DIY led anti-collision, position lights?

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

cluttonfred

Well-Known Member
HBA Supporter
Joined
Feb 13, 2010
Messages
7,292
Location
World traveler
With the wide availability of very bright LED lighting, it would seem pretty straightforward to adapt an LED-compatible turn signal flasher and a bright LED bulb to make a DIY LED anti-collision beacon. Basic navigation lights also seem like they should not be too hard, whether for integral installation inside the wingtip or for external surface mounting. Does anyone know of any such projects and have any sketches or circuit diagrams? Cheers, Matthew
 

goldrush

Well-Known Member
Joined
Apr 11, 2010
Messages
165
Location
Aberdeenshire UK
If you are not aware, you may find the information in FAA Advisory Circular 20-74 dated July 1971, which gives details of requirements for Anti collision strobes nav lights etc... light output, position etc of interest.

Unfortunately, it appears that very few strobe manufacturers, or indeed LED manufacturers themselves give light output figures... and those that do, often do not measure this in an approved way... so "caveat emptor"
 

Blue Chips

Well-Known Member
Joined
Oct 15, 2014
Messages
370
Location
Mount Gilead Ohio, USA
Certified strobes are terribly expensive so today's very inexpensive and incredibly bright LED and CREE type light have a great deal of merit for the experimental builder.
Anti-collision strobes are not required for day time VFR are they?
If that is the case seems anything would be better then nothing, worse case being they would have you remove them.
 

Aerowerx

Well-Known Member
Joined
Dec 1, 2011
Messages
5,536
Location
Marion, Ohio
Certified strobes are terribly expensive so today's very inexpensive and incredibly bright LED and CREE type light have a great deal of merit for the experimental builder.
Anti-collision strobes are not required for day time VFR are they?
If that is the case seems anything would be better then nothing, worse case being they would have you remove them.
I was going to mention Cree.

I have a little flashlight that has one. Amazing little thing. Runs off of a single AA battery, and you do not want to look directly at it. It is amazingly bright. Even if I shine it on my hand it is too bright to look at. I don't know how they do it, but it also has a blinky mode and two brightness levels.

I would think that a half dozen or so white Cree LED would work well for a strobe.
 

Toobuilder

Well-Known Member
Log Member
Joined
Jan 19, 2010
Messages
4,725
Location
Mojave, Ca
Position lights are easy. I use nothing more than a multi element SMD green or red LED bulb in a standard automotive 1156 bulb socket, and the whole thing lives under my wing tip lens. In this case, the overall output is easily achieved, as is the masking of the lens to get the required coverage pattern.

A compliant anti collision system is a different story. The required light output is very tough to meet with LEDs even for a dedicated manufacturer. Sure, you can do a "bright" anti collision light DIY, but it's really just likely going to be a blinking light. That said, I'm running trustfire 11000 LM flashlights as landing lights and they completely blow away the "approved" strobes I'm using for anti collision.
 

kent Ashton

Well-Known Member
Joined
Aug 15, 2008
Messages
739
Location
Concord, NC
Are you flying in the U.S. or Kenya? If you're licensing in the U.S. it's just easier to buy some non-TSO'd lights than brew your own and have problems with inspectors. However, here is a kit sold for the RV crowd that might give you some ideas.
LED POSITION LIGHTS

There are plenty of net tutorials on how to wire LEDs and how to make flashers.
 

fredoyster

Well-Known Member
Joined
Mar 20, 2012
Messages
333
Location
Monterey Bay, CA
Not to confuse the lighting product itself being "approved" vs the installation of the product being "approved"
Right. TSO is irrelevant for experimental aircraft, and many lights are TSO'd to older standards that wouldn't work for newly certificated Part 23 aircraft anyway. And it is still up to the FSDO, unless a DAR is willing (I haven't found one, yet) to issue the specific authorization required by 91.319(d)(2). It's not an airworthiness signoff at all, but an amendment to the operating limitations to permit night flight (aircraft with older op lims may have different ways to accomplish this.)

