Nice experiments. Would you like to post your schematic? One picture tell more than 10 000 words.
I suggest you to first try with the coils wounded over cardboard or plastic tube with unequal number of turns (assymetric configuration). Winding ratio could be any for example 1.618:1. Current coil (CW) 1,618 x lenght of Voltage coil (CCW). Put each coil on separate independent plastic tube with sligtly bigger diameter than main spport tube so you could easy tune their position. Also put resistive load across both ends, but not lower than 1K. Put ferrite kolbasa inside support tube. I prefer exciation coil like Floyd Sweet have done, with 3rd additional coil. For the first time put 3rd coil coaxialy to POC coils (on the same axis). Also put 1K resistor between his ends and pump a square wave pulses on it. But be careful. Don't attach signal generator directly on 3rd coil because if you trigger POC effect, due to extreme magnetic resonance, voltage in your excitation coil may rise extremely, and some high flyback pulse could easily burn your signal generator and oscillograph. Attach signal generator output only via RF choke taken from old PC power supply and block one side of choke, that side which going to excitation coil with some quality HV capacitor for example german wima hv polypropylene capactor with voltage ratio minimum 1kV. On that way you will guard your signal generator from the burn failure.
Permeability of ferrite kolbasa should be chosen according to pulse frequency. For HF pulsing use low permeability for LF high permeability ferrites.
Use 1:100 2,5KV HV probes for monitoring the circit. One channel over POC load, other over excitation coil load.
If you have some two U-I cores you can easily wound POC coils over 2 U halves.
Good Luck and Happy Hunting.