Utter Stutter - stutter pedal with recordable patterns and variable tempo

[ElectricDruid]

Hi all,

I've been working on something new over the last few days, and I'd like to get your feedback on it before I go much further. That's a polite way of saying I want you to spot my stupid mistakes

It's a simple stutter pedal, but controlled by a PIC microcontroller. Rather than just having a simple square wave LFO, you hold down the "Record" button and then tap a pattern of On's and Off's on the Bypass switch. Once you release the record button, that's your pattern.
There's also a "Tempo" knob which lets you speed up or slow down your pattern. It covers a 1:16 range, so if you record your pattern with a medium tempo, you can go four times slower or four times faster. If you record on the fastest tempo, then you can slow it down all the way to 1/16th of the original speed, and similarly in reverse; record at the slowest speed and you can speed up by 16 times.
That's about it. I wanted something simple to operate, but flexible because what pattern you enter is entirely open-ended.

Here's the schematic:
http://electricdruid.net/wp-content/uploads/2017/09/UtterStutterSchematic.jpg


There's enough RAM on the chip to store 60 buttons presses, so that's how complex your pattern can be. There's no EEPROM on this chip, so it loses the pattern on power-down, but if I could find a similar chip with 256bytes of EEPROM I might try and add that.

I struggled for ages to find a simple mute circuit that worked well, and I tried several. The one I've gone for in the end is from the ESP audio pages:

http://sound.whsites.net/articles/muting.html

I haven't tweaked it much beyond using 2N5485 (what I had) and dropping the 10K resistor on the way in (PIC only gives a 5V signal). It seems to work pretty well, doesn't thump with no signal, and isn't too noisy.
I tried many variations on the two-FETs audio mute, including R.G.'s one from the Geofex pages:

http://www.geofex.com/FX_images/stutter.pdf

I didn't get on so well with this, and it's more complicated, so it has to justify the extra parts. I also experimented making several changes, trying to organise the two FETs more like a voltage divider, so that with one on/one off you get full volume, and with one off/one on, you get no volume. R.G.'s has the grounded FET first, and the series FET after, which is a bit weird in my view. IT worked, but I found I had problems with distortion from the FETs when the signal was off, and I needed to add a transistor level shifter to the PIC's output signal to avoid this - more bits.

Anyway, enough chat - what do you think?

Thanks,
Tom

[EBK]

you hold down the "Record" button and then tap a pattern of On's and Off's on the Bypass switch. Once you release the record button, that's your pattern.

Is this a two feet operation?  I would get hurt trying that! 

[ElectricDruid]

you hold down the "Record" button and then tap a pattern of On's and Off's on the Bypass switch. Once you release the record button, that's your pattern.

Is this a two feet operation?  I would get hurt trying that! 

I dunno. I haven't decided. It'd be possible to use momentary buttons, in which case, you'd just set up a pattern with a couple of fingers.But you could use momentary foot switches for both too, in which case you could put a new pattern in standing up, assuming you can balance on your heels a little bit.

<edit>My stripboard prototype for code development has little push buttons, but I was thinking footswtiches for a stage version. Maybe arcade buttons would be nice?</edit>

T.

[KarenColumbo]

This is utterly crazy!

[bean]

Why not just go optical? An LDR works really well for square wave trem like this and you can get incredibly fast speeds (into ring mod territory) without thump if you do it right. Plus, you could also incorporate a depth pot for the LED driver so it doesn't have to be full on/off which can be really useful. I think it would work with 5v output from the PIC.

This setup works well. It's different than yours (inverted with gain recovery) but many people have used variations of this successfully. I drew the depth for how I think it would work for your circuit (I would add a cap in parallel with T1 which this example doesn't have)



In this case, though, LED2 on your schem might become troublesome as a noise source in parallel with the new LED driver. It might be better to up R4 (your schem) to 100k and then use a transistor to turn the output led2 on and off. Honestly, not sure.

[bean]

Whoops - one thing I just realized with the example I gave. You'd need 0v from the PIC to create the pulse not 5v. So your pin would have to be 5v nominal the switch to ground for the pattern. So, for an indicator LED you would need to put it parallel with the LED driver instead. But, I think that will work. Sorry!

[anotherjim]

Nice to see an example shunt mute.
I was about to call you out on bias/DC issues until I realized you have a ground symbol for Vref!
Does a stutter have to be a perfect silent mute, or does a significant drop suffice I wonder?

[ElectricDruid]

Thanks Bean. It's true that this time I didn't try an optical solution, though I could have done. I've used such a circuit in the past - the Tap Tempo Tremolo design that is flying around here (and which became a PCB at MusicPCB) is done that way. I've been pretty lucky with my builds of it, but some people have had issues with that trem ticking. I never understood exactly where the ticking came from. I might have a think about it and do some experiments with it. As I said this isn't set in stone yet, and I'm looking for any improvements I can make, so your suggestions are welcomed.

The gain recovery is one thing that I haven't worked out yet. I'd like the option of having a software-bypass, in which case, when you bypass the pedal, the FETs switch off (or the LED/LDR switches on) and you get a solid signal straight through. The trouble is doing it like that makes it difficult to add make-up gain on that second amp, since it'll affect the bypassed signal too. Dunno what to do about that. I'd like to keep the option of soft-bypass, since you can always ignore it and wire the pedal as true-bypass if you want to, so it doesn't limit anything if done right. Also it allows the use of momentary SPST switches for bypass, and I like both the feel of those (clickless) and also the much simpler wiring (impossible to wire up wrong - two wires, one to the board, one to ground, and it doesn't matter which way around.)

