounds like more treble. So, from your description it sounds completely normal.

Except I'm talking about a very large increase in volume, enough that I need to turn the Volume control down. I should have stated that earlier, sorry.

[John Lyons]

Since the BB is using a two diodes in series per side and anti parallel, there is a lot of gain that is making it past the diodes and bringing up the level. Diodes in series will have a larger threshold etc etc.
You could drop the gain amount and reconfigure the diodes to clip sooner...

For the treble increase one thing you could do is use a dual pot and when the gain knob is turned up the treble gets dropped.
Or use a cap across the wiper and ground, but that would only work from 0-50% ish of the pots travel.

There's got to be something simpler though.

John

[johngreene]

The gain control not only increases the gain, but also increases the overall volume and seems to add a little treble as well.

This happens in every single distortion/overdrive/fuzz I've ever seen. Gain is essentially the same as adding volume, until you get massive amounts of clipping. With distortion comes added overtones which sounds like more treble. So, from your description it sounds completely normal.

Except I'm talking about a very large increase in volume, enough that I need to turn the Volume control down. I should have stated that earlier, sorry.

The reason for the volume increase is the 6.8k resistor in series with the diodes. This allows the output 'clipped' level to continue to increase with an increase in input level. The diodes have a fixed voltage drop, the resistor is forming a voltage divider with the input resistance and the voltage across it will be proportional to the input level. This is also the primary difference between diodes in the feedback loop of an inverting amplifier as opposed to a non-inverting amplifier. If there is no resistance in series with the diodes, Once the diodes conduct, the '-' input is clamped hard in the inverting amp, while with a non-inverting amp, the '-' input will exactly follow the '+' input. You can actually make an attenuator with an inverting amp but you will always have at least unity gain with a non-inverting amp. This is readily seen from the gain equations: inverting gain = Rf/Rin  non-inverting = 1 + Rf/Rin. That added '1' in the non-inverting amp means it will always pass signal. 

A good way of thinking about this is that the opamp is going to do whatever necessary to keep the '-' input equal to the '+' input. So if the input increases in level, the opamp is going to increase the output level to cancel it completely at the '-' input (so it matches the 4.5V DC level at the '+' input). Once the diodes conduct, the output follows the input in portion to the 6.8k Rf and Rin, below the conduction threshold of the diodes, the voltage at the output needs to be much higher because of the 220K/Rin voltage divider. As Mark points out, Rin changes depending on the Gain pot setting.

The more treble thing might just be a perception thing having to do with driving the diodes harder. The 47pF across the gain pot should roll off at a lower and lower frequency as the gain is increased. You could try increasing the 47pF as you have the gain set to a low gain setting until you start noticing it rolling off the highs. Back off one value so now it will roll more highs off as you turn the gain further up. The point at which you set this 'cross-over' is up to you.

--john

[skiraly017]

The gain control not only increases the gain, but also increases the overall volume and seems to add a little treble as well.

This happens in every single distortion/overdrive/fuzz I've ever seen. Gain is essentially the same as adding volume, until you get massive amounts of clipping. With distortion comes added overtones which sounds like more treble. So, from your description it sounds completely normal.

Except I'm talking about a very large increase in volume, enough that I need to turn the Volume control down. I should have stated that earlier, sorry.

The reason for the volume increase is the 6.8k resistor in series with the diodes. This allows the output 'clipped' level to continue to increase with an increase in input level. The diodes have a fixed voltage drop, the resistor is forming a voltage divider with the input resistance and the voltage across it will be proportional to the input level. This is also the primary difference between diodes in the feedback loop of an inverting amplifier as opposed to a non-inverting amplifier. If there is no resistance in series with the diodes, Once the diodes conduct, the '-' input is clamped hard in the inverting amp, while with a non-inverting amp, the '-' input will exactly follow the '+' input. You can actually make an attenuator with an inverting amp but you will always have at least unity gain with a non-inverting amp. This is readily seen from the gain equations: inverting gain = Rf/Rin  non-inverting = 1 + Rf/Rin. That added '1' in the non-inverting amp means it will always pass signal. 

A good way of thinking about this is that the opamp is going to do whatever necessary to keep the '-' input equal to the '+' input. So if the input increases in level, the opamp is going to increase the output level to cancel it completely at the '-' input (so it matches the 4.5V DC level at the '+' input). Once the diodes conduct, the output follows the input in portion to the 6.8k Rf and Rin, below the conduction threshold of the diodes, the voltage at the output needs to be much higher because of the 220K/Rin voltage divider. As Mark points out, Rin changes depending on the Gain pot setting.

The more treble thing might just be a perception thing having to do with driving the diodes harder. The 47pF across the gain pot should roll off at a lower and lower frequency as the gain is increased. You could try increasing the 47pF as you have the gain set to a low gain setting until you start noticing it rolling off the highs. Back off one value so now it will roll more highs off as you turn the gain further up. The point at which you set this 'cross-over' is up to you.

--john

Thanks John. So if I understand this correctly (and there's a good chance I don't) then the 6.8k resistor would have to be variable and increase/decrease as needed to truly attenuate the drive/volume properly? My second question...is it worth messing around with the 6.8k and if so can I do it safely with blowing something up? Thanks.

Edit - I should also say that the increased treble is okay. I wish I could keep it at the perceived level as I decrese the gain.

[johngreene]

Thanks John. So if I understand this correctly (and there's a good chance I don't) then the 6.8k resistor would have to be variable and increase/decrease as needed to truly attenuate the drive/volume properly? My second question...is it worth messing around with the 6.8k and if so can I do it safely with blowing something up? Thanks.

Edit - I should also say that the increased treble is okay. I wish I could keep it at the perceived level as I decrese the gain.

Unfortunately not quite that simple. The only setting that will give you no increase in volume with increase in gain is zero. Also, the character of the clipping will change depending on the value. So in getting the volume to be constant, you will end up changing a lot of things you probably don't want to. It's kind of like squeezing a balloon. When you squeeze it in one place, it is going to bulge out somewhere else.

To achieve what you want, a better choice would be a dual-ganged pot for the gain control that has the second pot in series with the volume control. You could parallel the compensation pot with a resistor to get the right amount of change. If it was me, I was just accept that this is the way the circuit works and deal with it. A little time with the controls and you will know that with 'x' amount increase in the gain control, the volume pot needs to drop by 'y'.

Try messing around with the 6.8K. I don't think you can blow anything up, even if you short it out altogether. If you short it out, it will be almost the same as having the diodes shorted to ground on the output. The only difference being with the diodes in feedback loop, the opamp is never going to hit its own rails. But you might find a different value that you like the sound of more.

--john

[John G]

Hi,
As already mooted by Mark in the past try adding caps in the 100pF-1000pF range across the gain pot lugs 2-1 for boost at lower gain settings, lugs 3-2 for hi cut at higher settings. (did I get hat the right way around )You can also put a resistor in series with the cap to control the maximum amount of boost or cut. Try half the pot value (47k)to start with.
Cheers
John G

[DougH]

As already mooted by Mark in the past try adding caps in the 100pF-1000pF range across the gain pot lugs 2-1 for boost at lower gain settings

I've done this before with the BB and it works fine. It's a simple brite-cap mod.

[John G]

Hi,
Then again it could have been Doug doing the "mooting"........my bad, my only excuse is that I live at the bottom of the globe, and as you would expect we walk around upside down, and the blood tends to gather around the head....clear tinking is always a struggle.
Cheers
John G