Need an assist on a weird voltage problem - flanger

[bean]

I'm working on a BF-2 flanger style project. I've run into an issue I simply cannot solve. I have a stray voltage on the CP1 clock pin that shouldn't be there. CP1 is connected to pin2 of an MN3102 and pin2 of an MN3207. There are no other components connected to these pins or PCB trace. Even with both chips removed I'm reading about 2.5v on CP1. This is kind of baffling. I've removed both sockets, cleaned the pads, even removed the power decoupling caps etc. Checked for continuity errors on the PCB and there are none that I can find that would result in DC being read on these pins.

I'm wondering if my design is bad. Often with modulation pedals I'll put the LFO and audio on separate ground pours then connect them with a small bridge on top and bottom. I did that on this layout, but I also placed the clock further away from the BBD. I'm wondering if that could induce some weird impedance thing here or parasitic capacitance. I'm kind of wracking my brains on it.

Illustrations below if anyone has input. The power and ground on the BBD are selected by jumpers according to bbd/clock type used but that should be irrelevant to the issue here.

[bean]

One thing I forgot: even with the 5v regulator and 1n914 removed I'm reading about 250mV on that same pin. Just bizarre. I do read proper ground on grounded pads, though, and approx. ground on pin4 (CP2).

[Mark Hammer]

Do the CP2 pins on both chips read as they should?

EDIT:  Ooops should have read your 2nd message! 

[Rob Strand]

Even with both chips removed I'm reading about 2.5v on CP1. This is kind of baffling.

With both chips removed there's nothing on that line.   I'd be thinking PCB leakage.
Perhaps the PCB has a *very* fine layer of copper still there, or, perhaps a salt residue.
Cleaning will remove the salt but not the copper.

With the chips out try adding a resistor to ground and/or +5V.    Try 100k, 1M, 10M that type of thing.

If you can see 2.5V with a 10M input impedance meter there could be 10M to +5V (or higher to 9V).

On the other hand it could be leakage from 5V (and/or 9V) and to 0V forming a 2.5V voltage divider.
In this case the impedance of the leakage path could be anything.  That's where loading the line
comes in.

Maybe if you get down to 10k you can start to realize there's a short to another track, perhaps through
an invisible connection of copper on a mis-etched board.

You have to start somewhere to narrow down the problem, even if it's only a stepping stone.  The problem could end-up being something completely different.

[anotherjim]

Measure for AC also. A DMM on DC can try to make something of it. If there is some AC, do something to stop the LFO and ground the signal input.

[bean]

Do the CP2 pins on both chips read as they should?

EDIT:  Ooops should have read your 2nd message! 

Actually now that I think about it, I mis-rememberd: I was seeing some weird stuff on pin4 under power. It was reading at the supply level which is about 5.7v and it should have been around half that, IIRC. I know the chip is good though. I checked for frequency output as well but got some inconsistent results there.

With both chips removed there's nothing on that line.   I'd be thinking PCB leakage.
Perhaps the PCB has a *very* fine layer of copper still there, or, perhaps a salt residue.
Cleaning will remove the salt but not the copper.

With the chips out try adding a resistor to ground and/or +5V.    Try 100k, 1M, 10M that type of thing.

If you can see 2.5V with a 10M input impedance meter there could be 10M to +5V (or higher to 9V).

On the other hand it could be leakage from 5V (and/or 9V) and to 0V forming a 2.5V voltage divider.
In this case the impedance of the leakage path could be anything.  That's where loading the line
comes in.

Maybe if you get down to 10k you can start to realize there's a short to another track, perhaps through
an invisible connection of copper on a mis-etched board.

You have to start somewhere to narrow down the problem, even if it's only a stepping stone.  The problem could end-up being something completely different.

Thanks that gives me something else to try. I believe I did try just a bare reading between pin2 and ground and saw something like 5-6 MOhm but I need to check again and be certain.

Measure for AC also. A DMM on DC can try to make something of it. If there is some AC, do something to stop the LFO and ground the signal input.

Will do, thanks.

[Rob Strand]

Thanks that gives me something else to try. I believe I did try just a bare reading between pin2 and ground and saw something like 5-6 MOhm but I need to check again and be certain.

If you still have the issue perhaps you could contact the PCB manufacturer to see if they have had any problems recently.   If you mention the resistivity issue they might know what goes bung in their process from time to time.  You might have to clean the whole board.