4up MC-E proposal / idea.

[marcopolo]

I have a few ideas on a new 4up MC-E bicycle light. Purely speculative. My XR-E dual 4ups are excellent and more than enough light but forget common sense - it's now an obsession! There would be a major benefit going to 16 emitters (4 MC-E): At low drive levels you are getting hugely efficient light from 16 emitters.

So will a new slug/case design be able to pull enough heat away? The case can be a bit larger than the XR-E 4up standard footprint as the new hipFlex driver is 1.4in diameter but I'd want to keep the LED footprint the same size to use the 4up XR-E optics.

Also instead of a slug/case interferance fit I would just machine out of solid rod and have the front face of the rod as the emitter interface - milling off material to leave 4 square pills on which the bare emitters would be fixed to. The wires could be routed round the channels cut (as there would be no MCPCB). Cut threads on the side and tread on a lager diameter front to hold the optics on. This would surely pull heat away faster than a slug to case fit.

Any thoughts on the matter?

Marco.

 

[2xTrinity]

http://www.dealextreme.com/details.dx/sku.1915 -- This quad optic, meant for the XR-Es, will also accept MC-E's. I once made a triple MC-E light that used a one-piece triple reflector. I later tried out a triple optic with good results.

With multiple MC-Es -- you can actually negate the little "dark cross" in the hotspot, formed by imaging the LED which has 4 separate dice. When you have multiple closely packed MC-Es, even in smooth reflectors, the hotspots all overlap each other with a slight angualr offset. As a result, the "bright" spots from one MC-E manage to fill in the "dark" regions from others, and you have a hotspot (more like a "wall" than a spot...) where the gaps in output that would be totally objectionable with a single MC-E, are largely masked by having multiple.

Since you have the ability to machine a heatsink, I would first start with a stock optic that will work -- like the one I linked, then design your custom piece around it. That will eliminate some headache of trying to use multiple separate optics or reflectors.

 

[marcopolo]

So you think it would be more sensible to go with the MR16 size, rather than the cutter MR11 4up? Maybe MR11 is wishful thinking for so much heat.

I've just found some nice 6086 2in diameter 100mm rod on ebay. Emitters are available too ( might go warm instead of WH) so it's just a case of waiting for availablility on the hipFlex.

 

[Geir68]

People are struggling to get rid of the heat in dual/tripple builds, and you are planning to squeeze four of those into a MR11 housing!! You need roughly 1 in^2 surface area for each watt of heat, so even a MR16 housing may be to small. You will need lots of fins.

How are you planning on driving it with the Hipflex? Four parallel connected MC-E(serial stars) at 2.8A would suck 35W out of the battery.
This would be the new leader in the ongoing lumen war among the bicycle lights...

 

[marcopolo]

You need roughly 1 in^2 surface area for each watt of heat, so even a MR16 housing may be to small. You will need lots of fins....

Yes I was wanting to try out a different technique - cutting the slots along the length of the tube rather than round the circumference. Using a bandsaw/large hacksaw with 1mm thick cuts (2mm deep), 3mm tube wall thickness and. 140mm circumference = 70 cuts down the length of the tube all 2mm deep. Say the tube was 60mm long this yields a surface area of 140mm+70*2mm [fins]* 60mm = 16800mm^2 = 168cm ^2 = 26in^2 ie 26 square inches of surface area - should be enough for Scotland here where my winter riding is regularly -5degc - 5 deg C air rushing past me at a not too slow speed!! My XR-E's dropping 14W on max to my case are barely warm while riding. Plus they are EMITTER-MCPCB-SLUG-CASE whereas i'll be using for the MC-E's EMITTER-CASE - just 1 thermal interface as oppesed to 3 thermal interfaces - far better heat draw.

Four parallel connected MC-E(serial stars) at 2.8A would suck 35W out of the battery.
This would be the new leader in the ongoing lumen war among the bicycle lights...

Remeber 35W is the top line and the hipFlex has different power settings. My dual XR-E 4ups are only on full for fast downhill technical sections. Usually they are on 1/2 or 1/4 power. 35W from a 4S or 5S lipo is only around 2-3A which is very easy for a LIPO battery. let the lumens battle commence!!!LOL.

Marco.

 

[marcopolo]

OK, i've just purchased a few of these on ebay to play with; 2in OD solid aluminium:

http://tinyurl.com/acmugu

Hopefully should have something machined up within a few weeks (work permitting!!!).

