Question(s) about GEO CCA volume pedal

[David]

I have been doing some research and experimentation in preparation for breadboarding the GEOFEX CCA volume pedal implementation.  So far, I haven't gotten past the op-amp buffer that receives the wah pot as input and sends current to the 13700 chip.  I was able to breadboard the necessary buffer and bias it with a resistor divider so I could use a single-ended supply.  By way of a test mechanism, I used an LED.  Yup, moving the wah pot changes the brightness of (and, I assume, the current going to) the LED.  I got it to work as diagrammed in the GEOFEX article -- maybe.

The "maybe" is because I noticed something odd.  I noticed that, near the end of the pot's travel toward minimum input (therefore maximum dimness of the LED), it hits a point where it suddenly goes from what is probably sufficiently dark to consider as zero input, to quite bright again.  And it stays bright to the end of the pot's travel.  Thinking this was a glitch in the pot, I replaced it for a test with a 100K linear taper pot.  Same result.  Now, this is probably not the end of the world because in the real world, you could cover this up just by leaving the footswitch in place in a Vox or Crybaby pedal.  The footswitch would stop the pedal's travel before the "Surprise!" point occurred.  I'm looking for more than this, though, because I'm trying to come up with an implementation that has a longer pedal travel than the traditional Vox/Crybaby configuration.

Has anyone else encountered this event?  If so, is there any way to work around it?

Thanks!

[slacker]

Do you mean the one in the "Applications of an audio Current Controlled Amplifier" document?

Some opamps like TL072s freak out if the input gets too close to the negative supply and like you've found the output voltage shoots up. Someone will probably come along with a more technical explanation, but it sounds like that's what the problem is.
You could try putting a resistor say 10k or between the wiper of the pot and input, that might stop it. Or put a smaller resistor between the bottom of the pot and ground to limit the minimum voltage.

[David]

Yes, that's the one I mean.  I didn't use a TL072, but I DID use a TL071.  I infer that, due to its being in the 07x "family", it will exhibit the same "freak out" characteristics?  Will another op-amp deal with this better -- a 741, maybe?

[slacker]

Yeah any TL07X type opamp will do the same. LM324s work fine for this sort of thing, not sure about other types.

[01370022]

It's got something to do with the input circuit on the opamp. I was reading it in the Art of electronics last time I got it from the library but can't remember the name of the glitch or the technical reasons for it happening. Fat load of good it was me reading that then

The short answer is that slackers right, and you're right about the same family displaying the same glitch. If someone who actually remembers their EE education can give the technical details and the opamps internal configuration that causes the problem, you should be able to look through the data sheet for any given opamp and see if it will display the same glitch.

To solve your problem either try slackers answer or try a rail to rail opamp. I think a 741 will do the same thing due to it being an older opamp design. I could be wrong about that tho.

Edit: Slacker got there first

[cpm]

as slacker said, it may be related to "latch-up"

http://www.analog.com/library/analogDialogue/Anniversary/6.html
(at the end of the page)

datasheet says TL07x is "latch-up free" which means you wont ruin the chip, but it still would probably do that kind of strange things.