Learned a bit more, changed some things, starting to get a good feeling about the whole thing.
The “budget light” seems to be out the window, 25 bucks worth of parts it should be for the PCBs, the LED, metal PCBs and parts, excluding the battery, for a single unit.
The circuit looks like this now, swapped some pins, removed the fancy LEDs and put some regular ones on it.
I would have used a RGB led but i only have seperate ones:
The program memory of the AtTiny44 is too small to get all the code in there to drive them properly.
Before i etch new boards i want to wait till the new LEDs have arrived, that should be tomorrow.
Since i only have one battery holder and want to make two test units i have to figure out how to hold the thing together, good excuse to get more comfortable with my new CAD program.
For years and years i only used Sketchup but this was too much for Sketchup, so i had to learn to use Fusion 360.
Slowly it is no longer a pain in the neck to use it…
Took a bit of time to measure and put all the components into it, lots of time spend with my calipers.
But it should make it simpler to figure the rest out.
Not happy with this at all, i think i am gonna start over.
Not pictured are the covers for the whole thing but i learned a bunch of new stuff so this exercise was not in vain, getting more comfortable with fusion 360.
Meanwhile, i got a bunch of LEDs and lenses. Had to solder one of the emitters myself, it was surprisingly easy, another thing to cross of the list.
They are 2700K and 5000K with 169 and 197 lumen at 250mA current per LED.
The lenses are most likely not very suitable for the LEDs but those were the only small ones i could get from the same source as the LEDs and they did not cost too much money.
The next few minutes will be spend in a dark room comparing the LEDs at different current levels to finally settle on a dual/single led design.
Yea… actually testing things out works way better in the end. Playing with virtual numbers only goes so far.
When compared against each other there is no real difference.
One LED at 250mA runs about as bright as two in series at 130mA to my eye, of course, when measured there will be a difference.
But in the end, the eye is the benchmark.
Even two in series at 250mA do not produce not that much more light, to justify the additional cost and effort.
The only real difference that would make things worth it, would be four in series at 250mA but that is way too much heat.
When compared to my Led Lenser P7 at nominal brightness, i am more then happy with the amount of light coming out of one of these Nichia LEDs at 250mA.
Just swapped out the sense resistor on my prototype and it runs a single LED at 250mA without complaining.
Another funny thing i learned, i would like to put one of each color temp in the light and run them in series, one 2700K and one 5000K together make a very nice light.
For you all that does not mean that much but i am pleased to finally be able to nail down one detail, from here on in, the two choices are:
169 Lumen at 2700K or 197 lumen at 5000K as per datasheet of the LEDs, not measured numbers.
I mean, i knew that already but it is interesting to see that two times the power is not that much brighter and only four times the power produces a significant difference.
10x10mm rectangular Ledil 54° i can not find the manufacturer website for
The Ledil looks the best but is 1mm too tall to fit on the LED.
Both Carclo would fit well on a PCB with the right footprint and do produce a noticeable focus that is nowere near as agressive as i thought.
The narrow frosted spot (24.9) looks not all that bad, it looks nice and soft and looks not the least bit like 25°.
I will have to print a few holders for them though to mount them properly to get a better idea.
If you put both colors in the light ght you may want adjustable color temperature, by having 2 channels controlled
If you need an Attiny 85, I got one spare, costs not much to put in a letter from Germany
I thought about that, maybe some time in the future, for now i can not allow myself to put in more features.
Otherwise i will keep tinkering with it forever.
Danke, hab noch 14 von den AtTiny44, zwei Tiny26 und nen Tiny13A, damit komm ich ne weile hin
Just a bit of boring math today, everything checked out so far until now so i did not bother creating a set of data points.
Does the thing actually work like intended?
The thing works with about 80% efficiency over the nominal input voltage range, that is not stellar but within reason.
Sadly i ordered the wrong inductors to run a comparison, i got 4.7µH instead of 10µH.
I planned the whole thing around a 22µH inductor and that is what i am using at the moment.
The idea was to use 10µH if possible but i do not want to order new parts once again, 10µH would be smaller and a bit cheaper.
Everything has gone a bit tits up while drawing the enclosure for the rest unit, in an ideal world there are two ways to do it.
You either design the electronics for the enclosure.
Or you design the enclosure for the electronics.
In this case, i have neither and that means, you have to settle on a few things, go from there and hope it all works out.
But when that initial guesstimation turns out to be not suitable, you have to start over from the beginning.
Current status, the circuitry is still the same and working fine, the optics have been figured out and how they will be mounted.
How the LED will be mounted and cooled has been figured out.
It turned out that i could not get the wires for the battery, led and ntc routed in a way that i liked while maintaining a small footprint.
