Madness: airplane landing light in LEDs?

[IanJ]

So, I'm thinking about building an airplane. Nothing too fancy, just a little biplane. The one I've picked (a Ragwing Special II, most likely) is light weight and lightly powered, so anywhere I can save weight is worth it.

This brings me to areas where I have more experience. Specifically, electronics.

One thing I want to do is fly at night. As far as I know right now, the FAA doesn't have any requirements for landing lights for Experimental Homebuilt class planes (which is what this would be). In fact, if I recall correctly, to be legal for night flight, you don't even have to have a landing light. But I want one.

You see where I'm going with all this.

What's the best choice for emitters these days? Optics? Controllers? I will have a 12v system, just like on a car (using many similar parts). I was thinking a set of N (maybe 6?) Crees driven at 700 mA each, on a strip shooting through the prop. That guarantees a heavy flow of air, so cooling is basically a no-brainer -- put 'em on something that conducts heat, and make sure it has a couple fins. I'll never fly over about 5000 feet, so the air will never thin appreciably enough to worry about.

A miniscule amount of research suggests that landing lights typically produce 650 to 3000 lumens, more light obviously being better. How much light does a typical Cree produce? I read Fenix claiming 135 lumens, but I don't know if that's true or not (the difference between a "135 lumen" Fenix P1D-CE and my 42 lumen HDS EDC wasn't very dramatic). If I can really get 135 lumens out of a Cree, then 6 emitters with 15 degree optics (do you use optics with Crees? seems like you would, but that's why I'm asking here) would produce enough light to be practical (810 lumens), especially on a slow biplane.

Thoughts? Am I crazy?

(For reference, a typical landing light weighs in around a couple pounds, and the bulb lasts ~10 hours of operation, costing $20-100 to replace, typically. An LED system would not only save weight on bulbs, but also on wires (less current to pass), and would have a replacement time around the time the sun goes out.)

 

[RCatR]

I love running a cree through an optic! You might want to consider the optics that give a panorama beam; so that you can have light spreading to just the sides as you land. With a q4 cree you can expect 200bulb-lumens or around 140 out the front(running at 1A) if you used a finned copper heatsink(or a copper plate with holes drilled in it) and soldered the crees directly onto it you could easily dissipate the heat from running at 1A; and then epoxy the optics in place(or bolt them behind a lens)

 

[IanJ]

If I have success with the landing light, I'm very likely to next pursue a wide-n-flat beam for a taxi light (important on a taildragger, since the landing light will be pointing up at the sky when the plane is on all three wheels!). I'm also planning on Cree or Luxeon position lights (red, green and white lights at the appropriate points), because the savings on wire weight adds up.

I was actually just thinking that if I can get a fairly streamlined enclosure, I could do the wingtip lights on standoffs, to get airflow under the lights as well as over them. The wingtips have to be done more carefully, since they might be on for 10-20 minutes while sitting still, with no wind. I've seen at least one design for wingtip lights using Luxeons which incorporated an always-on fan, which is certainly another way to do it.

On the Crees, what's a good source? I see emitters on theledlight.com, and optics for Luxeons, but no indication whether those same optics will work with the (larger, as I recall) Cree. Thoughts?

 

[yellow]

while a 5 emitter setup might be too much for biking, I heavily doubt its enough for such an airplane (at 1st I thought You mean a remote controlled, there such a light would be enough).

Now, for this here, IMMEDIATELY SKIP THE IDEA OF 15 DEGREE OPTICS, they are useless for such!!!! A bright, floody light will be much worse than a less intense with throw.
Better get the 8 deg Cree ones, but imho skip optics totally. Way to go are the 19mm or 27mm Cree reflectors from the Sandwich Shoppe.
I have used the 16, 17, 19 and 27mm models and like the 19mm best, the 27 gives an even better throw but placing is a pain (and there wont fit more than one in a Mag-head, but 4 of the 19mm)

the best way to build lights are the usual: as many emitters as possible (depending of the size of the reflectors) into a MAG or similar flashlight head (see the diverse bike light pics), all parts available from the shoppe

 

[IanJ]

Yellow, the 15 degree idea came from actual landing lights, which are 16 degrees. Taxi lights are 38 degrees. It sounds wide to me too, but I figure the people making landing lights have some idea what they're talking about. If I need more emitters to throw enough light, so be it.

