The SYNIOSBEAM - CFT90 recoil thrower

Looking forward to the build, nice work Enderman! :+1:

This is a very cool pic!

I agree.
Although Enderman’s light wont look anything like that with just 400 Lumens. :slight_smile:
It will be more laser like and the beam will be less visible.

I still think the Luminus CFT-90 is by far the best suited LED for this project.

Couldn’t agree more.

I have done research which suggests a laser phosphor system could get much higher than 300cd/mm^2, but have done no real world tests yet due to the high cost of lasers as well as the danger to the eye.
In the future I will definitely be exploring that option, but it was just not something I wanted to put into this light.
The goal for this was LED only :slight_smile:
I don’t really consider the laser phosphor technology an LED.

Thanks :smiley:

Great idea, I will work on a comparison of different lamps.
I also have to add some stuff to the optics comparison, stuff like multi-lens is missing.

I have designed a focusing system that is very precise, some spring-loaded screws can be turned to move the entire reflector forward and backwards by fractions of a millimeter :slight_smile:
You will see in the renders I post soon.

Thanks :slight_smile:

Yeah I was surprised when I heard about that W10 getting such high lux from such small diameter.
Laser+phosphor may be the future for high intensity!

Thanks! I have’t shows any of the real work yet since I want to post all of the preliminary stuff first like parts list and renders, but I have begun working on the physical thing.

If I can find a compact 30-50A driver that fits in the light I will mod it with a CFT90.
Currently the only drivers I have found that can do that run on 120/240VAC.
I do agree the beam brightness will be much more impressive even if I get less throw.
Worst case scenario I will mod it with an XHP70.2 and see how it performs :slight_smile:
For now I just need to get the thing built first haha

See here. Can be ordered here. You can choose the current (up to 40A).

Also, Lexel recently developed a 20A driver. You could ask him if it’s possible to make a 30A one or combine two of them.

Would this light be able to shed that amount of heat? after all the led needs to be in the beam path incl. the which includes the heat sink.

Maybe a heat pipe or water cooled solution to have less stuff in the beam and shed the heat at some place not affecting the beam.

Looking forward to the build.

A 6x6 array of XD16 dies would be perfect. ~28000 lumens and 80-100 cd/mm^2. :open_mouth:

BEAM SHOTS!! :partying_face:

Too large to fit inside the light, unfortunately.
I only have a few cubic inches of space due to how compact everything is inside.
When the light is finished, if there is no 30-40A driver out yet I will contact Lexel and see if he can custom build one for me.

You’re correct, LED cooling is very important but also hard to do when the LED is floating in front of the reflector and trying to block the least amount of light possible.
It is liquid cooled :slight_smile:
I’ll post the parts list tonight and renders tomorrow.
If I used the Synios DMLN31.SG I might have been able to get away with simply using heatpipes or directly attaching it to the aluminum, since it only uses 4W.
The higher temperature will probably reduce the candela though, even if it doesn’t overheat.
I might test it anyway though.

That’s not a lot of CD :frowning: but 28000 lumens in a spot does sound kinda cool :slight_smile:

Use your imagination :slight_smile: It isn’t even built yet! xD

Should be interesting :+1:

Not a lot of cd, but it should be a bit more than the sliced XHP70.2.

The XHP70.2 without slicing is almost at 100cd/mm^2 though?

At least near the peak of its performance.

I still think the CFT90 is the best choice for throw+lumen output :slight_smile:

That is not taking into account the effect of the cross on throw from a reflector. The cross between the dies always takes up some area in the reflector, reducing the effective luminance. Though the XHP70.2 is better in that respect than the first generation. Texas_ace measured the sliced XHP70.2 at 6A (12V) in the BLF GT at around 65 CD/mm^2. I present you with Giggles, the most hardcore BLF thrower in the World. - #101 by Texas_Ace

I agree the CFT90 is in another league of brightness and output.

The cross doesn’t affect anything, all the light from the dies goes to every point on the reflector.
There is no “taking up part of the reflector”.
The result is simply that the lux in the center of the hotspot is not as high as the lux slightly to the side diagonally, where one of the square dies would be projected.
This is only for short distances though, far enough away the four beams from each individual die will have diverged enough that all the points in the hotspot have the same intensity.

