MS-20 filter fun!

[Matthew Sanford]

May not be much help, but in making this LM358 scheme I found an Rf back to the Vbias input on the Schmitt helped. It does get weird at longer times...

[duck_arse]

orright then. if you change the R1 to 150k, the output swings about 5V peak to peak triangle at pin 7. [a little more can be got, but needs a trimmer and a rewire on the bias section.] it is 2V short of V+ and 2V above ground, so which way you wire the led won't difference.

with the 150 k, I got 0.5 ~ 2Hz, which is probably too slow for you. change the 10uF pair to a single 1uF poly or a pair of 2u2 electros, and it will run as fast as my shadow.

100k pot and 22k stop gives about 5:1 frequency range on the pot. when you go up with the pot value, you have to squash more ratio in to the same rotation angle, making setting a speed needs finer control. how's my english. same goes for the stopper, if you go lower value you increase the ratio, but going faster.

diodes are 1N914 or 1N4148. when you get it working right, and you switch in a diode, the speed will jump, as you have cut-out half of the sweep time. this is where the second pot comes in, as it adds some sweep time in the diode direction. with a 100k pot at max slow speed, the 10k value won't make much obvious difference. perhaps 50k would be better for the diodes sweepings - but, obvs, there's a catch. if you set a fast speed on the rate pot, say 10k speed, and then a long time on the diode pot, say all of the new 50k, the diode pot setting will appear to be the time/sweep setting. if you see where I'm driving to.

and as for how you drive the MS20 thing -- I don't know. I'm just showing you an osc for driving a  led. carry on.

[duck_arse]

Picture time! Hopefully this makes sense... at the moment I wonder if decreasing the 390K resistor will give me more depth (higher values turned the circuit off), so maybe 330K or slightly less would be the thing to try? If not that then I'm not sure. No idea if that will fix the slope diode issues. Not sure if the diodes should be a different type?

yer diodes [the whole diodes loop] need to go between pins 1 and 6, not 1 and 7. AND the caps go between pins 6 and 7, not 6 and the 100k.

[moid]

May not be much help, but in making this LM358 scheme I found an Rf back to the Vbias input on the Schmitt helped. It does get weird at longer times...

Thanks for the suggestion Matthew  - I wasn't sure how to use it though. Luckily Duck's suggestions have done the trick So all's well that wah's well

[moid]

orright then. if you change the R1 to 150k, the output swings about 5V peak to peak triangle at pin 7. [a little more can be got, but needs a trimmer and a rewire on the bias section.] it is 2V short of V+ and 2V above ground, so which way you wire the led won't difference.

Wow, what a difference! You're a genius 150K made all the difference - nice deep sweep, not quite as much as is possible on the original pedal without the LFO, but that goes too deep and removes all frequencies at points; your suggestion sounds great on guitar and always leaves some audio! My son walked into the room where I was testing / making strange noises and said "no, don't play it like that, play it like this" took the guitar off me and proceeded to make awesome sounds for the next 45 minutes! Your modification has achieved the impossible - it separated a teenager from Whatsapp for 45 minutes! That's a record.

I told my son I hadn't made all your changes yet but he said "don't change anything else, it's awesome!" so I think the version we currently have is staying.

with the 150 k, I got 0.5 ~ 2Hz, which is probably too slow for you. change the 10uF pair to a single 1uF poly or a pair of 2u2 electros, and it will run as fast as my shadow.

Now here's the weird thing - the circuit seems to be doing the opposite? With the caps on 470uF and normal triangle wave engaged and the rate pot set to slowest it takes the LFO 6 seconds to sweep up and down once (which sounds great). If I engage either sawtooth, the sweep varies - forward sawtooth is 2.2 seconds long, while reverse sawtooth is 2.9 seconds long. My son liked the sawtooth settings being faster... so I'm inclined to leave them where they are. I prefer the original super slow triangle mode, but he loved the sawtooth on both guitar and bass... and it sounds better when he plays it anyway! I would say that the part of the 10K pot that adjust slope from triangle to sawtooth is at a narrow part of the end of the range - I can see it working visually before that, but the overall audio effect is not that obvious until you get near or at maximum sawtooth. I'm not changing this, but anyone else reading this might prefer this pot to be a fixed resistor instead.

