MS-20 filter fun!

[moid]

Hello everyone

I hope your Summer is going well... mine was going well until Moid Junior came home from a camp and gave my wife and I COVID... urghhh... so apologies if this is really rambling or boring; I'm not feeling 100% sadly. Today I managed to finish building an MS20 filter pedal that I started last year and worked out how to fix it (without having to annoy you chaps, so I feel slightly smug... although if I had just learned to attach certain cables to the right holes then it would've worked first time...). Anyway the pedal works, it oscillates somewhat when the resonance pot is at maximum and the frequency pot is near or at maximum, but from what I've read elsewhere, that is what this filter does.

While playing with it I realised that this circuit badly needs an automated sweep effect across the Frequency pot - because if i sweep it manually while strumming guitar it makes a great phaser-y filter sound. So... (you can see where this is going can't you) what would be the best way to achieve this? I have built a 555 circuit with an LED/LDR combo for a Park Sounds Autowah last year which sounded good, but the pulse of the LED/LDR combo makes the LFO have a strange shape to it where the troughs of the cycle are much narrower than the peaks (very broad rounded peaks with sudden drops to a very sharp trough/valley with steep sides that shoots up to become a broad rounded peak again) and this does sound odd at times. I was wondering is there a way to generate some sort of triangle /sine wave so I could get a smoother gradient - where the peaks and troughs are more equal in width? What sort of circuit should I look at?

I did find a very interesting circuit that allows triangle and sawtooth / reversed sawtooth waves, but it outputs to an LED so would presumably need an LDR after that to get varying levels of resistance? I was going to attch the output of such a circuit to lug 2 of the frequency pot of the MS20 (doing the same thing on the Park Sounds Autowah worked a treat last year) but I wonder if this circuit would have the same strange shape wave if it had an LED / LDR combination attached?

The interesting circuit I found that generates triangle/ sawtooth and reversed sawtooth is here:
http://www.pcbheaven.com/circuitpages/555_Breathing_Pulsing_LED/

and the schematic is here:


There's a video about it here:

If it matters, here's the vero layout of the MS20 I used:


Which is based on this MS20 schematic


I used an LM13700 IC if that matters.

Thanks for your thoughts!

[Focalized]

I have that layout hoping to maybe try it in a Wah pedal.

[Elektrojänis]

That filter is Voltage controlled in the synth that it's from. Based on that, I quess it doesn't necessarily need a varying resistance. Varying voltage should be fine.

I'm not an expert by any means, but it looks like your LFO shoild have a varying voltage available at the collectors of those two transistors. Maybe try connecting a resistor (try 1k - 10k) from the collector of T1 or T2 to Point A on the filter?

You might have to insert similar resistor between the frequency pot pin 2 and Point A to mix the two control voltages better.

[ElectricDruid]

Does that filter thump when you turn the Frequency control right down?

I ask because the LM13700's control input is a current, not a voltage, and the pin is at (iirc) two diode drops above ground (so 1.2V, roughly). While the input voltage is above that, current flows through the 100Ks and it's happy. Below that, and it shuts off, and I've found it thumps a bit when it does. I'd have expected to see a resistor below the frequency control (and perhaps above it too) to limit the range to the useful part.

[Ben N]

I thought I recognized that circuit--or should I saw circuit snippet. That's one of Tim Escobedo's. https://nexp.pt/escobedo.html

[moid]

Hi everyone

Sorry for the slow reply, COVID really hit me hard I thought the second time around would be easier, but it doesn't seem that way. I've managed to make the LFO LED circuit on breadboard inbetween coughing furiously, and it lights up the LED when I apply power - but only once and then turns off and doesn't repeat. I have to pull the power leads from the breadboard, wait a few minutes and stick them back in again and then the LED lights up once and turns off. Really odd. Maybe just my bad breadboarding skills? And possibly this would work if it was soldered? Or maybe I should go round the board with a DMM and see if there is a point that there is no voltage to be found? I have tried touching various components to see if doing so shorts them and resets the circuit but haven't found anything that specifically does this.

