Roswell Bob
Solar Enthusiast
Sorry I haven't posted any news on the osi design. Looks like we will be doing a pc board for the inverter. The inverter will be very compact.
The inverter may best laid out with a beefy 4 layer board. or a pair of 2 Layer boards. The inverter has 4 serious wires going into it. Best is to put 4 planes on a 4 layer board, or.... If anyone is familiar with copper plating then please step up. The cheepest way to do it may be some DIY copper plating. The output transistors can be any style. I think the TO-247 package lend's itself well to IGBTs Mosfets or Sic or maybe even possibly Gan. What is the state of the art is the question to be asked. I suppose the the board will be good for at least 240vac so run whatever you want to. If you want to drive a LF transformer than that is the way to go. Inverter will probably end up on a small module sans heatsink. So basically a half inch thick piece of aluminum. It's called a spreader plate. End user could determine how many of these inverter modules (S)he wanted to put together. They can be paralleled and put in series too. That would best be done at the output on the other side of the EMI filters. Just takes more dc-dc converters to power them. In the case of the LF design just take more iron and copper.
I have a design for a resonant push-pull converter that will run somewhere between 300kHz and 1MHz. I have a LTspice file if anyone wants the file. Still playing with it. Converter will be capable of three distinct transformer designs. Output voltages are strictly on the transformer design - which is easily changed. Each operating voltage / power level would require selecting the proper switches and a quick transformer assemble. These modules can also be paralleled and put in series. 240v split phase is easy. Same deal as inverter - the chassis is a block of aluminum.
I'm sorry that my webpage doesn't have all the files up there yet. Maybe tomorrow.
The inverter may best laid out with a beefy 4 layer board. or a pair of 2 Layer boards. The inverter has 4 serious wires going into it. Best is to put 4 planes on a 4 layer board, or.... If anyone is familiar with copper plating then please step up. The cheepest way to do it may be some DIY copper plating. The output transistors can be any style. I think the TO-247 package lend's itself well to IGBTs Mosfets or Sic or maybe even possibly Gan. What is the state of the art is the question to be asked. I suppose the the board will be good for at least 240vac so run whatever you want to. If you want to drive a LF transformer than that is the way to go. Inverter will probably end up on a small module sans heatsink. So basically a half inch thick piece of aluminum. It's called a spreader plate. End user could determine how many of these inverter modules (S)he wanted to put together. They can be paralleled and put in series too. That would best be done at the output on the other side of the EMI filters. Just takes more dc-dc converters to power them. In the case of the LF design just take more iron and copper.
I have a design for a resonant push-pull converter that will run somewhere between 300kHz and 1MHz. I have a LTspice file if anyone wants the file. Still playing with it. Converter will be capable of three distinct transformer designs. Output voltages are strictly on the transformer design - which is easily changed. Each operating voltage / power level would require selecting the proper switches and a quick transformer assemble. These modules can also be paralleled and put in series. 240v split phase is easy. Same deal as inverter - the chassis is a block of aluminum.
I'm sorry that my webpage doesn't have all the files up there yet. Maybe tomorrow.