I wasn't aware that anyone was selling reflectors, I'll have to look into that.

There's absolutely no way I'm building anything into a Mag head. This isn't a flashlight, it's a vehicle light. It's going to put out 10x more light than any normal flashlight, with cooling requirements to match. It's going to get its own custom-made housing for sure, with special care taken to ensure water tightness and cooling (which will be an interesting challenge, for sure).

 

[Gryloc]

IanJ,

I am so glad someone is doing a project like this. The possibilities are endless when it comes to the design of the LED landing lights. I am still working on a LED headlight system for a friend's truck, but that has been going slowly. I have to be careful with the overall brightness and the tilt angle so my friend does not get pulled over. You do not have to worry about that, as long as the landing lights work!

However, I do have a few questions for you. I have been trying to look at a few small pictures of the plane that you are getting. It is very nice and looks like a fun craft to fly. What is the largest size that this landing light can be? I mean this like the total area (dimensions?) of the front end of all of the optics or reflectors. Is there a specific brand and model of landing lights that would normally work for this plane that can be used as a reference for size? This will aid in how many optics or reflectors you can even fit. What is the maximum power output of the electrical system? Is there a maximum amperage or a battery capacity that you have to work around?

I am not a pilot, but I was wondering about the beam angle, also. When landing, do you need the 16 degree spot? At what altitude do your lights have to be effective to light up the runway? At what angle do you usually descend at? How much area do you need lit? The designers of the 16 degree incandescent landing lights have the luxury of using a very high lumen bulb (with a very small emitting source) and a large reflector to redirect all of the light. With LEDs, you have to use many of them to get about the same brightness, and because of that, you have to use many separate, smaller, and less effective reflectors. Maybe 15 degrees is ideal for landing (you would know). I was just worried that with 15 degree optics, not enough light would be hitting the runway to be useful.

With my headlights, I used 18 of the Luxeon K2s for each headlight cluster. When I first got the K2s, they were acceptable for brightness, but they now pale in comparison to the Cree XR-Es and the Seoul P4s. I paired the K2s with 18 of Fraen's 27mm FHS TIR optics (with the 10 degree beam angle). I estimated that each light cluster of 18 LEDs produced about 1800 lm (when each K2 is powered at 1500mA), and with those optics (with 22lux/lm efficiency at the beam center), delivered about 40,000 lux at 1m. This estimation is conservative and based of the specs of the LEDs and optics. The beam is powerful, and bright! However, even with the "10 degree" beam, at very long distances (past ~500ft), the beam produced by one headlight starts to lose its greatness. I chose optics because there is very little spill to blind oncoming drivers. I did simple experiments involving people standing facing me, like in left lane, when comparing 27mm optics and 27mm reflectors. Reflectors might work better for you, giving you a brighter projected beam on the ground.

I was doing some thinking, and I am unsure how to get the Crees to work very well for you. The XR-Es are awesome LEDs, but because of its odd built-on optics (glass aspheric dome), you will need the correct, matching optics, or just a deep reflector. The original optics made for the XR-E is expensive, and (IMHO) the beam produced may not work since it seems so diffused and wide. I do not know the beam angle of the standard XR-E TIR optic. The reflectors mentioned before (the McR series from the Sandwich Shoppe), are made from aluminum and will cost you $16 to $18 each, depending on the size. They are very nice quality reflectors, but you can get cheaper reflectors elsewhere. I always worried about throw with these because they all have a slight orange peel.

The Seoul P4s are nice because they are slightly brighter than the XR-E, and because of the Lambertian radiation pattern, will work better with reflectors. The Seoul P4 will direct more light sideways to be reflected properly by the reflector. Optics would work better with the XR-E because the light exits out the front (within 70 degrees, versus 120degrees of a Luxeon or Seoul P4).