So, i have made the board 17mm longer and now have to figure out how to mount the USB socket in a way that that works somehow without glue or screws.
Random thought: something like this with 3D-printed mounting posts to go through the holes?
As I understand it, micro USB sockets in smartphones and the like just rely on soldering to the board to keep them in place. That’s why you hear phone repairers complaining about people who ram USB charging cables in wrong way up and push the socket right off the board, instead of stopping, checking the cable and gently doing it right…
I had to remove the USB port on my Klarus XT12GTS mod, it was a pain to get off.
I desoldered the 2 side pads with solder wick, but still sticking on there
I did trash the whole USB port to get it off the board finally
If its soldered right its almost impossible to damage the mechanical connection to the board
Thanks for the reply!
I thought about something like that, i can solder a USB connector, that is not the problem. You get them in quite a few different footprints.
With holes, pegs, nubs and even to mount into the board.
My problem is more, i need to mount the board somehow and do not want to use screws, but on the other hand i need to get the USB port to the edge of the enclosure so i can actually plug things into it.
I want just one PCB were everything sits on to made it tidy, but the PCB is in the way when the connector needs to poke through the enclosure.
If i use vertical mounting slots, the USB port is in the way, if i use horizontal ones, i need to screw/glue something so the board does not move.
Its a bit hard to explain.
I think i finally got an idea, just did not have time yet to draw it, spend the last two days modifying my printer.
Small intermission, i want to do what i have sworn never to do again at home, etch a double sided PCB for the light.
For that i need a light panel to align both masks since i do not want to stand at the window and tape things to the glass.
The first big proper thing im gonna make with the printer, i am curious how it will turn out, all that will take a few days to print, lots of parts.
Two sheets of opaque plastic stuffed with led light strips, i got 30 feet of the best chinesium tape i could get, i want to run them at 1/2 power, that should make them last a bit.
It’s been an interesting project to follow. Heat transfer through the plates might improve by tapping the holes in the plates as well as going into the backing nut. Maybe a brass screw would be better than steel. A threaded copper post better still. Your picture of the pcb with USB and cell looks like it slides into an oval tube with the USB port at the end, I don’t see where the conflict is there.
Thanks for the reply!
It is a long way still to go but things are progressing, the next update will be a bigger one.
By now it looks like there is a high probability that i will add some metal into the head like you suggested, some through holes with aluminium rods or inserts, i will have to test that out once i got the new head figured out.
That is how it started out, but by now it is a h-shaped insert in the tube where the PCB can not slot into because something protrudes over the edge.
I wanted to mount the USB socket vertically so it was not a problem until now but i want to go with a horizontal one instead, the vertical mount plugs are a bit exotic.
There has been some progress with the new PCB and the 3D drawing but unless someone really wants to see that, i’ll spare you booring Screenshots.
Meanwhile i have ordered a spool of filament:
That is PETG, i have not used that before, there will have to be some Experiments.
Until now everything was just done in PLA but that is not feaseable for something that might get smacked around a abit.
PETF is less brittle then PLA and can withstand 20°C more temperature, that means the led can run a bit warmer then the 50°C that it is now.
But mainly i have to use PETG for the mechanical strengh.
The Alternative would be ABS or Nylon but both are not suited for my printer.
Even the PETG is pushing it a bit.
If you’ve never printed with PETG you may find there is a considerable learning curve. You may have to re-adjust all the printing parameters you’ve become used to with PLA.
PLA is so easy to print with. PETG requires considerably more effort and attention. At least that’s what I found.
I have a lot of eSun PETG on the shelves but PLA is still my go to filament whenever possible because it’s just so easy to print.
That said, the stuff is VERY hard to damage once printed. Exactly as it claims to be….partially between PLA and ABS.
Temperature wise……PETG is slightly more tolerant of heat……which is what makes it harder to print
Adhesion to the print bed was one of the bigger issues I had with PETG. That and lifting and curling on edges.
If you master it quickly please pass on your techniques…I still haven’t
Good luck with your project. Very cool stuff.
Best Regards
Z
Par for the PETG course.
I used Blue painters tape. Some use hairspray. Some recommend using a slurry of PETG melted in an acetone bath and wiping that on the bed before printing with PETG.
You might also try adjusting your print speed to a slower speed.
PETG seems to like a higher temp than PLA…around 230…but then you can get those nasty blob boogers if it gets too hot.
It’s a dance and the song is different for every printer
PETG is not soulable in acetone, at least im fairly sure.
Looked up quite a few things before i decided to use PETG instead of acetone, one con was, that i can not vapor smooth is with acetone.
I might give the painters tape a try.
My first layers i always print at 10mm/s, that is horribly slow but that gives me a perfectly nice and squished first layer, at least in PLA that is.
