What indication of scalability/stackability did you find in the paper?

I think most inverters use some MOS transistors. These things are made with ridilously high amperages and small power losses at reasonable prices these days.

The most power loss comes when you switch the transistors on or off.

A simple square wave inverter is usually cheaper. I think because you only need to switch on and off 2×50 (2×60 in US) times per second. So you can get by on smaller transistors and less cooling.

A sine wave needs to be “chopped” or PWM:ed to look like a sine wave so it needs more on-and-off per second to approximate a sine wave.

An arduino or similar would probably handle the controller circuit in a versatile way (why not make one for both Europe an US for a change .

Good news might be that only sensitive low power units, computers, tv’s etc. need sine wave.

Another thing that brings the price up, I think, is the transformer. You need to go up in voltage, generally because these things are designed to run on 24 or 12V car-systems.

Maybe the batteries could be connected in series to form the correct voltage 12×20=240v (12×10=120V for US)?

This would of course mean you have letal volatages on both sides of the inverter!!!

For sinewave you need slightly more since the peek voltage of the grid is larger than the effective voltage

This might be cheaper since you don’t need upwards conversion. Also you get to work with 10-20 times less the current, giving you cheaper transistors. (It is much easier to make a transistor withstand high voltage than high current)

For the welding, I just found there are tons of suicidal people out there using the car battery directly as a welder, it also happened with me once, it made my wrench into a permanent 1/2″ *grin*

In the long run, that might be a way to go, make the welder without the transformer? Don’t really know a lot about welders only that they use a transformer to produce high currents. We already got that in the batteries!

Here it might be best to connect batteries in parallel, thus gaining huge amperages. But since they are often unevenly charged, I’d suggest a big diode on each cell and perhaps individual current limiters. Individual current limiters could also be used to regulate the current to the weld.

We are looking for a schematic of an inverter, scalable in units of 1 kW. We want to be able to build 1 kW or 10 kW inverters. If you have any insight on this – or friends that do – let us know. Right now we’re dependent on either throwaway or very expensive inverters. We want to build a 10 kW inverter to power welders. We have the battery bank to support that. Off-the-shelf, it is $10k for 10 kW inverters.

But I’ve also wondered about inverters – and I just wanted to give you a heads-up – I googled “power inverter design” and the very first page result looks like it has some basic schematics linked to it – I think power inverter design may be already a decent way down the open-source road.

Best of luck!

- Mike