However, I thought that the Seoul, mated with an IMS 27mm reflector (a smooth reflector with a 10 degree beam angle), would be the ideal way to go. The Seoul P4 LED is very similar to the Lumileds Luxeon Star, so there is a better variety of optics and reflectors available for sale. The Seoul P4 does have its LED die sit 0.030in lower than that of the Luxeon, however, but in many cases, the stock reflectors for the Luxeons work just fine with the P4. Some prefer to sand down the back side of the reflector to focus it perfectly, but that is a simple task you can do if you want. These smooth reflectors are available from Future Electronics for only $1.73 each, and they are a pretty decent quality aluminized plastic reflector that will work well for you.

As for the LED, you can find them at many different prices. For example, the XR-E can be found at Cutter Electronics for $11.00, the Sandwich Shoppe for $13.00, and Deal Extreme for $6.13. Meanwhile, the Seoul P4 can be found at Mouser Electronics for $14.18, the Sandwich Shoppe for $12.50, and Deal Extreme for $7.18. There are other places to get them, but I haven't looked around enough. Deal Extreme might be your best bet. They are cheaper, even if you have to wait another 2 weeks for delivery. You do still have to build the plane, anyways, so I bet you can wait some.

I wish that I could help you with finding the true lumen output by any of today's LED emitters. I go by the specs, then after finding the lumen output at a specific current level. From there, you can consider losses. You will get different results of finding the lumen output, depending on bin codes, reflector and optic losses, differences in measuring equipment, and other variation. Comparing your "130lm" Fenix (with the XR-E powered at less than 700mA, most likely) with another 1W flashlight with 45 lumens is a bad idea. Each reflector is different and flashlight is different. Your little 1AA light might reduce current quickly after you turn it on, due to its heat-sinking and battery source. Those little 17mm reflectors are crap, anyway, compared to any 27mm reflector. Don't forget that 2X increase in lumen output only appears as a 33% increase in brightness to our eyes, also. Because all of these weird variations, I stick with the semi-BS ratings from the manufacturer. Cree did testing of their LEDs in real-world, warm environment, just to prove that their 160lm at 700mA rating is very close. Expect the same with Seoul, I am sure. All that matters is that both the XR-E and the Seoul P4 is very bright, and either would be better than nothing at all.

So, let's say you try the Seoul P4 and IMS 27mm reflector. You can have a nice cluster of 9 LEDs that fit in a neat square (3.2in X 3.2). Power each bank of three LEDs with its own driver (at 1000mA), and you can expect ~2160lm (if you go by the specs of 240lm per emitter). The whole array would consume about 32W. You can always reduce the current if desired. You can add more LEDs and reflectors, or take some away. I go by multiples of 3 or 4 per bank just because of the Vf of the emitters and the supply voltage.

So, what do you think? Sorry about the length of this message. I cannot help it. I hope you are as patient to read this as I was typing it. This is really exciting, and I do not think that I could have compresses this into 2 paragraphs.


-Tony

 

[KingGlamis]

I know you are wanting to build your own, but I thought you might be interested in this. This is the latest in LED lighting for the off-road vehicle market. Something similar would work great for landing lights I think. I believe they come in 10, 20, 30, and 40" widths. I have seen these in person and they are crazy-bright.

This is a comparison of off-road-style HID lights (typically 8" diameter) on the left and the LED light bar on the right.

Here is the link: http://www.holderoffroad.com/products.htm

 

[Gryloc]

Whoa! That looks interesting! Nice to see a production made light like that. Unfortunately, the tint looks uncomfortably blue, even compared to HID! I wonder which LEDs they used. I bet the reflectors (not optics-look up close) are custom made for them. Each 10" segment contains 20 LEDs, and according to the web site, consume "38 watts with a 3.3 amp draw at 12 volts". That is 1.9W each LED. Hmmmm, interesting. Nice, but comparable light can be built for cheaper though...

