Morning
I am looking for a torch/flashlight that emmits only 480nm to mimic this rather expensive tool
Ideally to mod a torch / flashlight or to be used as such
Any help much appreciated (Online I only found torches as close as 470nm.
Best wishes
Simon
They apparently have these LEDs at 475 to 480nm and 480 to 485nm. I don’t know how much of this is true but it’s worth a shot.
Easiest to get a drop-in type light, and swap out the LED.
No idea where you at, but Newark has a bunch of 480nm XP-Es.
Thanks a lot!
I saw that there are single LEDs that emit in that wavelength but I wouldn’t know how to make/fabricate a focal Lightsource out of these. Is there a easy way? Guide? Building instructions?
Sorry. I am a complete noob. I am Veterinary Neurologist and just don’t want to spend 1000GBP for the product above which practically just emits two specific wavelength.
Best wishes
Simon
a usually recommend a Convoy S2+ host for a build like this. Find someone who can swap the emitter with an XP-E 480nm and total cost can be under $50
Just like everyone is saying, just get a S2+ and swap the LED, that’s the easiest way. Depending on your country there might be a modder that could help you on this project.
Amazing! Thank you so much everyone!!!
Question is, how bright do you want it? You don’t want to scorch any retinas with 1.4A drivers and such…
If you use a drop-in host, or S2+, or other light, you might want to look into something that limits output to only moonlight or so.
And if pupil response differs in, say, “low” being a LED shining at a constant 20mA, or PWMed at 1% duty-cycle at 2A, that needs to be considered, too.
And if it’s to shine the light up, down, up, down, etc., to test pupil-reflex, you probably want to use a TIR lens to avoid hotspot-vs-spill issues. Or maybe a light with a momentary-on switch (forward-clicky).
Lotta things to consider…
The information I have regarding that is: Nominal light intensity 200 kcd/m2
I guess kcd/m2 depends on A: how “bundled” the light is by lens and distance to the retina and B: the output of the LED?
I know its difficult as I have 0 knowledge regarding the physics behind that. I contacted the infamous “hank” if they can sort something for me. If not I have to get my head around building and trialling myself.
Thanks
Simon
kcd/m^2 is luminance, the brightness of the LED itself when looking directly at it. Some LEDs go up to 275 cd/mm^2, which is 275,000 kcd/m^2.
1.4A would be too high, that’s for sure, but I was surprised to see it in action, it’s brighter than I thought.
Hard to judge by the video but it looks like it is being driven at about 1W, so, not much power but way more than I was expecting. I don’t really know what the 200kcd figure means in terms of brightness or optical power but for me it seems that this value might be too high.
True, using a frosted TIR just like the original is the way to go.
As him if he can make a dual channel D3AA with one 480nm LED on one channel and two 630nm LEDs on the other channel, just like the original setup of this equipment. I don’t really know how much this retina tester costs (can’t see the price without logging in) but a custom made D3AA would cost way less than a purpose built medical equipment.
It is kcd/m^2, not kcd. cd/m^2 is nits, as in how bright monitors look. If you point an SFT-25R straight into your eye at 5 mA (0.1% mode with 5A driver), that’s about how bright it looks.
It costs ÂŁ1194, Hank or Convoy has real market disrupting potential.
Thanks again for the input! I am definitely lacking sufficient knowledge at this stage to custom make my own. I went with this for now:
I think thats as close as I can get to the wished for wavelengths.
I can test it on myself and compare to the 1100GBP product and then gain some experience regarding clinical application if it is safe.
Best wishes
Simon
That’s a very good choice. Just be ware that the power output is orders of magnitude higher that the medical device. Don’t blind your patients by turning on high mode accidentally.
Good shout but don’t worry! I will carefully test on myself with different settings/distances!
From a legal point of view obviously this is all on me and my responsibility so will be very careful!
Disclaimer: I wrote the following content as a hobby and am not responsible for any possible legal or medical risks.
“200Kcd/m2” is a strange spec for me, because it doesn’t seem to be a direct property of the light, like lm(lumen) and cd(candela). It’s more like an observed luminance property of the light source (LED emitter) behind the lens inside the light. It probably doesn’t represent the luminance of the observed flashlight itself, because if it does. Multiplying by the approximate area of a usual flashlight lens (r=10mm) yields a light intensity of roughly 62.8cd, which is very harsh to the eye at close range.
Considering the use in eye check, this spec seems to limit the brightness not being too high.
To prevent the brightness from being too high, the best thing to do is to record the Lx value that the product (if your friend, coworker owns it) gets from shining the lux meter (a phone app might work as well) at a certain distance, and adjust your flashlight to output a similar value as well.
If the above is not available, you may need to know some more info, such as the emitter produced Lm. According to the variant of the luminance formula Lm=Luminance* SolidAngle * Area, the LED light-emitting solid angle is often about 2Pi. The “Area” actually needs to know the refractive index of the XPE emitter dome and the chip area, but those are actually unknown, I assume the chip area is 3.38mm2 based on the emitter dome R = 1.3mm, and the refractive index is guessed to be 1.5 since this is common.
lm = 200k * (2*3.14) * ((3.38*1.5)/1000k) , the result is about 6.36 lm. For reference XPE blue a certain bin is 45.7lm at 350mA, if you want 6.36lm you need to control the current at roughly 48mA.
Actually, up to this point, the effect of the lens hasn’t been taken into account, because that product isn’t marked with info about the lens. But it doesn’t matter, a lot of the data above is hypothetical and rough anyway, and it’s not like your flashlight can really control the current that accurately. So, just try not to make it too bright.
CSLNM1.23 (617 nm) will be closer to the 630 nm red of the medical device than 660 nm deep red. I’m afraid eyes aren’t that sensitive to 660 nm so it may not work as intended. For true 630 nm I recommend this SST-10-R. Hank may not have these emitters but he should be able to sort something out for you if you’re willing to pay more; it should still turn out much cheaper than the medical device. Also you may want dual channel instead of triple channel unless you really need white.
in addition to what others said, I want to point out the XP-E cyan is not a monochromatic emitter, meaning it doesn’t just put out a single peak when looking at the optical spectrum. It’s pretty broad band, probably using a regular 450nm pump just like a white LED but with a different phosphor composition on top. Now there will hardly be any red wavelengths in that cyan light so it might work for the intended purpose, who knows. If this neurological test is more qualitative than quantitative maybe the exact values don’t matter as much?