That filter is Voltage controlled in the synth that it's from. Based on that, I quess it doesn't necessarily need a varying resistance. Varying voltage should be fine.

I'm not an expert by any means, but it looks like your LFO shoild have a varying voltage available at the collectors of those two transistors. Maybe try connecting a resistor (try 1k - 10k) from the collector of T1 or T2 to Point A on the filter?

You might have to insert similar resistor between the frequency pot pin 2 and Point A to mix the two control voltages better.

Thanks Elektrojänis - I will certainly try your ideas as soon as I get the breadbloard to flash repeatedly. That way I can use the LED as a rate LED on the outside of the enclosure and take the resistance from the circuit and not have to worry about an LDR. The circuit definitely needs an LFO or it's just a %^&*ed wah (nice, but nothing special) so I'd rather get the LFO working. If this one refuses to work I will have to look at finding another LFO.

EDIT the forum is censoring C o c ked wah. I think. Either that or COVID is ruining my brain.

[moid]

Does that filter thump when you turn the Frequency control right down?

I ask because the LM13700's control input is a current, not a voltage, and the pin is at (iirc) two diode drops above ground (so 1.2V, roughly). While the input voltage is above that, current flows through the 100Ks and it's happy. Below that, and it shuts off, and I've found it thumps a bit when it does. I'd have expected to see a resistor below the frequency control (and perhaps above it too) to limit the range to the useful part.

Hello ElectricDruid
Thankfully no, it doesn't thump! It does however cut all volume from 7 o'clock to about 9 o'clock on the Frequency pot in the low pass mode (presumably the filter is passing frequencies a guitar doesn't generate?). I'm somewhat deaf so I asked my son to play through the pedal with Resonance at different settings and neither of us could hear a thumping sound, but we both agreed that low frequency settings removed all volume. The high end of the pot (3 oclock 5 o clock) also does nothing on the low pass mode, but does have some effect on the high pass mode. I think the filter is really designed for the huge range of frequencies a synth puts out and would benefit from some adjustment to put the Frequency pot in guitar range better perhaps? Not sure if that helps your analysis of the circuit though - it certainly sounds good to me and similar to recordings other people have made online of this pedal.

EDIT: I forgot to say the pedal isn't boxed yet and is running off a 9v battery instead of from a wall plug; not sure if that matter though.

[moid]

I thought I recognized that circuit--or should I saw circuit snippet. That's one of Tim Escobedo's. https://nexp.pt/escobedo.html

I reckon you are right! I didn't know it was his work - I just saw the vero layout and thought ooh that doesn't look very complicated and so built it.

[moid]

OK weirdness time. I worked out that if i unplugged the ground cable from the board and touched pin 3 of the IC that would short the circuit - so if I plugged the ground cable back into the board it would flicker once and then stop. I then watched the video of the chap who designed this schematic and on his schematic there is nothing attached to pin 3 of the IC, but in the video there is a resistor attached to it... but that resistor isn't attached to anything?

To make my life that extra bit more fun, I'm colourblind - could someone please tell me what resistor that one is (I can't read colour bands on resistors). I have highlighted the legs of the resistor in yellow. And if you have anyway of knowing how that circuit works when parts of it are not connected I would love to know!

I have removed a link on my breadboard from pin 2 of the IC to ground (which was a mistake). The LED now lights up permanently, but there is no flash... my guess is this is something to do with pin 3 and that weird resistor.

[moid]

OK weirdness time. I worked out that if i unplugged the ground cable from the board and touched pin 3 of the IC that would short the circuit - so if I plugged the ground cable back into the board it would flicker once and then stop. I then watched the video of the chap who designed this schematic and on his schematic there is nothing attached to pin 3 of the IC, but in the video there is a resistor attached to it... but that resistor isn't attached to anything?

To make my life that extra bit more fun, I'm colourblind - could someone please tell me what resistor that one is (I can't read colour bands on resistors). I have highlighted the legs of the resistor in yellow. And if you have anyway of knowing how that circuit works when parts of it are not connected I would love to know!

