i comment but i am so far from a real EE its rediculous , so take it with a grain of salt, or a whole sea of salt
just things i learnt the hard way
as long as mosfets (in general) go from full open, to full closed they handle a lot of power, its when they are Partly closed or open that they become resistive items heat up and could die immediatly in this shown ammount of power.
if you were really ever going to run 400full watts switched through it, then 480W possible mosfet isnt giving you much leeway, EVEN if it is snapping quickly from full on to full off. if you expect better longevity and handling of switch burping (manuel or switch flaws) you should have a handling capacity of more than 2x that , or at least some good heat removal should a burst hit the mosfet.
The Upper Resister:
to trigger the mosfet gate , you need very tiny ammounts of "current" hence the use of the resisters on the gate, to avoid frying the gate.
to FULLY trigger the gate it will need the full trigger voltage of the gate (at very low currents) if it does not get the full voltage of the gate trigger, then it will be part open (see above). Note: full voltage FOR the gate (prob about 4v), not full voltage of 36v battery.
The lower resister:
once the gate thing closes the mosfet, via hitting it with the correct voltage, it (sorta) stays locked closed, even when your switch here is off, it takes Very little to open it again, hence the second resister, the other resister pulls the gate back down (in voltage) and stops the flow through the mosfet.
the resister RATIO will set the voltage of the "voltage divider" that the pair of resisters are, setting the voltage to the gate.
got it?
with such (conciderably) small pull-up pull-down resistances as shown you should get a very fast SNAP to gate triggering, both up and down. potentially a lighter version of that curcuit would use much higher ammount of resistances for the pulldown resistance, as only the smallest ammount of Drain off the gate is actually needed.
This very fast snap back down, should assist in keeping the mosfet fully opened or fully closed, and not in-between which would be bad.
plus you could overvoltage the gate (i think around 24v), so the ratio would be important.
sooo
Although your going to get some power consumption (resistive short) between the poles of the battery, and maby more than is nessisary, you will get a good fast switching, which you need to keep the mosfet from handling resistive power.
Most of the power consumption your worried about would be minimal (comparative to total on load), and only when it was switched on.
As far as the current limiting resister to keep bulb pop from occuring, somehow it seems like a waste to me. although it can "bend" a bit when the filiment is cold, and reduce total current flow to achieve a starting, its going to be the biggest waste of power to the whole curcuit, and will only provide a bit of soft starting, and a lot of wasting.
higher level of experts would trigger the mosfet in sharp (fast) pulses (PWM) for the soft starting, once the mosfet was an integral part of the curcuit, of course complicating the whole thing by a lot.
another thing, is a FUSE, a fuse will have some resistance itself as they must to heat up and break, so some of your inrush current control resistance could come from a fuse, which could also assist in the case of other issues, sort of kill 2 birds with one stone with a fuse there instead, or Also.
You may need to selectivly add resistance in ANYWAY just to get to the desired power to keep from popping your bulb over time,but myself i would more likly to drop out a cell of the battery, instead of adding in resistive inneficency.
to calculate (easily) your losses through a "inrush" resistance put in, just check the voltage drop there relative to the total voltage, it will give you a quick and dirty idea of its losses.
so all you need to do is "test" what you end up with, by using a smaller wattage of 36V bulb thing (or 3x12v) then switch your switch on, and test the heat hitting the mosfet, and/or checking the voltage drop from source to drain (on the mosfet). Checking the voltage drop on the mosfets power switching area BEFORE putting the full amp load on will tell you if the triggering of the gate is sufficient both on and off. that would be a good test prior to putting on the full load, using the same voltage but less amps for testing.
if you observed to much drop through the mosfet still, or to much heat hitting the mosfet while on or off, then you would adjust the pull-up pull-down resistances, prior to putting your full load on and blowing the mosfet.
so with that, you can test with 3x12V festoon (dome) bulbs or something, and check for yourself and adjust if nessisary.
Notes: if this had not already been demonstrated as working, i would try putting much higher resistances on the gate, just seems to low more like 50K is what i would have attempted, specially with such a high inital voltage on the battery. i was just pointing out the positive ramifications of doing that.
