We have all at some point have made a flashlight. It used to be a staple of childhood electronics, the screw-in bulb in a holder, and a cycle lamp battery. If you were a particularly accomplished youthful hacker you might even have fitted a proper switch, otherwise, you probably made do with a bent paperclip and a drawing pin.
So you might think that flashlights offer no challenges, after all, how many ways can you connect a bulb or an LED to a battery? [Peter Fröhlich] though has a project that should put those thoughts out of your mind. It uses a power LED driven by a TI TPS61165 boost driver, with an ATTiny44 microcontroller providing control, battery sensing, and button interface. The result is a dimmable flashlight in a 3D printed case housing both control circuitry and a single 18650 cell which he sourced from a dead laptop. Suddenly that bent paperclip doesn’t cut it anymore.
The result is a flashlight that is the equal of any commercial offering, and quite possibly better than most of them. You can build one yourself, given that he’s published the physical files necessary, but probably because this is a work in progress there are as yet no software files.
We’ve featured a lot of flashlights over the years, but it’s fair to say they usually tend towards the more powerful. Back in 2015 we published a round-up of flashlight projects if it’s a subject that captures your interest.
In case someone missed it:
https://hackaday.io/project/11527-lampounette
Hey, that looks like my proof of concept :)
http://i.imgur.com/3PA32PP.jpg
It has a Murphy friendly switch. No fun having a dead light, when you wanted it charged up and ready. Battery operated devices should be designed so as to prevent Murphy from meddling with operation.
Yes, this is the first and rough prototype, the switch design is just cobbled together to have something physical to play with.
It will turn on in my pocket, turn on on its own when bumped and coughed at in the general direction.
This will be addressed in the next design.
Recessed power switch, high activation force switches and either a multiple or long press klick will be necessary to turn it on, to be on the safe side.
I really dislike that so many LED flashlights can be turned on (i.e. discharged) while being carried in a pocket/backpack/etc.
So many times i took the keychain light out of my pocket just to notice it is empty.
Because for the umpteenth time it turned on, accidentally.
No worries, this issue will be addressed :)
B^)
You just need a bigger keychain light. When this happened to me, I suddenly wondered what stings (gets hot) in my pocket, pulled the light out and switched it off again. :-) With an AA LiIon it has some power, around 5W, with a 1,5V AA it works but is quite weak. Luckily the switch is not too touchy, so the above happened only once.
I feel like maybe they should have a little neon telltale or something to let you know when they are on… ;-)
I made a 555 LED strobe/flashlight in an Altoids tin many years ago. It served me well until about a year and a half ago, after suffering one too many drops. I still have plans brewing as to how I can make a better version, but thumbwheel potentiometers with switches don’t seem to be common anymore. A pity, as I would much rather not have to include a microcontroller to read a rotary encoder.
Less common than I would have thought, but out there: http://www.mouser.com/Alpha-Taiwan/Potentiometers/RV121F-Series/_/N-1z0zlgnZ9q0ypZ1yzuzcq
I’d check thrift shops and rummage sales for an old transistor radio to scrounge one from.
What I’d like to see, is a headlight that automatically switches off when you step inside and can turn itself back on when you walk back into the darkness.
I made such a lamp nearly 10 years ago, when bright LED lights weren’t so common: https://stuartl.longlandclan.id.au/blog/2007/09/05/hatlamp/
The automatic switching used a NXP 74HC14 inverting Schmitt trigger with a LDR mounted on top of the hard hat. Yes I drilled holes in it, so no it won’t do on a construction site. But that wasn’t its purpose. One gate controlled a reed relay (v1) then later a IRF9540N (v2). Another gate made half the rim LEDs blink, a third gate inverted the output of that previous gate and made the other rim LEDs blink. The remaining 3 gates were unused.
Commercial headlights have well and truly surpassed this thing for brightness, the “high beam” (1W Luxeon star module) on this thing doesn’t put up much of a fight against a $2 cheap headlight these days… but I found the automatic aspect of this headlight useful, and miss that feature with the brighter ones.
I had one of those automatically dimming headlamps from W-mart. Worked great! If you put something up close to your face it dims. A friend dropped it into the pond…
I’ll post the model # later if I can find it online.
It was a cheap Walmart Ozark Trail Auto-Dimming Headlamp 200-Lumens. The spotlight dims to minimum when the sensor detects enough light. Great ‘throwaway’ headlamp.
I did something similar but with a pic mcu, a 230V 3W led spotlight with the 230V bits removed and using the mcu to boost 3x aaa size batteties to 13ish volts. Also dimmable with a touch pad. The beam is very wide which is very handy if you want to light a large area.
Funny how it has become hip to mix english and german words. “BRIGHTENMACHER” translates to something like “brighter maker”. However if you instead germanize the first word and anglicize the last, it would be “HELLMAKER” a much more metal name!
So… here is the thing about that, i wanted a name that i can track.
There were a few other choices, one was the one you mentioned.
You are right, it sounds way better, but… only Brightenmacher returned 0 results in google when i tried :)
I want a flashlight that strobes and/or beeps for a few dozen seconds when dropped while walking in the woods. ;)
Sucks not having light on the way back because the ‘backup’ flashlight is no good.
Oddly enough, that was one feature that got axed along the way, it started of with accelerometer and drop protection was one of the reasons.
The first version had a OLED screen, buzzer and a maxim fuel gauge/protection circuit.
Exact remaining runtime calculation on the fly, depending on the brightness.
Maybe i will get back to that once this is done.