What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

New (October 2013) LED wingtip landings light discovered

On my santa wish list is a spectrum analyzer to determine exactly what frequencies things like these LED lights bark at.

Bob,

Have you ever looked at using a Software Defined Radio (SDR) spectrum analyzer? I was quite impressed with using the free Windows program SDRSharp(http://sdrsharp.com/) with an inexpensive USB SDR dongle (RTL2832U & R820T). You can pick up the USB SDR Dongle for about $15 shipped on Ebay. It works quite well - I have been able to detect 978Mhz and 1090Mhz ADS-B signals, AM, FM, Digital television and APRS signals. I am sure it would be good enough to find the frequency of the noise coming off these LED's.
 
One of my EE guys suggested a .1uF capacitor between power and ground, close to the light as possible.

How about using a shielded wire for the landing light? Do you guys think that would supperss all RFI?

I used the capacitor for my HID with no effect, not to say that it may not be useful in this situation.

I really like these flashlight type for LL if I find a solution for the RFI.
 
Bob,

Have you ever looked at using a Software Defined Radio (SDR) spectrum analyzer? I was quite impressed with using the free Windows program SDRSharp(http://sdrsharp.com/) with an inexpensive USB SDR dongle (RTL2832U & R820T). You can pick up the USB SDR Dongle for about $15 shipped on Ebay. It works quite well - I have been able to detect 978Mhz and 1090Mhz ADS-B signals, AM, FM, Digital television and APRS signals. I am sure it would be good enough to find the frequency of the noise coming off these LED's.

Yep I have been playing around with SDR#. Very cool indeed. I have thought it could be used as a spectrum analyzer, albeit one without any filtering.
 
How about bringing it over to my house (I'm off this Friday) and we can compare the two versions of this light?
 
I don't have need for one of these right now, but I do have access to a anechoic EMI room and test gear. I can do a full DO-160 radiated and conducted emissions screen on one if somebody wants to drop ship one to me. I'm kind of the resident EMI "fixer", at least on the stuff I don't design...:)
 
RFI

I haven't done board level radio in a long time but I did stay in a Holiday Inn last night!
Sorry, couldn't resist.
Is it possible the RFI is coming from the light itself. Wrap that thing with aluminum foil and turn it on for a few seconds. Just long enough to see. If I read correctly, it's pretty close to the antenna.
 
Why not try installing a simple series dropping resistor in the power line so the lights aren't being hit with full aircraft 13.8V, then if necessary add a couple of capacitors between the ends of the resistor and aircraft ground? Poor man's filter. You could also substitute some turns of wire as an inductor, make it small enough gauge that it will have some resistance value, plus will add some inductance, and again a filter capacitor on each side of it to ground.
 
Tried using shielded wire today, no change. Bummer. Might have to abandon the trustfire.

Yeah I figured that wasn't going to work. Basically the lights dump a large amount of AC noise and the wire feeding the lights acts as an antenna. Shielding it will have minimal effect. A ferrite is essentially an AC resistor so the more times you wrap the wires thru it, and the more substantial it is, the more it will attenuate AC frequencies. The ferrites I mentioned earlier get the radio noise down to a very acceptable level.
 
Did you ground the shielded portion only on one side? Grounding it on both side will create a ground loop?

Yes, only grounded on the panel side.

My immediate need for night flight seems to have gone away, I will revisit this landing light thing later. I need to get my logged RV time out of the single digits, so I am working on buttoning things back up so I can fly a bit (hopefully: one SB bullet dodged, will check #2 later today).
 
Last edited:
Quick teaser!

Just a quick teaser, I plan to do a better write up later. For now, here are the cliff notes:

After talking to Rocket Bob, I purchased 4 of these LEDs ($26 total):

http://www.kaidomain.com/Product/Details.S009985

I installed them in the right leading edge in place of my 55W HID lights.

Here are the LEDs: (0.7amps each / so total of 2.8amps)



For comparison, here is the left wing 55W HIDs: (4.5amps each, I know 55W divided by 14.2 volts is 3.87amps, but it measures a higher draw)



And both on:



The HIDs are super bright and will light the runway at half a mile, but are a very narrow beam. The LEDs are bright as well, but the beam width is much wider. I have only flown once with the new lights and these pictures were taken before I re-positioned them(they are aimed too low), but the results are promising. I will post a full write up once I replace the left wing. No RF noise.

Dan

Not sure if the video will work, but here is a try:

 
They do look small...

