More help concerning Geofex panner?

[egasimus]

Hi, my apologies for starting a new topic, but my problems are now of a completely different nature, and I figured I have a greater chance to get your attention like this.

I'm building a circuit that allows me to blend an effect in and out, based on this (page 2, bottom circuit). I have, however, replaced the 10k pot and the 15k and 51k resistors with 100k, 150k and 510k.

The blending works fine, however the volume is too low, and the blend pot is prone to noise when in the middle position, which makes me think of an impedance problem. Do I have to replace the 1M resistors at FX in and FX out with 10M ones?

Also, the circuit clips quite early. I checked the voltages at the opamp's pins, and, while the second half is OK, the first half has 3.5 at the non-inverting input and 4.5 at the inverting one and at the output. Those should all be within 10mV from each other, right? Try as I might, I couldn't make it right.

I'm sorry for entering your community head first with such persistent pleas for help, but I'm in kind of a hurry, and this pedal has already taken me more time than I could possibly give. I'll be immensely thankful if somebody pointed me in the right direction.

[egasimus]

here's my schematic. a bit crammed, sorry for that.

[slacker]

I have, however, replaced the 10k pot and the 15k and 51k resistors with 100k, 150k and 510k.

Those changes should be fine, they might add a little noise compared to the original values but I wouldn't think it would be anything to worry about. You don't need to scale up any of the other resistors.

Also, the circuit clips quite early. I checked the voltages at the opamp's pins, and, while the second half is OK, the first half has 3.5 at the non-inverting input and 4.5 at the inverting one and at the output. Those should all be within 10mV from each other, right? Try as I might, I couldn't make it right.

It's normal to measure less than 4.5 volts on the non inverting input of opamps when they have large value resistors from them to VCC/2. What happens is the impedance of your meter acts like a resistor connected from the non inverting input to ground, this makes a voltage divider together with the resistor going to VCC/2 and gives a reading that will be lower than expected.
If you know the input impedance of your meter you can work out roughly what measured value equates to 4.5 volts. Say your meter has an impedance of 1M, which is common, then the voltage divider effect will drop the voltage to 1000k/(1000k+100k) * 4.5 = 4.09 volts. It will be slightly lower than that, because the 47k resistors used to generate the VCC/2 come into play as well but it's close enough. So if you measure around 4 volts it means the voltage at the non inverting input is really 4.5 volts.
This means your 3.5 volts is probably fine, the fact that the other pins have 4.5 volts on also indicates that everything is fine.
What sort of levels does it clip at?

Thanks for posting that panning article, I hadn't seen that before, there's a bit in it that would be really useful for something I'm doing at the minute

[egasimus]

This morning, the device works, as far as I can tell, just as expected. I think it has to do with extensively bathing the PCB in alcohol. I use this flux which is conductive and has proven quite hard to remove, and it's also called "FU". No, really. The 3.5V at the + input of the opamp is probably due to my DMM's input impedance. Clipping is because of my amp, and the volume is OK. Made a lot of silly mistakes, haven't I? There's still a lot to learn before I can call myself an electronics engineer.

Thanks for the advice, you've given me some useful insights. Now, going back to the bottom half of the schematic (relays), which I built beforehand, which previously worked, and which is probably going to give me some more hell until I get it to do what it has to...

[R.G.]

Sorry I didn't get to this sooner. Yesterday was kinda eventful.

It makes sense that conductive flux could have messed you up. Some do, some don't. The worst ones are the ones which pull in air humidity so the thing works on dry days but not humid ones.

Any time you start messing with resistances of 1M and over, there is the possibility that meters, humidity, fluxes, house dust, and in general Maxwell's Demon will obscure and obfuscate what's really happening.

Glad you got it working. Good work.