My PWM3F & Cone Plate Stack idea

[Bob Boyce]

[ElectroGasMan]

[nickle989]

[ElectroGasMan]

[Bob Boyce]

[ElectroGasMan]

[slash1]

[ElectroGasMan]

[chemelec]

[Bob Boyce]

Gary is correct.

One of the first attempts by another to replicate my results, without considering my complete imformation, was through the use of a divider network to produce a single frequency. This resulted in the lack of ability to tune the unit to a specific frequency as required. Instead, it offered a single channel list of fixed frquencies to choose from. I tried telling people it would not work, but people ignored me and wasted tons of money anyways. If it had been adjustable, then 3 of them may have worked, but "channelized" operation is not going to work.

My microcontroller based system (HexController) uses a crystal or ceramic resonator for the 20 Mhz clock, then all references to timing are based upon that. This produces very accurate timing results.

Bob

[ElectroGasMan]

[ElectroGasMan]

[slash1]

[Bob Boyce]

I'm sorry, I didn't catch the frequency question before. The primary target is near 43 Khz, however depending on actual plate size and spacing, this can vary as much as several Khz. My early transformers were laminated iron, which were much less efficient at these frequencies. The addition of harmonically related lower frequencies does add to the overall effect, much more so than mixing 2 or more signals of the same frequency slightly out of phase. The complex blend of rich harmonics yields so much more energy to harvest.

When working out what you need for a core, you have to consider the effect of a steady state DC current. A typical RF core will saturate or come too close to saturation under these conditions.

Bob

[coffeyw]