Tweet
Great Results
Glen!
Fantastic - really good work there. Great record keeping to boot
This is exactly what I was expecting on the Drain and Source, I can't express my thanks enough for what you've shown. And that negative spike perfectly aligned with the gate falling edge
, so glad you moved the marker over on those shots so we could really see it's full impact. That one is the capacitive gate-source feed through that occurs 1.3 divisions before the inductive collapse. Very interesting how well that passes through there. But the Drain and Source activity after the FET is off, that is the really interesting thing.
As the magnetic field collapses around the load resistor the Drain spikes to what, 220V? That is a positive going voltage. At the exact same time this is happening the source is going negative. Where does that negative energy come from? Could it be the 220V positive spike punching through the battery and sneaking up from the backside? Nope, wrong polarity. Could it be inductive energy punching through the capacitance in the FET? Nope, again, wrong polarity. Could it be the timer chip sucking it down below ground somehow through the gate-source capacitance? Nope, the gate is already low and the pulse is too wide. So where then? Perhaps it is the 'shunt' resistor acting like an inductor...
well, yes it does, a little, but this is more energy than that so something is adding to it. The load resistor is inductively coupled to the 'shunt' leads, and possibly the resistor itself. This means that when the field collapses it induces voltage in that 'shunt' the same as it does in the load resistor. But why is the potential inverted? Because of where we are looking at it. The energy is the same and the current is flowing through each in the same direction but we are looking at different ends of each device. And as a side note, none of the current is flowing through the FET right there during that part of the pulse - that might make Kirchhoff squirm a bit - but that's how it is. Each inductance is now acting as a current source. The return pulse is a little different, the drain doesn't get to see it drop below zero - instead, it is clamped by the body diode which promptly converts 220V of negative energy to about 0.7V real quick. Where does the energy go? Surely we should see a negative current flowing in the 'shunt' here when this is happening, right?. That would show up as a negative voltage across the 'shunt'...but it is not negative here ... instead, what we see is a reduction of the BEMF that would be present on the Source if the Drain was at a higher potential. But since the Drain is at a low potential, the positive going Source cancels it's energy across the diode. So all that is wasted on the body diode. But then the diode stops conducting and we see an inductively coupled ringing both at the source and the drain until all the energy present is dissipated in the load resistor - good stuff, and good heat BTW, 48°C over ambient and all with a 22% duty cycle. IMHO, aperiodic operation was probably above that in the original circuit.
My anniversary today, I need to get some sleep and I want to spend time with my Wife. But when I return I will DL the CSV's and see how things look. Now if I could just figure out how to convert degrees to watts....
Keep up the good work, and many thanks to you, Aaron and Lisa for getting us this far.
@Groundloop,
That self osc. looks great, was thinking of doing that myself - even tried some feeble attempts at feedback on my modified circuit. I wonder if there is a way to sharpen those edges? The better the dV/dt the stronger the magnetic field
Cheers,
Glen!
Fantastic - really good work there. Great record keeping to boot
This is exactly what I was expecting on the Drain and Source, I can't express my thanks enough for what you've shown. And that negative spike perfectly aligned with the gate falling edge
, so glad you moved the marker over on those shots so we could really see it's full impact. That one is the capacitive gate-source feed through that occurs 1.3 divisions before the inductive collapse. Very interesting how well that passes through there. But the Drain and Source activity after the FET is off, that is the really interesting thing.As the magnetic field collapses around the load resistor the Drain spikes to what, 220V? That is a positive going voltage. At the exact same time this is happening the source is going negative. Where does that negative energy come from? Could it be the 220V positive spike punching through the battery and sneaking up from the backside? Nope, wrong polarity. Could it be inductive energy punching through the capacitance in the FET? Nope, again, wrong polarity. Could it be the timer chip sucking it down below ground somehow through the gate-source capacitance? Nope, the gate is already low and the pulse is too wide. So where then? Perhaps it is the 'shunt' resistor acting like an inductor...
well, yes it does, a little, but this is more energy than that so something is adding to it. The load resistor is inductively coupled to the 'shunt' leads, and possibly the resistor itself. This means that when the field collapses it induces voltage in that 'shunt' the same as it does in the load resistor. But why is the potential inverted? Because of where we are looking at it. The energy is the same and the current is flowing through each in the same direction but we are looking at different ends of each device. And as a side note, none of the current is flowing through the FET right there during that part of the pulse - that might make Kirchhoff squirm a bit - but that's how it is. Each inductance is now acting as a current source. The return pulse is a little different, the drain doesn't get to see it drop below zero - instead, it is clamped by the body diode which promptly converts 220V of negative energy to about 0.7V real quick. Where does the energy go? Surely we should see a negative current flowing in the 'shunt' here when this is happening, right?. That would show up as a negative voltage across the 'shunt'...but it is not negative here ... instead, what we see is a reduction of the BEMF that would be present on the Source if the Drain was at a higher potential. But since the Drain is at a low potential, the positive going Source cancels it's energy across the diode. So all that is wasted on the body diode. But then the diode stops conducting and we see an inductively coupled ringing both at the source and the drain until all the energy present is dissipated in the load resistor - good stuff, and good heat BTW, 48°C over ambient and all with a 22% duty cycle. IMHO, aperiodic operation was probably above that in the original circuit.My anniversary today, I need to get some sleep and I want to spend time with my Wife. But when I return I will DL the CSV's and see how things look. Now if I could just figure out how to convert degrees to watts....
Keep up the good work, and many thanks to you, Aaron and Lisa for getting us this far.
@Groundloop,
That self osc. looks great, was thinking of doing that myself - even tried some feeble attempts at feedback on my modified circuit. I wonder if there is a way to sharpen those edges? The better the dV/dt the stronger the magnetic field
Cheers,
Alex thanks for the advice Bro, thats easy to try, winding Glens inductor now. ill get onto it ASAP my friend.I have a contact at the university, with some equipment will try and match Aaron's and Glens test for Rose/ALL.
Comment