...Small enough to use as dedicated taxi lights mounted somewhere clever.

Hmmmm...
 
The reflectors are 1" at the widest part and just a hair under 7/8" at the base. They are really small, but pack a punch. Check back a few post and you can see where RocketBob uses a single one in each wingtip as his taxi lights. After talking to Bob offline I figured if 1 is good, 4 must be better, right? :cool:
 
Finally got the Rocket back in the air and put the dual Trustfire lights to the test. I landed at dusk so have not tested the "landing light" aspect, but did make several runs against a ground observer with the multiple modes and the reports are good. The "SOS" function seems to be the most conspicuous, but the "steady high" setting is still considerably more visible than the 100 w GE still mounted on the other wingtip. The modes are easier to select with the power switch than I anticipated so that's also a plus. The main problem remains the noise. I have about 20 turns through a big ferrite and I think it actually increased the noise! There will definitely be some tuning involved, but this installation seems to be worth the trouble.

More to follow.
 
Hey Mike are you using a airframe ground? I was told to start running a ground wire back to the buss and put in 7-9 twists of the light/taxi wires in bundle per foot. This should reduce induced noise thru the wiring. I used this in my professional line of work. Can't see why it shouldn't work here. Worth a try. Let know if it works.
Ron in Oregon
 
Just a quick teaser, I plan to do a better write up later. For now, here are the cliff notes:

After talking to Rocket Bob, I purchased 4 of these LEDs ($26 total):

http://www.kaidomain.com/Product/Details.S009985

QUOTE]

Dan,
I bought the same LEDs after a posting from Rocket Bob some time ago. The first set I got looked like the picture in the link, specifically with the red ring at the back. A second order shipped LED's, some with red and some with black ring at the back. They look exactly the same apart from the red vs black ring. The ones with the black ring blew right off the bat when connected to 12V battery.. The red ones would light up, stay on for about 30 sec then start flashing. They also got VERY hot, to the point that it will burn your fingers. Too hot too handle. I have tried a variety of in line resistors, 5 - 100 ohm, in which case the resistor got too hot to touch. I used resistors up to 100W. I eventually gave up. I made a mount that held three of the small LED's you've got, plus one of the three LED CREE from Dealextreme, ones that Rocket Bob mentioned as well. I now have two of the 3 LED CREEs (as landing lights) as well as the 75W VANS Halogen (wider angle for taxi) installed in the wingtip in a home made bracket. I am feeding power to the LEDs through a 5 ohm resistor, and no excessive heat anywhere. Not flying yet, but had them on for hours, and seems to work well. Very little heat palpable through wingtip.

So, I was wondering how hot does yours get? Did you drive them through a resistor? Would like to know more about the setup please.

Keep an eye out for the ones with the black ring at the back.

Johan
 
Hey Mike are you using a airframe ground?

Local ground on the tip rib only.

Another interesting thing is the fact that these things sing externally now. Meaning, they make audible noise when you are standing next to them. They chirp, beep and click, depending on their mode. Weird.
 
Another interesting thing is the fact that these things sing externally now. Meaning, they make audible noise when you are standing next to them. They chirp, beep and click, depending on their mode. Weird.

Chinese nanobot mating call? :p
 
I have about 20 turns through a big ferrite and I think it actually increased the noise! There will definitely be some tuning involved, but this installation seems to be worth the trouble.

More to follow.

Entirely possible. Inductive reactance. Are both the power and ground looped thru the ferrite? Mine are.
 
FIXED!

Took the wingtip off today, disconnected the ferrite lead and went straight to the battery. They got quiet externally, but still showed up on the comm. OK, let's make less turns though the ferrite and see where it starts to sing. Turns out, ONE 360 loop makes them squawk.... OK, how about just running the wire straight through the ferrite with no turns? That did it. Just flew it, and yep, they are dead silent now.

Happy days!

The ferrite I'm using came with my flat screen TV. They are about an inch in diameter and a little more than an inch in length. They are supposed to snap over the various cables in the TV.

Planning on getting some movies of the airplane on YouTube soon so you all can see the various modes in action and the comparison to the 100w GE bulb on the other side.
 
Sorry it has taken me awhile to get back to you guys, I took the RV on a snowboarding trip to Mammoth, CA.


Dan, did you add any additional heatsinks to these bulbs? love to see how you mounted them.

