Rat gain pot sounds like a wah

[Beyer160]

I just added a clipping diode switch (1N4148s and LEDs) and a Reutz mod switch (1K pot in/out of the circuit) to my Rat2. It all works and sounds fine, but the gain pot sounds like a wah when you adjust it. I've never run into this one before, anyone have any thoughts?

[dschwartz]

It haven't happened to me. Bu i remember simulating a rat type distortion on lt spice and noticed a sharp peak near 1kHz..like a wah..this happened when the feedback shunt resistor was too low

[stonerbox]

I just added a clipping diode switch (1N4148s and LEDs) and a Reutz mod switch (1K pot in/out of the circuit) to my Rat2. It all works and sounds fine, but the gain pot sounds like a wah when you adjust it. I've never run into this one before, anyone have any thoughts?

Yep, this is the way the RAT works.. the higher the gain the higher the high pass filter goes up.

[GGBB]

Yep, this is the way the RAT works.. the higher the gain the higher the high pass filter goes up.

This is true for ac-coupled non-inverting op-amp circuit designs that change gain by varying ac resistance between the inverting input (pin 2) and ground while maintaining a constant resistance between the inverting input and the op-amp output (pin 3). However, the RAT uses the opposite configuration: the inverting input to ground ac resistance is constant and gain is adjusted by varying the resistance between output and inverting input. So in the RAT, changing gain does not change the high pass filter.

Beyer160 - something else is at play here - you shouldn't have this behavior. Can you describe it more (what conditions cause it)? Can you post pics of your mod work?

[anotherjim]

Unless the gain pot is larger than stock 100k, or the feedback cap larger than stock 100pF, the filter these make is well above guitar amp bandwidth.
A guess -  because the Rat is non-inverting, there is a risk of positive feedback from out to in via some of the wires being close to each other. Is it wired like this?

[stonerbox]

Yep, this is the way the RAT works.. the higher the gain the higher the high pass filter goes up.

This is true for ac-coupled non-inverting op-amp circuit designs that change gain by varying ac resistance between the inverting input (pin 2) and ground while maintaining a constant resistance between the inverting input and the op-amp output (pin 3). However, the RAT uses the opposite configuration: the inverting input to ground ac resistance is constant and gain is adjusted by varying the resistance between output and inverting input. So in the RAT, changing gain does not change the high pass filter.

What.. well I guess I was wrong then!  My bad! Strange cause my RAT acts just as Beyer160 describes his. When the gain is low the tone is natural/bassier and as you turn it up the sound gets thinner, not really thin but just tighter.

[GGBB]

Strange cause my RAT acts just as Beyer160 describes his. When the gain is low the tone is natural/bassier and as you turn it up the sound gets thinner, not really thin but just tighter.

Yes - my RATs do that too. But I attribute that to increased harmonic content due to higher distortion levels, not to the high pass filter. Beyer160 needs to explain exactly what he's hearing. I have been assuming that what he is hearing started after he modded the pedal.

[ashcat_lt]

What.. well I guess I was wrong then!  My bad! Strange cause my RAT acts just as Beyer160 describes his. When the gain is low the tone is natural/bassier and as you turn it up the sound gets thinner, not really thin but just tighter.

No, you were right.  This cap-to-ground (with resistor in series) is actually a shelving low-pass filter (and actually there are two of them in a Rat), and the cutoff quite definitely does move with the gain pot! 

See:


Likewise, the cap across the feedback loop is a shelving HPF, and its cutoff also moves.  That one doesn't touch quite so much of the audible frequencies, but at high gain settings there is some roll off of the highest frequencies coming out of the guitar.

[Beyer160]

To describe the problem a little better, at the lowest gain settings the sound was thin and nasal. As you raised the gain, it sounded like rocking a wah pedal- a narrow boost that got lower in frequency as you turned the knob up. Turning the knob up and down quickly produced the classic "wah-wah" sound.

It didn't have this problem before so I knew I must have screwed something up, I just couldn't figure out what would make it do that. I got back into it today intending to take some pictures and post some more detail, but while I was in there I figured out the problem. I'd used a 1K trimpot on a switch for the Reutz mod, which I think is way too sensitive in this application- just touching the pot noticeably changes the value, and forget about setting it to a precise value. I'd set it around the stock value of 47r, figuring I'd play with it when it was in the circuit. Somehow in installing the switch, I must have bumped it and changed the value to somewhere around 5r. Replacing the trimpot with a 47r resistor fixed the problem as far as I can tell on my crappy bench amp- I'll know for sure once I can crank it up. Now all I gotta do is fix the intermittent cutting out, which I'm fairly sure is the switch. 

Thanks for all the feedback!

[GGBB]

No, you were right.  This cap-to-ground (with resistor in series) is actually a shelving low-pass filter (and actually there are two of them in a Rat), and the cutoff quite definitely does move with the gain pot! 