The requirements for position light color, intensity and coverage are easy to meet with single 10-watt LEDs (running at about 5 watts or less for the position lights) now. It's approval of the lighting system that people don't get. The anticollision light system has to deliver 400 effective candelas with no more than half a steradian of blockage. I am not done with measurements but think that one 10 watt white LED on each side of the aircraft, run at little higher power at about 20% duty cycle will meet the requirement without any extra optics. All of the calcs are in AC 20-74 and AC 20-30B. My hope is that showing compliance with current Part 23 requirements for light output per those AC's will be enough.
 

cluttonfred

Well-Known Member
HBA Supporter
Joined
Feb 13, 2010
Messages
7,292
Location
World traveler
As a first step, I would be thrilled with just an affordable DIY LED single rudder-mounted anticollision beacon.
 

Jan Olieslagers

Well-Known Member
Joined
May 28, 2014
Messages
134
Location
Haacht, Belgium
Forget it. To achieve any degree of useful visibility, a single LED needs to be severely limited in its angle of projection, both horizontally and vertically. I've discussed this subject to death with a fellow aviator, as yet we agree only up to there. The next arguments are "how severe must one limit the angle, to achieve a minimum of visibility distance" and "given a maximal dispersion angle, how to share it out between the vertical and the horizontal plane". On the latter subject, we have developed a consensus that vertical beam can be rather narrow, to the advantage of the horizontal coverage.

Beacons that simulate rotation by flashing a lot of LED's can be found in abundance at spruce et al, but I've no idea how effective the products on offer are. But they do apply mirrors/reflectors to concentrate the lightbeams.

Worst of all, for a www discussion: progress is quick in the domain of power LEDs, so today's truth is tomorrow's ridicule.
 

fredoyster

Well-Known Member
Joined
Mar 20, 2012
Messages
333
Location
Monterey Bay, CA
Forget it. To achieve any degree of useful visibility, a single LED needs to be severely limited in its angle of projection, both horizontally and vertically. I've discussed this subject to death with a fellow aviator, as yet we agree only up to there.
Like you say, progress is quick. No ridicule is necessary. What's becoming common in industry is the matrix LED with some number of LED chips bonded to a 20 mm square substrate. These easily handle 10 watts continuous at about 100 lumens/watt. There are 20 watt and bigger ones but these typically are wired in series for more than 12 volts. I am planning to mount four on a piece of 25 mm square tubing. We'll see how it works. Some typical ones are available from a surplus outlet, 10 Watt LEDs | MPJA.COM Being bare LEDs with a common phosphor they are approximately a diffuse surface radiator that makes useful light over about 140 degrees.
 

cluttonfred

Well-Known Member
HBA Supporter
Joined
Feb 13, 2010
Messages
7,292
Location
World traveler
Jan, I am not sure where the idea came from that this would be limited to a single LED. What I am looking for is a design for an LED anti-collision beacon which could be adapted to single- or multiple-light installations, but there could certainly be more than one LED per light.

For example, let's take the basic requirements outlined in the attached quick reference sheet from Whelen.

As far as I can tell, something like this 540 lumen (white) multi-LED bulb would be sufficient for the 400 candlepower +/- 30 degree (60 degree apex angle) requirement with an appropriate lens/reflector according to this calculator.

Something like this 2250 lumen dual CREE LED headlight bulb ought to be sufficent for the 400 candlepower +/- 75 degree (150 degree apex angle) requirement.

Am I missing something?

Cheers,

Matthew
 

Attachments

goldrush

Well-Known Member
Joined
Apr 11, 2010
Messages
165
Location
Aberdeenshire UK
Like you say, progress is quick. No ridicule is necessary. What's becoming common in industry is the matrix LED with some number of LED chips bonded to a 20 mm square substrate. These easily handle 10 watts continuous at about 100 lumens/watt. There are 20 watt and bigger ones but these typically are wired in series for more than 12 volts. I am planning to mount four on a piece of 25 mm square tubing. We'll see how it works. Some typical ones are available from a surplus outlet, 10 Watt LEDs | MPJA.COM Being bare LEDs with a common phosphor they are approximately a diffuse surface radiator that makes useful light over about 140 degrees.
I have been running just that for about 18 months now (non certified aircraft) with good results.. reported as visible at about 3/4 to 1 mile in daylight.....(we don't get too much Sun in the UK:)..........4 "cheapie" 10W 900 Lumen LEDs
although I actually over run them. Being pulsed with a "low" duty cycle, they stand it and heat sink requirements are low. equivalent to about 2 watts and really only required when stationary.


