eco, if the unit was left running at full output, it put out way too much hydroxy gas for engine idle. I had to cycle it on and off in order to reduce average production on those early resonance units. I had no way if dialing in production on the fly.
Through the more recent research, I have solved much of the resonance stability issues of those earlier research years. You are already working in the region that allows this to happen. Highly efficient production without the high waste heat, that is the whole key to stability. The higher waste heat production of my early units, while still much lower than other electrolysers, was still too high for good production stability. My newer generator/driver design also allows tuning of output levels electronically. The PWM3E does not automate this function as it is a manually tuned generator/driver, but it does allow these anomolies to be explored manually.
I don't see where spinning a contact point to chop the voltage would really gain anything, because unlike PWM, as the frequency increases, pulse width decreases, so the average on time v off time stays the same.
Even varying voltage a bit does not have the same effect at resonance as it does at brute force electrolysis. If the electron supply drops below a critical level, then gas production will suddenly decrease as hydrogen ions become attracted to the more negative oxygen atoms and reform into H2O molecules prior to being expelled from solution. Sort of like a dog chasing its tail scenario, spinning in circles but not providing usable results
Look at the picture of the toroid transformer on my project pages and then compare that to the output stage of the PWM3E and you may see the corrolation.
Bob