The 7135 is regulating the current to 350mA or 380mA, and the PWM is being applied to that. The fet is letting the voltage straight through to the LED so the led is seeing a higher pulse current. There’s probably more going on and this doesn’t really answer the question.
… just measured the current also with an old analogue amp meter (this should annihilate potential DMM inaccuracy because of PWM pulsed current), and got the same results. With the FET the LED it is few times brighter and uses a significantly bit less current.
If this is so, why use a less efficient 7135 to get few times less light.
The reason for the difference is that the 7135s are slow to turn on and the FET is very very fast. With the 7135s it takes more time for the voltage to rise above the LED’s threshold voltage where it starts making light. This is why a 7135 driver won’t make any light at all with a PWM value of 4 (& sometimes not at 5, either), but the FET is too bright even with PWM=1.
Perhaps because you want less light? I finally received my programming hardware today and one of the things I want to try is to get a lower moonlight on the FET+1 driver. It’s not always about blinding everyone
I usually only have one true moon mode, somewhere in the range of 1 to 2 lumens and then I have a low mode that is about 50% of a single 7135. I set up all of my lights, with the exception of big throwers with these 2 modes. I like to be able to grab any light and carry it to bed with me and I find that much more than a candle wakes me up to much. I also like the idea of being able to get light for as long as possible with any light that I tend to use with any regularity. You never know when you might be stuck with 1 and need to conserve battery.
Generally I put firmware on and build the light with long led wires and fine tune the modes while I can still get the driver out and then when I am happy I button her up!
Thanks for feedback & explanations everybody. After spending the night with my new light the conclusion is that FET PWM level 2 is too bright for a night mode when my eyes are fully adjusted to night vision. So still 7135 will be used for 1 or two moonlight modes.
Which DMM did you use?
Trying to measure average value of a very short and fast spike(few Amps,10kHz+) isn’t that easy. Most budget/midrage multimeters have 10-50Ohms of series resistance at “mA” range setting,so current won’t never reach even 1Amp on high,and that means measuring moonlight current/lux could be far from reality.
Easiest way to check real drain from battery is to put big low ESR el. cap+few uF MLCC in parallel to driver(wires as short as possible) and then measure current between battery + and driver +. Capacitors would smooth spikes,so current from battery should be almost pure DC of few mA(and this is easy to measure with any DMM).