SkyRay King 7x TIR / XML2, 8x 18650

Hello again, I would like to show you my SkyRay King, my favorite host. Unfortunately I dont have photos from building prosess of the head, I made however some photos from the conversion to 8x 18650.

To begin with the head, I wanted 7 XML2s on Sinkpad IIs. After testing some TIR on single emitter lights I was really impressed with their performance and decided to use 7 of those on this project. So there was a need for a bigger head.
I cut part of the original head off and created a really thick shelf of 2mm copper discs and some 4mm thick aluminum pieces I had. All together I think I have 8mm alu and 8-10mm copper for a total of 16-18mm shelf!! A look at the photo will help…The discs are press-fited inside the head.
To host the reflectors (total width 3x21=63mm) I used an aluminum lens hood with inner diameter of 62mm, and filed down each TIR as much as possible. This helps finding a nice filter cover to protect the leds.

I always wanted to convert the skyray in 8x 18650 but I recently got this idea that I am showing you here.

First the head:

Then the ready double tube 8x18650 SkyRay

I started by insulating the spring board of the first tube. I am using some polyamid washers I cut myself from a 10mm round polyamid bar. I am using this material as its heat-resistant up to +350C from what I have read. A short at this point of the tube would be catastrophic/explosive so I made sure it wont happen. I also placed a polamid cylinder to insulate the screw that brings the positive of the second tube, you will see in the next photos…. You can also see how I have bridged the springs, works fine for me.

Small black silicone pieces and tape added for extra protection against shorts….


Then I prepared the copper disc that will transport the positive pole of the second tube up to the head alongside with the 5mm screw. Copper disc and spring board are strongly glued together with epoxy glue. It is immportant that all washers are the same hight as this will affect the centering of the screw.

I had to cut the tail screw of a skyray to use it as a connecting part for the 2 battery tubes:

This ring was glued with epoxy at the tail of the first tube so than the second one can be screwed and unscrewed easily. I can adjust the pressure the screw is applying on the head by screwing/unscrewing the long screw at the tail, see photo. Spring board was screwed in placed giving the following result:


This little head between the batteries is bringing the positive poles of the back batteries up to the head. I used a small amount of solder to create a soft spot on the head board where the screw can create a perfect contact point.

Here is the King:

Output meassured to 9000lm with 10A LG cells at 4.05 volts, I will test with 4.2v later. With a maximum output of 2000lm per emitter at 6A(test from this forum) I think I still have room upwards (14000lm possible), but I am limiting the current with a 22AWG bridge wire at the moment. Wouldnt like to run them at the maximum… I need to mention that I added an additional Mosfet in parallel with the stock one to handle the higher currents.

Only thing that becomes hot on the driver is the current limiting snake trace and the 22AWG wire I am using to bridge it, I think it is possible that it can unsolder itself after some min in high mode. Here is where I need help, how to limit the current in another way? If I use thicker wire I will burn the leds out.

I will be adding a 105D driver to control the mosfets to have 3 modes as my current lowest mode is 3000lm, which is extreme for some cases…
Wires are silicon heat resistant. From the driver to the leds double 16AWG, between the leds 22AWG. High quality thermal paste was used everywhere.

Some notes/thoughts/ideas:
-Conversion from 1s8p to 2s4p very easy by just removing the middle board.
-It is possible to use different degree TIRs to adjust throw and flood according to anyones needs.
-Light can be still used even with 1 cell
-Back tube can be used if wished as storage for fresh cells, or used to host charging borads, powerbank borad, or even a small spare flashlight! By removing the long screw basically one can disable the second compartment in seconds!
-It is very immportant to meassure output with only the front tube and only the back tube independently to make sure both have the same current trasnporting ability. In case the back tube cant transport as much current as the front then we have uneven cell discharge which is pretty bad for a number of reasons… I meassured many times and got exaclty the same output, will repeat the meassurmetns often! The long screw pressure to the head must be enough but not too much…find adjustment needed!
-The handling and grip of the flashlight was dramatically improved with the second tube.
-The flashlight has 2 modes at the moment, high 9000lm and mid 3000lm. At mid I can use it for hours without thinking about temperatures. On high it becomes a little hot but not too much. It has just too much mass on it because of the copper.

Thank you for reading!!! I hope you like the project, personally I am very satisfied with the results!

Nice nice nice!
Really cool light you made.
8P talk about runtimes!

:+1:

Very creative Kourtinoksylo. You have one very unique light. Well done. :+1:

Nice mod! And I like your engineering that went in to converting to 8p. How is the weight balance when holding? :smiley:

if you ever consider 2S/4P I got still one board that is suitable for it with little modification

limiting the current by modifying the firmware not to do 100% FET is a good option to control current
the FET can handle from datasheet 26.6A at 70°C ambient temperature

For my hands the light is very good balanced. The head is really heavy with all the heatsinking and the extra tube balanced it out. A lighter head would bring the center of mass further back I think and the grip would be also further back.

New meassurments with fresh 4.2v cells have shown 10.000lm with one tube and 11.800lm with two tubes!!! The voltage drop is much less with 8 in parallel resulting in much higher output than 4p. I am a little afraid to use the light like this, I might consider steping it down to 4 cells again, I dont know…. The currents the light is handling(near to 40A) is higher than my experience can allow, plus the risk of uneven discharge is a problem.

My meassurment method is a light box that has been calibrated as good as I could, my meassurment of near to 12.000lm seem to me absulutely possible considering the wires I am using, 8p cells and SinkPads….

The idea of a driver with 90% mode is very nice….I will search for one…

Narsil drivers can be programmed specifically as you need it, if you get one from me
or you build your own and alter the firmware

I got 46mm a 1S4P Narsil driver as well from last batch if you want one, just the boards came with wrong solder mask on ground ring, so I removed it works great but does not look perfect

Dang, how hot does it get?

I let it run for 15sec in the light box and got from 24C to around 35C, I didnt want to let it more as I wasnt expecting so much output and got afraid I am gonna fry smth :smiley: . I am sure it will be too hot to touch after 2mins.