Lockdown Breadboard Snippets: Easy Tremolo is Easy

[Fancy Lime]

Hi there,

a few posts by different people have made me realize that we all need a little distraction. So here is what's on my breadboard at the moment:



Disclaimer: This may contain errors! Please let me know if you spot them.

It's a simple, fast square-wave-only tremolo. It started out as just some experimentation for the signal path part of a PWM tremolo I intend to use as a way of getting my feet wet with microcontroller programming. So the plan was to just use the simplest square wave LFO I could think of, and I decided to go with something CMOS inverter based. Then I found it sounded really rather good and decided to run with it and turn it into a fully functional pedal design. The basic topology is sound, but some of the R and C values need tweaking. But the way it is, it works and sounds good and should serve others as a starting point or idea source for similar projects.

Some explanations:
- I did not want the LFO ticking away while the pedal is on bypass, so the Schmitt trigger flip-flop around U2E and U2F stops the LFO via Q3 in the state that leaves Q2 in it's minimum resistance state (as well as reducing the gain to 1 via Q1).

- The Lambda pot is essentially speed wired backwards but I did not want to call it Deeps in fear of that being confusing. And Wavelength or Period were also weird so I went with the symbol for wavelength. Never be concise when you can just as well be cryptic The reason the pot is wired backwards is that you can use a log taper instead of a revlog. If you use a revlog, you should call it Speed.

- The ratio of the Lambda pot to R10 defines the total speed range ratio. I experimented with up to 50/1 but found that with anything over 10/1, it was too difficult to dial in the speed with sufficient precision. If you want more range, switching in different values for R6 is the way to go. Another way would be to have a 1M coarse and a 100k fine pot in series and reduce R10 to 10k or so. As it is, the Lambda range is from somewhere around 1Hz to somewhere around 10Hz, I did not bother measuring or calculating. I personally would not be using square waves at lower speeds very much, and wandering into audio frequency territory would mean that one would effectively get a pseudo-ringmodulator and that there would probably be the need for some kind of carrier nulling trim, and I did not want to bother with that. Even as it is, with low notes on a bass guitar you get some interference and pseudo-ringmod-ish sounds at the fastest speeds. At just under the fastest speeds, there is some nice gurgle and warble to be had. The slower range is standard rhythmic tremolo territory but definitely one of the better ones I've played.

- Instead of the three J113s, you could use a bilateral CMOS switch like the CD4066. I just went with the JFETs because I find it a bit easier to breadboard this way. In a proper pedal design, I might go with the CD4066.

Credit where it's due, the switching circuit is shamelessly stolen from R.G.'s geofex article on soft switching, the LFO is a bog-standard textbook square wave oscillator (no idea where I first saw that or who originally invented it), using Q1 to reduce gain to 1 is something that I have rarely seen anywhere other than Merlin's designs, which is also where I saw it first. The rest is basically what first came to mind when thinking "how does one build a tremolo" but I don't know if there is any design that is very similar. There probably is, it's pretty obvious.

Have fun with it,
Andy

P.S.: I'm not normally the competitive type, but if you insist on reading this post as a challenge to share your Lockdown Breadboard Snippets:, there is nothing I can do to stop you. Wink wink, nudge nudge.

[anotherjim]

Chopper trem is cool in my book. I think it works best when used prominently and the rhythm of it is integral to the tunage.

I do take it as a challenge and this might be a good place to stick a few of the things I've been doodling with.

[iainpunk]

i'm currently designing the Harmonic Instant Coffee, i have the schematic and the Falstad simulation here

percolators make good high quality coffee with nice soft flavors.
instant coffee is not very nice and has some harsh flavors.
Harmonic Percolator has nice and soft overtones and harmonic content
Harmonic Instant Coffee is harsh fizzy and sharp ''octave'' fuzz

copy and paste this in to a .txt and open that file via http://falstad.com/circuit/circuitjs.html

Code Select Expand


$ 0 0.000005 27.727228452313398 50 5 43
t 32 256 96 256 0 -1 -0.30483889177585644 -0.5594991129858706 100
t 64 160 96 160 0 1 0.5223599707946267 0.5568334814083791 100
w 96 272 96 288 0
c 16 160 16 240 0 0.000001 1.9134475029497846 0.001
w 96 144 96 128 0
c 16 256 -48 256 0 0.000001 4.033369387719034 0.001
g 96 336 96 352 0 0
R 96 -16 96 -48 0 0 40 9 0 0 0.5
r -96 256 -48 256 0 10000
c 208 208 96 208 0 0.00009999999999999999 -4.9235460520479695 -3
g 208 208 208 240 0 0
c 176 112 240 112 0 1.0000000000000001e-7 5.249562603811742 -5
w 240 208 208 208 0
w 32 160 32 128 0
w 64 160 32 160 0
r 176 112 96 112 0 1
j 160 0 96 0 32 -4 0.00125
r 96 16 96 64 0 10000
r 96 304 96 336 0 22000
w 96 112 96 128 0
t 32 128 64 128 0 -1 1.1077658434905437 -0.5382361812451562 100
w 64 144 64 208 0
w 176 112 160 0 0
w 32 288 32 256 0
w 16 256 32 256 0
w 96 304 16 304 0
w 16 304 16 240 0
w 32 160 16 160 0
w 96 288 96 304 0
w 96 208 64 208 0
r 96 64 96 112 0 10000
r 96 288 32 288 0 470000
174 240 112 272 192 1 500000 0.005 Resistance
w 272 160 304 160 0
r 96 240 96 208 0 2200
w 96 64 64 64 0
w 64 112 64 64 0
r 96 176 96 208 0 3300
200 -96 256 -128 256 0 50 3 0.6
o 33 16 0 12290 0.6846059652061619 0.0001 0 2 38 0



cheers, Iain

[anotherjim]

This is not a pitch follower but more of a weird Outerspace noise & wibble thing.
The VCO is set to a high supersonic frequency. It runs all the time, but without an input -  you can't hear it.
The audio input modulates the VCO.
Audible sideband products are from the VCO output and low pass filtered. Very strong input can push the VCO down to audio.
Pin 14 is only connected to provide a half supply voltage bias to the VCO CV input. The phase comparators are not used. A variation that occurs to me (just now) is to use a DC bias around 1v below supply and set the timing resistor so it's just ultrasonic with no input. Give it a rectified signal input so that it only moves the VCO down into audible range. This ought to provide a really strong effect. The advantage of going ultrasonic though is that it needs no gate on the output to shut it up. That said, I've heard the odd random tweet from it with no input - probably a random strike from a tachyon in the VCO core.

There is a nice smooth fuzz from pin2 but as it's just the old squared up by logic fuzz, it's got glitchy note on-off so it's nothing special, and not the point of this.

Drive the input with your favourite high gain front end. I had a 741 on the BB but some like the LM386 for this kind of thing. With a 9v supply, up to 6v or so swing on the CV input gives a stronger effect.

With a guitar, try choppy rhythms with a varying attack or staccato single notes. Thrashed chords result in solar fission noise.
Synth pitch sweeps give good results.