Input Caps Question

[Steve Mavronis]

I'm wondering about this and want your thoughts. What effect does the placement of an low pass filter input cap to ground "before or after" a signal cap have on a circuit?

For example:

1.) In a 70's MXR Distortion+ right after the input wire pad there is a 0.001uF cap to ground, followed by a 0.01uf signal path cap and 10K resistor to the op-amp.

2.) In the older 76/77 year DOD 250 Overdrives after the wire input pad there was a 2.2nF cap to ground, but it was 'placed after' the signal input cap, in between it and the resistor to the op-amp. In later year models this was removed completely. You can still see the old part holes for it.

It just seems odd to have the low pass filter located after the signal input cap instead of before it as in the later case. Maybe it doesn't matter or was done for different reasons. What do you think?

[R.G.]

It is most likely to be there to shunt RF to ground. Well, or anything above a reasonable guess at the desired guitar signal.

As such, if the resistance after it to ground is much bigger than the cap impedance at guitar frequencies, it doesn't matter whether it's before or after the input DC blocking cap. It still shunts even higher frequencies to ground, at a first approximation.

[Toney]

I built the DOD '77 with the 2.2 on a socket as I thought it might be a bit too much - it is. It darkens the tone considerably and I pulled it out and left it.

[joegagan]

when i prepped the dinosaur fuzz for production, i found a 2200 pf cap did a great job of cutting out RF, with no ill effects on tone. but that was a slightly trebley circ to begin with.

[Steve Mavronis]

Hmm, good info and experiences. Interesting what Toney said about the 2.2nF value. Maybe that's why DOD had done away with it, yet Joegagan liked it and MXR kept their 0.001uF value. I'm now in the process of prepping a test for this by revising one of my board layouts to see if this helps any at all. See the forum topic "Hybrid 250+" link for a graphic - http://www.diystompboxes.com/smfforum/index.php?topic=97064.0

I'm including sockets for the low pass input cap and feedback cap to experiment with different values and op-amp combos to prototype it. Until I try that I won't know exactly what is the most effective low pass input cap value to 'reduce' high end gain noise hiss as much as practical without affecting the so-called 'grey' tone that I like. That's why those cap values are blank right now on my ExpressPCB screen capture. When everything looks right I'll freely post the layout pattern if anyone else wants to play with etching it. My patterns are mirrored due to how I print to transparency for UV surface contact exposure onto PCB. But I can easily flip it for the toner transfer method guys.

[Steve Mavronis]

It is most likely to be there to shunt RF to ground. Well, or anything above a reasonable guess at the desired guitar signal. As such, if the resistance after it to ground is much bigger than the cap impedance at guitar frequencies, it doesn't matter whether it's before or after the input DC blocking cap. It still shunts even higher frequencies to ground, at a first approximation.

I'm still wondering in a circuit like this having a low-pass filter cap to ground right before the signal path input cap, how do you calculate the best value for dersired high frequency cutoff? For example the MXR Distortion+ choose a 0.001uF filter cap while the very first grey DOD 250 Overdrive version used a 0.0022uF cap, basically doubling the value that MXR used.

To decide this in a classic RC-Filter configuration, you account for the desired frequency cutoff, the filter cap, and the resistor used so you can plug the numbers into the formula to calculate everything right:



In this case there is no resistor there to have an affect on it, so how do you figure out the optimal value of the filter cap, as a starting point, besides by trial and error listening for the result? Or in this configuration as MXR implemented is the filter cap just there to mitigate any initial high frequency spikes from pick attack?

[kurtlives]

R would be your input impedance roughly.

Your cutoff range would ideally be something out of range of the guitar's frequency spectrum. 10K-20K maybe? That should account for some higher order harmonics.

[Gus]

If you want to do this more exactly don't forget about the guitar and cable before the input of the effect. 
Model the guitar and cable
Draw the schematic of the guitar/bass then make a spice model of the guitar/bass to use before the effect in the spice model.

[Steve Mavronis]

Do they make small 'variable capacitors' in the 1000pF to 2200pF or so range for testing and any datasheets for component dimensions and pin spacing? I'd like to maybe experiment with something like this instead of manually swapping sizes or regular caps. Might not be practical though.

[nocentelli]

I think variable caps ONLY come in very low (picofarad) values. The ones I've seen were pretty large (physically), like a 24mm pot size, and the value was very low, like under a nanofarad.