Also if the MS20 resonance is set near to maximum and either sawtooth mode is active and you unplug a guitar... wow are you about to discover a very loud siren synth You can tune with the LFO rate pot and the slope pot and the resonance pot of the MS20... it's seriously loud!

100k pot and 22k stop gives about 5:1 frequency range on the pot. when you go up with the pot value, you have to squash more ratio in to the same rotation angle, making setting a speed needs finer control. how's my english. same goes for the stopper, if you go lower value you increase the ratio, but going faster.

Well I love your English, whether its fast or slow or squashed I'm not changing the 10K pot / 100ohm combination though; it sounds wonderful. There seems to be a really good speed range on the pot, with a little range overlap from the 470uF and the 22uF capacitors - on the fast setting I'm still geting  pseudo ring modulation with a constant drone note, while I can get really slow (6 second) pulses on the slowest setting.

diodes are 1N914 or 1N4148. when you get it working right, and you switch in a diode, the speed will jump, as you have cut-out half of the sweep time. this is where the second pot comes in, as it adds some sweep time in the diode direction. with a 100k pot at max slow speed, the 10k value won't make much obvious difference. perhaps 50k would be better for the diodes sweepings - but, obvs, there's a catch. if you set a fast speed on the rate pot, say 10k speed, and then a long time on the diode pot, say all of the new 50k, the diode pot setting will appear to be the time/sweep setting. if you see where I'm driving to.

And the other best thing about diodes is once you had corrected my drawing I looked at the breadboard, cursed myself and shifted the legs of the diodes to the right place, and guess what? They started working! Amazing. I'm beginning to suspect that real circuit design isn't just throwing a random selection of components at a PCB and hoping something happens...

and as for how you drive the MS20 thing -- I don't know. I'm just showing you an osc for driving a  led. carry on.

It drives it beautifully - and the best part is not having to find an LDR that works with the LED and has a linear response (because none of the ones I own seem to have a linear response). And I get a proper rate LED as well! THanks for your advice; awesome as ever.

I've redrawn the schematic - would you mind checking this is right? I want to make a vero version so I can start building this ASAP! Thanks.

[amptramp]

It would be interesting to see if an LM566 voltage-controlled oscillator would be usable for an LFO.  It needs 10 - 24 volts for Vcc but it gives linear sawtooth and square waves.

[duck_arse]

"no, don't play it like that, play it like this"

when can we hear the master at work?

I've redrawn the schematic - would you mind checking this is right?

errors. pin 3 connects to two 100k resistors, you only have one. and, that 100k is well lost, because there should be one connecting hard-to pin 7, not via switches and things.

[PRR]

It needs 10 - 24 volts for Vcc but it gives linear sawtooth and square waves.

I bet a buck it will oscillate somewhat below 10V; just that it won't necessarily meet its "very linear" spec. Slick for remote temperature readings, maybe not essential for guitar sweep?

In the 1970s we would just try it. I guess if we still have any they are deep in the bottom box in the bottom of the closet? It seems to be N/A at reputable suppliers.

[ElectricDruid]

It would be interesting to see if an LM566 voltage-controlled oscillator would be usable for an LFO.  It needs 10 - 24 volts for Vcc but it gives linear sawtooth and square waves.

My first thought is "Does it tick as badly as the 555?".

[amptramp]

It would be interesting to see if an LM566 voltage-controlled oscillator would be usable for an LFO.  It needs 10 - 24 volts for Vcc but it gives linear sawtooth and square waves.

My first thought is "Does it tick as badly as the 555?".

Probably - it doesn't even give a power consumption.

[moid]

when can we hear the master at work?

