Archive for the ‘Square Earth Theory’ Category

Over the Christmas break I’ve been pretty busy playing with 2 new pedals I have been working on. The first I have mentioned a few times before – the Square Earth Theory, which still isn’t working too well to be honest. Here’s the schematic I have so far, trying to filter out the popping is proving to be such an enormous pain in the ass you wouldn’t believe. This uses a FET to switch the signal, in the past I used a 4053 and 4066, with OK results. The 4053 worked the best of the 2 CMOS IC’s, set up so that the signal is shunted to the Vref rail, and is also broken, but the FET seems to be the quietest option when set up to allow a slower switch on time as in the schematic. If you have any luck getting this to work better than I have please let me know!

Square Earth Theory

I have also been playing with an op-amp clean boost circuit that uses a charge pump to internally double the +9v supply run the op-amp that does the boosting at +18v which allows a decent amount of headroom before the signal starts clipping when boosting. I’m pretty sure the Klon used a charge pump to get ~18v without at 18v supply, and a clean boost seems like a great application for using a charge pump.

When you talk about clean boosts everyone throws the term ‘transparent’ around as if it is the holy grail, and it kind of annoys me as what most people are really after when they think of a clean boost is an all around ‘tone fattener’, which is far from transparent! Op-amps get discarded as being slightly sterile when used as clean boosts, but I don’t necessarily agree. Mosfets do tend to be slightly ‘warmer’ than op-amps and do other things to the tone as they boost, but I actually really like the way the op-amps stay so clean (especially when you have a the increased headroom that is provided from a larger voltage than the standard 9v) even when the boost is many decibels.

Anyway, here’s the schematic.


The amount of boost the pedal can kick out is set by the trimmer, which sounds great around +24dB (trimmer set to 8k8). This worked well for the pick ups of my PRS Custom 22 as it allowed loads of boost without any clipping. I made adjustable with a trimmer so that it can be adjusted to suit your guitar, as some with really hot pickups, especially active, may clip earlier. The available range is between 9-35dB. This could have been a pot, but I would rather have it set internally and then just forget about it. With the setup I have here the signal is always boosted by x amount, and then attenuated with the volume control. I think it was Jack Orman (who incidentally designed the Mosfet Booster, which is the basis for many commercially available clean boost pedals) who said that this is like driving a car and always having your foot on the accelerator and jamming on the brakes to slow down, which is a great analogy of what is going on here. I opted to do it this way as in practice it felt more natural to me than adjusting the gain when using it. You could substitute the 100k trimmer (R7) with a 100k pot and do away with the volume control to control the gain directly, which, on paper at least, makes more sense (and actually my original plan). If you are thinking about doing this I would consider changing R9 to 510k, R3 to 10k and using a 500k pot, which will give you around +34dB of gain on tap, but not all of which will be clean.

I think I’m going to call this pedal the Scallywag, I’ll get it built up and boxed asap, and may make a small batch.


Here’s a quick demo of the SET. It’s got some pretty bad playing in it, but it was a bit of a rush job just to get something done so it is all done off the cuff, just to give people an idea of what it is all about.

The circuit still isn’t finished and has a few popping issues, so go easy on it!

I’ve quickly made this diagram to show how the square wave outputs from the two oscillators a and b can be added together using the various gates on the SET.

By the nature of the logic and with the oscillations both having 50% duty cycles the or gate will always allow the signal through more than the and gate. A only is simply the output of just a. In this mode the B led will not be flashing to avoid confusion. You could actually create the output wave from these examples with a single oscillator with a variable duty cycle, but in practice the dual oscillator approach creates many unique patterns you couldn’t create with a single oscillator.

Square Earth Theory Wave Examples

Just a quick update on the SET audio slicer I’m developing.

The logic side of the circuit that is used to control the stuttering is all up and running well, but I’m still having trouble slicing the audio without introducing noise. I’ve removed the 4066 chip and redesigned the circuit to use a JFET as the switch. By having this in a shunt arrangement where the source and drain are biased to a reference voltage I can eliminate most of the popping, but there is still a little bit of ticking on the audio line. I’ve also tried using a LED and LDR to shunt the signal without allowing any noise from the oscillators, but it isn’t perfect either way. It may be that the noise is on the supply rail and I need to filter the supplies, but I’m not sure with very limited time to spend on it at the moment.

The controls will also have changed by the time this comes out. At present they stand at Rate A, Rate B pots, long/short range  toggle, A/AB toggle, and/or toggle.

Whatever happens, I have learnt a lot from this project, one being that nothing is as easy as it first seems!

Anyone who has followed the progress of any of my pedals within the last year may remember this post I made way back in January of 2012. Or you may not, hell I hardly remember writing it! But anyway, in that post I started harping on about this idea for an auto stutter/killswitch type device that is based on two 555 timers, and allows you to set 2 independent frequencies, decide how to combine them to make one pulse and use that to stutter your guitars signal. The last week or si I’ve been working on the circuit and it is nearly at a point where I’m going to start making a few, just as soon as I can iron out a few small issues (thanks to the guys at DIY Stompboxes). It is not actually based on 555 timers but 2 4047 oscillators operating in astable mode, which cause either a 4066 or 4053 to switch the signal, creating the stutter. I won’t go into the components I’m using now as it may well change, and when I have finalised the circuit I will post it up with an explanation anyway.
The pedal will feature 2 pots – rate a and rate b (can’t tell you the range of the frequency yet as even that isn’t fully decided!) and 3 toggle switches. A/ab allows you to select either just oscillation a or oscillation a and b. And/or controls how frequencies a and b combined – either both on or either on, and invert which flips the output of a/ab. A, b and the combined logic will all have their own LED. The pedal will be momentary, so your signal is unaffected until you hit the stomp at which point stuttering begins and continues, until you release the switch. By combining the 2 different oscillators you can create some really cool patterns, and it actually has a very mathematical feel especially when using ab and using the and setting. That probably makes little sense, here’s an idea of what the pedal may look like, I’ll get a video up soon once I iron out the few small problems with the design.

Imagine it as a square wave tremolo with the depth up full and you will get some kind of idea as to how it sounds in ‘a’ mode, ‘b’ mode sounds much more pattern based, although the patterns are only very simple with the 2 oscillators. I wanted to add a blendable fx loop so the pedal could stutter between dry and wet, but that was way too big for the 1590B version, maybe if it is a cool design I’ll build a 1590BB version with more oscillators, a blend control and a fx loop!