 

[2xTrinity]

So you think it would be more sensible to go with the MR16 size, rather than the cutter MR11 4up? Maybe MR11 is wishful thinking for so much heat.

I've just found some nice 6086 2in diameter 100mm rod on ebay. Emitters are available too ( might go warm instead of WH) so it's just a case of waiting for availablility on the hipFlex.

The MR-16 quad-optic size is the smallest reflector size I would use with the MC-E. Even this will be a "wall of light" with little throw. Any smaller and I think there may be problems with much of the light from the MC-E not making it in at the right angle to undergo internal reflection (for a TIR optic, that is), or simply not being focused enough (aluminum reflector)

Yes I was wanting to try out a different technique - cutting the slots along the length of the tube rather than round the circumference. Using a bandsaw/large hacksaw with 1mm thick cuts (2mm deep), 3mm tube wall thickness and. 140mm circumference = 70 cuts down the length of the tube all 2mm deep. Say the tube was 60mm long this yields a surface area of 140mm+70*2mm [fins]* 60mm = 16800mm^2 = 168cm ^2 = 26in^2 ie 26 square inches of surface area - should be enough for Scotland here where my winter riding is regularly -5degc - 5 deg C air rushing past me at a not too slow speed!! My XR-E's dropping 14W on max to my case are barely warm while riding. Plus they are EMITTER-MCPCB-SLUG-CASE whereas i'll be using for the MC-E's EMITTER-CASE - just 1 thermal interface as oppesed to 3 thermal interfaces - far better heat draw.

For a bicycle light in open air, this makes sense -- you want the air to flow between the fins when you are moving forward on the bike, so the slots should run down the length of the device. Also, thick more widely spaced fins makes sense here as well, whereas for example computer heatsink fins use extremely small fins as they are forcibly cooled by a ducted fan (usually), not just open air. I'm interested in seeing the results.

 

[trout]

I have had fantastic success using the stem and bars as heatsink
for a triple MCE light and am also planning some thing similar with 4 MCE
now the Hipflex is close
The stem and bars just suck the heat away and there was the option of putting a small amount of coolant inside the bars which as they are constantly moving would move the heat also.

The maxflex driver also lived in the hollow stem but I think the Hipflex is to big this time so I need to think of something else.

 

[jtr1962]

I would think the biggest problem would be powering it, not heatsinking it. 4 MC-Es is going to require up to 40 watts, even without overdriving. To get that amount of power for any length of time will require a huge battery pack. Right now I'm in the planning stages of making a 4 XR-E bike light. Biggest problem is the battery. Driving the emitters at 1 amp each will use 14+ watts. I'll be lucky to get much over an hour runtime from 8 Eneloops. This is about the largest pack which can physically be mounted in self-contained light (i.e. batteries and emitters in the same enclosure). Most of the time I'll be running at lower power to get 3+ hours from the batteries. I can't comprehend trying to power 4 MC-Es at full bore for any reasonable length of time. Even going to a water bottle cage mounted battery, I'm not sure if you could get enough battery to physically fit. Unless of course you're happy with only 15 or 20 minutes run time at full power.

 

[marcopolo]

Trout: really great idea - I saw you mention this idea in your other thread. I'm keeping a close eye on your threads too see how your "throw" is.

JTR1962: No problemo. New NiMH but especially Li-PO will handle it no problem.

5 cell LIPO 18.5 V, 3200mAH = 61WattHours and weighs only 500g and is 42x42x132mm.

This will run this big boy at full power for well over an hour even with temperature dip. You uare missing the most important point though. You only need to run at full for the fast blasts (I'm mountain bike some moderately technical downhill). For most of the climbs / level trails you are running medium or low power giving you stupid long runtimes. I don't think i've ver done a ride where I was on technical/fast descent for longer than 30mins. Remember these are NOT road lights. At 3000+ lumens you are going to be arrested by traffic cops! Or have aliens from AlphaCentauri punching out replies to you.

2Xtrinity:

Even this will be a "wall of light" with little throw

- I may have trick up my sleeve but i'm going to wait until ive got mounted LED's before I spill the beans. My other hobby is telescope making so I know a small bit about optics, excuse the spam but here is my telescope making website: www.geocities.com/telescopiman

Marco.