65cd/mm^2 at 6A definitely suggests over 100cd/mm^2 at close to 20A.

That’s 6A in the 12V configuration.

When I look into the beam of a C8 with sliced XHP50 from 7m away there is always some area of the reflector filled with the dark cross, in the beam center and anywhere in the beam. What you say about going slightly to the diagonal makes sense, but have you observed this? In my experience with the four die LEDs it is always like I describe above.

Oh I see, so 12A 6v.
I haven’t tested my XHP70.2 in a reflector yet, but from the modded GT beamshots there is pretty much no cross visible.
Anyway, I’m just concerned about lux, not how visible the cross is.

Since the light will still hit the entire reflector the lux will be the same if you measure it where the die projection is.
As you can probably imagine, if you just take a single die LED and move it a bit to the side you will still get the same lux because the LED has the same surface brightness and the reflector has the same area.
It’s just that the spot with the highest lux will be slightly over to the side off centre.

Well, sometimes our imaginations are wrong. :wink: I’ll just let you try it out.

:+1:

PARTS LIST

The original LED chosen for this flashlight was a Osram Synios P2720 KW DMLQ31.SG, hence the name of the flashlight. Unfortunately a flaw in the testing method made this LED appear very high eprforcming, despite it not actually being that great. For this reason, the Osram Oslon Black Flat is still the highest intensity LED available currently at ~260cd/mm^2. Tests of this LED can be found here: LED Test The price of one LED is $4 CAD.



The driver that will be used to power the LED is a 27.5mm MTN-MAX 1A-6A Buck Driver from Mountain Electronics. This driver can take 5V-18V making it perfect for the use of 3S lithium batteries. It will also be configured to output 6A just like the OptoFire which is about the maximum that the Black Flat LED can take. The driver costs $26 CAD.



The reflector is the most expensive part of this flashlight. It is made out of electroformed nickel plated in aluminum, a very large and high precision optic. There were several options for this reflector, such as 1 or 2 inch bottom hole and 11 or 13” diameter. In order to maximize the available area, the option with a 1” hole was chosen. Even though the 13” option would also increase the area, the LED only emits light in a 180 degree hemisphere so anything over the 12” mark would be unused area. A $100 USD fee could be paid to have the 13” reflector cut down to 12”, or an almost identical reflector could be purchased from Optiforms, the P76. The P76 is 12” diameter, and has an even smaller center hole at only .75”, however this reflector costs almost double what the Phoenix reflector costs, so for price/performance purposes it will not be purchased. The $100 option to cut down the Phoenix reflector will also not be taken in order to fit the reflector more easily inside the aluminum tube.

Overall, the 11” reflector will still perform extremely well and collect almost all of the light (96.5%) and the price is very reasonable compared to the competition. One reflector costs $402 CAD.



No radiators or fans will be used in the cooling of this flashlight, thanks to the aluminum body design. In order to transfer the heat efficiently from the LED in the middle of the reflector to the sides of the searchlight, liquid cooling will be used. Copper tubing will be used to transfer the heat from the fluid to the aluminum body. 25 feet of 3/16” copper tubing cost $42 CAD.



To move the fluid through the loop, a compact self-priming pump will be used. The pump has a flow rate of 1.3L/min and a pressure of 0.3MPa which should be enough for this small loop. The pump costs $16 CAD.



Flexible FKM tubing will be used to connect all the cooling components such as the LED block, pump, and copper coil. This tubing has a 1/4” outer diameter which will fit perfectly below the aluminum arms holding the LED block. A 3m length will be more than enough for this project as well as future projects. The tubing costs $31 CAD.



Small tubing barbs will be used to connect the tubing to the LED cooling block. The barbs have an outer diameter of 5mm which will fit tightly inside the tubing. The set of 10 barbs will cost $6 CAD.



