Lumiled's Rebel: some protos

[McGizmo]

Hi guys,
I have not had the time or energy to get cracking on some Rebel ideas that I had planned on but I am now making some headway. Wayne Y. showed a light he made using a section of his Rebel "flower" thin PCB and three of the Rebels behind the McR-10R reflectors. I had intended on doing a single Rebel light but just couldn't come up with a quick and easy host or solution so I took a path inspired by Wayne and did a Rx3 proto:

I had commented previously that I noticed that the spill and spot distribution of the Rebel with McR-10R is very similar to that of a LuxV in the SF KL4 head. I decided to do some comparisons between this proto above and a L4 I have on hand.

This Rx3 proto has a New Bad Boy driver and it's constant current output to the series string of Rebels is 88 mA on Low and 465 mA on high. I also measured the current into the converter with it powered by my power supply set at 6 V. Input currents: .16 A on low and .89 A on high. The SF L4 pulled .93 A from the power supply.

I introduced both lights to my 6" integrating sphere and measured the following flux:

Rx3 - Low - 68 Lumens
Rx3- High - 286 Lumens

SF L4 sample - 85 Lumens

I would guess these measures are within 10% of reality.

Some beam shot comparisons:

I am very pleased with the flood beam the Rebel provides behind the McR-10R. The transition from spot to spill is not too high in contrast, IMHO and this gives you the similar "wall of light" many enjoy with the SF L4.

I think there is some real potential with clustering some of these Rebels and I have a "flower" thin PCB presently loaded with a red Rebel in center and 6 whites surrounding it. I have a rare SF proto that is an E3 with turbo (M3) head that I believe will make an interesting host for this array but the geometry of the sink and converter looks to be quite a PITA to me. :green: It will be a good test bed for proof of concept on a couple ideas but I don't look forward to the floor scraps I expect to make in the process. :shrug:

One way or the other, I plan on sharing some additional cluster ideas here in this thread.....

 

[matrixshaman]

Wow - that's a ton of light compared to the L4 and the L4 is certainly no slouch in putting out a wall of light. The one Rebel I've got in a smooth reflector puts out a surprisingly narrow long throwing beam although it's got quite a large circle of spill it's a very fast dropoff from the hot spot and enough artifacts to drive Monk (TV detective) crazy. Those McR-10R's look like they'd love to be in a new yet to be released McGizmo McMini Rebel

 

[jch79]

Very cool Don! :naughty:

With 465mA on high, I'd love to see a runtime test of what this light looks like... as well as the New Bad Boy driver!

:wave: john

 

[Nebula]

Wow (snip) Those McR-10R's look like they'd love to be in a new yet to be released McGizmo McMini Rebel

:thumbsup: Excellent idea.

 

[griff]

I'll take it..... I call Shotgun.. First in line ..Pretty Please....

 

[koala]

COol one Don, the Rebels is really superior. Is there enough space to fit 4 or 5 in there? They make the light look so big .

 

[yclo]

Excellent, I picked up a McR-10R along with my last order at the shoppe and was thinking about how nice it would be to fit 3 into an E2e head. The beam distribution is amazing for its size.

Amazing how led technology has improved in the last couple years!

Maybe put in a red emitter for a broader spectrum of illumination?

-YC

 

[Manzerick]

wow!!

That does look very nice!

 

[greenstuffs]

This is just like computers double perfomance every 18 months

 

[luxlunatic]

!!!

 

[Nedtheshred]

Wow, very nice Don.

 

[McGizmo]

Well I just spent 2 hours at the lathe turning a nice bar of 1.5" Al into scrap. :green: I was able to glean a good lesson from it so tomorrow I might be successful in making a part instead.

The 3xR cluster just doesn't make it in a 1" OD head unfortunately. The 27 head might well host 4xR but it would require custom PCB work or mounting of 4 descreete PCB's, neither of which I am interested in messing with. The other consideration is battery supply and drive current. Some of these clusters would be better suited with a larger host with greater thermal capacity as well as current capacity.

The single Rebel with McR-10-R would be well suited for a AAA sized light but I leave that to those who have properly sized drivers and the inclination.....

 

[supes]

!!!

+100% I love my L4's output but this light is simply amazing, pure shock and awe to me! I've been impressed by the new LEDs but this is giving a tingling feeling all over, like when I first discovered the L4.

So that Rx3, is it running 2 X CR123a batteries like the L4?

Keep up the good work, Don! Time to save the pennies.....

 

[kenster]

NICE! The L4 has a great beam so it is wonderful to see your Rx3 beam is so similiar. Well, only on low. High at 465 ma looks incredible but did you try an extended run with high? Just wondering about heat issues. I haven`t got my hands on a Rebel to play with yet so..... I know nuttin?!?!

But I do have a host that looks like a nice candidate for an Rx4 light build. Too bad I only ordered 3 Mcr10. BACK TO THE SHOPPE!!!!

Ken :wave:

 

[Kiessling]

This is the death wound for the LuxV lights it seems
VERY NICE !!!
I am drowning in drool ...

bernie


P.S.: there's no better thing on CPF than "proto show time" at Don's place :wave:

 

[LED61]

How about a 27 clickie with Rx3 ?

 

[billybright]

Don, looks like another Great light! :thumbsup:

 

[McGizmo]

Hi guys,
Let me throw out some random thoughts in response.

The Rx3 could be hosted in a light about the same size as the L4 and possibly a bit shorter. The OD of the L4 head is greater than 1" and I suspect a Rx3 head would be similar in OD but not taper back to the 1" bezel cap of the L4.