- I did not add any additional heatsink. I did mount them in .04 aluminum attached to the standard duckworks mount in the leading edge.

So, I was wondering how hot does yours get? Did you drive them through a resistor? Would like to know more about the setup please.

- I had them on a bench running off and one for over an hour. Most continuous was 30 mins. They did get warm to the touch, you wouldn't hold them in your hands, but they weren't "HOT". I have borrowed a laser thermometer and will take measurements as part of my write up.
- I am not driving them through a resistor, just unpluged the HID lights and plug in LEDs.
- The four lights I have all have the red rings on the back like you mentioned. I will keep an eye out for the black ring lights. I did have one light, that after 20 mins of run time began to flash. I will be replacing that one soon. I believe I received one of the 3 mode LED lights by error and accidentally put it in a flashing mode.

I have ordered 4 more of the LEDs with different reflectors. I will take more pictures and do a better write up in a few weeks. (Lights take a couple weeks to arrive from Hong Kong)

Dan
 
Trustfire more like TrustPffffffft!

Bought a pair of the 3-led Trustfire lights as recommended way back in this post. They are the "other" version with the additional heatsink fin and no spring for the negative (-) power connection, only a single spring for the positive (+) power connection.

Connected the first one to a bench power supply and immediately drove the power supply into current limit mode. Yup, the light was a dead short, straight from the factory. ****.

Second light is NOT a dead short, so at least I'll have one to play with.

At this rate, it will likely be cheaper in the long run for me to either stick with my original plan of using an LED Maglight head or go with the Baja Designs lights. One can't keep dropping 25 bucks at a time for very long before the additional cost of a Baja light is covered!
 
Fortunately, Amazon has a very liberal return policy.... Knowing these lights have questionable quality, I test them all before I start modification. I'll also likely buy a 5th unit to keep around as a spare.
 
Mike/Michael/Catmandu

Thank you all for your great write ups and carrying out the pioneering work that will help others who follow and to Rocket Bob for supporting them.

Like many I have the Vans MR16 light cluster and the ideal solution would be drop in MR16's. I tried this but I had RF issues above 126.0 and ferrites actually reduced that to above 125.0 (see my post)

Michael you say no RF issues with the wire running through the clamp-on ferrite but can I ask have you tested the radio over the full frequency range?
 
Last edited:
I have not checked the full range, but will run through the scale next time I fly. Thanks for the heads up.

Update: Flew the airplane today and no noise anywhere in the band, from top to bottom.
 
Last edited:
For those who continue to have interest in the Trustfire lights, I've been doing some measurements on my one working unit. It seems that in the range of 6-6.5VDC the lights are actually quite quiet and produce full brightness. Above this voltage the internal power circuitry becomes more active and starts to spew lots of RF.

I'm working on a filter solution which may resolve the RF noise issue. It's my hope to have it a modular unit which would fit within the short battery tube, thus requiring no external components. Will report back as progress is made.
 
Interesting. I know Tikicarver is comparing the two versions of this light side by side and is seeing some significant differences. I expect he'll drop in and report the results before too long.
 
Mike let me borrow his light and I compared it to mine. Here is what I found.

First let me give a little explanation about how these lights work. I first found out about them while reading a forum dedicated to modifying LED flashlights.
(hey everyone has to have a hobby and the guys on that forum know a lot about LED flashlight inner workings)

All the LED lights out there consist of 3 basic components. The LED(s), the driver board and the power source.

LEDs are a current driven device, (leaving out a lot of the technical stuff) that means they are designed to be driven at a specific current. Less than that and they get dim and higher they burn out. So you can't just hook them up to a power source, they will instantly burn up. The simple way to handle this is to put a resistor in line with LED and power source. For low power LEDs this is usually the easiest and best way but for a lot of reasons, it does not work for high power LEDs.

The LEDs used in all the latest high power flashlights are mostly the Cree XML.
This LED is designed to run at 3 amps!!. You can run it at a lower current, but the lumen output will be lower. Also, if you run it at high current, you must have a good heatsink or you will destroy the LED.
When you have a light that has multiple LEDs, they can be connected in series or in parallel. Most of the time they are connected in series because if you connect them in parallel it is hard to feed them all with the same current. The guys on the flashlight forum have dissected our light and the LEDs are connected in series. The Cree XML runs at 3.3 volts when driven with 3 amps.
So, we have 3 LEDs in series, that means we need 9.9 volts @ 3 amps to drive them at their maximum design output.