I have removed a link on my breadboard from pin 2 of the IC to ground (which was a mistake). The LED now lights up permanently, but there is no flash... my guess is this is something to do with pin 3 and that weird resistor.

EDIT: Just read the description of the circuit from the author's website and at the bottom of the page he says:
"Finally, something about the LED voltage and current. Different LEDs require different forward voltages and draws different current. You may need to change the value of R5 to change the voltage across the LED and R6 to change the current."
So should I try swapping random values for these resistors instead and that might trigger the flashing effect? I've tried a standard Red LED and a bright blue one and they both light up but don't flash. I'll try some different colours next.

[moid]

There was no difference if I changed the LED colours. I then discovered I could pull out the pots and the circuit still lights up so I must have screwed the layout up somehow. I am going to dismantle it and try again; hopefully my covid brain is getting better.

[moid]

OK some progress... rebuilt the breadboard circuit and the LED now blinks - but only On / off. The two pots seem to set either the length of ON or OFF in the pulse they emit (I assume this is squarewave). I cannot get the damn thing to fade up / down though. I'm wondering if I should just look for a triangle wave circuit instead.

[duck_arse]

it's probably your corvid that's the problem, squire. twice? that's just beeing greedy.

I'm sure we belted the life out of this flasher circuit once before, didn't we? pre-corvid? are you using T1 and T2? because T2 will snap-action from the T1 gain, and you will only get on/off from there. if you delete T2 and shift that led to the emitter to ground of T1, what happens?

[moid]

I know! Normally I'm a share and share-a-like kind of guy, but when it comes to infectious diseases I just gotta catch 'em all!

Blimey! I had forgotten about that one! I was thinking of the one I built last year into the Park Autowah... (which now that I look at my records is almost identical) - that flasher worked but has a problem - it generates a very odd shape LFO - I would describe it as looking like this:



Which worked fine on the Comet filter from 4 years ago (yikes, 4 years ago? how?) but on the Park Autowah gives it a bit too strong of a character - it sucks the frequencies out as it drops into that narrow channel and then suddenly launches back up into a broad almost flattened off peak - I don't have a scope so I drew this based on what it sounds like (and if I look at spectrograms of audio recorded with it, it kind of looks like this). What I really wanted was a triangle LFO, and when I saw the circuit I'm trying to get working I thought that was exactly the sort of thing to build - the bonus being it can do sawtooth and reverse sawtooth as well, so it seemed a no brainer! (Except to me who doesn't have much brain right now).

If I remove T2 I still get squarewave only output.

By the way if you know of a good (easy?) to build triangle LFO then I'm happy to try something else! I did find a simple looking schematic on this page
https://electro-music.com/forum/topic-58087.html The IC is a LM358 so I just need to add connections to 9V and ground? plus a voltage divider to get 4.5V - maybe this is worth a try? Thanks for your help!

[rutabaga bob]

Martin!  Liked the song sample!

[ElectricDruid]

That's the standard schmitt trigger-integrator LFO that you see a lot of places. It's a classic. You get a triangle wave from one op-amp output, and a square wave from the other.

There's been a discussion over on another thread about one variation of it which uses a pot instead of those two 10K/8K2 resistors. That provides an alternative way to change the frequency which also alters the depth. This can be handy for chorus circuits.

It's definitely a good circuit to get on the breadboard and play with so you really understand it.

[duck_arse]

... and, you can add a pair of diodes, a switch and a pot, and have variable slope selectable forward or reverse saw.

[ElectricDruid]

... and, you can add a pair of diodes, a switch and a pot, and have variable slope selectable forward or reverse saw.

+1  You can indeed!

[moid]

That's the standard schmitt trigger-integrator LFO that you see a lot of places. It's a classic. You get a triangle wave from one op-amp output, and a square wave from the other.

