Strongest beam ever :)

MaximeH · 2018-05-18T16:59:09.000Z

Hello to everyone !

I'm trying to build a led beam, the same as moving heads on stage lights. It's for a light art installation.This is my job, and you can have a look at collectif-coin.com

(Ok. It's not a flashlight, but you guys know how to build throwers. Right ?)

Objectives:

- very parallel beam (0° - 1.5°)

- very strong, very white beam

- very cold

My first attempt is with a CFT-90-W, an aspheric lens from Optolife (A380 : diam=120mm, f=161.5mm), and a cheap condenser lens taken from a led pin spot.

Above you have a beam from a Martin MH3 (with a philips 5r inside :)on the left, and my beam on the right.

In real life, the results are not so good. My beam is not "white" enought. You see throught it.

I'm pretty sure I have to find the right first condenser lens, but I don't know which one could fit.

Somebody to help me ?

How can I throw a maximum of light to the 120mm lens ?

Thank you !

zak.wilson · 2018-05-18T17:04:50.000Z

The MH3 produces 8,786,000 candela. That’s… a lot. Matching it without similarly fancy optics might be tricky. The BLF GT makes about 1.2 Mcd from a similar lens diameter, though it’s using a reflector.

Enderman · 2018-05-18T17:31:40.000Z

There is no LED that will give that much candela from that optic.
The technology isn’t here yet.
Also a 0 degree beam is impossible, unless you round down below 0.5.

lawallac · 2018-05-18T17:40:00.000Z

What current are is your driver supplying the CFT-90? Are you using a wavian collar on this because you need light recycling?

Enderman · 2018-05-18T17:56:15.000Z

lawallac:

What current are is your driver supplying the CFT-90? Are you using a wavian collar on this because you need light recycling?

Even with a collar and a perfect focal length lens the most he will get out of it is ~4mcd and 2500 lumens.
Moving heads often do 10-20mcd and about 20000 lumens.
if he wants something that makes a similar beam he will need to use short arc.

zak.wilson · 2018-05-19T01:17:40.000Z

This moving head is a bit under 9 Mcd. 4 Mcd might be close enough visually when it comes to beam effects like this.

Enderman · 2018-05-19T01:26:51.000Z

zak.wilson:

This moving head is a bit under 9 Mcd. 4 Mcd might be close enough visually when it comes to beam effects like this.

Not just candela, also lumens matter.
If you have a 1mm^2 LED making 1000 lumens or a 2mm^2 LED making 2000lm they will have the same intensty and cd/mm^2 giving the same throw (assuming same polar distribution and other stuff).
The second LED will have a beam 2x brighter (initially) because of double the lumens.
The down side is that the beam will diverge more too due to the larger LED area.

ma_tumba · 2018-05-19T11:11:16.000Z

What would the light scatter on? Just air with natural dust? Are you allowed to use smoke to make the beam denser?

MaximeH · 2018-05-19T11:57:41.000Z

Thank you for all your answers !

Iawallac:

My driver is supplying 4.0v / 15A. I can’t get more right now. The LED should handle 4.2V / 27A … I’m not using any wavian collar. We can’t find them anymore, can we ?

ma tumba:

It would be used in a concert hall. Rather dark and with smoke. It would be quite optimal.

Just a quick picture of the first optic. It's on this strange position that I get the narrowest beam (with the 120mm aspheric lens on top)

So what do you think I should do ?

find a more powerfull driver ? (That I'll do)
find some kind of wavien collar ?
work on the first optic ? (maybe a full spheric one ?)

djozz · 2018-05-19T12:13:52.000Z

My feeling is that you should check leaving the first optic out and see what you get with just the 120mm lens.

But if a condensor lens is used to get more light in the hotspot (mind that with it the hotspot will be wider, not brighter), in this position it does not work as a condensor, it is way too far from the led to collect a significant amount of light. I reckon it should be right on top of the led.

Zelda · 2018-05-19T12:16:07.000Z

The connected wire in the picture from first optic looks very thin to me, I estimate around 0.75mm2.
Thats perhaps not enough for suppling 15A+ current to the led? (Depends also on driver type…)

Enderman · 2018-05-19T19:04:00.000Z

Nice setup, here’s what you can do to get the maximum performance in the beam:

1) replace that aluminum heatsink with a NH-D15, you will get MUCH better cooling performance.

2) Use some MX-4 thermal paste between the heatsink and the LED (if you want to go all out, use some Thermal Grizzly Conductonaut liquid metal)

3) Use a more powerful driver. Even though the LED is rated at 27A you can get to 40-50A if you have good cooling like I mentioned above.
https://www.mouser.ca/ProductDetail/MEAN-WELL/UHP-350-42?qs=%2Fha2pyFaduiKYJ07tagxAKyrsCEh8RRLF8%2FREeUojk8YCOnKM04Pfw%3D%3D
Just an example, up to 60A power supply.

4) Remove the first optic and use a wavien collar instead, you can buy this flashlight and rip the collar out of it:

Marinebeam

Marinebeam Ultra Long Range LED Illuminator Flashlight

High Lumen, 700 Meter ANSI FL1 Beam Distance | Over 100,000 Candela

Average Rating: 5

Price: USD 129.95

5) Use an A037 lens instead of the A380. (beam will be brighter but not as tight)

ma_tumba · 2018-05-20T11:33:54.000Z

I wonder if there are “smokes” that would fluoresce under uv light? In this case you could use an uv laser instead of a flashlight. More efficient in terms of beam divergence, brightness and there could be a choice of color, too

Enderman · 2018-05-20T19:29:23.000Z

No there is no fluorescent hazer fluid.
Also a lazer has a far thinner beam making it less visible, even if it is brighter.

lawallac · 2018-05-21T17:12:06.000Z

What color temp of the led is this? I see mouser lists instock the 7800k, stage white, and even 8500k. I would expect these bluer whites to provide greater scatter and a more visible beam.

The_Driver1 · 2018-05-21T20:15:37.000Z

Enderman:

Remove the first optic and use a wavien collar instead, you can buy this flashlight and rip the collar out of it:
Marinebeam Ultra Long Range CREE RLT Illuminator

The collar in that light (probably the small one) might be too small for this LED.

Enderman · 2018-05-21T21:49:51.000Z

The_Driver:

The collar in that light (probably the small one) might be too small for this LED.

What would it be too small?
It has an inside diameter of 25mm, the LED is only 3mmx3mm.
It will have the exact same effect as a larger collar, projecting an inverted image on the die.

MaximeH · 2018-05-22T07:34:56.000Z

Thanks Enderman for all these advices ! I’m gonna test these.
Iawallac, mine is 7800k

Enderman · 2018-05-22T07:45:57.000Z

Good luck, please post results I love moving heads and strong LED beams.

Let me know if you need any help with the collar stuff.

This is the difference between A037 and A380: Optolife review - the best aspheric flashlight lens

The_Driver1 · 2018-05-22T10:13:28.000Z

Enderman:

The_Driver:

The collar in that light (probably the small one) might be too small for this LED.

What would it be too small?
It has an inside diameter of 25mm, the LED is only 3mmx3mm.
It will have the exact same effect as a larger collar, projecting an inverted image on the die.

Because sma, the only guy who actually studied them in detail here , noted this possibilty. He says that they only focus good in the center which might lead to a reduced effect with large LEDs and/or small collars.