Envelope State Variable Filter - No Funking Quack

[patrick398]

So i've just breadboarded this state variable filter suggested to me in my last thread about getting crazy sweeps out of a Mutron V.

http://www.sabrotone.com/wp-content/uploads/2012/12/EnvelopeSVF.png

It passes signal which was a nice surprise but there's no quack...not a noticeable amount anyway. The pots all seem to be working correctly except for the attack where i gradually lose signal as the pot is turned up. The leds light up when a play the guitar and the decay pot affects the decay of the leds correctly. I'm starting to wonder whether the ldrs i'm using may not be suitable? I'm using these:

http://mklec.com/components/resistors/ldr-light-dependent-resistors/GL5539-ldr-light-dependent-resistors

Also, this is probably due to my lack of knowledge but i was surprised to find 4.82v on so many pins of IC1...

IC1
1 - 4.82
2 - 4.82
3 - 2.36
4 - 9.73
5 - 4.60
6 - 4.82
7 - 4.82
8 - 4.82
9 - 4.82
9 - 4.82
10 - 4.60
11 - 0.00
12 - 4.59
13 - 5.04
14 - 4.83

IC2
1 - 5.02
2 - 2.87
3 - 2.35
4 - 0.00
5 - 4.09
6 - 4.11
7 - 4.11
8 - 9.72

Any thoughts or suggestions would be gratefully received

[PRR]

> i was surprised to find 4.82v on so many pins of IC1...
> IC1 .... 4 - 9.73

R23 R24 derive a half-9V reference so all the opamp stages will "center", be able to swing both ways like audio.

Your "9V" is 9.73V. Half of 9.73V is 4.865V. Which is "exactly" your 4.82V (1% difference, which is closer than we deserve with 5% parts).

> IC1 ... 3 - 2.36

This is half of what it should be, and is probably correct. Pin 3 is held to 4.82V through a 1 Meg resistor. Your meter is evidently 1 Meg loading (not uncommon). When you are not poking it, pin 3 is surely at 4.82V, as seen by your reading on pin 1 (which "follows" pin 3 but with negligible loading).

> The leds light up

I suspect you know this, but: the LEDs and LDRs must be inside light-tight boxes to work right. You can peek for debugging, but room-light will spoil the expected action.

> the ldrs i'm using may not be suitable?

90K light resistance is really high! You want to overwhelm R9 220K; 90K is not a huge change. Is that spec even right?? Use your ohm-meter and hold the LDR to a bright light. If really over 10K, get a juicier LDR. Most will go close to 1K in very bright light. Modern LEDs right against the LDR will be very-bright.

[Mark Hammer]

Assuming R23 and R24 are precisely 100k each, or otherwise identical in value, yes the battery should measure 9.7ish, and Vref should be 4.82V.  If they are 5% resistors, then R23 might be 95k and R24 is 105k, in which case V+ would be less than 9.7V and Vref a bit more than half of V+.

Not a problem, but if Vref measures at 4.82 and V+ doesn't seem to be exactly twice that, you shouldn't worry; dividers don't always divide perfectly.

[patrick398]

Thanks for the replies. I'm using 1% tolerance resistors so that figures
I've got the led/ldr wrapped up securly in some electrical tape and also turned all the lights off last night so there's no risk of other light source influence.
I just turned the leds to full brightness (i'm using ultra bright 3mm red) and measured the ldrs, one measures 7.82k and the other 0.56k. Not sure why they'd have different resistances, maybe one has slipped out fractionally.
I'll go back through the circuit today and tripple check everything, i was really hoping from i'd be able to localise the problem though.
Am i right in thinking the envelope detection part of the circuit is functioning correctly and it's a problem somewhere around IC1?

Thanks

[patrick398]

Ok so i've had some progress here. Seems to be working well now. I changed the LEDs for super bright pink 3mms which made a small improvement.

I changed R10 and R12 to 220R which seems drastic but sounds good and i'm guessing it's compensating for the LDRs i'm using.

For R9 and R11 i'm using the same GL5539 LDRs (only ones i have) so when i wiggle a torch in front of them i get the crazy big sweeps i'm after.

I tried using another LED and pairing it with these LDRs in the same way that the other optocouplers are set up but didn't quite get the results i wanted. As i see it i now have three options;
1. set up another LED which goes from fully on the fully off and pair that with the LDRs.
2. Use an LDR which works in the opposite way to the one i'm using now so that the resistance decreases with less light...does that exist?
3. Use some kind foot operated pressure pad so i can vary the resistances of R9 and R11

Maybe i'm making this overly complicated...but i guess that's kind of the point isn't it?

[patrick398]

Update: Just tried a 555 controlled LED to control the LDRs on R9 and R11 but it introduced a lot of clicking into the signal, don't think it's quite what i'm after anyway. I'm starting to lean towards the pressure pad thingy majig. Something like the 'Igor' that comes with Rainger FX Dr Freakenstein.

http://www.raingerfx.com/product/dr-freakenstein-fuzz-drff-3-and-igor/

I'm thinking one of these could work:
https://cdn-learn.adafruit.com/downloads/pdf/force-sensitive-resistor-fsr.pdf

'Resistance Range - Infinite/open circuit (no pressure), 100KΩ (light pressure) to 200Ω
(max. pressure)'

I guess if there's infinite resistance with no pressure it will be necessary to have a resistor in paralel? Say 220k?