 

[jtr1962]

JTR1962: No problemo. New NiMH but especially Li-PO will handle it no problem.

5 cell LIPO 18.5 V, 3200mAH = 61WattHours and weighs only 500g and is 42x42x132mm.

Sounds good. 18.5V is enough to hook the MC-E dies in series. If you'll only be using full for fast descents then you have adequate runtime.

Remember these are NOT road lights. At 3000+ lumens you are going to be arrested by traffic cops! Or have aliens from AlphaCentauri punching out replies to you.

Very true. I'm worried enough that my 4 XR-E setup will attract attention on the road. 4 MC-Es might have people calling in about UFOs. 😱

 

[trout]

Marco Have you thought about 16 XPEs then
they would fit in the same space and throw

 

[marcopolo]

I don't know Trout. 16 individualy LEDS - thats going to cause problems all of is't own!! I really like your intergrated design but it's too bespoke to be useful to other people and I couldn't sell it on / make for others very easily. I don't actually plan on selling, but sometimes my projects have a habit of costing too much and I have to punt them on to fund the next project and sometimes friends ask me to "do one for them"!

With 16 individual LEDS though you could do something like a light bar - LED directly on the handlebar face in a row, maybe with some sort of perspex front.....but thats another project!!

If I wing it with the 4up MC-E there is always the option of using different current table if there is a heat problem or playing with some custom made optics if there is a throw problem. I'm absolutely determined :thinking: to crack this throw problem the MC-E's inherantly have and if I manage i'll post full details.

Marco.

 

[trout]

Marco this is what I have been chewing out of solid bar for a while now
I was not going to post it till finished but wanted to share.

I am making the bar clamps also from ali as I have found that even those attached to the bars helps greatly heat wise .

I am using the ledil CMC RS / SS and awaiting a HipFlex

power wise I am going to use a 3 ah 18 v Li ION Makita drill battery .
it just fits in the bottle cage .

 

[marcopolo]

Jesus H Ch*^%!!!! Trout, I have a wife and a kid but you make me want to spend more time in my garage.!! LOL. Incredible work.

 

[trout]

I have just copied this from my post on MTBR




This has been ongoing for a while now waiting for a hipflex and got put aside while the XPE lights were done , and Deestas sort of developed .
The Hipflex will be available very soon so it was time to get it ready to recieve it.
The housing was turned from a 38 mm solid bar for the fins and then the cavity milled out
I am really supprised that my mill has kept going after all the abuse it has had.

Then I glued 4 of the CMC holders together with some plastic epoxy these were then used to get the spacing correct for the 4 MCE M bin
Each MCE was to be wired paralel and then the 4 mce in series.
I milled 2 grooves to isolate the mce contacts and for the wires to run in

Then it was mix some AA epoxy and apply to the back of the MCEs and fit them into the housing
The holders did not press on the MCEs so I had to replace the optics to get some pressure on the leds while the AA set.

Then a quick test on my Nflex test rig it all lit up great .
and this is the Hipflex house all ready , they are a bit large 35.6 mm dia
I have left this oversize and will trim it down whae I see how much space it requires
I am going to have a remote switch so no space needed just cable entries .

Next we come to the bar mounts this is using the bars as heat sink too so they were milled to slot in to the fins for a good contact and will be bolted and AA to get good transfer.

It is just waiting for a Hipflex from George and it will be up and running
It does seem to be pretty bright running off the Nflex @ 1000ma
so each led is seeing 250 ma
and the VF of the lot was 11.67v
How many lumens for 16 emmitters @ 250ma
I will be tempted to get a beam shot even @ 250 ma
Oh The optics Have been chosen because they were the only ones I had left.
and are
1 CMC-RS
and 3 CMC- SS


and a Beam shot with all emitters @ 250 ma

 

[Illum]

:wow:

just :wow:

thats all I can say at the moment

 

[Gaereth]

Great new light Trout.

Can someone explain the wiring on those to a noob? I am trying to wrap my head around all this new fangly electrogigity stuff and the few brain cells I have left keep screaming for mercy.

 

[TONY M]

Trout you are the man! :twothumbs

Super work. :bow:

 

[trout]

Thanks guys .

I have been very rude and please accept my apologys Marco for thread hi-jack.

It was just coincidence my light nearly finished , when you posted this thread and in my excitement I just jumped in with both feet.

Very sorry , Chris