ENERGY AMPLIFICATION

[TinselKoala]

.3ohm(approximately. possibly lower or a bit higher) 12v in.  Even at 50% duty cycle could draw nearly 20A(near 40a with constant 12v).

This is why I had leaned toward the function of the diode and coil freewheeling causing 4A instead of the diode being reversed, and now as it shows, I leaned 'right'.

(snip)
mags

Yes, if the schematic is what he's using and shows the diode in correctly, then it's a matter of the diode's speed I think.

[MarkE]

Probably forest is just pumping up his inductor.  For any given frequency and duty cycle the integral of V*dt across the coil when the MOSFET is on, (red) increases the current through the inductor while the integral of V*dt across the coil when the MOSFET is off, (blue) decreases the current through the inductor.  With a linear inductor, when the two areas are equal, the current stabilizes.

[forest]

Probably forest is just pumping up his inductor.  For any given frequency and duty cycle the integral of V*dt across the coil when the MOSFET is on, (red) increases the current through the inductor while the integral of V*dt across the coil when the MOSFET is off, (blue) decreases the current through the inductor.  With a linear inductor, when the two areas are equal, the current stabilizes.

Thank You MarkE. Yes it seems that I'm pumping my coil with sharp pulses with exactly that circuit, minus diode . I don't use diode currently, I only pointed that WITH diode  placed as that last picture, I got 4Amps from my small battery, while (which is surprising !) it takes continuously only 0.2 or slightly more (0.24A) with this circuit. It is anomaly and I'm now just pretty sure it is due to this excellent mosfet avalanche mode .
Well,if you think about it more , and how many inventors talked about superiority of mechanical switch, I think this avalanche mode works very close to it. Unfortunately I must go now but when I return I will try to get scopeshots, I wish I could have digital scope to take exactly synchronized waves from driver output and from circuit after the coil, which will prove me wrong or right. All I can do now is to turn on some ugle "A+B channel" mode on my 10Mhz analog 2 channels scope, because displaying alternate channels do not guarantee synchronization. Maybe it's all just my mistake and mosfet has much more resistance then stated in datasheet (maybe is damaged right now) , because what is astonishing is the minimal current consumption. I computed that in best mode it should take 0.5Amps at 15us time due to L/R constant.


Some detailed explanations : 15us, because I use own made circuit, which was used for other purpose and which is Attiny45 based board with 16Mhz crystal oscillator. So I programmed it as CTC mode square wave 50% duty cycle and I have to set timing. It is set to 239 cycles of 62.5ns or roughly 15us/50% duty cycle, very stable output . I don't know but this stable output may also help a bit .

[MarkE]

Thank You MarkE. Yes it seems that I'm pumping my coil with sharp pulses with exactly that circuit, minus diode . I don't use diode currently, I only pointed that WITH diode  placed as that last picture, I got 4Amps from my small battery, while (which is surprising !) it takes continuously only 0.2 or slightly more (0.24A) with this circuit. It is anomaly and I'm now just pretty sure it is due to this excellent mosfet avalanche mode .
Well,if you think about it more , and how many inventors talked about superiority of mechanical switch, I think this avalanche mode works very close to it. Unfortunately I must go now but when I return I will try to get scopeshots, I wish I could have digital scope to take exactly synchronized waves from driver output and from circuit after the coil, which will prove me wrong or right. All I can do now is to turn on some ugle "A+B channel" mode on my 10Mhz analog 2 channels scope, because displaying alternate channels do not guarantee synchronization. Maybe it's all just my mistake and mosfet has much more resistance then stated in datasheet (maybe is damaged right now) , because what is astonishing is the minimal current consumption. I computed that in best mode it should take 0.5Amps at 15us time due to L/R constant.


Some detailed explanations : 15us, because I use own made circuit, which was used for other purpose and which is Attiny45 based board with 16Mhz crystal oscillator. So I programmed it as CTC mode square wave 50% duty cycle and I have to set timing. It is set to 239 cycles of 62.5ns or roughly 15us/50% duty cycle, very stable output . I don't know but this stable output may also help a bit .

It's been awhile since I had to wrestle with analog scopes.  If you trigger on the output you are using for the gate drive, then whether you use chopped or alternate sweep you should get clean results for a constant frequency drive.  Don't use separate A and B triggers.

When the recirculation diode is in place the current decay cycle to cycle is less and it takes more current and I*R loss to make the blue area under the curve match the red area above the curve.   As it is the circuit doesn't do anything other than heat up parts in the circuit.  If the coil were acting as the armature magnet in a relay or the winding in a solenoid, then mechanical work would get done, the current attack rate would be slower and/or the decay rate would be faster and you wouldn't have such a severe difference in power consumption with and without the diode.  The kind of diode also matters.  A Schottky or at least ultrafast makes a big difference above a couple of kHz compared to an ordinary 1N400x diode.

[zdiblo]

Please keep focused on the topic ....
It´s simpler ....

Zd