In Reply to: Re: single phase inverter posted by Denis O'Brien on 11/05/02 at 6:36 AM:
Well, this isn't easy.
I try to explain the problem:
A MOS-FET is a realy useful unit. The Rds-ON resistance is very low, so the maximum switchable power is most times over 50A.
But only if the Drain-Source-Path is fully swichted "ON".
But the only way to use a FET this way, is to create a "switched power supply", wiche is driven by a "PWM" signal.
But this signal could not be used to create an AC inverter. (Or couldn't be used in an easy way)
If you try to make a real sine wave, the FET flows throuh *iternity* points of "R-DS resistance". The resault is a very high power dissipation.
For example: The Drain-Source resistance sets the voltage over Drain-Source-Path to 10V. The current is 1A for example. So the power is: 10V * 1A = 10W of power dissipation (for this moment).
It's a little bit difficul to calculate the exact sine wave power dissipation on an inductiv load (like a transformer), couz of the phase shift.
The best way to create a "sine-wave" with a minimum of power dissipation on the FET(s), is to use a "trapezium signal". The phase shift of the inductiv load and a possible additional capacitance could create a signal wiche looks nearly like a "sine-wave" - and is realy good to use in AC inverters.
But's very tricky to regulate the possible power changes on the output of inverter. An unregulated inverter is also possible, the voltage drops are not significant on power less than 100W.
If you want to know how to rebuild a sine-wave with a PWM ("swiched power supply") mail me. But's not easy... and i have to paint some diagrams ;-)