Coil Resonance and Bench Testing.

[Anonymous]

Thought I would have another look at this to make an interesting end to the week. I tried various coils on a signal generator using the method described by BBSailor and some results emerged which will need more detailed study. In general, both the Goldquest and Deepstar coils start off with shielded coil only, at between 600 and 700kHz. Add 2.5m of coaxial cable and the frequency drops to 400

 

[Anonymous]

Hi Eric,
Thanks for the excellent information. Now I need to look very carefully at each stage of my coil building to see just what happens.
Reg

 

[Anonymous]

Eric,
I was thinking and tinkering with what you said about various coils and resonance.
If my conclusions are correct, this may help others who tinker with PI machines.
This analysis is based on the a coil having a constant inductance of 300 uH and demonstrating various resonances, as you reported above, both in and out of the circuit. I calculated an approximate load imposed by your PI Detector of about the equivalent of 250 pf. But this may not all be capacitance as it may be the combined effect of the MOSFET and other parasitic circiry board components. The MOSFET accounts for most of the the capacitive effects but the series input resistor connected to the back-to-back diodes imposes another effective load, most resistive and a little capacitive. The net result is that the coil sees a lower self resonant frequency. If this were calaculated with several different coils, the equivalent circuit loading would should stay reather constant.
What does all this mean? If you have a coil, shield and wire with a combined capacitance much higher than 250pf (or what ever a particular PI machine imposes on a coil) then reducing coil capacitance will have a greater effect on performance than if the coil, shield and cable are at or close to 250pf (for that particular PI machine). It operates like the classic S shaped graph for parallel resistors or inductors. The formula is (R1 X R2) divided by (R1 + R2). For inductors, just substitute L for the R in the formula.
Thanks for stimulating my thinking by providing your coil resonance values.
Another interesting tidbit that you mentioned is that even using a 10 X O'scope probe can impose significant loading on coil resonances measurements because the probe measureably reduces the resonance compared to using another pulsed coil to induce ringing in a coil under test. Once the coil is loaded down by the PI circuit and damping resistor, the scope probe has much less of a loading impact on any measurements; probably less than .5uS on the optimum damping resistance adjustment.
When coils have a much higher combined capacitance than the load imposed by a PI machine, each 100 pf reduction in capacitance improves the potential performance by approximately 1us. If the coil is at or near the equivialent load imposed by the PI machine, coil capacitance reductions have less effect in improving performance.
I have done all these experiments with only a 20MHz O'scope and a 2 MHZ function generator.
Thanks for stimulating my experimentation.
bbsailor