Fet+1 into a Wuben E05 with a 4000k SST20. I sanded down a Mountain Electronics e-switch board and 3D printed a spacer that supports the switch. The switch board is then epoxied to the spacer. Really liking this one. I’ll be building a couple more.
I got a attiny 412 swapped on to one of the convoy t2 boost drivers I have been messing with. Huge thanks to gchart for sharing his clicky firmware. I tweaked it to pwm on pin PA2 to match the layout of the convoy driver and I increased the PWM frequency as that allows for a lower low mode on these drivers. The OTC is across pins 7 and 8.
After the pic was taken I put a jumper across the 0.2 ohm resistor that sits between the fet and led -.
With the resistor jumper in place I took current measurements at the LED (nichia 319) and got 1.8 mA on low and 1.01 A on high. I used a cheap meter with standard leads so actual current in normal use is probably slightly higher. Freshly charged energizer NiMH.
Welcome to the AVR 1-Series club. Glad to have a another member on board. May I ask, which programmer did you use? And let me know if you think there is anything I should add to my guide thread.
Swapped to a LH351D 5000k 90 cri. Measured at the led with the same crappy setup I get 2 mA on low and 1.2 A on high but drops quickly. I’ll do some torture testing to see if the bypassed resistor becomes a problem but I think it’s excellent output for a single AA boost driver.
I have some suggestions and questions. I’ll be PMing you when I have some more time. Thanks again!
I just did a Wuben E05 with a fet+1, but I’ve been on a real boost driver kick lately. This is a E05 with one of the red convoy T2 boost drivers I have been experimenting with recently. A lot of the basic boost drivers you can buy are based on the PAM 2803 boost controller which lists a quiescent current of 1-3 mA. I measured a couple myself and they were both right at 2 mA. That’s a bit to high for a AA e-switch light IMO. The convoy driver uses a different boost controller and what I discovered is that ~5 seconds after power is applied the parasitic current drops to ~0.1 mA with a 14500 and around .45 mA with a NiMH. That’s not incredibly low but it’s actually lower than some production lights I’ve seen measured, and plenty good enough for me. Fully charged to fully discharged eneloop would take around 6 months if my math is right.
I recycled my filed switch board with 3d printed spacer technique for this one. I used a attiny85 rotated 180 degrees with the PWM trace cut and a jumper to the PWM pin on the attiny85. Switch connections were made directly to the legs on the attiny. I used Narsil, green switch LED’s and a LH351D 5000k 90cri.
You’re “killing” me with these AA lights mods :o
I’am an AA/14500 fan, but there is no single “stock” UI that makes my day yet (well, the Skilhunt M150 V2, eventually), and you getting all these done :o
Long story short, board in photos and 351 were testbed. When emitter turned angry blue on turbo i realized it needs Cu board.
So i took one from SP10S which is perfect fit BTW and after test with 351D i went with 219C.
Warrior Mini has 0.7A on high and around 3A on turbo so big fat 351D beam loose a lot on 0.7A. That’s the reason i choose narrower 219C
An update on my Sofirn SP10S with Anduril2 via an attiny1616 swap. My original board used an airwire because I wasn’t paying attention when I picked the pins and I used and ADC1 pin instead of an ADC0 pin (and coding around it would have been a pain). So… new board with no airwires. I also went with the 0.8mm Oshpark service. It seems like the castellations rip off a bit easier, so you’ve gotta be careful with that.