In terms of run time and heat output, in my comparison here, the Rx3 pulls slightly less power on high than the L4 and it puts out considerably more light. It should have similar run times then and produce measurably less heat due to LED efficiencies giving you light instead of heat. The relative efficiencies of the converters would also need factoring in.

We measure the light output in terms of lumens but in regards to thermal considerations, we might be better off measuring the relative light output in these lights in radiometricpower instead. An increase in X miliwatts of radiometric power would mean a decrease in X miliwats of thermal waste.

I have no plans at present relating to the Rebel but R&D and prototyping provides the building blocks to be considered in future ideas. I believe the advantage the Rebel has over other LED's is its small footprint. To exploit this advantage seems to imply a small host or small overall cluster of multiple Rebels. That a small optic lends itself to a flood type of divergent collimation needs to be considered and exploited as well.

Back to the thermal comparison of the Rx3 VS L4, the lumen maintenance of the LuxV has always been poor and elevated temperature is a primary factor. The Rebel is reported to handle higher JT with better lumen maintenance so this is all on the plus side for the Rebel here.

I have drawn up a "L" version of the McR-10-R but I haven't ordered any since I have no need for them. Heck, I don't have a need for the McR-10-R at present but curiousity got the best of me and it may on a L version as well. A deeper reflector will take some of the perimiter spill and center weight it into the spot portion of the beam; consider a 27L beam VS a McR-20 beam as an example.

I am off to see if there are still some waves breaking from some south swells that have been providing wet refreshment now for more than a week. Hopefully this will put me in the proper mind to succeed in the proto based on the flower Rx7 package...... :wave:

EDIT:

Oh yeah, I forgot to mention that the Ti 27 head hosting the Rx3 is much larger and deeper than necessary and as a result, there is a 3 tiered and substantial "wedding cake" heat sink hosting the thin PCB of the LED's. This sink no doubt aids in slowing down the rise to thermal steady state as well as provides a good transfer of the heat to the Ti host. :shrug:

 

[easilyled]

Oh boy! Please bring them on

The thought of all that light in an L4-sized package is almost too
much to take.

 

[McGizmo]

Hi guys,

After a great and relaxing surf session, I was able to finish up the Rebel x7 "RedEye" proto.

The light engine was constructed completely outside of the host.

I had initially intended using a BadBoy in ES-can but the thought of one speed on the 7 Rebels was just disappointing and too limiting. From the time of the first PD and X2 converters, I knew that the X2 converter could be used in an E series compatible head but refersing the male and female thread assembly on head and pak as in the PD was the only way to go with the piston drive and I never got around to building an E-series head with a x2 converter. This Rebel x7 was now a chance to ptove out the concept and would allow me to use a two stage New Bad Boy.

The contact ring and Kilroy are smaller in diameter than the minor diameter of the thread and this allows them to be electrically isolated from the head. I should point out that such a configuration does waste up to 1//4" in "Z" length given that the only component that goes into the battery tube at the front end is the anode contact. (Consider that in the case of the Aleph LE's that the converter and can are digested by the battery tube)

Going with the NewBadBoy, I was able to set two drive levels to the series string of 7 Rebels (Red LED in center surrounded by 6 white LEDs). I went for a sense resistor that gave me 55 mA on low and 315 mA on high. The 3xCR123 of the E3 proto battery pak allowed me to go for the 315 mA high. At 9 volts in, low draws 170 mA and high pulls 920 mA from the power source.

The x7 LE was profiled to be a slip fit and mate with the head on both cylindrical surfaces as well as the 45 degree chamfer face. I bonded the LE into the head using Artic Alumina Ceramic epoxy and post cured the assembly for an hour at 140 F. The array of LED's with reflector "just" fit in this head and the head was tall enough to allow for inclusion of the NBB x2 converter.

This light is not nerly as large as one might expect. Here is a shot of it next to the Rx3 proto:

As anticipated, the inclusion of a red LED in the cluster warms up the tint noticibly to the naked eye and relative to a all white LED source, it looks more incandescent in nature but up against an incandescent source, it looks white. I noticed a deep purple back pack I have that under LED illumination looks blue but under the Rx7 RedEye, it looks purple. I want to play with this light more to see just how well it enhances color rendition but so far, I have seen a marked improvement.

In the beam shots below, the LowBeam shot is close to proper exposure with the manual settings used and the color seen in the image is close to theat the naked eye sees. In the HighBeam images, the over exposure results in a washed out image which is not what ones eyes sees as we can adapt to the higher intensity (as could the camera had it been left on auto)

In the combined image below, both shots represent equal exposure settings but unlike the combined images of the Rx3 VS L4 above, I did not increase the saturation level in PhotoShop to give a better feel of what the eye actually sees. In both cases of the Rx3 and Rx7 on high output, the entire field of view in the yard is easily seen and to a much greater extent than the photos would suggest.

I grabbed a proto SF regulated incan light I have which is like a steroid version of the M6 but with 9xCR123 cells driving the M6 HOLA lamp. It has a seemingly brighter but much smaller spot beam than this Rx7 but beyond the smaller and brighter spot of this incan, the Rx7 has much greater intensity. In a white wall comparison, the tint of the regulated HOLA looks greenish compared to a white tint from the Rx7.

All in all, this has been very interesting and I sense more questions and considerations will arise. For one thing, this cluster is being driven at a little less than 1/3 of its rated power. I would imagine that in a host like a big mag where more thermal and power capacity were available and the LED's driven harder, a very impressive output might be viewed!