LED exterior aircraft lights 20 times the life

Honeywell's Astreon products have been tested to 20 times greater life than traditional lights and carry extended warranties. The contract provides for incorporation of Astreon anti-collision and navigation lighting systems on all 7E7 aircraft.

The systems represent a potential total program value in excess of $2.6 billion for Honeywell over the life of the 7E7 program.


http://ledsmagazine.com/articles/news/1/12/3

I read a little further, this article is linked from the above one and talks about the actual design of the modules: http://www.ledsmagazine.com/articles/features/1/10/4/1

The design team determined that it could achieve the desired optical performance using for each light three 1-W Luxeon LEDs from Lumileds. A single lambertian emitter was sufficient to meet the requirements of the region between 20 and 110 deg from the forward direction. The much higher intensity requirements in the 0-20 deg region were realized with two side-emitting LEDs and a specially designed reflector.

My gawd... THREE one watters? I've always thought of the one-watt LEDs as wimps, but I might just have to rethink that. Not only that, based on one of their pictures, I think they're using the lambertian LED WITHOUT a collimator/reflector.

I'm impressed, I've seen large aircraft flying overhead from very far away (must be tens of thousands of feet) at night and I can still see the green and red blinkers from the ground. I always thought it took hundreds of watts to produce light that bright!

Based on an operational life of 3000 hours minimum and a nominal 28-V DC input voltage, a low cost resistive circuit was chosen, despite this solution's inefficiency (which impacted thermal management) and the need to design different circuits for the red and green lights.

Them regulated folks are gonna freak out when the pros like Honeywell go for resistored.... /ubbthreads/images/graemlins/grin.gif

LOL.

Thats Honeywell for ya!

There are two driving issues in aircraft applications. ONe is weight, and the other is simplicity (KISS). Compare the MTBF and cost of a regulated supply with a resistor, then compare the weight. The parts that are not there cannot fail. Here is a number for you (and I can provide the details about how it is calculated), each extra pound you carry around in commercial aircraft will burn about $33 worth of fuel per year at current prices. Resistor is simple (and ultra reliable), and it doesn't weigh much. The hallmark of the 7E7 program is weight reduction. If you can get rid of a pound by spending $100, it has a 3 year payback, and is well worth doing.

I figure they are using 3 LEDs to cover three overlapping areas instead of trying to design a fancy optic to do that.
Also wiring 3 in series would reduce the power dissipation of the resistor used vs one 3 watter.

Actually I like the simplicity of resistors when I have a constant-voltage supply that is not subject to massive fluctations (as opposed to automotive 12V systems). If the aircraft can already regulate power reasonably well, a resistor should work ok. A desk lamp I designed for fixed voltage input works great with a resistor, whether I'm driving it off a 5V feed tapped from a computer, 5V wallwart, or as a last resort, 4.8V NiMH battery pack (dimmer, but that's meant to be a power failure light, not its main power supply).

Someone may want to point Honeywell to the work that DougS did, looking at the Vf shift over time. The Vf shifted enough to significantly change the current flow and thus the heat sink requirements.

Doug's Work

Daniel

[ QUOTE ]
KevinL said:
Actually I like the simplicity of resistors when I have a constant-voltage supply that is not subject to massive fluctations (as opposed to automotive 12V systems). If the aircraft can already regulate power reasonably well, a resistor should work ok. A desk lamp I designed for fixed voltage input works great with a resistor, whether I'm driving it off a 5V feed tapped from a computer, 5V wallwart, or as a last resort, 4.8V NiMH battery pack (dimmer, but that's meant to be a power failure light, not its main power supply).

[/ QUOTE ]


Actually, aircraft power buses are subject to fluctuations and significant pulses, in addition to lightning hits. In fact, depending on the category you certify the equipment to, you have to take 60V to 100V hits on the bus, as well as thousands of volts of static, and various other things. In an aircraft, the buses often see subtantial pulses when switching between generators.

I've been working on military, commercial, and private aviation electronics (Avionics) for 18 years, and have seen alot of failed equipment from stuff just like this. I don't work on exterior lighting though, maybe the requirements are alot lower for exterior lighting vs. Avionics that the pilot utilizes to fly the plane with.

However, I do remember a certain master lights panel in the A6E Intruder and EA-6B Prowler that used to get blown *alot* due to fluctuations in the aircraft power bus...

Though, when thinking about it further, if you design it to fail, then there is more money in your pocket from replacement when it fails...

Actually with the resistor pack in this configuraion, the VF shift downward is far more likely to LOWER the power dissipation requirements on the Luxeons. It is the resitor pack that takes the beating.. On most of the aircraft I have dealt with, the lighting is from a 28V source. So with 3 Luxeons in series, the total Vf is something in the 10-11 volt range, so
17-18 volts will be dropped in the resistor pack. My guess is the changes in Vf are actually small compared variations in the 28V supply to start with. So if you set up the Resistor pack to deliver 800ma in the luxeons with a nominal 11 volt drop across the luxeons the resistor pack is 17/.8 or 21.25 ohms.
If the Vf falls to say 3 volts from 3.67 per Luxeon how much does the current rise? Well, we are now going to need to drop 19 volts instead of 17. 19/21.25= 895ma, so yes the current went up, , but not by much .800 x 3.67=2.96 watts, .895 x 3=2.68 Actual power dissipation went down in the Luxeons (it was the resistor pack that takes the beating in this situation, not the LED). Vf shifts may cause a problem in the resistor pack, but probably not in the luxeons.

Although I don't work in the airline industry, I imagine that cost is also a big driver - especially in larger aircraft - and this is primarily labour related - if you only have to replace the light 20 times less frequently (say 15 minutes each time), at $40 per hour, plus the big killer, non-operational aircraft time, then the cost savings can be huge. Especially when you add this up over the entire aircraft - I don't know what the scheduled maintenance on these things is like - imagine if they always replace them on a regular basis - even if they don't cease working.

Probably much the same reason that truck fleets are often going for LED units for brakes, indicaters, etc - you don't have to replace them anywhere near as often - when labour costs are considered, they are really really cheap.

Cheers.

[ QUOTE ]
NewBie said:
Actually, aircraft power buses are subject to fluctuations and significant pulses, in addition to lightning hits. In fact, depending on the category you certify the equipment to, you have to take 60V to 100V hits on the bus, as well as thousands of volts of static, and various other things. In an aircraft, the buses often see subtantial pulses when switching between generators.


[/ QUOTE ]

Ouch! For some reason I thought the power might be cleaner than automotive systems, as it's well known that car DC voltage tends to fluctuate and many a newbie has been advised to avoid a resistored Luxeon setup on a car. Ooops.. /ubbthreads/images/graemlins/blush.gif

You raised a good point about their design, who knows, I hope these don't end up like the traffic lights, fried after how few hours of operation.

The funny thing is look who is the chair...

http://pr.erau.edu/~lyallj/SC135/



You can see the test requirements for the voltages here:
http://pr.erau.edu/~lyallj/SC135/D-to-E.pdf

See Section 16, subsection 16.5.4.4 b for an example.

You'll note even the lowest category (if you went for that), is a hit of 46V....