[bluebunny]

If you're playing with 555s, go for the CMOS version (7555).  It draws much less current and therefore introduces much less LFO wobble back into the power supply lines.

[patrick398]

Thanks for the tip

[Kipper4]

Well done you for getting this far.
That SVF works. I have built more than four variations.

If when you shine a bright light source at the ldrs you get the sweep you wanted. Then I suggest  you get more light to the ldrs.
I dont like the CLR/Gnd pot1 (2k2) arrangement.
If you firstly make sure the ldr is getting max possible light source by pointing the lens at the ldr.
Theres more than enough roll your own vactrol blogs in internet land.

If it where me i'd use seperate fixed resistance to eachled.

You can carry on with the pot1 if you want.

You did well to diagnose it passing signal for one then figuring out the envelope detector is working.
Then the issue with the leds.

I'll go look at my old schematics and see what i did with the envelope and led driver.
I likely used an lm358 for those dutys.
Probably used a diode charge pump Envelope detector.

Rich

[PRR]

> LDR which works in the opposite way to the one i'm using now so that the resistance decreases with less light...does that exist?

No. Can't happen. Light excitation stirs-up the sluggish electrons so they conduct better. It seems unlikely we can have the electrons stirred-up normally, and calm-down when excited by light.

> Use some kind foot operated pressure pad

? ? ? I thought you wanted "the envelope detection"? Now you want manual (foot) control?

I'd bet 10 cents your troubles are the 10-cent LDRs. That price is suspiciously low. Smells like rejects.

[patrick398]

Thanks guys!
I'm happy with how it sounds now but i do have a feeling i can get it to sweep a bit more. I am using the envelope detection in the normal configuration i guess but i want to have the option for getting a bit crazy and introduce mad sweeps that i can control (hopefully via an FSR). Damn physics getting in the way of my plans for a LDR that works backwards. You have to admit a Dark Dependant Resistor sounds pretty cool...for all you goths out there.
PRR i think you may be right about the LDRs, although the link i posted wasn't where i got them from, i did get them cheap off ebay so could be crummy.
One problem i am having is that is the envelope sweep dies out it gets pretty boomy with certain pot positions. Like if i play an open A string and let it ring out, as the note dies away it sounds like the lower frequencies are feeding back or something.

Stay Tuned

[Kipper4]

You can probably get the leds to work the other way. Simple fix to the dark dependant resistor idea.
It might mean altering the E.D. and led driver biasing, some switching.

"One problem i am having is that is the envelope sweep dies out it gets pretty boomy with certain pot positions. Like if i play an open A string and let it ring out, as the note dies away it sounds like the lower frequencies are feeding back or something."

Might be the result of envelope ripple.

I aint no Goth........



Have fun with the pressure pad. Make something we'd be able to build.
That reminds me I should get some FSR's

[patrick398]

I'll be sure to put a schematic together when i'm finished. You'll certainly all be able to build it, the question is whether you'd want to. I think my idea of what sounds nice differs quite drastically from other people haha.

What exactly roughly is envelope ripple? Is it indicative of a problem somewhere? And how do it kill it?

[Kipper4]

Envelope ripple is an artefact from the envelope filter that might explain what your hearing.
Might not be.

You might find out more here.
http://www.diystompboxes.com/smfforum/index.php?topic=49246.0

Mark Hammer has some very succinct explanations on the forum.

Search and you will find.


Rich

[patrick398]

Hey guys,
The FSR arrived and all my strange little dreams have come true. Landed somewhere pretty close to what i wanted on the breadboard so will put it on vero soon. Will also draw up a schematic but it only deviates very slightly from the Sabro SVF schem to which i am eternally indebted.

Here's a short sound clip. The input jack on my mac has shat the bed; just had to record through the in-built mic so it sounds pretty crummy but you get the idea. There's some obligatory funk chords in there which i can only apologise for.
Chords at the beginning are just the envelope and then the riffs at the end are using the FSR with an overdrive before the breadboard.

https://soundcloud.com/user-418904944/svf-with-fsr-14092017-1221

[Kipper4]

Yep nice Pat.
Very nice vocal tone to it, but that might be the amp too.
Can't wait for the drawing.
Please link us to the fsr.
Did you do an up down mod?

[patrick398]

Thanks
The amp i was playing through is truly awful. It's a horrible little keyboard amp. Really, indescribably awful haha. I do need to get a decent amp to test these pedals through.
I'll try and get a schem up online tomorrow, did a hand drawing today but no way of uploading it.
FSR is an interlink FSR406:
http://www.ebay.co.uk/itm/Interlink-Electronics-FSR406-1-5-Diameter-Force-Sensing-Resistor/381939766453?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2057872.m2749.l2649

Didn't do an up down mod...i think the interaction of the attack and decay pots in this set up are just pleasingly versatile