I have plans for a swap coming up anywhere from 1-3 months, and I can’t seem to find this Cree XP-G3 Photo Red 660nm LED bare; it is aways on a star.
I guess I should just start from the beginning. I’d like to have a Zebralight SC5w II modded with 660nm LED, preferably Cree. I’ve had help determining with research the light will draw 1.25A on its high mode, and the XP-G3 Photo Red can handle up to 1.5A, so that is a match.
However, SC5w II is a single AA light, and the XP-G3 specs say it wants 2.65V at 1A. This could be a problem, but maybe the boost driver is giving the LED more than 1.5V?
Help me out BLF.
I need to know what voltage Zebralight SC5w II is giving its stock Cree XP-L2 EasyWhite 4500K LED, and I need to find a suitable Photo Red LED bare that will both fire on a single AA in Zebralight, but also not be destroyed.
TIA
Edit: I guess I wasn’t looking hard enough, because I found one in Australia, and another in Texas.
Now I just need help making XP-G3 Photo Red will work and fire in a Zebralight SC5w II.
FYI, in case you are not aware, Sofirn is planning the SP10R AA flashlight with a 660-670nm LED, you can easily find the interest check thread. The Zebralights are excellent flashlights but also a bit expensive. If you don’t need their UI using one for the mod may be bit of a waste.
I am sorry but it sounds like you are way over your head for this project. If removing an LED from an MCPCB is too much for you then swapping the LED on a Zebralight will not be possible. (How do you plan to remove the existing LED from the Zebralight’s MCPCB?)
It could be possible that your looking for a XP-G3 660nm dominate photo red that doesn’t exist. I dont see it listed in the datasheet.
They list the XP-G3 as 645 typical dominate, which if you look at their spectral graph which includes 660nm but it is just not the peak produced by the led.
I used one here in a modded Thrunite Ti3. ~~ENDED~~ [US Only] Free Thrunite Ti3 v2 on Thrunite.com (with $5.95 shipping fee) - #112 by moderator007
They can be found loose at Digkey, link is in that post.
Peak Wavelength - Peak wavelength is defined as the single wavelength where the radiometric emission spectrum of the light source reaches its maximum. More simply, it does not represent any perceived emission of the light source by the human eye, but rather by photo-detectors.
Dominant Wavelength - Dominant wavelength is defined as the single wavelength that is perceived by the human eye. Generally one light source consists of multiple wavelength spectrums from the light source rather than one single wavelength. Our brains turn those multiple spectrums into a single color of light consistent with a single specific wavelength which is what we see when we look at the light. That’s the light source’s Dominant wavelength.
ggf31416, I’d like to mod the Zebralight because it is so versatile, lots of modes, AA, and I think it will get very bright on one AA, and very low. I don’t believe Zebrlight is expensive, just the opposite for a premium quality high end flashlight. I was not aware of Sofirn’s plans nor the interest thread, and I’ll watch out for it. I can always use another red flashlight.
Parametrek, I never intended to nor specifically indicated I was doing this myself. The modder will be familiar with getting inside Zebralight and fully experienced with swaps otherwise. I think it would only make sense to get the bare LED if available, and it is, to reduce the labor. I am familiar with the H502pr, which is not a flashlight, it is a headlamp with the emitter facing 90° rather than forward, doesn’t put out amps as high as I’d like to get as bright as I’d like, it is an older model without the fully programmable mode groups, and its runtime is surprisingly poor for Zebralight at the given brightnesses (100Lm for less than an hour?). I have a couple Sofirn C01R on the way, but these are AAA twisty and won’t have the runtime I’d like nor get as bright or as dim as I’d like, and only have a few modes of brightness from which to choose. You see, I’d like a recent generation AA Zebralight flashlight in Photo Red, with all the modes and programmable interface, with the very high brightnesses and very low brightnesses, and all the others in between. I have corresponded with Zebralight recently, and they indicated they do not plan on producing any red side clicky flashlights, nor even any more red headlamps.
lightdecay, the reason for not choosing a 3A LED is because the SC5w II does not put out 3A. I have to match the LED to the SC5w II driver, and it only gives about 1.25A from a ~1.5V cell (though I don’t know what voltage is reaching the LED), so I’ll need an LED that will still fire and yet not fry with specs in that range.
moderator007, I am reasonably sure they do exist in 660nm, as I have linked to two of that variety in my initial post. The XPGDPR-L1-0000-00F01 peaks at 645nm, but the XPGDPR-L1-0000-00G01 peaks at 660nm. I also have a TiS that is modded with a Cree XP-E 630nm, and the runtimes I am getting on a AAA are nothing short of amazing. I may choose to use a 630nm Red rather than 660nm Photo Red for this ZL mod in the end.
