The SYNIOSBEAM - CFT90 recoil thrower

Almost all of them are 2d exported images from sketchup (free)
The last one is a real render done in solidworks.
I made a solidworks model after sketchup because I needed a way to make the drawings for the waterjetting and assembling the components more accurately.

No fins inside the water block?

Way too complex to do, also not necessary for such low power LEDs.
Making microfins in there would probably make the blocks cost like $1000 rather than $100.
It might have helped if I decide to put a CFT90 in there later, but the project was already nearly 1.5-2x more expensive than I wanted it to be.
I could have probably gone with aluminum instead of copper to lower cost and still get decent performance, but what’s done is done :slight_smile:

You don’t need microfins. I remember people were cutting them in CPU blocks with a hacksaw. Finer ones would work better here, but it’s not something very complex to make…

Why does the reflector need a hole?

He does not need it, but the manufacturer delivers it.

While microfins or channels do reduce temps by some degrees, they are not needed. especially not if you are cooling just 10-15W.

I know that in the beginning of PC Watercooling we were cooling heavily overclocked CPUs which were also over volted to oblivion with simple aluminum blocks with just a round S-channel.

@Enderman
There is no reservoir visible in you CAD pictures. While bleeding is surely possible without one, it makes this task so much easier.

So awesome!
What about a handle or a tripod?

I like the space optimization, especially the copper pipe wound dissipation surface. Do you plan to secure the batteries in the carriers (i.e. velcro strap) to ensure they do not pop out if they encounter a sudden jolt?

The block is 1” diameter. That channel is only 1/8” wide.
Any fins that fit in there would be called microfins IMO.
The smaller the details, the more expensive the machining cost gets :frowning:


It’s part of the manufacturing process, can’t get rid of it.
It also doesn’t matter since that part of the reflector is blocked by the LED MCPCB and cooling block anyway, so it would be unused even if there was no hole.


Correct :slight_smile:


Yeah I wanted the light to be usable in any orientation so I skipped a res (which could introduce bubbles) and will just fill/bleed it using an external container, then close the loop.
I’m not sure how much of an issue evaporation is going to be, but I guess we’ll find out with this prototype :stuck_out_tongue:


Thanks!
It will have a handle on the side as well as two 1/4” tripod mounts, one on the side and one on the bottom for vertical or horizontal use :slight_smile:


That’s something I will need to consider, it would be bad if one fell out and hit the reflector.
I just checked and the batteries fit extremely tightly inside the carriers, so I don’t think there’s any risk of them falling out, but for any long plane or road trips I think I would recommend removing batteries.
The battery carriers do have a small slot on the side that seems like it would fit a zip tie though, so maybe a small design modification by adding small washers between the carriers and the base plate will allow a zip tie to go around the carrier and make sure the battery is 100% secure :slight_smile:

With a reflector you have some axial averaging going on. If you put a pinhole in front of the reflector, like the method you used to focus your lens, you will see that the projected image of the LED rotates as you move the pinhole around the reflector. So at every point in the beam there are some areas of the reflector that are reflecting the dark cross (in four-die LEDs). This is what causes the beam intensity to be reduced, even outside of the donut hole. Also, I think for most LEDs the die luminance is a bit less near the edges, and averaging these areas will further reduce the beam intensity. On the other hand, with a single die in a reflector the center of the beam consists of images of the center of the die so you get high beam intensity.

The projected image of the LED does not rotate.
It gets skewed/stretched depending on what angle the reflector “sees” it from, but it does not rotate.
With a lens, all the points are almost perfectly in front of the LED which is why you get a clear die projection, all the squares remain square.

I just tried it and it does rotate.

Edit: as you move in a circle around the LED.

I like your style Enderman. Very exciting project. I probably missed it, but are you having the reflector custom made? If yes, maybe they can quote a price for say 5 or 10 units. Maybe it would be low enough for some members to want to by their own reflector so that they could make their own version of this monster thrower. Of course you would want the manufacture to ship the individual units straight to the purchasers.

If someone wanted to go with a laser and remote phosphor, that center hole would be a wanted feature. Powerful blue (445nm) lasers are cheap, but they don't emit a pinpoint of light. Violet (405nm) would emit a nice pinpoint, but last I checked there are not any cheap high power versions of those available.

If you’re going to be using an XHP70.2, don’t forget that all 4 pins are brought out on the chip, +/~~ A and +/~~ B.

So you can have one driver power the A LEDs and another driver power the B LEDs.

Was gonna suggest…

If you want a nice stiff and fairly lightweight frame for the reflector, you can go back to the cake-pan, but use what’s called “gorilla hair” and epoxy resin, and make your own fiberglass casing for it.

I know people who are into car audio, and instead of using ¾” thick wood slabs for speaker-enclosures, would use thin plywood, then GH+ER to build-up the wood and make it rock-hard, no vibrations.

Messy as Hell, though. Staple/Glue the sheet(s) of GH to the wood (depending how thick you want it), then get it sopping wet with ER, and let it all dry.

Dunno if you have anything close to you that can supply it, but something like this:

https://www.clausenautobody.com/proddetail.php?prod=claw-glas-mat

https://www.clausenautobody.com/proddetail.php?prod=gorilla-hair

And the resin, hardener, anything else you might need.

If you want the outside to still look nice and slick, apply the stuff to the inside of the pan.

I’m not sure if I understand what you mean? As you move a point around the optic the LED always remains right side up:

If it rotated there would always be a flat side towards the center and the result would be a round spot when they all overlap.
As you know, the spot isn’t round but actually square.
If you put a quad-die LED behind a lens you get 4 individual die projections.


It’s a stock option from Phoenix Electroforms, but they didn’t have any in stock so they did need to produce a batch for me.
Even for 10 units I doubt you would get a deal below $250 USD each, especially since they need to add some sort of mount to it.

With remote phosphor there are two configurations, through and reflected.
With the ‘reflected’ option, the laser would be behind the reflector and shine through the hole.
With the ‘through’ option, the laser would be in front of the reflector shining through the phosphor.
According to Crytur the ‘through’ option yields higher intensity, but I have not yet found any test data for laser-phosphor intensity.


You’re right :slight_smile:
Would be great if the CFT90 had that feature huh…


Composites are a good idea, but I think the choice to go with aluminum was good because mounting things inside is far easier when you can simply drill and tap a hole for a screw anywhere.
In the prototype revisions I considered carbon fiber for the front spider but decided against it because it would just be added complication to cut and mount.

Yes, but I thought we were talking about reflectors. I would think your recoil reflector will behave similarly to my C8 reflector in that it will average the dies and dark cross. Just trying to save you some time and energy because you will be disappointed with the lux of a four die led in a reflector if you are expecting 100+ effective cd/mm^2.

How about mounting the water block directly to the lens, leaving just 1 arm with colling/power connections to reduce the blocked surface?

That was my point, that you could even keep the (flimsy) cake-pan as the inside or outside, and just coat the outside or inside with the ad hoc fiberglass.

But a nice thick coating of fiberglass is hard without being too brittle, so you can absolutely drill holes in it just fine. (That’s how they mount crossovers, spring-terminals, etc., on the speaker boxes.)