Help debugging Bi-Filter

[captntasty]

I hope someone can help me on this one... I put together the Craig Anderton Bi-Filter Follower and am not getting a whole lot of "funk".  I initially had two roll-yer-own vactrols in there from RatShack LDR's (the tiny ones) and 3mm red LED's and it "wahhed" a tiny bit... So I threw in some CLM600's that are in functional but bad shape (leads had been cut short and I needed to solder some new legs on)...  I got even less "funk" - it now sounds like a static wah pedal heel almost completely down - there's definitely a honk to it but no sweep.  The sensitivity and sweep pots are unresponsive.  I have to guess the LED/LDR's are not doing their thing. 

I've been through the usual stuff with no luck (solder bridges, traces cut, etc).  I am using TL072's for IC's and before I started I didn't realize this had a 10pf cap in it - I have a 22pf in there now and I've tried 100pf without any difference.  I've read every shred of literature out there about the Bi-Filter (I might add there's not that much!) and on this site.  Please let me know if I have left any info out.

I hope someone can help me out on this one - is this a job for Mark Hammer? 

Peace
P

IC1
P1 = .002v
P2 = .002v
P3 = 0v
P4 = -8.76v
P5 = 0v
P6 = .003v
P7 = .003v
P8 = 9.03v

IC2
P1 = .115v - .40v w/ input
P2 = .005v - .015v w/ input
P3 = 0v
P4 = -8.76v
P5 = 0v
P6 = .002v
P7 = .012v
P8 = 9.02v

D1
A = 9.69v
K = 9.01v

D2
A = -8.73v
K = -9.43v

D3
A = .11v
K = 0v at rest - 4v w/ input

CLM600A
+ = 0v - 1.7v w/ input
R = 55k at rest, +2M w/ input

CLM600B
+ = 0v - .80v w/ input
R = 40k at rest, 90K w/ input

[Mark Hammer]

Well first verify that the signal path through the input and filter stages is actually a viable audio path.  Disconnect the 47k input resistor from ONE of the filter sections and sub a 100k or greater pot (wired as a simple variable resistance to ground) for the optoisolator in the remaining filter.  Strum your guitar and twiddle the pot.  You should hear a wah-wah effect at the output.  If not, then the screwup is somewhere in that path.  If you DO hear a wah-wah like effect, then connect the 2nd filter section and repeat the same kind of test, leaving the first filter at a fixed frequency-setting.  If you hear wah-wah, you're clearly in possession of a solid audio path and the "problem" lies somewhere in the envelope follower portion.

You can use the AC setting on your DMM and check at the outputs of the relevant op-amps to see if there is an appropriate signal level in response to your strumming.  The stock circuit has a gain of around 4 in the first stage IIRC, so the output of that first op-amp shold be around 4x the input signal level.  The rectifier/detector circuit has a substantial gain, but it also has a DC bias voltage added on to the audio signal, so you'll need to verify the DC shift at the output of that op-amp as well as the AC "flutter" that rides on top of it.

You can consider measuring the DC voltage at the output of that op-amp AND on the other side of the LEDs as a way of verifying that they introduce the voltage drop expected of an LED that isn't fried (should be about 1.5v below what comes out of the op-amp).

These are a number of strategies for trying to troubleshoot the circuit.  There are more, I suspect, but try these on for size and report back.  This IS a circuit worth getting off the ground.

[StephenGiles]

An analog volt meter would be useful to look at the dc sweep voltage at the output of the rectifier, as you can see the voltage change as the needle moves accross the dial.

[captntasty]

OK, I did my best on this but I'm a little confused about a couple things...

I replaced the LDR's with pots and both filters wah as I move the pot.

Here's where I get a bit confused - I measured pins 1 & 7 on both IC's (the outputs) here are the results in VAC.

IC1 (closest to optos)
1 = 0v at rest - up to 0.7v when strummed
7 = 0v at rest - up to 0.2v when strummed
the only "output" that reacts is on pin 3
3 = 0v at rest - up to 0.4v when strummed

IC2 (closest to 10pf cap and D3)
1 = 0v at rest - 0.7v when strummed
7 = 0v at rest - 0.7v when strummed
none of the other pins react

I can't identify the rectifier/detector circuitry - a bit beyond my knowledge.

I did measure "the other side of the LEDs" (positive leg of the LED)...
they both behave the same, they rest at about 1.25v (which starts to drop once the DMM probe touches) and rise to about 1.65v max when strummed.

I would, of course, like to know what the rectifier/detector circuitry is so I can provide you measurements.

Thanks so much for the replies.
P

[Mark Hammer]

Okay.  Clearly the problem is in the path between IC2b (I'm using JD Sleep's drawing over at GGG....temporarily closed) and the ground side of the LEDs.  In other words, there is no sweep because the signal is not being rectified properly, or the rectified signal is not creating the appropriate change in the LDR.

I'm trying to leave work shortly, so I'm going to be superficial rather than thoughtful.  My apologies in advance.  Consider the following possibilities:

1) Your diode is backwards.
2)  Your optoisolators are installed backwards (the dot should be connected/going to ground)
3) The TL072 is somehow fried
4) The chip is good but not seated properly in some fashion.  Maybe a pin is bent underneath.  Maybe the socket is not making good contact with one or more pins.
5) The LED on one or both optoisolators was fried on installation.
6) The usual stuff like solder bridges and things being connected in a manner different than what you thought.

Note that, even without a signal, adjusting the "Sweep/Calibrate" pot will result in variations in illumination of the LED ionside the opto and consequently the LDR resistance.  Just for the heck of it, you can temporarily remove one of the optos from the circuit (just lift one side of its LED) and replace that one that you can't see with a normal LED that you can see, and check to see if varying the sweep pot produces visible change in the illumination of that LED.  If you ain't getting anything, then there is a good chance the problem is not with the optos but with the circuit driving them.

[StephenGiles]

I'll go with all of Mark's post, having thrashed this beast out several years ago. Unfortunately my board is not to be found at the moment - never boxed of course!

[captntasty]

I tested to see if the LEDs light and the did... afterwards I got to thinking    maybe the LED's were not bright enough - tada!  I reduced the current limiting resistors R1 and R2 (1k's) to 470R and lowered the resistor that feeds that network, R11 down to 82R.  I also dropped the input cap, C6, down to .001uf to allow for more highs (with the original value .01uf it was very low frequency).  I also played around with other spots but these tweaks seem to be the keys.  It sounds pretty good but I had another idea    what if I drove the front a bit more?  I was playing single coils and know that their output isn't huge... I put a BOSS CS-2 Compression Sustainer with the level turned almost to max in front and the Bi-Filter became even more responsive!  I'm sure I will tweak some more and may try rolling other optos with larger LED's, maybe Hi-Brightness Red's.  I'm also going to give some Mark's tweaks a try - first the switch 3 where the 1200pf and 2200pf caps are split.  Are there any layouts for the inverted/non-inverted signal mod?

Thanks guy for all your help  I couldn't have gotten this one up and running without your input.  This is going to work great for what I have in mind - a chain of effects that equal a sort of guitar synth... I think this approach has been taken before - Blue Box, UglyFace into filters (I'm using the FSH1 for sample & hold) of various sorts and clean Octave down effects (I have an Arion Octave that works pretty well).  Well, I'm babbling so again, Thank you...
P

p.s. I forgot to mention that I tried the CLM600's after the effect was working and was puzzled that the filter effect was pretty much gone (I have confirmed that the optos do work).  Hmmm   I put my hand-rolled optos back in and voila - funky filter!

[StephenGiles]

That is diy electronics my friend!