For the IR, you have to look out for the heat damage to the eyes, as IR radiation causes a gradual but irreversible opacity of the lens.
That’s also the reason, you don’t want to look at the security cameras, with leds pointing at you (usually the beam is too weak to do damage in seconds but the risk is there, you don’t know what emitter they used)
I really don’t think you have to worry about IR from security cameras unless you’re holding the thing to your eyeball all day. Anything over a few hundred degrees is going to be emitting near-IR. A regular 60w incandescent bulb is putting out a few watts of it. A toaster is probably putting out a few hundred watts. A security camera can’t be putting out more than a couple watts or itd overheat.
Even this $1300, 8lb, IR illuminator with 260m range is only 39w max.
1 Thank
the thing is, it’s not the overall power, but the Irradiance mW/cm2 that is dangerous. It’s like the power and candelas, the candelas are what’s better to note the danger…
“The ICNIRP commission therefore recommended that to avoid the thermal injury of the cornea and the possible cataractogenesis, IR exposure (770 nm–3 µm) should be limited to 10 mW/cm2 for lengthy exposures (> 1000 seconds), and to 1.8 t –3/4 W/cm2 for shorter exposure durations.”
10 mW/cm2 = 0.1 W/m2 ~ 12lx (calculated from table of solar irradiance to lux illuminance)
Am I calculating this wrong?
I think 10 mW/cm2 is 100 W/m2.
And I think lux would be zero because it only applies to visible light. If the light isn’t visible its zero lumens and so zero lux
1 Thank
Yes you are right. I should have calculated it manually not on the online calculator.
10 mW/cm2 = 0.01 W/cm2 = 100 W/m2 well that’s actually quite high… probably dozen times more than brightest IR on security camera. That is, if you are not 10 cm from it.
And yes, the table probably just mean the lx of the visible solar light, to the solar IR irradiance power.
It’s available now,
5 Thanks
Too bad there are not much color options for D1V2…red would be great with 719A emitter.
@Hank_Wang is it OP reflector with this emitter option?
1 Thank
Oh dear MT-G2 on a single cell boost light. This is one of those things I would have killed to have back in 2015. It’s too bad now that 6V boost drivers can be found everywhere the MT-G2 only exists in unsold old stock. I don’t care if it’s only 70 CRI, the tint and beam looked great. Is there a modern LED that could be considered an equivalent to this thing?
2 Thanks
I’d probably buy a D1K with MT-G2 too
4 Thanks
+1 there.
5 Thanks
3+ here!
Also @Hank_Wang or anyone who knows, what is the lumens candela rating for the 719a? Was thinking about this in a D1!!!
Now you did it, as it seems. Many thanks!!!
Yes, it’s working now… but I’m still not entirely happy with it. I prefer the no-FET version, which is much simpler.
On my D4Sv2 dual channel, the DD FET makes no visible difference unless I use a completely full high-amp cell, which is not how I use it… so I prefer the no-FET firmware instead.
On a DM1.12, I’d imagine the FET has a much bigger effect. I’d recommend using the K9.3 firmware on it (0261), because it is tuned for a much bigger gap between the regulated and direct-drive output levels. However, I don’t have a DM1.12 to test with.
1 Thank
Thanks for the suggestion. You are right. On my D4V2 Edit D4SV2 Dual Channel with the r747 m0136 2ch hex, TURBO makes hardly a visible difference to HIGH. Full battery. I also have a DM1.12. With that, same hex, same configuration, TURBO does make a significant difference. I don’t fully understand that. I’ll try it out with the K9.3 hex. Would you suggest to use that for the D4SV2 as well?
Both lights use the same regulation circuit. I think it regulates up to 9A on most models, if I recall correctly… though it varies depending on which LEDs are chosen at time of purchase. It could be anywhere from 2A to 12A depending on LED type. Anyway, both lights have two of those plus a direct-drive FET. I’ll give some very rough ballpark numbers to show how that works out in practice.
D4S dual channel:
- ch1: 9A linear on LEDs 1+2 (~2000 lm)
- ch2: 9A linear on LEDs 3+4 (~2000 lm)
- ch3: DD FET on LEDs 1+2 (up to 2500 lm)
- Max regulated: 4000 lm (2000 lm + 2000 lm)
- Max turbo: 4500 lm (2500 lm + 2000 lm)
DM1.12:
- ch1: 9A linear on 12 flood LEDs (~2000 lm)
- ch2: 6A? linear on throw LED (~1300 lm)
- ch3: DD FET on 12 flood LEDs (up to ~7000 lm)
- Max regulated flood: 2000 lm
- Max turbo flood: 7000 lm
So on the D4S with all 4 LEDs on, high to turbo goes from like 4000 lm to 4500 lm when using a completely full high-amp cell. This doesn’t make much visible difference. However, on the DM1.12 in flood mode, high to turbo goes from 2000 lm to 7000 lm. This is pretty visible.
That’s why the emisar-2ch-fet build only uses a small part of the ramp for the DD FET… because the FET only adds a small amount of brightness. It’s regulated for almost its entire range, and simply can’t get much brighter than that.
Meanwhile, the k9.3 firmware uses a larger part of the ramp for the DD FET, because it gets so much brighter when direct drive is used. The regulation circuit doesn’t get anywhere close to full power, so when you short the battery to the LEDs (i.e. direct drive), it gets a lot brighter.
… as a side note, a DM1.12 with SBT90 LED is a special case. It has the DD FET on the throw LED instead of flood. It should still use the same firmware though.
Nope, I would definitely not use the K9.3 firmware on a D4S. It won’t make the light brighter… it would only stretch that “4000 lm to 4500 lm” range out across a wider segment of the ramp, which would make the “elbow” worse.
Personally, instead of extending the FET range, I go the opposite direction and use the use the no-FET firmware on the D4S (emisar-2ch.hex). The DD FET doesn’t really increase brightness by a perceptible amount there, and on a not-full cell it doesn’t increase brightness at all, so there’s not much point using it. Instead, the no-fet firmware uses the entire ramp for the regulated brightness levels, for a smoother ramp with lower lows and more precise tint ramping, and more consistent behavior between different channel modes.
TL;DR:
- D4S dual channel: I recommend the emisar-2ch.hex firmware. No DD FET.
- DM1.12: I recommend the k9.3.hex firmware.
1 Thank
Many thanks for the detailed explanantion. Got it. 
1 Thank
Sort of depends what LEDs you have, I guess. I can definitely see a perceptible increase with a D4S with 519A on the FET channel, W2 would be even more.
1 Thank
Certainly not due to any kind of intoxication I’ve used my D1 in a pool for about 10 minutes. I’m not sure how deep I’ve gone though the pool was about 1.5 m deep so above the 1 m specified for IPx7 testing and the light has had major water ingress through the button and the tailcap.
I blame nobody but myself but I guess be mindful of that…
ETA obviously the light was being put in and out of turbo repeatedly which might have contributed to the issue due to internal pressure swings. Same thing about the button presses.
2 Thanks
I’m not sure I trust any of my lights to be fully amphibious, not even the Zebras. Did your D1 survive?
1 Thank
It started behaving oddly then and there, it would only turn on at a high level or switch to turbo, that persisted despite immediate tailcap+1H factory reset. After shaking out some of the water and leaving the light dismantled overnight it seems to work as expected, now it’s baking at 50°C to drive out all moisture, then we’ll find out. I’m not into soldering heatsinked PCBs at weird angles to very short wires so I’m not planning on taking out the driver and doing a proper inspection unless it dies.