Very cool. I really admire your talent. If I could do half of what youāre capable of, Iād be making all kinds of sci-fi ādevicesā that just happen to have emitters for flashlight use.
Really nice job, would somebody knowledgeable explain what is going on in this mod :Dā¦ for serious I am interested in some explanation or picture description :).
Nicely done on that C01S fix! Considering how tiny this thing is, thatās very impressive.
It looks like the stock board was stripped and the main components of a ātypicalā linear driver have been fixed on top of it (with an insulating layer in between). A attiny13a to run the show (the 8 leg chip), a single AMC7135 for current regulation (the second largest black chip), a capacitor to help smooth out the incoming current for the attiny13a (usually called a decoupling or bypass capacitor, itās the small tan-colored? chip), and two resistors (small, black) that form a voltage divider so that the attiny13a can read the battery voltage.
Thatās all the components youād find on a Nanjg105C / Qlite driver. But instead of 4-8 of the AMC7135 regulators (350 or 380mA each), thereās one. At ~350 mA, heād probably see 120 or so lumens. BTW, this makes the light 10440 only, no AAA.
A quick and dirty one. Iāve been meaning to do this to a light for ages but just not to this light. I shrink wrapped my D25c ti. It probably wonāt last long.
Yes. 7135ās have three pins: VDD (positive input), Ground, and Out (low level / negative output). When voltage is supplied to VDD, the Ground is allow to flow to Out* (restricted to ~350mA). When voltage is removed from VDD, current stops flowing. By pulsing VDD (via PWM) you can vary the effective (not actual) current being sent to the LED. If you use the MCU (attiny13a in this case) to supply a 10% duty cycle PWM signal, the 7135 is enabled for 10% of the time. Over the course of a second, youād get 350mA for 0.1 second followed by 0.9 second of no current flow. Of course we use a much faster frequency than 1 Hz (Hz is cycles per second)ā¦ more like 8 - 16 kHz. Because of it switching so fast, our eyes usually cannot detect the on-off switching (unless a lower frequency is used) and we just see the average state of the LED, so we see it as if 35mA had been constantly applied to the LED.
Iām sure there are some nuances here that I may have mis-stated, but thatās the gist of it.
That made me wonder how this light level compares with trits. And the level is even lower! My integrating sphere could only just measure it, take it as a ballpark number but my LT1 measured 0.001 lumen output on the aux leds alone (switch led off).
Nitecore D10 anduril (14500 powered, sort of a mini fw1a)
Hardest part wasnāt the custom driver, it was sourcing the 3 hosts I purchased. This particular one was NIB and I paid a good bit for it, I was also thankfully able to scrounge 2 more used ones for a much better deal.
Luxeon V2 5000k
Oshpark boards in the new black fr4/clear mask (ugly, stick with purple!). FET+1, uses GT Mini FW target.
I only have 1 super old Efest 14500 for testing untill my li ion wholesale order gets here next week.
I donāt hate the black/clear but IMO itās kinda boring looking compared to purple (itās what oshpark is known for after all).
I think itād look better with a hatched ground pour instead of solid Cu. If I run across any more hosts and need to order more boards I think Iāll update it so itās prettier.
I also donāt like how hard the silk screen is to read with this color scheme.
This is the first reflector in which Iāve actually been pleased with the luxeon v2, LOVE the emitter behind a triple optic but IMO the beam from a reflector was lackingā¦ Not in this case, itās perfect!
As to the plain body, Iām really hoping to find a Balder HD1
Awesome! I am willing to do that in the one I bought a while ago, but havenāt ordered the driver yet!
Did you just swapped the driver or did you have to do any trick inside the pill (because of the wires,etc)? If that is the case,fo you have any photos that may help the mod?
I think it has to do with the reflectors depth:width. I tried them in an S2+ with OP shorty 20mm reflector and an Olight S15 (also very short overall but Ć=20mm) and both of them were ringy and had a pretty strong tint shift from spill to Corona to Hotspot.
The D10 reflector (also Ć=~20mm) has a much better depth:width ratio, Iād guess thatās why.
Also note the centering ring setup of the D10 keeps the reflector quite high off the 3535 emitter (originally XR-E).