The IRF3205 is tested for ON resistance. 1N4007 diodes are added to the FET output to snub any voltage spikes. Common negative wires are added to the center of the board to eliminate possible terminal failure, insure lower resistance and prevent inductive coupling to the voltage sensing circuit. The two green wires are 14ga with silicone insulation. This wire is so nice to work with. Instead of drilling some holes and a tywrap, I soldered a wire jumper to hold the wires in place. A 600F iron does not even make a dent in this wire and they are super flexible. It is best to limit any current path on the board if possible. The FET on resistance was quite good being less than 9 milliohms a 5A. This board will be tested at 40V with a 13A pulse current into a 3 ohm load. My latest 100W panels are 20.5Vpp, much higher than the older 17-18V panels. We will see how well it works with a bunch of used under rated capacitors. Even these should last a couple years. Next will be an actual bench test. Diodes are general purpose 1N4007 or faster SR5100. All diodes are fast turn on. Turn off is slower for these general purpose diodes, still much faster than the deadtime of this inverter. Band on diode goes to the positive terminal. Some metal sheath heating elements which conduct to the base act as a shorting winding of a transformer and dampen spikes. Others are insulated and there is no dampening. This board has 4.7 ohm gate resistors so it will be fast. In my other designs I use a 27 ohm gate resistor to slow turn on and off. This eliminates many spikes from forming. Slow frequency of inverter makes transition heating minimal.
The IRF3205 is tested for ON resistance. 1N4007 diodes are added to the FET output to snub any voltage spikes. Common negative wires are added to the center of the board to eliminate possible terminal failure, insure lower resistance and prevent inductive coupling to the voltage sensing circuit. The two green wires are 14ga with silicone insulation. This wire is so nice to work with. Instead of drilling some holes and a tywrap, I soldered a wire jumper to hold the wires in place. A 600F iron does not even make a dent in this wire and they are super flexible. It is best to limit any current path on the board if possible.
The FET on resistance was quite good being less than 9 milliohms a 5A. This board will be tested at 40V with a 13A pulse current into a 3 ohm load. My latest 100W panels are 20.5Vpp, much higher than the older 17-18V panels. We will see how well it works with a bunch of used under rated capacitors. Even these should last a couple years. Next will be an actual bench test.
Diodes are general purpose 1N4007 or faster SR5100. All diodes are fast turn on. Turn off is slower for these general purpose diodes, still much faster than the deadtime of this inverter. Band on diode goes to the positive terminal. Some metal sheath heating elements which conduct to the base act as a shorting winding of a transformer and dampen spikes. Others are insulated and there is no dampening. This board has 4.7 ohm gate resistors so it will be fast. In my other designs I use a 27 ohm gate resistor to slow turn on and off. This eliminates many spikes from forming. Slow frequency of inverter makes transition heating minimal.