Pitch/Detune Warble Pedal

[patrick398]

Ok as requested, here's a quick demo going through some of the sounds. Not necessarily the most 'usable' ones but my favourites. There are 8 sections in the track which are:
1. Sine wave lfo, medium depth, medium mix, slow warble/detune (switch on 100n cap between pin 9 and 10)

2.Fast sine wave, medium depth, medium mix: chorus/vibrato (switch on 100n cap between pin 9 and 10)

3. Ramp up lfo, medium depth, full mix (short clip) (switch on 900n cap between pin 9 and 10)

4. Ramp down lfo, high depth, full mix, min speed: slow detune (switch on 900n cap between pin 9 and 10)

5. Square wave lfo, full mix, depth set so it hits octaves of what i'm paying, i love this setting (switch on 900n cap between pin 9 and 10)

6. Square wave lfo, full mix, depth set a bit higher so it's hitting harmonies of what i'm playing (switch on 900n cap between pin 9 and 10)

7. Random slopes lfo,  medium speed, full mix, medium depth, warbling randomly: broken walkman vibes (switch on 900n cap between pin 9 and 10)

8. Random steps lfo: 70s sci-fi computer. I slow the depth down so you can hear it stepping, nice for glitchy space stuff (switch on 900n cap between pin 9 and 10) I increase the depth at the end so it kind of glitches the PT2399 which sounds cool

Let me know if the link works:
https://soundcloud.com/user-418904944/fideleater

[Boba7]

Wow that sounds absolutely amazing, thanks a lot for sharing!

So the 1uf between pin 9 and 10 adds treble and volume?

[patrick398]

Glad you like it
I think the caps at pin 9 and 10 and 11 and 12 are to do with filtering, reducing high frequency noise. For some reason increasing the size of the cap between 9 and 10 introduces a sort of resonance/repeats thing but i'm not entirely sure why

EDIT: I should mention, all credit to Electric Druid and Valve Wizard. The StompLFO is such an amazing chip, it's doing all the hard work. The pt2399 circuit is based around the jenny greenteeth so i take little credit for this

[ElectricDruid]

+1 what Boba said. That really does sound amazing.

Nice work. I'm glad to see the StompLFO adding some more outlandish touches to some modulation effects! Personally, I love a  nice random waveform, but sometimes I feel like the rest of the world is happy with just woowoowoowoowoo

[Elijah-Baley]

Crazy effect!

I didn't get if some of the veroboard layouts in the previous page are verified.
Thanks for sharing!

[patrick398]

+1 what Boba said. That really does sound amazing.

Nice work. I'm glad to see the StompLFO adding some more outlandish touches to some modulation effects! Personally, I love a  nice random waveform, but sometimes I feel like the rest of the world is happy with just woowoowoowoowoo

Yeah i very rarely use the tri/sine waves, in comparison to what the others can do they seem very vanilla.

Crazy effect!

I didn't get if some of the veroboard layouts in the previous page are verified.
Thanks for sharing!

I think the last vero layout i posted is verified. I've built it twice off that layout but the schematic now differs from it a little bit.
Before i was using a 'bright/dark' switch which changed between 10n and 100n cap at pin 11 and 12. It's just a 100n cap in there now and the switch instead adds 820n in parallel with 100n between pin 9 and 10. The vero will work but i'd recommend making adjustments for these changes

[patrick398]

Well i finally got around to making and ordering the PCBs for this, which arrived last week. Much to my surprise everything works perfectly. I think i was slightly confused by the back mounting pots so i'll have to mount them upside down so that they don't function backwards.
The optimum value for C24 is definitely 940n which meant having two 470n's in parallel in the same spot. It might be best to use 1uf caps with 20% tolerance and sort for closest value. If i can't make that work i'll have space for two 470n caps on the next revision of the board.

I only ordered 5 boards because it was my first attempt at PCB design so wasn't sure if it would work. If anyone really wants one though i'm happy to sell for a minimal fee



The smaller board in this photo is a relay bypass using 555 chip, it is the Incandenza Bypass by Steve Demedash and it works really well. The is a slight audible pop when using it on very high gain circuits but it's nothing i can't live with.

[Addy Bart]

This sounds great!

[octave]

Awesome work here! I was wondering if I could omit the PT2399 if I wanted to use the STOMPLFO just for frequency modulation (pitch modulation)? I could use the PT2399 with the shortest delay (30ms), but I was wondering if there would be a simpler way that would be realtime and/or maybe analog?

[ElectricDruid]

Anything done with the PT2399 is going to hit that 30msec limit as you've noted. There's not much getting away from that - it's as fast as it can go.

If you want something without the slight delay for realtime pitch effects, you probably have to look at a BBD like the 3207. The MULTIFLANGE chip provides a similar set of wacky waveforms to the StompLFO, but outputs a biphase clock signal that can directly drive a MN3207 or V3207. This gets the delay down into the 1-10msecs range (depending on settings), below perceptible as a delay. MN3209 would give an even shorter delay, but those short BBDs are hard to find, and it's maybe not worth it.

Alternatively, you could use a VC-clock design from a chorus or flanger circuit and paste the StompLFO in place of the original LFO. I've been playing with the EH Small Clone clock design because it only needs one extra chip and is pretty simple, but there are lots of others.

For absolute-minimum parts though, the MULTIFLANGE or 4KNOBFLANGE chips from the Flangelicious project are the way to go, since these chips wrap up the LFO modulation and the VC-clock onto a single 8-pin chip.

https://electricdruid.net/flangelicious-a-super-dooper-flanger/

[octave]

Oh nice I need to try that! Thank you!

