Next I reflowed XHP70.3 HI D2 M2 90cri on 20mm 7070 board. Then I added thermal paste and dropped it in flashlight head.
Next I mounted the driver and soldered the wires.
I assembled the head. I used Convoy M21B green AR coated lens for tint improvement.
Light is ready for use. I also added bypass to tail spring.
Next I’ll do some beamshots and Opple measurements.
To be continued…
Very nice! I hope to build one for my FW1A when components are available. Is the height of the inductor 3mm?
Yep
It can take 6*6mm and 7*7mm inductors.
Here I took some Opple readings.
First stepped mode stock step 4
Then turbo
This led has very neutral tint for whole output range. DUV 0.0029. Beam is very pure white. It seems to sustain top of the ramp easily. If checked from XHP70.3 datasheet it corresponds to about 1000 lumens after reflector and lens losses. Not bad for 90cri led in small flashlight(led bin: 2D M2 90cri). Turbo output corresponds to little over 4000 lumens after losses.
Here’s still one photo of the lowest mode.
Next I planned to took some beamshots and maybe compare moonlight modes of different flashlights.
To be continued…
Quick moonlight comparison between this KR1 and Convoy M21B. Both in lowest mode and both using a boost driver of similar max output. Both also have same led.
Can you guess which is which?
Some beamshots
2m distance
5m
15m
Some piping in darkish enviroment ~20m away.
Control photo
Turbo
That’s it for now. I’m happy with this build. Only con of this light is the capacity of 18650. This driver will allow me to use every drop of it.
I suggest everyone to read TheFreeman’s driver thread. He’s closing the point of making these public. Still some UI things to solve. Anyone with skills in that regard is welcome to offer some help for him.
I will continue in near future with KR4 and 12V build.
Thanks for reading.
Outstanding design and execution!
I applaud your impressive skills.
Now if only Hank would adopt your modifications into the regular production line...
I'd love to have one exactly like it.
Enjoy your well earned success with this gem!
Thanks 
I am eager to see the final release. I’d love to put one in my Jaxman X2
That seems to be good looking flashlight, but not available anymore.
It would be good if Hank can make a KR series flashlight with 21700 battery!
Truly
Very nice to see the first of my HDR boost drivers assembled and used by someone other than me 
, Haukkeli was very motivated to go through the whole process and he provided valuable feedback, he’s a great beta tester :+1: .
In this configuration (6.4A6V) the lowest mode is ~75uA, for comparison the lowest output (li-ion) Zebralight is the SC700D at 50uA (6V), with some tweaks it can go lower but I think this is low enough.
That light looks pretty good, what driver size is that though ? it looks very small for a 26650 light.
About 22mm. I was looking at D4v2 boost driver in it. It’s very compact, though not the most compact. The machining is excellent and I have a soft spot for Jaxman. But figure if you buy Jaxman you need to plan a driver swap before you purchase.
Here’s a comparison to some familiar friends:
I have XHP50.3 HI 90cri that I’d love for it.
I have one Jaxman. Their S2+ equivalent. E2 I think. I baked it, modded it to triple that uses 18500 tube but takes 18650 battery and has H17Fx driver.
Nice little hotrod.
Do we see a PWM in 2m distance pic? Have you tested it for PWM (stability) in various modes and batt voltages?
It couldn’t be PWM but it could be oscillations due to instability, Haukkeli can check with the Opple though we don’t know how high it can go in frequency and since there is no noise it must be higher than audible frequency if indeed there are oscillations.
I never tested it above 6A so it’s possible the compensation values are not adequate for 6.4A. The values from the suggested schematic in the datasheet for example are only good below ~4A (at 6Vout).
I’ll assemble one with the same values as Haukkeli’s and check with the scope.
Im just asking because of Lume1 project. There was no any info of instability in nicely made data sheets, but that drivers was flickering in some cases. Stability in switching drivers is important.
Very true, even if it’s not visible/audible it can decrease the efficiency.