This was an exciting test. We took water to boil 0.7 litres. There was absolutely NO evident discharge of energy from the supply. And if you look closely you'll see that during the 'on' time of the duty cycle the voltage across the shunt was only ever 'fractionally' above zero - indicating that the current flow from the battery during the on time was next to nothing.
The following copied over from the forum
Guys What followed on from here is a series of shots to show the results from a minor adjustment to the offset. The first screen shot shows the limit of that adjustment. There was a second adjustment during the test procedure to re-adjust the offset to return the positive back to this position. I'll down load a few of these because I actually took 15 downloads showing the same thing, essentially, with an ever but slow increase in the rise of temperature. It rose from 66.9 degrees C to it's final temperature of 240 degrees C. That's when I put it in water. And then it took the water up to 80 degrees centigrade where it pretty well stabilised. I'll post in the time it took when I've checked the time on those downloads.
Also, NOTA BENE guys, the voltage across the batteries that I mentioned in my earlier post is wrong. But just note that the battery voltage both climbs and falls - on these slower frequencies. This is most clearly evident at the very slowest fequency which is when we get that delicious oscillation that just goes on forever.
Also. I've not posted all the downloads - 15 in all - as they're too repetitive. I downloaded a screen shot at certain intervals just to relate it to temperature rise. I think I'll post three as that should be fairly representative. So. The first is to show the 'offset' detail. And two others - at the beginning and near the end - when I had to immerse it in water. Again. The water temperature then stabilsed after an hour or so - at a little over 80 degrees centigrade.
ALSO Please note. The actual level of oscillation across the CSR reduced at those higher wattage levels. But it did nothing to stop the temperature rise. The offset was pretty well stable but I think it may have ended up fractionally 'higher' than the start of these tests.