I like the placement of a choke there, that at least provides a good feed from positive to ground for recycling purposes...
I've been considering akula's screenshots. 3ns rise time... but with a 10ohm resistor before the gate in the KT805, it's only going to be only just barely on... enough that when Primary conducts, secondary also reacts, and causes a raise in the ground voltage, which closes the KT805 (base-emitter not high enough now).
The pulses that result from such a condition are increasing frequency, and staggered at a 'fixed' distance... it's not really 'ferroresonance' more like 'ferro-feedback' ... but then the pulse from 7414 goes off, so only a couple of these spikes happen.
which means primary feed of secondary must be from VD4,VD3,VD(unlabeled and choke) side to ground ... or from ground to these?
The issue would be, as drawn, all grounds are the same (which in theory they are), but relative placement of the parts is going to matter also... for instance if L2 is tied directly to the emitter of KT805, and then by a junction to another, because there is more space, the junction is slightly lagged... so maybe the mass ground of 1000u's and 100u's scattered all over don't apply as much....?
But the impedance of choke is going to make L2 go to a high voltage against Low MSR3545 and VD4 as capacitances...
So for that to happen, current 1 is to ground, then the high voltage and ground cause a sharp spike through L2 which is seen in the collector of KT805 from L1... partially becuase of ground distrtion partially just the pulse in the coil...
which will pull on unlabled, unparametered choke, which should feed from positive to ground... actually the capcitance of VD3,VD4 causes the coil to revese quickly and conduct from ground with a very high voltage, which is the distortion on the feedback....
if the pulse width was wider it would be more spikes with a decaying time and magnitude
I dunno it's what the pulses look like to me, from playing with lots of inductances and chained transistors....
the transistor must be off for a high voltage to develop, and that's a forward EMF that drives that (continues sucking electrons)... in a diminishing current, which should reverse the other coil also which is already wanting to itself reverse conduct... ... I dunno I'm losing the tought...
One can make similar pulses just adding a small inductance to the emitter to ground.
|