This is interesting, but when we’re talking about an LED with a narrow emission spanning 10nm of wavelength and peaking in the middle, is the distinction between peak and dominant wavelengths all that critical? I suspect they may be identical in many instances of single color LED, or at the very least, indistinguishable by our eyes. While it is not so difficult to see the difference between 630nm and 660nm wavelengths, are you sure you tell the difference between, say, 655nm and 660nm wavelengths?
I really titled my post incorrectly. The questions I need answered really have to do with specs of the Zebralight SC5w II, which Zebralight doesn’t share. Someone (I don’t have the reference handy) figured out it was pulling 1.25A at the LED in its highest mode. What is the voltage of the highest mode at the LED? What range could it possibly be if we don’t know for certain? I don’t think finding an appropriately specced 660nm LED will be too difficult once that is known, whether a Cree or some other manufacturer.
I do appreciate the responses and the discussion. I am sure I will not be the last person wishing to do this, so let’s keep them coming, please.
I did a little more digging, looks like they list in the datasheet 660nm under Group Code P4.
Looking at Cree’s site for the XP-G3 list it as a photo red 660nm, but when you go to the datasheet. The two bins you listed.
.
.
Then the Spectral graph looks like its peak is around 675nm.
.
.
When you scroll down near the middle of the datasheet it list group codes for the peak wavelength.
.
.
So how do you know what group code you get?
Am I missing something, not sure if I’m figure this out or getting further from the truth.
The driver should be current regulated. The run time graphs I have seen have all been pretty flat.
Judging by Texas Ace’s Test with the xp-l2 given the 550 lumen output of the Zebralight, my guess is 3vf at 1.25 amps or slightly more.
.
Djozz tested the XP-G3 photo red here.
He suggested the sweet spot to be around 2 amps, its in that thread two post down from the linked post.
If the driver is current regulated, then it should work.
But if its not then I guess there is only one way to find out if it will survive.
I appreciate the research you’ve done, very generous. I get tripped up in Cree data sheets, so I just go buy the reseller data on their sales page… which lists 660nm. Would both sites be exaggerating the same wavelength? There are so many different Cree LED that more or less have similar specs, I can’t handle it… which is why I asked for help.
I think that is a good guess! I figured voltage would be boosted, but had no idea how much. Doubling to 3V sounds reasonable, now that you say it. How much more could it be? I think you have to be correct, and seems like it should fire that LED I picked, and yet still come in below its max of 1.5A.
I know the driver must be current regulated, because it keeps constant brightness in all modes. I don’t know how else that could be achieved. Kind of blows driver won’t make it to 2A, but maybe it can be tweaked. Also think it might not be the same (I may have linked to the wrong LED), if the spec I saw was 1.5A max for the LED: “xp-g3 660nm Red, max 1.5amp”… unfortunately, for another week I am locked out of where I saved that information (in my CPF private messages. Could be I saw this, which is XPGDPR-L1-0000-00F01, which says 660nm there, but not above in the comments, says 645nm elsewhere… but if I have to go with a lower frequency, 645nm isn’t too bad, shouldn’t be orangey).
Thank you moderator007! Looks like you’ve done all the work, and made it look easy.
You are correct with the 1.5 amp max spec. Thats what Cree suggest but at BLF we push things a bit harder to find out what they can take and where failure is.
Cree’s suggested max for a xp-l2 is 3 amps but look at what it will actually do if you use a copper mcpcb with good thermal transfer.
On that point you must be aware that each wavelength have different eye sensibility i.e. luminous efficacy, for the same amount of luminous power (watts) the perceived luminous output (in lumens) peaks for green monochromatic light and falls very quickly for red and violet light near the edge of visible light, see this table . LEDs are not fully monochromatic, you would need to do integration to get the precise conversion factor between lumens and watts of luminous energy but I think you could approximate it by the dominant wavelength if I’m not mistaken, math hurts my brain at this hour.