The "driver" is the circuit board that is used to give the LEDs a constant current. The drivers are designed for a specific current output. So the designer of a light has to decide what current he wants to drive the LEDs at and then find or build a driver to do it. There are a lot of factors in picking a driver. For example if you don't have a really good heatsink you might decide to drive the LEDs at a lower current and accept the lower lumen output. All depends what you goal is. We will assume that the designer of these flash lights are looking for the highest lumen output that doesn't destroy the LEDs.

It gets a little more complicated because there are two types of driver circuits.The Boost driver and the Buck driver. (there is also a combination boost/buck but it is not used much)
The Boost Driver will step up the input voltage to a higher level so it can drive a bunch of series LEDs. For example, say you want to drive our 3 LEDs with one lithium battery. The Li battery is 3.7v so by itself it will not light up the LEDs. But, if you get the right Boost driver you can power 3 LEDs with one Li battery. Now remember the output of the driver is a constant so the boost driver will draw higher current form the battery than it delivers to the LEDs.

The other driver is a Buck driver. It is designed to have a input voltage higher than the LED voltage you need ( in our case 9.9v) It will draw less current than it delivers to the LEDs.

So what does all this mean with our lights? Well simple story is the two lights I tested have very different drivers in them.
First thing I did was take them apart and compare them.
I found that Mike's light was an original Trustfire and mine is a clone.
Look at the photos below.
They both have the Cree XMLs, that's good. But you can see the outer housing is different. Mine (on the left) has one more heatsink fin. The inner core (it holds the LEDs and the driver) is the same basic shape but they are machined different. Both have threads for connecting the inner housing, but the threads are different pitch. On the other side mine has a spring for connection of the positive input and uses the threading for the negative connection. The Trustfire has a spring soldered to the driver for the negative connection. You can't see the details of the drive board because it is inside the housing and is hard to open up. (I will do it on mine and post pics because I'm using mine as a prototyping platform)
So from a physical look we know they come from a different place so it is probably a good assumption they are not using the same driver. And in fact the the testing proves that. More one that in the next post.

trustfire-2.jpg

trustfire-1.jpg
 
Last edited:
So now the testing....

First I took a handheld near the lights and tuned different freqs. I didn't hear any noise. So the noise people are getting is not radiated RF from the light itself. That makes sense because the driver is contained inside the solid aluminum housing.

Next, I hooked up a scope to the positive lead and powered the lights up with a power supply set at 13.8v.
First I tried the real Trustfire. The noise was obvious.
TEK00018.PNG

The noise is about 2.5 volts peak to peak. and zooming in the freq of the noise is approx 416 KHz.
TEK00022.PNG

Next I tried my light. It has a definite different noise signature.
TEK00020.PNG

The noise on mine is 5.5 v peak to peak and the freq is 111 Khz

But there is more to the story.
I measured the current each light drew as I varied the input voltage.
Without knowing what driver they are using there is no way to know what current the LEDs are being driven at, but this will give a little insight on the driver. I started at 8v and increased one volt at each step.
here is the results

Input Mine Trustfire
8... 2.03... 2.0
9... 1.8... 1.8
10... 1.64... 1.6
11... 2.4... 1.5
12... 3.6... 1.4
13... -... 1.3
13.8... -... 1.25

I didn't like how the current was increasing on mine so I stopped at 12 volts.
But you can see the Trustfire was pretty linear. It looks like a Buck type driver
Mine goes out of regulation when the voltage goes above approx 10.5v
I'm guessing this driver was designed for 3 lithium batteries ( what my light uses). But this test shows mine will not work with a bus voltage of 13.8v.

I didn't have any ferrite cores to test to see how they reduce the noise on the bus. But from what others said, it works.
It seems clear, you need to have one of the real Trustfire lights to be able to use it in our planes with a bus voltage of 13.8v. You can do a simple test with a power supply and a meter like above and see how the driver in your light works. I guess the real question is how to buy one of these and make sure you get the real Trustfire and not a clone. To be sure you could buy direct from Trustfire, but they sell it for $48.
http://www.trustfire.com/en/flashlight-tr-3t6-3x-cree-xm-l-t6-3800-lumens-5-modes.html
 
Last edited:
Tikicarver, Those are some excellent tests. Guess that's why they call this Experimental. It would be interesting to run the same tests with a LuxMeter to see if they are reducing the light output at the higher voltage.

