My first beamshot.
This was taken with a Samsung point and shoot camera.
Left: OL Mini-Mag with Nichia 219
Right: OL Mini-Mag with XP-G2
Actually Nichia rates it at 1.5A max. Just like Cree rates the XPG. 350mA is just typical and they use it because the LEDs are very efficient at that drive level.
And if you look at the measurements over at CPF you will see that the stated CRI of 92 is achieved from 1.4A upwards.
750-1000mA with XM-L T6 is usable for me indoors.
From DS: Around 3.0V/1A, ~380lm.
From DS: XP-G2 R5, 2.95V/1A, ~330lm.
From DS: Nichia 219 B10, 3.5V/1A, ~210lm.
@3.8V/1.5A it’s ~290lm.
An idea, with a single 4.2V Lion, switch the output (PVM) when the voltage/amps is too high for the Nichia, when it gets fine just direct drive it. = high mode. Other modes just PVM it. Though with a fresh cell the Nichia might get a little abuse 0:)
Sure buck boost would be nice but there are none (1.5A) and none are compact (low profile).
A nichia tripple looks nice, but on the other hand I start thinking about XM-L triple.
Since it’s a triple I could use warm, neutral and cold emitter at once, control each separately to adjust output tint. Since cool will over power the warm.
Anybody ever made a balanced XM-L triple? Or is everybody just putting in 3 XM-Ls of the same tint?
Thanks, what tints are they supposed to be?
And what amperage are they each driven?
@JackCY: What benefit would switching bring? I mean, what you described is pretty much what AMCs do. As long as Vin > Vf, they burn off the excess voltage. Once Vin < Vf, it just goes DD.
XML triple idea is great, problem is: driver. There are so few drivers to power LEDs individually..
Both lights are Old-Lumens mini-mag mods
The Nichia 219 is a neutral white 4500k 92CRI
The XP-G2 were mistakenly sent to Old-Lumens instead of Nichia 219 by illumination supply, and the switch was not discovered until after he had already modded several lights. I think it might be a 1D tint.
Link to Nichia 918 Mini-Mag.
Link to XP-G2 Mini-Mag
I don’t know what amperage they are driven at, but if you PM Old-Lumens he could probably tell you the amperage and what drivers he used.
Yeah, I know that's what the datasheet says too. I'm talking about actual use. Just because it's rated for 1.5A max on the datasheet doesn't mean that it's best suited to be put to actual use at 1.5A all the time.
But its only up to specs CRI-wise at 1.4A and above.. so it should be made for that. ;)
The Nichia 219 is already horribly inefficient at 1.5A, a mere 60lm/W (at 600mA it is 90, at 100mA it is 125), but most people care less about efficiency than about output (I run mine at 1.4A as well 
). The led does not seem to suffer from that.
The CRI is very good at all currents, it does not vary so much that you would notice I think. Example: at 800mA CRI is 90, don't think anyone will have a problem with that.
You wont notice a difference in brightness from 1.2A to 1.5A either, I was just talking about the specs.
And who cares about efficiency. I have low and medium for that. :P
Similar yes, but not exactly. The AMC7135 drops 0.1V because it has high impedance and what it does as far as I was able to understand is limit the current. So the excess gets burned out as heat somehow, such as it increases the resistance of the switch and it heats up and limits the current throughput.
Where as using a power mosfet, it’s just switching full power on and off the battery, with ~2mOhm switch resistance there is 0.01V drop on the switch at 5A. So if the Nichia can handle 4.2V being switched over it, which should yield close to or little over 2A through it.
Maybe it could handle, maybe not, it seems simple and delivers nearly full voltage as it is nearly direct driving the LED. There is just a little resistance added by the mosfet.
Triple yes, custom driver. If the ATTiny can control 3 outputs or some other Atmel or other chip, it seems quite simple to make. In he worst case and need of full power, just use 3 Nanjg 105C each to drive one LED and program/modify them in such a way that it results in a desired color output. Some big hosts seem to have room for big drivers.
Buck-boost would be nice but I couldn’t find any on Digikey that could do over 1A, one seemed to be 4.5A but I think it can do only 1A and that 4.2A means something else in the DS. LTC3478
60mOhm is heck a lot for a switch. Means 0.1V drop at 1.5A.
Still it seems it can only put 1A on the output. So the 4.5A switch there is useless. Plus effectivity at lower voltage seems to be poor.
Driver with LTC3454 is what is being built right now in one other thread, 1A only.
Bare Nichia enroute from IS, they are stock again 
Can somebody share their experience of putting an XP-G or Nichia 219 into a light with an XM-L reflector?
What is the change compared to using an XP-G/Nichia 219 reflector for it?
You can also replace the input diode with a jumper and pick up a 1/10 volt on output run time, just be careful with your battery. :zipper_mouth_face: Is it just me or do all “C” tints seem greenish? I always try to get 1D or 3D instead. The 1D by itself appears very white to me and only next to a Nichia(which is yellow/magenta) does it appear bluish and the 3D is even better.
You are not alone. I want to find some 3D XM-Ls.
Cutter has them.
3C is yellowish, 1C and 2C are probably greenish. 3D should be good, but no XM-L2 is available in 3D, XP-G2 is, should have bought it maybe (along with 20mm TIR Carclo facepalm), not only the Nichia.
XM-L T6 3C is nice, will see if the XM-L2 T6 3C will be a bit less driven high in tint as it’s higher temperature binned. Could be a bit more in the red/yellow than it’s predecessor.
Still waiting for them to arrive. XM-L2 T6 3C and Nichia 219.
I tested one from IS and values are almost spot on as in datasheet. I measured 3,3V@0,8A / 3,5V@1A / 3,7V@1,5A