Red Llama input cap question

[Redvers]

I've heard people say that you cant just increase the cap size for more bass. I want to do the opposite though, remove some low end flub and introduce a bit more treble so it fits with my other pedals a bit better. Can I just reduce the cap size? Any good reason not to do this?

[bean]

You can reduce it down to 22n...that should help a bit. You can also reduce the 100pF cap to 47pFto give yourself a bit more highs out.

[swever]

You want more treble in red llama? 

My red llama build has tons of treble. In fact, I think it's the least "bandpassed" od/dist/fuzz circuit that I tried. At the low gain settings it seems to be very transparent timbre-wise compared to bypassed signal. Of course, the higher the gain the more of it's own coloration it adds.

Answering your question: I think experimenting with input or output caps is a safe bet 

[Djentronio]

He actually said he wanted less bass, not more treble 

Cap changes are the simplest way to do this in the signal path. If you want to allow the user to decide to have or cut bass, you could add a tone stack or a simple switch to cut bass.

[swever]

He actually said he wanted to "remove some low end flub and introduce a bit more treble" 

[Fancy Lime]

Hi Redvers,

reducing the input cap is one way of reducing the bass and there is no particular reason not to do it. However, if you consider why this works, I think you will see that there are better (i.e. more predictable and reproducible) ways to do this. The input cap forms a simple RC high pass with the input impedance of the inverter. Because these CMOS inverters are not actually built and specified to work as analog inverters (the way they are forced to in the Red Llama), their input impedance may vary considerably depending on manufacturer or even from batch to batch. So every Llama is going to sound different even with the same component values. One way of reducing the effect of this, is to switch the positions of the resistor and the cap between the input jack and the input of the first inverter stage. This way, the resistor should become the dominant factor in the input impedance, if I'm not mistaken (please someone correct me if I'm wrong on this, this is only based on inference from op amp impedance and may not be 100% applicable although my own experimentation seems to confirm it). Now you can calculate the corner frequency of the high pass. For 68nF and 100kΩ that would be a 23Hz cutoff. Changing the cap to 10n would make it 159Hz.

If you want an actual bass control (or rather a "tight" control), I would try adding a log 1M pot from the input pin of the second inverter via a cap (which must be much larger than the cap between the inverters, so anything >330n should be fine) to ground. Without the additional cap that would mess up the bias of the second inverter.

For more treble you can reduce the cap values in the negative feedback loops but be advised that that can get nasty and noisy real quick, depending on the particular inverter chip you use. However, reducing bass automatically increases the perceived highs, so I would change one thing at a time and check the results in between.

Cheers,
Andy

[Redvers]

Thanks for the replies guys. I guess the way I use it is a factor. I exclusively use it at very low gain usually with a middle or neck pickup for a bit of a grit. Trouble is the neck pickup flubs out in the lower end and compared to my other pedals and pickup configurations the red Llama changes the sound just a bit too much. I'll change the cap first of all, I really want to keep the sound I have, just change it ever so slightly in eq.

At some point I'm going to breadboard one up and experiment.

[duck_arse]

fancy lime - for two components in series, their order doesn't matter. one in, one out, no branches in between, doesn't matter cap first or resistor first.

[Mark Hammer]

This topic is one of the reasons why I prefer to precede the two invertor stages with a single op-amp stage.  It allows one to do more pre-clip tone-shaping, to yield the desired final flavour.

That's not any sort of admonishment to anyone.  And certainly it takes more parts.  But, even though the only-two-invertors form factor is elegant and economical, I think the final product of using an op-amp (or even JFET) input stage is preferable.

[Plexi]

Nothing to add to what the guys says.
All the Red Llamas that I've built, was extremely highs, nice mids and loud volume.
Thinking in a Bassman cranked...

What I suggest if you want the Llama don't be SO Llama?
Add a Big Muff tone stack, as many suggest.
The Llama is full of high-mid's per se, so..this scooped tonestack fits nicely.

[Fancy Lime]

fancy lime - for two components in series, their order doesn't matter. one in, one out, no branches in between, doesn't matter cap first or resistor first.

Huh, how about that? You're right of course, that should make no difference for two series components. But then what is the resistance that makes the RC high pass with the cap? I assumed it to be the input resistor if it comes after the cap because in this case the resistor connect to virtual ground, the other way around, the cap goes to virtual ground. But since the virtual ground point is also the input of the inverter... Thinking about CMOS inverters as amplifiers always makes my brain hurt. Can someone clarify what I'm getting wrong?

[Djentronio]

To make it analogous, a cap is like a bucket in the middle of a line. With DC or near-DC signal, the bucket stays full because you're only shaking it a little bit, but with high AC, the bucket spills out because you're shaking it a lot so more water can always get into the bucket from the water line.

Taking signal from between the cap and the resistor would be high pass because the bucket is disturbed easily by AC higher than the cutoff frequency, but not by AC that is near-DC. The attenuates lows, which is how reducing the cap at the input works to reduce lows.

If you put a cap to ground from an input, that is a low pass filter, because the bucket is easily shaken up and spills the AC on the ground, but the gentle shake of the lower frequencies does not spill on the ground.

[anotherjim]

CMOS inverters. The input impedance can certainly be uncertain. Up to a point, the inverting op-amp analogy works. Low open-loop gain with an "S" transfer curve is significant. If you go for unity gain Rin=Rf then it holds up as you would hope.

What I think (so far) is...
Try for x20 gain. As the output swings closer to the power rails, it gets well into the flattening region of the S curve. This means the Rf path isn't able to balance the Rin path to keep the gate input at bias voltage (virtual ground). If you scope it, you should see some signal at the input. In the op-amp analogy, the virtual ground input should not show any signal voltage. If Rf was a potentiometer, you could scope the wiper and turn it to a point where the virtual ground has moved to, at some point in Rf! For the Rin path, there is now Rin + some percentage of Rf. Furthermore, when output swings away from the rail back into the straighter portion of the S curve, feedback increasingly reasserts itself and puts the virtual ground back at the input pin where it belongs.
Actually, an inverting op-amp will also loose virtual ground at the input once the output swing has reached its limits, but I'm not sure if any tonal effect would be apparent with the hard clipping that ensues.

So, input impedance at high gains is dynamic, actually increasing at the peaks. I think this has tonal implications when used as the direct input from your guitar in respect of pickup loading effects. If the input is fed from a buffered low impedance source instead, then that tonal effect is negated.