T.

[ElectricDruid]

Nice to see an example shunt mute.
I was about to call you out on bias/DC issues until I realized you have a ground symbol for Vref!
Does a stutter have to be a perfect silent mute, or does a significant drop suffice I wonder?

That's not a ground symbol, that's a "virtual ground" symbol! At least, that's how I think about it...

I think the closer to silent you get, the better. The "double shunt" design above does pretty well - I can't hear signal when it's off.

Another potential improvement I can see is some way to add a "depth" control which alters exactly how extreme the stutter is, from all the way off up to only 50% drop or some such. Not sure how best to do that with the current circuit.

Tom

[EBK]

Another potential improvement I can see is some way to add a "depth" control which alters exactly how extreme the stutter is, from all the way off up to only 50% drop or some such. Not sure how best to do that with the current circuit.

Couldn't you just put a pot between your FETs and that "virtual ground"?

[bean]

Thanks Bean. It's true that this time I didn't try an optical solution, though I could have done. I've used such a circuit in the past - the Tap Tempo Tremolo design that is flying around here (and which became a PCB at MusicPCB) is done that way. I've been pretty lucky with my builds of it, but some people have had issues with that trem ticking. I never understood exactly where the ticking came from. I might have a think about it and do some experiments with it. As I said this isn't set in stone yet, and I'm looking for any improvements I can make, so your suggestions are welcomed.

The gain recovery is one thing that I haven't worked out yet. I'd like the option of having a software-bypass, in which case, when you bypass the pedal, the FETs switch off (or the LED/LDR switches on) and you get a solid signal straight through. The trouble is doing it like that makes it difficult to add make-up gain on that second amp, since it'll affect the bypassed signal too. Dunno what to do about that. I'd like to keep the option of soft-bypass, since you can always ignore it and wire the pedal as true-bypass if you want to, so it doesn't limit anything if done right. Also it allows the use of momentary SPST switches for bypass, and I like both the feel of those (clickless) and also the much simpler wiring (impossible to wire up wrong - two wires, one to the board, one to ground, and it doesn't matter which way around.)

In the LED/LDR case as long as your LED doesn't increase its brightness on bypass from the maximum "on" level I don't think makeup gain will matter. But, I think there might be a "perceived" volume increase because our ears are detecting relative volume changes when the effect is on (which they are really good at)...not absolute changes (which ears don't do well). Just a thought....I might be talking out my butt on this.

[ElectricDruid]

Another potential improvement I can see is some way to add a "depth" control which alters exactly how extreme the stutter is, from all the way off up to only 50% drop or some such. Not sure how best to do that with the current circuit.

Couldn't you just put a pot between your FETs and that "virtual ground"?

Yeah, I suppose that would do it. Why hadn't I spotted that?! Do you suppose I need a dual pot, or could I get away with tying the bottoms of the FETs together and then having a variable resistance to virtual ground? Try it, I suppose.

Tom

[bartimaeus]

This is a really cool idea, I'm very interested! Here's some of my initial reactions:

Requiring two feet would be fatal imho... The only way that I can think of this working in a band context is if I can tap out the pattern with one foot while we're playing. Maybe make the "record" footswitch latching?

I imagine keeping something like this in sync with other instruments will be pretty challenging, though. A "reset" switch to tap on the downbeat every couple of measures would make for a great live performance tool. The "record" footswitch could become a "reset" after a pattern has been recorded, and maybe you can hold down the switch to clear your pattern and start over?

[ElectricDruid]

Thanks Bartimaeus. Good thoughts. I hadn't really considered keeping it in time, but you're probably right. On the Tap Tempo Tremolo, a single tap resets the LFO to the start of the waveform so you can keep it in time if it starts to go out (or the band do!). This pedal could probably use something similar, I agree.

T.

[R.G.]

R.G.'s has the grounded FET first, and the series FET after, which is a bit weird in my view. IT worked, but I found I had problems with distortion from the FETs when the signal was off, and I needed to add a transistor level shifter to the PIC's output signal to avoid this - more bits.

Actially that circuit is a volume control down first - that 10K and the shunt FET - and a series high impedance FET best looked at as running the gain on the second opamp down to as near zero as you can get with an almost upen circuit on its input. At least that was the idea.

[ElectricDruid]

R.G.'s has the grounded FET first, and the series FET after, which is a bit weird in my view. IT worked, but I found I had problems with distortion from the FETs when the signal was off, and I needed to add a transistor level shifter to the PIC's output signal to avoid this - more bits.

Actially that circuit is a volume control down first - that 10K and the shunt FET - and a series high impedance FET best looked at as running the gain on the second opamp down to as near zero as you can get with an almost upen circuit on its input. At least that was the idea.

Well, now you put it like that, it makes perfect sense! It didn't work any better when I swapped it around, so I'd leave it your way!

Incidentally, since I put this thread up, I've discovered there's already a pedal called the Utter Stutter (although it doesn't do anything very similar to this one):

http://ohnohofx.tumblr.com/post/64338711284/new-utter-stutters

So...name suggestions please?! What should I call a stutter/chopper-trem pedal with recordable rhythm/patterns?

I've looked at variations around Stutter, Stammer, Blipper, etc. Still looking.

Thanks,
Tom

[EBK]

P.p..p.program  p.p..p..pedal?

[Kipper4]

Stutter McStutterface.

[ElectricDruid]

That boat has a lot to answer for...

T.

[blackieNYC]

Very interested.
Skipper?