This is very interesting Ds
When i scan the delay in nS first off before i reach a delay that send all my digital equipment mad, i get the pulse that i also found this always occurs first on a very low phase delay between the pulses, i suspect this pulse is very sharp and goes way beyound even my 250mhz scope can manage, the reason i say this, with my sontroller 3 build i use a minimun etch style, and when i first ran it up i scanned and got my pulse on the scope but also got machine gun clicking from the arc over of the fet to the pcb ground, i was only driving with low Volts 30 i think it was, so the pulse was jumping at least a 1mm gap around the trace, i cut extra space around the fet and it stopped arcing.
When i had the loud explosions, i could not smell Ozone or if i put my finger on the coil could not feel discharges, being an old TV engineer i have a nose for electrical discharges, and tv repairs requires all senses, smell, hearing and good eyesight to diagnose a problem.
Ones ear gets trained to the type of discharge, wether DC AC ect, they all have their own characteristic envelope of sound and the loud explosion sound is best describded by me as an explosion, when i used the 3 coils around the ferrite i would call that a crackle inside the ferrite a different sound slightly.
The question i want to answer is for an explosion to occur, i think it must have been mechanical in nature, but how could this be, something happened faster than the speed of sound and was quiet violent, considering i am also getting laptops and digital chips behaving strange around this timing area i am edging towards a very large very fast EM burst, a large EM burst like this maybe enough to tug at the earths magnetic field and be enough if produced constantly to cause a wash board effect , you know where i am going with this
Once i get my new controller going tests will start, i have the pcb etched, just need to wait for the parts.
I will need to first track why my fets retrigger, crazy really but i have not checked previously, i think it's noise getting to the chip rails despite all the coupling i have, i have not been worried before as it was causing the effect i wanted to study, but if i can stop it then i know the nature of the retrigger, next i need to find a way of triggering my scope by the explosion so i can grab the scope data just before the event occurs.
Remember the explosion requires 2 fet stages, but all each fet does is pull one side of the coil to ground so why 2 stages?, if 1 fet if on then the coil is grounded, and the second fet wouldnt do a thing if it came on, so the only thing i can think of is that my fet takes about 16-25nS to turn and off, this is too long so using 2 stages allows for instance 1 fet to turn off and then allows the other fet to turn on within at least 1nS, maybe even faster if it's the retriggering thats driving it to turn on may even be 10 pS that it needs for the explosion to happen.
This is of course all guess work right now.
Peter
Author
Topic: Controller Version 3 - Hunting for more pulses using pulse train. (Read 8987 times)