 

[KingGlamis]

Whoa! That looks interesting! Nice to see a production made light like that. Unfortunately, the tint looks uncomfortably blue, even compared to HID! I wonder which LEDs they used. I bet the reflectors (not optics-look up close) are custom made for them. Each 10" segment contains 20 LEDs, and according to the web site, consume "38 watts with a 3.3 amp draw at 12 volts". That is 1.9W each LED. Hmmmm, interesting. Nice, but comparable light can be built for cheaper though...

I'm sure something better and/or cheaper will come along. No doubt. But for now I think it's a good sign of things to come.

One more pic:

 

[KingGlamis]

And yes, they had a custom reflector made. Here is the CAD drawing of it.

 

[IanJ]

Ah-hah. Further research shows that there are tighter beams (8 and 9 degrees) in landing lights. That's more like what I was thinking anyway...

And I ran across this thread on a 10-Cree battery light that gives me ideas on how to handle a landing light. Looks like 8 emitters (two 4-series runs) would be about right. He used MR17XR reflectors, and the result is certainly impressive.

 

[yaesumofo]

Since you are building this aircraft I would assume that you want to be seen as well as see.
There are always a flashing red and white lights on aircraft.
I would wire as Cree emitters using OPTICS into the leading edge of the lower wing. You could have 10 or more emitters set into both the right and left lover wings. you have plenty of room and would make you very visible.
you may need to place a couple of high output emitters into a 3 inch reflector and aim them carefully. you will gain throw with a larger reflector. when flying you have to have a light with throw because things are moving quickly towards you you need to be able to see far in front of you.
Personally I dont believe that Leds are up to the task for great throw.
You may want to look into HID lights from a car.

Georges80 makes car interior led systems You may want to look into a few of these to light your cockpit, I would.
Between multiple emitters on the forward wing surfaces, a red led flasher or 3 and a HID landing light you should be covered nicly.
Good luck happy flying.
yaesumofo

 

[Gryloc]

Even with the Cree (if that is what you prefer), you can use a larger reflector. Is cost an issue? I am the "scrounger" type, and I research the cheapest stuff of decent quality to use because of my limited budget. I just wanted to save you some big bucks. However, if you have a decent budget, you can go with the McR-27 XR-E reflector at the Sandwich Shoppe. The beam should be more intense than a light that used the 17m reflectors. I do not know how those reflectors would sit on the LED, though. The IMS 27mm reflector, for eample, has 3 supports to let it sit on a luxeon or Seoul P4 without pressing on the domes or becoming un-focued. The MCR-17 sits well on the XR-E, but I don't know about the bigger MCR-27 reflector. If you like smaller reflectors, the IMS 20mm works well with the Seoul, also.

So there is not a "preferred" beam angle of the landing lights? As long as it works? Could you still tell what altitude and angle of decent you would have when landing when you need the lights? That would be really interesting to know. Thanks. If you must illuminate very distant objects, LEDs may not be the way.

As yaesumofo said, HID would be a very viable alternative. It may not be much cheaper, but it would be much easier to mount and it would be simpler (one reflector, one really bright bulb). It would still be very efficient and you can use smaller, lighter wiring. It would be much lighter because you do not need all the aluminum or copper needed to heatsink all of the LEDs (unless you plan on using minimal heatsinking material due to the good air movement). This will be interesting either way.

-Tony

 

[IanJ]

Gryloc, thank you so much for your thoughts! I did read the whole thing, I'm just that way.

Knowing now (as I just posted) that a narrower beam can be had, I'm inclined to go with a narrow beam like 8 degrees. Reading through the aforementioned 10-emitter flashlight thread, it appears that a 12v nominal vehicle electrical system can safely power a series of 4 Crees (or, presumably, Seoul P4s) without a driver, although they'll need hearty heatsinking.

The most interesting part of this project for me is getting a blowtorch out in front of the plane for cheap, and weighing as little as possible.

