Befuddled by voltage inversion for positive ground pedals

[pb3000]

Hi, I've been trying on and off for a while to power my pnp Fuzzface and Rangemaster pedal builds off a power supply using 1044/7660 charge pumps to invert the voltage.  I was never successful, but I think I'm close and have a few questions.  I have both a cheap, non-isolated power supply, and my own adaptation of RG's Spyder power supply with 8 isolated outputs - it works great.

I can successfully build up a circuit powered off either of my power supplies with a 1044 charge pump that inverts the voltage.  when using the cheap non-isolated PSU, this only works if the positive ground effect is the other one there.  when I plug in a negative ground effect to the that PSU, the positive ground effect dies. 

When I use my isolated ground PSU, everything works fine, I can combine the positive ground pedals powered by the inverter circuit with other negative ground pedals powered by the same PSU.

Any ideas why this is, or what else I can test to figure out what's the matter?

thanks

[Aph]

Are you using the standard power jack in each pedal (center pin negative)?
And, are you using the inverter like this:

[pb3000]

Thanks for the reply.  answers:

1 - I am using the "standard" power jack in the negative ground pedals, but on my positive ground pedals, I do not have power jacks.  I built them with batter clips and I am alligator clipping the -9V output off the charge pump (sitting on a breadboard) to the negative side of the battery clip and the ground to the positive side of the battery clip.

2 - yes that's exactly how I'm using the charge pump.

[Aph]

Is the "cheap" non-isolated supply able to supply enough current/voltage to the pedals? Have you tried voltage measurements when everything is plugged in (input and output of charge pump, voltage to the circuit, etc.)?

[Rob Strand]

You need to draw a diagram showing the main elements such as power wires, ground connections and jack switches for a positive ground effect and a negative ground effect .  Then join the two drawings to represent how you are connecting the two together physically.

When you do that you will see something wrong.   If you don't maybe check your drawing does represent what you are *actually* doing.  Maybe match each wire on the board with the drawing and tick it off when you check it.

Another approach is to wire it all-up without the power present and buzz-out the connections with your continuity tester.

It takes some perseverance but at some point the problem will just pop-up and smack you in the forehead.