There's been a discussion over on another thread about one variation of it which uses a pot instead of those two 10K/8K2 resistors. That provides an alternative way to change the frequency which also alters the depth. This can be handy for chorus circuits.

It's definitely a good circuit to get on the breadboard and play with so you really understand it.

The chances of me really understanding anything are limited, although I appreciate the concept! Well I have some successful news and some weird news. Firstly I built the circuit on breadboard and it worked (kind of) but I had to make some changes. The pot was far too large (for this circuit anyway) - it didn't appear to do anything I could hear, and plugging an LED into it turned the LED on, but no flickering. I put my DMM on it and resistance wise it was jumping from 0 to 2.5M and back again rapidly! So I switched down to a 10K pot, which looked better on the DMM in terms of the maximum resistance was around 140K I think... but still no effect on the audio (I have it attached to lug 3 of the Frequency pot of the MS20 circuit). While swearing and generally testing stuff I accidentally touched the DMM ground cable to lug 2 of the Frequency pot and the circuit came alive!

So the obvious thing to do was work out how to make the world's first modulated MS20 + DMM pedal.... luckily common sense intervened (OK I admit it, I couldn't find an eclosure big enough for both circuits) and I attached a cable from Lug 2 of the frequency pot to the ground of the Triangle LFO (this is a separate ground to the ground that the MS20 is using - both circuits are running on individual 9V batteries at present).

At this point I did find the THUMP that you mentioned! So from previous experience of thumping tcking sounds I took on Duck's advice from past years and started to swap the pot and capacitor sizes about. This also allowed me to try different rates because a 10K pot and 10uF cap ranged from fast autowah to harmonic tremolo to ring modulator and also a ton of oscillating noise and loud thumping / ticking / weird water drop like sounds and occasionally some guitar. A wild ride But too wild. So after trying a lot of pots and caps, a 470uF cap and a 20K pot gives me a really good range of super slow to moderate speeds LFO (with ticking occurring in only the absolute last part of the pot and just before 5 o'clock), but I think I will make a switch to make the circuit flick between 470uF and the 10uF cap because they overlap in range, but the 10uF gets into ring modulation and only ticks (now) at the very extreme almost 5'o'clock position. I found the ticking can be removed by dialing down the resonance and frequency pots slightly as well, although that can take a bit of character out of the sound. I doubt this is fixable from what you said about the LM13700, so I'll live with it

I have a few problems now. If I build this how do I make a separate ground for the LFO? Or will that issue go away if the LFO shares both the same ground and power supply as the MS20?

I'd like to have a switch to go from the original %^&*ed wah MS20 to the LFO MS20, but the MS20 stops working the minute I removed the ground cable from lug 2 of frequency... maybe this is all a power issue... and I probably would always use the LFO on, it sounds great... not sure how to cope with this.

Is there anyway to get a rate LED for the LFO so I can see how fast it is going? I attached one to the LFO output and it lights up, but doesn't flicker or fade in any way. Not the end of the world if that isn't possible.

I did look at that other circuit you mentioned, and a depth control sounds useful, but I think I have that with the frequency pot of the MS20? If I start turning that clockwise it seems to having some affect on the maximum depth of the circuit (I think?) There were moments today when the triangle LFO cut out all volumes in parts of the sweep, but they seem to have vanished at the moment...

I suspect I need to draw a schematic to show what I've done to the circuit, so I'll do that very soon, hopefully later tonight. Thanks again!

... and, you can add a pair of diodes, a switch and a pot, and have variable slope selectable forward or reverse saw.

(My best Scarlett O'Hara impersonation) My Mr Duck! And we barely know each other! **blushes furiously** (sotto voce, aside) Ohhh you know I love it when you talk dirty to me like that

OK I have an idea about how to create that, I will draw something and then we can all have a laugh at it and hopefully you can show me the right way afterwards? Thankee!

[moid]

Martin!  Liked the song sample!

Thanks very much - I assume you mean the one in my signature? That's my son and I. I really need to upload another album; finished it last year and haven't got round to making the artwork since...