One of his requirements was no - low blue wavelengths emissions. The 4000K SST-20 should be a better fit as it should have less blue while the 3000K and 3500K should have even less. While the lower CCT may not be as advantageous there will be less blue, and 3500K is considered to be about the sweet spot for task lighting by lighting designers.
About the battery: here is a 14250 that’s 1200mah and rechargeable. https://www.amazon.com/dp/B01H1ZF33U/ref=cm_sw_r_cp_apa_i_immOCbC352MF1. Are the just saying 1200mah but dont actually have that capacity? The point is I would like to have a battery power this light for 3-4 hours. I would need about 1200mah right? I really like the dimensions of the 14250.
LED: Reason I say no/loe bluewave length is because these will be used close to people’s eyes so I’d like to minimize damage if possible. Is it something I should even worry about? Isnt 3500k more of a yellowish hue? Most people using lights in the dental world are using lights more in the 5000k-6500k range. Thise ranges seem to have a more blueish hue to them so maybe my stipulation of little blue wavelengths might be unwarranted. That Samsung LED that bluesword posted does seem nice and can be had for a good price.
I’d like thank everyone for their contributions again. Yall have been very helpful.
@afristch, you can get even the switch only if you want.
Just look for Omten 1288 on Kaidomain.
About that 14250 though, it does have 1200mAh capacity from a reputable manufacturer.
However, it is a non rechargeable lithium primary cell, so no recharging is possible.
And I would recommend high CRI 4000-5000k LEDs instead.
Not only do 5000k and below LEDs emit less blue light, but because they are high CRI, their spectral distribution have much smaller blue/green peaks, meaning less is emitted.
Here are some tests of the 4000K Panda 2M, actually closer to 3800K. It showed a saturated red color accuracy —R9— of 75, and overall red wavelength dominance. The 5000K 90 CRI LH-351D, however, showed a very present dominant blue spike and a slightly lower R9 of 70. The 4000K SST-20 in Yajiami optics, however, performs very well. It has about the same to less blue then the 4000K Panda light, 10 more R9 points and a couple more R12 points, or saturated blue. The 3500K SST-20/Osram Square will reduce the blue spike further , as you can see for the 3500K spectral power graphs. The 3500K Osram Square also has even higher R9 value of close to 95. I don’t currently have the SPD/CCT for the 3500K SST-20, but expect less blue then the 4000K varient. As you can see the blue-spike at that point is no longer dominant, and more like a blue-bump. Please be careful using beamshots as the absolute definition as to what these lights will be in person. Camera settings are really easy to tweak/fudge, and can dramatically skew the view of lighted environments.
3500K is more a neutral color commonly used in offices/classrooms/malls/stores in fluorescent lighting. You’ve probably seen 3500K multiple times in passing, but recognized it simply as office lighting. 4000K is also commonly used, but that general range—3500K-4100K—for interior lighting in public spaces is pretty common. I believe Target uses 3500K in some its stores. 3000K is like halogen light. It looks quite brilliant with a slight halogen incandescent tone when seen alongside 3500K. Most cars have white halogen lighting at around 3000K with no color cast.
Thanks both of yall. I’ll take a look at those recommended LEDs.
My next question. What type of optics should I look for. I’m looking for a really focused beam pattern without much of a corona. I’ll repost the link from above to give yall an idea of what I am looking for: LumaDent HL360 vs HLv3 - YouTube. What degree of optic would accomplish this best? To be a little more specific, let’s say at 18”, I’d like a 100mm beam. I’m assuming I will need a TIR lens but do yall have some decent guesses as to what degree is needed? I’m happy to buy a few and figure what works best.
Buy one of these and tune or replace the battery if it’s too heavy or too weak: Dental Product For Dentist Lamp/round Beam Light - Buy Dentist Lamp,Dental Product,Dental Lamp Product on Alibaba.com It runs an XP-G2 at 450m mA. The supplied battery (2x18650) lasts all day long. They also have a wireless one.
I’m using it with MO Optics 4x Vinkep loupes (highly recommend them) and a selfmade Nanjg AK-47A 1x 18650 Battery at 1 A. I wouldn’t want less light. The 3400 mAh 18650 drains in about 2:30 but is easy to change. I’m looking into upgrading to a 2S battery with a buck driver for more endurance though.
Why no blue?
I mean, i assume you’d want real daylight white, maybe a tad to the warmer side, so 4000 to 5000K color tempreature.
Now it’s true that most LEDs have a spike in the blue, and a lack of cyan, although that’s not really noticeable in real life, but there are also LEDs that have a more flat spectrum on that area, like the Nichia Optisolis, which happens to have a maximum rated output of what seems to be normal for dentist headlamps (around 50 Lumen).
With a pebbled 30 to 45° TIR optic, i think you’re very close to what you desire.
Not quite sure of that - he asked for help on a DIY light. My own DIY scuba diving light phase lasted 15 years and started back when 50 W incandescent bulbs with IR reflective coating were the hottest tech. (-;
@hitoakge asked about my current loupe light (wich is a modified version of this: YND headlamp) via pm. I thought I’d answer here, as it might be interesting to others too.
I did ask Mr. Liu from YND about the LED before I bought one - he told me it was an XP-G2 and I later confirmed that when I disassembled the light.
The stock LED driver is inside the battery housing and when I changed to the AK-47A I 3D printed a new housing for a single cell.
The light handles 1A reasonably well - I measured a maximum of 82°C on the housing - I added some thermal paste to spread the heat better as there was some hard white stuff with bubbles under the led and no thermal paste on the threads of the optics. I guess I’ll have to exchange the LED much earlier than usual but that’s easy.
I ordered a 20mm buck driver limited to 1A that’s using NarsilM firmware from @lexel and plan to print a 2x18650 or 2x21700 housing for it.
The current housing looks like this:
Thanks for the info! You said the right light had a new driver (1A) but also different optics? What optics are you using? Also that 3d printed battery holder is awesome. I wish I could do that, but I just started learning about making this light just recently lol. I was also looking at the luxdrive buckpuck driver, but it looks like the one you use is much smaller and can fit into the battery casing better?
It took some time to finish the new light but it turned out very satisfying. I redesigned the housing from scratch in onshape cloud CAD software, which is free and great. (onshape file)
I used @Lexel ’s custom 1A 20 mm buck driver, 2x 21700 batteries with a protection circuit from a 2s2p battery pack. LED connection and charging is both done via the USB-C port (non standard wiring)
@quics, Thanks for sharing your excellent results, that holder looks great.
i might suggest to add some buttress support to the pin holder since that is a highly loaded area based upon the many broken belt clips i have experienced—the clip gets torqued when the device snags on a chair or edge of table, etc.