PWM LED THROUGH MOSFET

[collatz]

I have a boost converter circuit which takes an input voltage from as low as 1.8 to as much as 5.5 and outputs 6 volts at 10 amps.

The output of the boost converter goes to a P channel mosfet, part number AON7423. What I want to be able to do is PWM this 6v, 10 amp max output by 0.1% duty cycle increments.

For the output capacitors on the boost converter I'm currently using rather expensive ceramics, about 330uF, with a recent additional aluminum capacitor of 1000uF. Circuit has to fit within a 0.8 square inch area. The majority of my troubles so far have been size related.

So in these three pictures I'm running a 5.0% duty cycle at around 1000hz. The turn on time of the mosfet appears to be almost instantaneous. The turn off time seems pretty fast but then kind of decays more gradually near the end. I don't understand why??

When the boost converter is enabled... The turn on time still is fast enough for my purposes (looks instantaneous!) but after turn off it goes super into some kind of crazy oscillation (*** this oscillatory behavior fades away before the next turn on event of the mosfet. Any ideas??

My MOSFET is driven by a bipolar transistor which brings the gate to ground. (So I think I understand why there is a fast turn on), I am not using a push pull type driver, could this be the cause?

Thank you so much for any suggestion, help, criticism, etc! I'll really appreciate it.

 

[Arouse1973]

Hello and welcome to E.P. Can I see the circuit diagram please.
Thanks
Adam

 

[collatz]

Hi Adam, hope this helps

 

[Arouse1973]

Ok one thing you could try is a Baker clamp, this prevents the BJT transistor from saturating fully and will be faster to turn off. This will only work if it's the BJT causing the issue, but it's worth a try.
Adam

 

[BobK]

I don't think it is the BJT that is causing the problem. It is the 100Ω resistance through which the gate must discharge. You need a two transistor driver (high, low side) so that there is less resistance for turn on and turn off.

Bob

 

[Alec_t]

I am not using a push pull type driver, could this be the cause?

Very likely. I agree with BobK.

 

[collatz]

Thanks everyone for the help! I've spent the last couple days considering these things. Today I built a push pull driver with a 12 volt input voltage to slam the gate capacitance. The mosfet now opens and closes very fastand clean on the oscilloscope. I measured my efficiency at 75%, which is pretty good. I was hoping for 90%... Maybe my board size constraints make this impossible? When I look for lower rdson or lower inductor resistance etc... The parts start growing in size fast. My test today was at 3.5 volts in,,, 6 volts - 8 amp out , best I've seen so far on this circuit. temperature rose and stabilized to 190 degrees Fahrenheit. =)