Once I've got this built as a circuit on vero - at the moment it is a chaotic mass of cables, crocodile clips and whispers of hope holding it together like some demented spiderweb of enthusiasm and lies... it keeps falling apart whenever I change any setting, so the need to solder it together is great! I'll record him doing a demo through it (another good reason to pry him away from Whatsapp).

errors. pin 3 connects to two 100k resistors, you only have one. and, that 100k is well lost, because there should be one connecting hard-to pin 7, not via switches and things.

Thanks Duck - I'm a bit confused by the bit about pin 3 connects to two 100K resistors. I can only see one 100K resistor in your schematic (R3). I changed your R1 to 150K, is that what you mean?. The 100K now connects (I hope) properly to pin 7 and I realised I had the capacitors and their switch drawn wrong so that is now corrected. Would you mind casting your eyes over the latest drawing please?

[moid]

It would be interesting to see if an LM566 voltage-controlled oscillator would be usable for an LFO.  It needs 10 - 24 volts for Vcc but it gives linear sawtooth and square waves.

That will have to remain a theoretical item for me I'm afraid - I don't have any of those chips. I have some LM567 but I doubt that is that similar.

[duck_arse]

yessir, sorrysir, it struck me after I'd turned off the machine that I'd misvalued that resistor.

but, still - errors. there should be a 100k between pin 3 and pin 7. there should be a 150k between pin 3 and pin 1. so you need two resistors [values unspecified] connecting to pin 3.

your caps. you now show both sets of caps connecting across pin 7 and pin 6. you should probably really wire the low value pair hard-to pins 7 and 6, and then switch the large pair in parallel with, for the slower range. so one end of the 470's to pin 7, the other end to the switch common, one end of the switch to pin 6. other end of switch remains unconnected.

and your 100R stopper resistor should be connected to pin 6, not through the switch.

AND - the other switch with the diodes - the diodes should be connecting to pin 6, you have them at 7. news at 10.

[PRR]

power consumption

[moid]

yessir, sorrysir, it struck me after I'd turned off the machine that I'd misvalued that resistor.

No need to apologise; I'm so grateful for your help (and patience) I suspect you must feel like you are talking to this cat at times...

but, still - errors. there should be a 100k between pin 3 and pin 7. there should be a 150k between pin 3 and pin 1. so you need two resistors [values unspecified] connecting to pin 3.

This bit really confuses me (sorry!) To me in the drawing there is a 150K resistor joining pins 1 and 3 - I drew a line from pin one that hops over the line going to pin 2 and then connects with the 150K resistor which goes to pin 3. Presumably my drawing doesn't mean that? The same goes for the 100K resistor - to me that is a line that connects pin 3 to the 100K resistor and then travels to pin 7. Sorry I don't seem to understand what I'm doing wrong here.

your caps. you now show both sets of caps connecting across pin 7 and pin 6. you should probably really wire the low value pair hard-to pins 7 and 6, and then switch the large pair in parallel with, for the slower range. so one end of the 470's to pin 7, the other end to the switch common, one end of the switch to pin 6. other end of switch remains unconnected.

and your 100R stopper resistor should be connected to pin 6, not through the switch.

AND - the other switch with the diodes - the diodes should be connecting to pin 6, you have them at 7. news at 10.

I think I've changed all the above bits so that they are now right? Here's my current version of the circuit:



Am I right in thinking that the 470uF caps (when selected via the switch) now have the 22uF caps added to them, so that the slow version of the circuit will be slightly slower? That's fine with me, I'm just tryng to make sure I understand some of this circuit!

Thanks Duck - I hope the rest of your day / night is less annoying!

[moid]

Wait just a minute I might have finally understood the bit about the 150K resistor and pins 1 and 3 - does this look better?

[duck_arse]

it does, yes, look right. also yes. the 470's go in parallel and add capacitance to the 22's, so you get a slightly lower speed than bare 470's.

[moid]

Brilliant! Thanks Duck Sorry it took me so long... "slightly lower" might be my middle name... would you mind looking at this vero drawing please? I drew something else as a vero layout recently and built it and it doesn't work so I'm wondering if my vero skils are not as good as I would have hoped...

[duck_arse]

to the best of my squinting, that looks ok.

[moid]

Thanks Duck I'll get assembling!