I am also using a 50 watt 3,500 lumen cheapie for a "landing light".... 12 to 32 volt inverter.

Obviously both would be better with something like CREE.
 
Last edited:

Jan Olieslagers

Well-Known Member
Joined
May 28, 2014
Messages
134
Location
Haacht, Belgium
I am not the one to tell you what you are missing, being none too sure myself. However, as you mention "candlepower" quite a few times, I wonder if your quest is mainly for something that meets legal criteria - my search (and my fellow's) is for something that will actually help to avoid collisions, i.e. make one visible to a potentially dangerously approaching pilot in sufficient time. We never managed to translate that criterion into a number of candlepower, more's the pity.

(and excuse me for reading a one-LED requirement where indeed none was mentioned)
 

cluttonfred

Well-Known Member
HBA Supporter
Joined
Feb 13, 2010
Messages
7,292
Location
World traveler
Well, far be it from me to praise the bureaucracy, but I do think that the regulations are often, though not always, based in some aspect of practice. If the original 100 candlepower was upped to 400, and the +\- 30 degrees to +\- 75 degrees, then I have to think that it was because of some real-life incidents or testing. Regardless, it does give something to aim for.
 

fredoyster

Well-Known Member
Joined
Mar 20, 2012
Messages
333
Location
Monterey Bay, CA
I am not the one to tell you what you are missing, being none too sure myself. However, as you mention "candlepower" quite a few times, I wonder if your quest is mainly for something that meets legal criteria - my search (and my fellow's) is for something that will actually help to avoid collisions, i.e. make one visible to a potentially dangerously approaching pilot in sufficient time. We never managed to translate that criterion into a number of candlepower, more's the pity.

(and excuse me for reading a one-LED requirement where indeed none was mentioned)
My quest is to do both. Being visible is more important to me than compliance, as I fly in densely populated airspace. As Matthew pointed out, FAA went from 100 or 150 effective candela to 400, probably for a reason. Less important, I would like as little guff from the FSDO as possible when asking them to sign off my homebuilt to fly at night (for people who say this isn't required, read 91.319(d)(2).) so compliance with the current requirements for Part 23 airplanes seems prudent. We've all seen TSO'd wingtip flashers with xenon strobe tubes that meet the 100 candela requirement, and they're not bright enough. The strobes on the belly of that 737 way over there are bright enough, and there are a few of those on eBay right now (but they need 115 V 400 Hz AC) ... they are typically 400-500 effective candela and red. There is a good discussion of the problem from Eric Jones ten years ago at www.periheliondesign.com/downloads/aircraft_beacons_using_leds.pdf I understand the theory behind the Blondel-Ray approximation of "effective intensity" but while it's a start I am not sure it's the last word.

The French article mentioned earlier in this thread seems to have winking wingtip lights and a red beacon at a set cadence. I think it would be interesting to research what is actually more visible. SAE has just published a revision of the AIR 1106B standard that addresses visibility, I just haven't decided to spend $70 to read it. SAE seems to be doing the work on all the new lighting standards that the FAA has adopted in the TSO process. One can hope that there might be some wisdom in this process.
 

Toobuilder

Well-Known Member
Log Member
Joined
Jan 19, 2010
Messages
4,725
Location
Mojave, Ca
... Less important, I would like as little guff from the FSDO as possible when asking them to sign off my homebuilt to fly at night (for people who say this isn't required, read 91.319(d)(2)...
91.319 says to fly in daylight unless otherwise authorized by the Administrator... And that points you to the ops lims for your airplane, and that in turn drives you to 91.205. 205 describes "approved" systems, but there's no requirement for anyone at the FSDO to inspect or "approve" them that I can see. Like the initial airworthiness certification, it's up to the builder to determine if the resulting aircraft meets the standard or not. And much for the same reasons, the FAA is not going to hang their butt out there by "approving" your particular installation.
 
2
Top