The thermal compound used between the LED MCPCB and cooling block will be liquid metal. The three best options are Coolaboratory Liquid Pro, Coolaboratory Liquid Ultra, and Thermal Grizzly Conductonaut. Liquid Pro has better thermal conductivity than Liquid Ultra, however Liquid Ultra was made specifically because the Pro was very difficult to apply. Conductonaut on the other hand, has almost identical thermal conductivity as Liquid Pro, but is as easy to apply as Liquid Ultra. For this reason, Thermal Grizzly Conductonaut is the best of the three options. This liquid metal paste also has a higher thermal conductivity than most solders, and makes it much easier to assemble and replace the LED in the future. One gram of liquid metal costs $22 CAD.



To protect the reflector, a custom diameter acrylic lens will be used. The lens is 3mm thick and will be slightly below 12” in diameter. This ultra-clear AR coated acrylic has a light transmission above 97%. The custom size lens will cost $54 CAD.



The large LiPO batteries used in some of my other projects will unfortunately not fit inside this compact searchlight. In order to keep this a portable “flashlight” it will still require an internal battery. The Shockly 5500mAh 26650 cells are extremely popular for their very high capacity, more than the rated 5.5Ah, and discharge rate. Six of these batteries will be used in a 3S2P configuration for a total of $60 CAD.



To make the batteries easy to remove for replacement, shipping, or using external power, six 26650 battery holders will be used. The battery holders are sold individually, so six will cost $9 CAD.



A custom cooling block for the LED will need to be machined by CNC. The piece will be made outof solid copper for the best heat transmission. The CNC job costs $133 CAD.



The body of the searchlight will be made out of a 6061 aluminum extruded tube. A 12” outer diameter witha .25” wall will be very rigid to protect the reflector, and will allow 1/4” of space around the reflector for protection. This space will also be used for tubing and wires to pass from behind to the front of the reflector where the LED is. The tube will be custom cut to 4” long and will cost $192 CAD.



1/4” thick 6061 aluminum sheets will be used for both the bottom of the searchlight as well as the ‘arms’ that will hold the LED and cooling block. The 1ft x 2ft aluminum plate will cost $72 CAD.



A blue LED voltage display will be used to monitor the battery’s voltage, even though the LED driver also has low-voltage protection. It is always useful to know how much battery there is left, and if it is getting close to finished. The voltmeter costs $10 CAD.



A blue LED temperature display will match the voltage display and also give information about the LED’s temperature. The temperature meter costs $17 CAD.



A compact black handle will be attached to the outside of the searchlight for easy portability. This rubber handle folds flat to take up less space when being transported, such as in a backpack. The handle costs $4 CAD.



The sheets of metal will be cut using a waterjet due to the complex 2d shapes and accuracy needed. The waterjet time costs $18 CAD.



For the highest durability, Type 3 hard anodizing will be used on the flashlight body. This is the most durable and scratch resistant coating, and will give the flashlight a nice dark grey appearance. The anodizing costs $150 CAD.

Part Brand Supplier Price (CAD)
LED Osram Mouser.ca $4
LED driver Mountain Electronics Mountain Electronics $26
Reflector Phoenix Electroforms Phoenix Electroforms $448
Copper tubing Connect Amazon.ca $42
Water pump Uxcell Amazon.ca $16
Tubing Evolution Horizon Hobby $31
Tubing barbs Uxcell Amazon.ca $6
Thermal compound Thermal Grizzly Amazon.ca $22
Lens Flashlightlens.com Flashlightlens.com $54
Batteries Shockli Mountain Electronics $60
Battery holders ? FastTech $9
Copper CNCd block Metal Supermarkets Metal Supermarkets $133
Aluminum pipe MetalsDepot MetalsDepot $192
Aluminum plate Metal Supermarkets Metal Supermarkets $72
Voltage display ? Amazon.ca $10
Temperature display ? Amazon.ca $17
Handle Reliable Hardware Company Amazon.ca $4
Waterjetting N/A N/A $18
Anodizing Altech Anodizing Altech Anodizing $150
Total $1315


Some small objects such as wire, solder, connectors, and switches have been omitted. Cost of shipping and taxes were not included into the budget. USD prices were converted to CAD on the date of writing. Current prices may differ. Certain products such as the copper block machining had a discounted rate for me as a member of the university. A rough estimate of the total cost including all small components, taxes, and shipping is $2000 CAD.