To answer your questions, I've only flown with (probably 16 degree) halogen landing lights. They start noticeably lighting the runway at perhaps 300 yards. A landing light also serves the purpose of signaling location to the tower or other airplanes, so even if it's not lighting up the runway, being bright is good. Given that I'm working with LEDs which don't suffer from switching, I'm likely to make a flashing circuit, so it's got a 1 second on-off pulse mode (more visible). I'm also likely to make at least one or two of the emitters wide beams to increase visibility.

The offroad lights look very interesting, I'll have to look over that link. I figured it was only a matter of time until people started making vehicle lights out of LEDs; I'm glad to see some progress being made.

The airplane itself will most likely have a custom cowling, made by me. I'd guess I'll have an area of about a square foot for landing lights and/or taxi lights (taxi lights will just be wide-beam, and set at a different angle, likely in the same assembly as the landing light). That should be more than enough to do what I'm imagining. For what it's worth, the lights can go damn near anywhere -- on a wing, on a strut, on the engine cowl, etc. The cowl makes the most sense, because that virtually eliminates cooling concerns (I'm never going to have the landing light on for any length of time unless the motor is also turning).

One of the "problems" I have is that I haven't actually decided to make the plane yet! And when I do start (it would be some time in the next year or two), it'll take me minimum 2 years of dedicated after-hours work to get it done. That means that the minimum time to completion is 3 years, with 4 or 5 more likely. The current pace of LED development suggests that waiting until I'm further along would be prudent, simply because the emitters will be noticeably brighter, and for less current.

Based on what I've seen, I think this project is completely feasible, and for less money than any comparable HID system (the current hot-stuff landing lights are all HID). I suspect it'll also come in about the same weight, or even lighter, since I don't need any kind of controller electronics for the most basic system.

 

[IanJ]

Yaesumofo, fortunately (or unfortunately, depending on how you view things) landing lights are not intended for anything more than giving you a clue for when to flare on landing -- that is, in the sub-100-yard area. I believe this is easily feasible with LEDs. They're also to let you be seen, which LEDs will be fine at.

For what it's worth, an airplane is required to have nav lights (red, green and white lights on left wingtip, right wingtip and tail respectively) and a "suitable beacon." There are companies selling LED beacons (Kunzelman Electronics, I think) and LED position/nav lights already. An airplane should also ideally have a strobe system (which is a better kind of beacon, ultimately), and a landing light.

I do like the HID systems I see for landing lights, but they're $500ish for the cheap ones, vs. $250 for a healthy handful of Crees and reflectors. I'm not interested in saving a few pennies and getting a huge tradeoff in performance, but in a $20k project (which is what this will be) saving $250 is something to be considered.

The other important thing to consider is that this airplane will be flown during the day 95+% of the time. Night-time equipment is something of a luxury, in the sense that it's not even remotely necessary to get in the air. Legally, I can operate at night without any landing light at all, so any light I can throw at the runway is gravy, particularly on this "light and simple" design.

I love all the comments, keep 'em coming!

 

[IanJ]

Gryloc, one more thing. A common fallacy I see here on CPF is the assumption that a heatsink must be heavy/massive to be effective -- not true in all cases. In my case in particular, where I'm practically guaranteed at least 40 mph of headwind at all times, the heatsink need only have surface area. Mass gives you some breathing room if you lose your breeze, but it doesn't contribute to effective cooling beyond that. (This, of course, is a simplification, but I hope I'm conveying my meaning well enough.)

Of course, that's positive for me -- it means I can go with a light weight heat sink with lots of surface area, keeping my total weight low.

 

[Gryloc]

In that case, you should build the landing lights now! As a matter of fact, design the airplane around the landing lights. I am just kidding. Yes there is always time! In two years, there will be another LED with better efficiency and brightness. Unfortunately, in two years, maybe LED will be the hot thing (HID old-school by then), and you wont be that cool cat with all LED landing lights. Oh well. I would suggest buying a few Cree XR-Es, some Seouls, and a variety of optics and reflectors to play with, just to get an idea of their brightness, and the beam produced. However, I would not recommend sticking around here for long -maybe come back when you need the landing lights made. You will not be able to afford your plane kit at all if you visit the CPF daily. Instead, you will have a hangar/garage full of flashlights instead. Let us hope that you have a good resistance to flashaholism. Some of us are pretty weak Well, keep us posted with your progress and any updates...


