MS-20 filter fun!

[moid]

Thread Resurrection! See Igor, it lives! Why, I've created a monster... well kind of... but first Happy New Year to you all, hope you had a nice break. I have a few days left before I have to return to work so of course my first thought was what shall I do with my time, shall I compose some epic poetry? Paint a masterpiece? Fix that dripping tap (faucet) in the bathroom? Nope! I'm gonna make this MS20 filter work dammit! So yesterday I spent a few merry hours soldering things, (because why build things on breadboard and test them when you can solder everything together and then discover you have problems... ughhh do I ever learn?) and then plugged a guitar into it and... it's a bit of a curate's egg. Parts of it are quite excellent; the rest, not so much.

The pedal only works if the LFO circuit is connected to it. With the LFO it makes good sounds (apart from a few problems, see below). Without the LFO no audio passes through the circuit (the true bypass on the footswitch works fine). I suspect the main MS20 circuit is not generating resistance correctly that should go to the Frequency pot. When I connect the LFO circuit to lug 1 of the MS20 frequency pot (I originally tried lug 2 as in my layout above, but that does nothing unless the pot was either fully counter clockwise or clockwise; I've forgotten which), this now allows a certain amount of control over the sweep in that turning the pot raises / lowers the frequencies being swept (I think?), although that tends to cut out all of the signal after some point on the pot (not instantly; it gradually allows less frequencies to pass to the point where nothing passes).

I've noticed that when the LED on the LFO is brightly lit no audio passes in the circuit - when the LED fades towards being off, the audio is audible and the frequency is being swept. The problem is that the LED is on for longer amounts of time than it is off, which if the LFO speed is low is very annoying (not noticeable at faster rates). I measured the resistance on the LFO output and when the LED is on it is outputting 26 - 38M! When the LED is off the amount drops to 0 ohms.

The LFO ticks when the Slope Direction switch is set to Up or Down. In the middle position there is some faint tick I think (not always, I suspect it has something to do with the Resonance pot setting on the MS20, but I might be wrong). Also the speed of the LFO is faster whenever I switch from the middle position to the Up or Down position. That's not a major issue, I'm just mentioning it in case that is not expected behaviour)

So it's question time! What should I check / measure on the MS20 part of the circuit to get that working? How do I stop the LFO from cutting out all audio? How do I stop the LFO from ticking? Why do dogs lick their balls? What is your favourite colour? Any helpful or just plain amusing replies are gratefully received

EDIT: Sorry another question - if I wanted a stronger Resonance Effect would that be as easy as swapping the current pot out for a smaller or larger pot? At the moment the resonance is fairly genteel and nothing as extreme as the wild sounds I once heard a guitar player make when he plugged his guitar straight into an MS20 keyboard.

[moid]

OK it occurred to me that I could do some simple tests in case the valuestell you anything useful. I had disconnected the LFO circuit from the MS20 and I'm measuring lug 2 of the Frequency pot on the MS20. When the pot is on minimum (7 o'clock) the DMM reads 2.1 ohms. When the pot is on maximum (5 o'clock) the resistance is around 58K, but there must be a loose connection or something wrong somewhere because when I pick up the PCB and adust the pot sometimes the measurement will switch to a crazy high value (35M) or go to 0 ohms and then not return to anything sensible unless I yank the power out - when I put it back in the values can fluctuate... so I've done something dumb somewhere...

[moid]

and for those of you who love pictures, here is the MS20 part of the pedal. One note the vero board has an extra (unused) strip at the bottom because I found this piece in my bag of scrap vero pieces and it happened to be exactly the right size (apart from being one row too large) and I was too lazy to cut that row off. I have checked the vero layout against my PCB and they are both the same, no missing cuts or components so I am definitely confused with this one!

[moid]

OK some more testing equals more weirdness. If I plug my phone into the circuit and play something from youtube the circuit works fine! I can also connect the LFO (to lug 2 of frequency) and that works fine as well. There is no loss of audio when the LED goes bright. I can't hear anything if I put a guitar through it though(without connecting the LFO)? That makes no sense... is this an impedance issue? The circuit is very sensitive to me adjusting / disconnecting parts of it - I'm not sure what parts I might accidentally touch when doing this, but whatever it is kills the audio until I remove the power and plug in again. Any guesses folks? Is thissomething that might only work properly once it has been boxed? I don't think I can hear a ticking noise anymore either!

[moid]

Ok embarrassing time... something must be shorting things out somewhere - I managed to get the guitar to play through the circuit without any LFO by holding some of the switches and pots at different angles to each other! I cannot figure out which one causes the problem, but it seems like this circuit might be better off inside an enclosure and then check if it works when nothing can flex about! Hope I didn't waste everyone's time. Will try to drill a box tomorrow and see what works then!

[duck_arse]

moid's back.

always always always never ever measure resistance while your circuit is powered. this is always, unless you have an ldr which is isolated from the circuit while it's being flashed or something. you can't measure resistance in a powered circuit - not with the tools we have anyway.

and about those direction diodes. there is some resistance and some capacitance, the opamp charges the cap via the resistance, and the opamp discharges the cap through the same resistance. but then you modify, and the opamp charges the cap thru the resistance, but discharges the cap thu the diode, whoosh. there goes half the time it took for charge/discharge thru the resistor.

flip/flip the switch, so now whoosh the cap charges via that diode, doesn't see any resistance until the opamp starts discharging the cap again. so each slope-selected shortens by half the set time constant of R and C, and you get faster oscillation rate.

and because whoosh, you might also get tick, because current. so, add a stop/softening resistance to the diodes leg, or rearrange that string so your stop resistor for the pot is also seen by the diodes when in-switched.

about your pics -- tee-hee -- I still can't see imgur imgs outside the page. so I'm off the hook.

[moid]

No, no, this is moid's front, see? it's the bit with the eyes and mouth on it. Not moid's back.

Aaah there I go again making stupid decisions and measurements... let's put it down to a sad lack of time to think or do anything electronic and my inability to recall basic information. First day of school again! Yes, that does make a lot more sense, and sadly I now recall you telling me not to do this in the past now that you've told me again... kind of deja view, but the sight to be seen is not front to back but back to front. Which is neither here nor there if you can't see my pictures. No wonder you thought you were talking to my back; maybe a visit to the opticians is required Duck? Says the man who is currently trying to use varifocals through his optical visor to see components... this aging process is highly overrated I think.

The ticking may not be an issue; it might be caused by dodgy connections / shorting something because the mess of vero board and cables keeps getting entangled and upsetting itself (it might be better for you Duck that you can't see it - it's a sight for sore eyes, not duck eyes). I've not been able to drill the enclosure and ram the mess inside because the weather has turned suddenly very cold (it hasn't got above -1 degrees C all day!) and all my drilling takes place outside (because there's no room inside for metal work sadly) and I'm not standing out in that cold while trying to hold a drill straight... so this might be a short visit from me alas...the weather will warm up again next week, but then I'm back at work from Monday Once I can get this enclosed I will test again and see if the tick has gone - I'm sure it disappeared at one point while I was playing through it, so hopefully it's not a problem. The pedal does sound good though (when it works!).

Anyway it is good to be back and to hear from you again, hope you had a good Christmas / New Year's and your weather is somewhat better than ours!

Now where's that hook probe you're off about? I'll use that to make some measurements with!