See:


Likewise, the cap across the feedback loop is a shelving HPF, and its cutoff also moves.  That one doesn't touch quite so much of the audible frequencies, but at high gain settings there is some roll off of the highest frequencies coming out of the guitar.

The pic you linked is the passive version of an active low-pass shelving filter. The full section is here: http://www.linkwitzlab.com/filters.htm#5. As you can see from the other diagram of the active version of the filter:



the RAT is quite a different circuit:



It is in fact a shelving high-pass filter: http://www.linkwitzlab.com/filters.htm#6.





Unfortunately the diagrams aren't labelled consistently and the formulas for the active version aren't shown (which 5.11k resistor is R1 and which is R2?). So that leaves me a little confused, but either the knee doesn't change or it goes down as gain goes up. I'm fairly certain it doesn't go up. I'm no expert in this stuff and I'm happy to be corrected when I'm wrong, but the linkwitzlab site seems pretty clear to me in that it doesn't go up. Maybe there's more to it. But we've strayed from the original topic.

[ashcat_lt]

I'm quite familiar with the Rat circuit, and have a pretty good idea of how it works. 

The filter in question is a shelving low pass filter just like the one I posted.  It attenuates the high frequencies coming from the output to the inverting input.  The fact that the opamp then does its thing does not change a thing about how the cutoffs are calculated.  Being in the Bizzaro World of the negative feedback loop does mean that up is down and down and is up, but left is still left, and right is still right. 

[GGBB]

The filter in question is a shelving low pass filter just like the one I posted.

You are correct - the passive circuit you posted is no doubt low-pass shelving. Comparing V2 to V1, V2 will have high frequencies attenuated. But when you look at that circuit within the RAT, it is in the opamp feedback loop, and V2 is not at the output, it is the negative input. Opamp output is actually at V1 which is the filter's input. The low-pass filter is applied to the feedback signal which affects gain, and since there is therefore less high frequency content at the inverting input, more gain is applied to the high frequencies, resulting in the overall operation of the opamp including the feedback filter being a high-pass shelving filter. Being familiar with the RAT, you know that's exactly what it does. So going back to the linkwitzlab diagrams, the one that matches the RAT is the active high-pass shelving filter, so the formulas for that are different than the ones you posted for the passive low-pass shelving filter.

[ashcat_lt]

...so the formulas for that are different than the ones you posted for the passive low-pass shelving filter...

No, they're really not. 

The link you posted doesn't actually show how to calculate the turnover frequencies for the active version, and the passive version is a completely different thing.  The frequencies of interest will be exactly the same as those in the low-pass thing I posted.  Always and every time.  Don't let the opamp confuse you.  Yes, it flips everything on its head and makes attenuation into gain, but it does not change the frequency dependent parts of the equation.  If you figure out the critical frequencies as though it was the low-pass shelf, it will give you the info you need.  It really is that simple, and I'm sure if you stopped to really think about it, you'd understand why.

[GGBB]

...so the formulas for that are different than the ones you posted for the passive low-pass shelving filter...

No, they're really not. 

The link you posted doesn't actually show how to calculate the turnover frequencies for the active version, and the passive version is a completely different thing.  The frequencies of interest will be exactly the same as those in the low-pass thing I posted.  Always and every time.  Don't let the opamp confuse you.  Yes, it flips everything on its head and makes attenuation into gain, but it does not change the frequency dependent parts of the equation.  If you figure out the critical frequencies as though it was the low-pass shelf, it will give you the info you need.  It really is that simple, and I'm sure if you stopped to really think about it, you'd understand why.

I apologize for not expressing myself well. My first post was a reply to the statement that the filter frequency goes up as gain goes up. You later replied to that person "you were right." I took that to mean you agreed that frequency goes up as gain goes up, but that's actually not what you were saying (my mistake again), you just said it changes. It does in fact change, not up but down, and I was wrong about it not changing. I looked at a sim of the RAT and didn't see the knee going up with gain, so jumped to a quick conclusion. I have since actually thought about it and know from the formulas that it goes down with gain, not up, which was what my instincts told me which prompted my response.

I think we've gotten out wires crossed with the passive low-pass versus active high-pass, and yes the formulas aren't different as far as the knee frequency is concerned, but the RAT opamp is not a low-pass shelf filter - it is a high-pass shelf filter and is so because the feedback loop contains a low-pass shelf filter. The opamp makes a difference, it isn't confusing me. Make the caps larger and you have more bass, not less highs.

[ashcat_lt]

OK.  I guess I didn't realize we were arguing which direction the filter moves.  I was just trying to say that it does move.  Glad we cleared that up.

The fun thing is that even though that lower turnover point gets lower in frequency, which should mean more bass, the actual shelf level also gets lower in level, or rather the higher frequencies get more gain in comparison, so it still ends up sounding like less bass at higher gain settings.

Course, it never really should sound much like a wah pedal, and the OP found the problem, so...