100lm of 660nm at 40lm/W factor is 2500mW of luminous power, that’s would be a lot, equivalent to over 600lm of 5800K white light or 1700lm of pure 555nm green light. 100lm of 660nm looks as bright as 100lm of green or white light but you need a lot more power in the light to get to 100lm of 660nm light. However I would be wary of listed lumen outputs beyond 620nm, manufacturers like Cree only list mW and nothing guarantees that lux meters/integration sphere are well calibrated for eye sensitivity at unusual wavelengths, after all ‘photo’ means photosynthesis to grow plants, not photography, and plants lack eyes.
Edit: wall plug efficiency is higher for photo red than for white light LEDs iirc so it doesn’t take as much power to get 1 mW of photo red as 1mW of white light, but the point still stands.
Edit: conversion factor is likely closer to the double, 80lm/W because of the contribution of other wavelengths, so luminous power is likely nearer to 1250mW, which makes a lot of sense given the 0.9hr runtime that translates into 2.7W of input power.
I have trouble swallowing that explanation. I realize that red at the same watts as white will appear dimmer, but I doubt that fully explains it. I think my perception of H502pr having poor runtime is more likely a factor of the state of the technology when the light was released, at least weighted towards that rather than it being red as opposed to white. Once upon a time, getting 40 minutes of 80Lm of white light out of a AA was incredible. That was about 10 years ago. I would guess that H502pr model is at least 5 years old. That’s really a long time for digital driver and LED tech to advance. I’ll have to look at stuff tomorrow, but no way I’m buying that 5 years ago Zebralight could have squeezed almost an hour of 600Lm of white out of an AA, but instead chose 100Lm of photo red.
-If you take LED efficiency into account,5700k xp-e2 gets at most 140lm/W. Photo red LED wall plug efficiency is 55%60%, so you get 550-600mW/W, at 40lm/W from 660nm that’s 22-24lm/W, the eye is just not very sensitive to photo red.
Edit: actually djozz got over 50lm/W, 122lm at 1A with the XP-G3. There is a >15% improvement in radiant output from XP-E2 to XP-G3 but I think the main difference from my theorical calculations may be the 645nm dominant wavelength, the eye is twice as sensitive at 645nm than at 660nm.
XP-E2 and SST-* dominant wavelength likely is lower than 660nm as well, but it is not listed anywhere, XP-G3 is the oddity in listing the dominant wavelength for a photo red LED, rather than just the peak wavelength, as plants don’t care about dominant wavelength.
Edit 2:
Actually if you look only at the graphs the XP-G3 may actually be the deeper and wider one, I aligned the graphs from the respective datasheets as best as I could 660nm - Album on Imgur . The actual spectrum will depend on the bin, but we can’t say that the XP-G3 is less deep than others just because it lists a dominant wavelenght.
i build several deep red lights, just for kicks, and 660nm are dimmer to the eye compared to 625-630nm aka regular red. f the same wattage, things are harder to see in deep red, they say the wavelength of 660 is exactly what plants need, and was originally designed specifically for that. it may also be good for illuminated displays, but other that than idk,
Really appreciate the work moderator007 and ggf31416 have done. I have no way of repaying you. Also appreciate jon_slider adding more data, and alpg88’s input as well.
I was firm on 660nm when I first posted, but now I am open to the idea of 645nm if it is expected 660nm will lose too much top brightness. The only red I currently have is a TiS with XP-E Red 620nm-630nm, and it is pretty bright on the high mode, which was 120Lm when stock white. It is very bright in red, but I can’t hazard a guess at the actual lumens, but obviously not 120Lm.
I am expecting 2 Sofirm C01R 660nm to arrive in the coming weeks, and I only bought them to experience 660nm first hand, to help with the decision on the ultimate goal, a Zebralight SC5r (from w). I expect their light will be radically different from my ~625nm red TiS. I may have to do a less expensive 645nm swap in something, maybe a ThruNite Ti, just so I can see what 645nm does for me before I make a final decision.
But clearly it is easy to get low lumens of red. It is getting the red bright enough in the highest mode with that driver and that single Aa that is going to be the trick.
Killing me that I currently have $11 to my name. But I did it to myself, and so I have more lights coming in than I have had in the previous 3 years… so it was about time. This project was always stated for the future, 1-3 months from now
Thanks again, fellas. I never expected this much attention… so used to trying to kickstart threads in another forum. I really do hope this interests someone else to do also, so I am determined to get to some similar goal as stated, whether it is 660nm or 645nm… and I am even open to going as low as 620nm-630nm.
