Video update revisiting the design theory behind our noise gates.
Original post below from 2021. This is specifically for the design of the “Noise Gator” and “The Paper Shredder”. This design theory additionally got implemented in “Speed Bumps”.
With distortion inevitably comes noise. The initial amount of noise cannot be heard at default volume with any good gear. When gain is introduced the standard way to reduce noise is with filtering. The Electric Fiend pedal original had a mild enough gain to be called a boost pedal, however the Electric Fiend Sr. is the exact same circuit with a gratuitous amount of gain. My personal pedal board has always been a quiescently noisy rig. While this is probably pretty typical, it’s not acceptable for a single pedal to amplify the amount of intrinsic noise that the Electric Fiend Sr. was initially outputting. This gives pedal builders limitations in their designs, with one of two options, to either filter the crap out of the signal or dampen the gain.
In attempting to find a solution for this issue on the Electric Fiend Sr. I tried the technique of multiple stage filtering. No matter what, I could not find the original distortion tone that it acquired through the natural unfiltered version of the op-amps breakup. So… filtering wasn’t going to work for this design.
I have always been an analog head. For audio applications I enjoy digital, I own digital, but I build analog. I have no qualms mixing the two entities which is the Manilogue mantra of 2Mult in all audio and life applications. That being said, how can I filter out the intrinsic noise floor of the Electric Fiend Sr. without losing tonal quality? My brain scattered off into the digital arena, which seemed reasonable having a conditional statement, “if magnitude greater than noise floor, turn OFF filter, else noise filter ON”, however why can’t I keep this analog? So, I had to try. Here is my initial design. The idea of this design is to have a noise filter built in, once the magnitude of the guitar signal passes a set threshold, a comparator will then turn on a JFET switch that will shunt the noise filter. This will allow the signal to pass unfiltered.
After, pulling up some initial scope readings I knew I needed to set an initial value for the filter threshold and I knew this needed to be variable due to the noise floor not being consistent from rig to rig. The next thing I added in was a low gain amplifier to feed into the comparator, this makes sure the initial start value of the comparator is consistent. Using a scope reading I was able to find the noise threshold. After that I was able to set a trigger point at around 7.6 Volts using a voltage divider. I used an 18K ohm resistor, later I moved to a thumb-wheel pot, but this was for my initial tests. After the comparator was set and tested, I set up the JFET switch. I tested the switch with a voltmeter in continuity mode over the drain and source while strumming the guitar string. Strum….Beep. Strum…Beep. The test worked perfectly. With my initial design I thought I could shunt the entire filter, but while testing I found out the impedance was too high over the transistor. Which lead to the filter, filtering regardless of the switches state. So instead I set the JFET switch to switch the filtering capacitor on or off. And just like that, it worked. Happily, the breadboard layout, to my surprise, had no heavy troubleshooting, I was assuming that this would take multiple days. (6 hours)
This is practically a distortion pedal with an analog noise gate. After doing a quick search, there aren’t many builders combining a noise gate with their distortion. It appears that most builders are going the digital route for this, but have not looked into it yet. Anyways, this is the way we do it. This will not be implemented with the Electric Fiend Sr. and that pedal is planning to move to a more conservative amount of gain. The above, however, is the makings of a new pedal.