From PFRC Wiki
Jump to: navigation, search
~7/8 in. spark indicates that the system is capable of roughly 6.5 KV

Project status

A new variable driver circuit has been built and is fairly successful. With some minor modifications to the timer module of the circuit, the system is now fully soldered and housed in an attractive mahogany box. (fig. 2)

fig. 2
The new box powered by a modified ATX power supply which is now unfit for computer use

There is little ventilation through the box, but recent tests in which the device was run for a few minutes indicate that no dangerous amounts of heat build up inside. The resistor that is in line with the primary coil is the biggest heat source, but the amount radiated is presumed to be within a safe limit (doesn't burn fingers). A small fan attached to a heatsink on the MOSFET provides sufficient colling, and the transistor is actually cool to the touch after several minutes of operation.

The unit requires a 12V power supply capable of delivering a steady current of four amperes. There is a standard four-pin Molex socket, like those that are found on most standard IDE hard disk drives on the left side of the box (the 5V wire is not connected). This allows for easy connection and removal of various power modules including a standard ATX power supply and a custom adapter that was made to facilitate the connection to a lead-acid battery with large alligator clips. Both forms of power have been tested successfully.

There are two knobs located on the front panel that allow the resistance between pins 6 and 7 and 7 and 8 to be increased individually. The controls are rough, from 1 kilo-ohm to 2 mega-ohms, but with careful adjustment, good efficiency can be achieved. Turning the leftmost knob to the right very far, however, decreases the timer's frequency very rapidly which causes more current to be drawn, tripping the short-circuit protection on the ATX power supply. This has not yet been tested on battery power.

A brand new feature in this latest design is the addition of a switch slightly below the knobs. It provides a much safer and easier way to cut power to the timer module and primary coil than the previous method of removing alligator clips or unplugging the device, which would continue to run if powered by a system with a significant amount of capacitance, such as the previously mentioned ATX power supply.

All modules are still connected with small alligator clips to allow for easy replacement of parts or rewinding of the primary coil.

In the last test, the 3 Ohm resistor was removed, leaving roughly 0.03 Ohms total in the coil curcuit. That's right, it's not a very long circuit. It was powered by a car battery (In a remote location and with the car removed) and performed beautifully for about a minute. Then it ended in a cloud of smoke which was later determined to have come from the MOSFET. Yes, the MOSFET that was rated for 49A. This thing made some awesome gobs of violet plasma.

Driver circuit details

I'd like to buy analog ammeter and/or voltmeter dials to embed into the box. Not only would they look awesome, but it'd be easier to measure the circuit's performance.

The driver circuit has been plotted to the best of my ability in this file: File:Flybackdriver1.pdf

If anyone can think of any improvements to make, please post. Also, if you notice that there are errors, such as the MOSFET wired incorrectly, post that as well.