[eeliyx]

Sorry for the necropost.
Have there been any advancements on this schematic?
I wanna give it a try and I'm about to order the parts but I still have some questions if Patrick is still around.

[patrick398]

Yep, just replied to your e-mail and sent the latest schematic, a few changes. New PCBs are on their way, fingers crossed they work

[eeliyx]

Hi Patrick, really cool circuit! Thanks for replying so quickly.
Some questions:
1. What effect does putting back the cap between pin 13&14 of the 2399 have? Less noisy delay?

2. Does you circuit tend latch up? I'm thinking about implementing the latch up protection.

3. What is a useful range for the feedback resistor you mentioned?

4. How did you go about choosing the transistors? As far as I understand, the flow of current from pin 6 to ground controls the delay time.
What does adding the second transistor do? It's in parallel so the overall resistance to ground decreases, leading to shorter delay times, correct?

5. Are you located in Europe? I would love to buy a PCB from you.

[eeliyx]

One more question:
6. What was the purpose of the highlighted part of the old schematic and why did you take it out?

[patrick398]

No problem,

Some questions:
1. What effect does putting back the cap between pin 13&14 of the 2399 have? Less noisy delay?

Just noticed that there's a 2n2 cap between 13&14 in the earliest schematic i have for the circuit but not on any of my later ones. I must have omitted it at some point but never noticed any difference. I probably made the decision and never wrote it down anywhere helpfully.


2. Does you circuit tend latch up? I'm thinking about implementing the latch up protection.

I've never had a single one latch up so that's up to you

3. What is a useful range for the feedback resistor you mentioned?

33k is what i like most. It allows the STOMPLFO to do so very interesting sounding things without be too subtle and without running away in oscillation. I've seen circuits with 100k feedback pot so maybe that's somewhere to start.

4. How did you go about choosing the transistors? As far as I understand, the flow of current from pin 6 to ground controls the delay time.

Originally i used a BC108, it allows quickest delay i could manage, so it circuit functions like a mad chorus/vibrato kind of thing. It's fun experimenting with different transistors, there's a lot of sounds in there.


What does adding the second transistor do? It's in parallel so the overall resistance to ground decreases, leading to shorter delay times, correct?

Yep, the second transistor (BC108) is the one i used originally. The 2N2369/2N2222 provides longer delay times. When the LFO is moving that around the pt2399 freaks out a bit, in a good way (to my ears). The two transistors in parallel is 'stock' because the lower resistance of the BC108 takes precedence. Switching out the BC108 just leaves the 2N2369/2N2222, which is the new mod.

5. Are you located in Europe? I would love to buy a PCB from you.

Yep i'm in UK. New PCBs should be arriving in a couple of weeks, there's a lot of changes to the original, including the rotary switch for the waveforms and relay bypassing so there's a lot to go wrong. Fingers crossed it works haha. But i'd be more than happy to sell you one

6. What was the purpose of the highlighted part of the old schematic and why did you take it out?

They're part of the STOMPLFO circuit, i believe they provide some filtering to the PWM output of the LFO. I seem to remember Tom (Electric Druid) telling me they could be omitted once...at least i hope he did haha

[patrick398]

Here's what Tom actually said:
'If you're really trying to squeeze it down, you could drop the 1K/100n's on the StompLFO. They're "best practice" rather than "essential" and it'll work fine without them. The chip uses digital filtering on the CV inputs anyway, so it's not as necessary to include external filters as it was on some of my earlier chips.'

So he didn't say to omit the caps on the output of the LFO. I just checked my old schematics and the first pcb layouts and it seemed i omitted them a while ago but have not noticed any drawbacks from this.

[ElectricDruid]

It's an interesting finding that it works fine without that output filtering.

The output from the StompLFO is a high-frequency pulse train at about 2MHz. The number of pulses indicate the average level, and you can regain that average level by smoothing the pulses with a simple RC filter. This is a "pulse density modulation" rather than a pulse width modulation", which is what my earlier chips used. The PDM gives better resolution and a much higher frequency output which makes the filtering easier.

I assume the PT2399's VCO must be integrating the pulses somehow.

As Patrick posted, I was actually talking about removing R5,R6,R9,C1,C5,C7, which is the LP filtering on the pot inputs.
If you want to get rid of more unnecessary parts, R12 doesn't really do anything now either (10K is only 1% of 1M, so it only adds a tiny bit).

[patrick398]

It's an interesting finding that it works fine without that output filtering.

The output from the StompLFO is a high-frequency pulse train at about 2MHz. The number of pulses indicate the average level, and you can regain that average level by smoothing the pulses with a simple RC filter. This is a "pulse density modulation" rather than a pulse width modulation", which is what my earlier chips used. The PDM gives better resolution and a much higher frequency output which makes the filtering easier.

I assume the PT2399's VCO must be integrating the pulses somehow.

As Patrick posted, I was actually talking about removing R5,R6,R9,C1,C5,C7, which is the LP filtering on the pot inputs.
If you want to get rid of more unnecessary parts, R12 doesn't really do anything now either (10K is only 1% of 1M, so it only adds a tiny bit).

I did have somebody tell me that they could hear some faint ticking when they plugged it straight into some kind of computer recording interface...i wondered whether it was because i removed that filtering. I did ask him to re-check and send me some recordings but he said he couldn't hear it anymore so perhaps it was something else. I've certainly never noticed anything of the sort

[Kipper4]

Personally I'd get rid of R16 and see if its needed too or not.