-Tony

EDIT: You have a good point about the heatsink. Sorry. I have three Crees on a small Pentium CPU heatsink in a slightly enclosed area. It gets hot when left alone. However, with the smallest 40mm chipset cooling fan, just enough to create a breeze, it stays about room temperature. My big 18 K2 LED headlights consume 100W and uses a standard 80mm temperature-sensing cooling fan. I forgot how cool even that thing stays. Yeah. You will be perfectly fine. Just mount everything to an aluminum sheet, and epoxy a few fins onto the back. Actually, if you fly on brisk nights, you will actually gain efficiency with the LEDs and they will get brighter (funny characteristics of the LED die).

 

[IanJ]

Ah, the flashaholism. I have an HDS EDC B42, which seems to be tiding me over pretty well. The new Novatacs look cool, but I just can't quite justify spending another $110 or $150 when the B42 still does everything I want it to. I've actually been on CPF for a while (as my belt, wallet and flashlight will attest ). The airplane thing is relatively new.

I'm actually very excited to see what the next few years will bring with LEDs. I think in two years, I might be able to build this landing light out of a few emitters, instead of 8-10 (or 20 or 30), if progress remains at its current rate. Certainly as fixed incandescent-replacement lighting starts to take off, the market will heat up with the volume available. Add vehicle lights to that, and we'll be in a boom time like no other, with 200 lumen flashlights that run for 10 hours on a CR123A. (Sounds stupid, but so did 80 MHz 486s at one point.)

And, you know, I might just build a landing light now anyway (well, not right now, but before the plane needs it) just because I like working with that kind of stuff. Good enough is good enough, even if the tech surpasses the design later (as my B42 will readily attest).

 

[TigerhawkT3]

Um, not to burst your bubble or anything, but I really don't think LEDs are enough for a landing light. I have a LarryK12 (using a Q4559X lamp assembly, the landing light for 747s), and LED is not even close. A single well-driven Cree is barely enough for a bike light, so unless your airplane is going to really be crawling through the air, even ten Crees might not be enough. Just go for an ordinary incan landing light - even the "expensive" ones are around $30, tops. Just do an Advanced Search at bulbconnection.com and look for bulbs around 12V, and you'll find several nice toys to research. These things have lots of lumens in a single source, for good throw.

I really wish LEDs were up to the task, but right now, I think they might only be enough for beacons (i.e., making yourself visible to others).

This could definitely change in a few years when you actually get around to building your plane.

 

[yellow]

pricing will be a point, surely.

just my thoughts on some things stated above:

SSC --> ONLY advantage so far: can be used with the ceap IMS reflectors.
Much cons like: color shift, thermal problems, ... (read in here about these problems)

CREE --> none of these problems typed in here.
EXPENSIVE because they NEED the Shoppe reflectors. Most any other reflector gives rings. The Cree optic (8 degree) ist way cheaper, with 3 optics costing the same as one McR-XR reflector, but its not as good. Most obvious difference: a bit less throw and NO sidespill at all.

As for McR-diameters: I personally have 2 lights here with nearly identical mods, the one with the 19 mm rocks, the one with the 17 mm sucks (almost). The 27 mm, while giving even more throw, is a pita to place correctly, the 19 mm is possible and 17 and 16 mm are very easy to place (feature a "seat" for the metal ring of the emitter)

pic of 19 mm (right) compared to Cree optic.
powered totally identical (wired in series and run with same current), difference not too obvious, non existent spill not noticeable (by now I was not able to make a pic showing this)

Till the plane is ready, the leds will be much better. Dont get anything now.
Maybe even Luxeon has some answer then (and You can skip the positive slug SSC)