I've been tempted to strip one down to the circuit board level to see if they can be driven by a simple linear current regulator. PWM is efficient, but with the square wave drivers it's hard to eliminate the noise. Also, the switching to "SOS","Flashing" etc. is really annoying. Makes a wig-wag system difficult.

Keep up the good work!

Roger Bloomfield
Buena Vista, Colorado
 
?... Also, the switching to "SOS","Flashing" etc. is really annoying. Makes a wig-wag system difficult...

That was my concern at first too. As it turns out, however, it's a real benefit. The "SOS" function is your wig wag... No extra box or wiring needed.
 
Add capacitor to prevent mode switching

Hi Michael and Ron

Michael, the SOS doesn't work so well with multiple lights on one wing, but adding a large capacitor will keep the voltage high enough between wigs (OR WAGS!) to prevent mode switching. It may also help with the RF.

I used a 15,000 mfd and a small ceramic capacitor, and can't detect any RF using a handheld radio.

Ron the light output is essentially constant with normal supply voltage, and your voltage vs. current figures are spot on.

Thank's again RocketBob for bring these to our attention.

Roger Bloomfield
Buena Vista, CO
 
Michael, the SOS doesn't work so well with multiple lights on one wing...

Well, I do have two per side, and the cool thing is that they will very quickly get out of sync, so the "SOS" pattern turns into a random pulse. The effect is quite dazzling, and is the mode I will fly in most of the time.
 
Trustfire

Great thread, thanks for sharing. I checked out the trustfire link, did anyone notice the 13000 lumen torch they have? Now that would make a nice landing light! Just sayin
 
Great thread, thanks for sharing. I checked out the trustfire link, did anyone notice the 13000 lumen torch they have? Now that would make a nice landing light! Just sayin
I had seen those and just received my 11000 LM Trustfire. I hooked it up to the power line in the Plane for some testing. It draws just less than 3A when hooked up to the battery. I was also quite surprised and happy to not detect any RF noise in my radios or any side effect on my Nav. The test was both on the ground and in the air (the light inside the cockpit)
This is much much brighter than my Bi-xnon light. I am planning on working on mounting them tomorrow and will take pictures and post them when completed.
 
I had seen those and just received my 11000 LM Trustfire. I hooked it up to the power line in the Plane for some testing. It draws just less than 3A when hooked up to the battery. I was also quite surprised and happy to not detect any RF noise in my radios or any side effect on my Nav. The test was both on the ground and in the air (the light inside the cockpit)
This is much much brighter than my Bi-xnon light. I am planning on working on mounting them tomorrow and will take pictures and post them when completed.

Great! Please keep us posted. did you note what the maximum voltage was on the Trustfire 11k you bought was, I noticed the 13K was rated at 12V max.

Looking forward to your report and photos.
 
Hmmm... this gets interestinger and interestinger! :)

I have a "clone" Trustfire, as identified by its physical features (no ground spring). My testing started at about 6V and ran to 13.8V and showed an almost linear decrease in input current. This is contrary to Tikidriver's findings, so there might be a 3rd combination of driver/enclosure out there.
 
Just at Costco and found a two pack of 500 lumen LED Duracell flashlights with Duracell batteries (6 batt) for $20 bucks. The lens in the front collar slides back and forth to either show the square LED light footprint or spread the light beam out PERFECTLY! They are what appears as all aluminum housing. Thats $10 a piece, heck 3 duracell batteries are $6 . I'm a flashlight guy and have NEVER seen a square pattern LED chip thats about a 1/4" square inside the lens.. In total darkness at 200' I put the narrow LED footprint of light on the side of a building and was able to take a picture of the footprint with my cell phone camera with no flash. The first person I showed it to was another RV pilot and said hes buying one too or three! I would imagine next year the 1000 lumen with be down in price to $30. Today I rate this flashlight the best for the buck!!!! Maybe if you went to Costco.com you might find them on their website. I thought about making a 4 lamp comb light. 3 for landing and one for taxi in each wing of my RV9A. Plus you get the 12 batteries to use for whatever from the two two packs.

Ron in Oregon

Ron in Oregon.
 
One thing I found after my last post. The Trustfire will actually turn on with an input voltage of 3 volts.
I have a LUX meter and I will run another test tomorrow and check the output vs input voltage.
 
Hmmm... this gets interestinger and interestinger! :)

I have a "clone" Trustfire, as identified by its physical features (no ground spring). My testing started at about 6V and ran to 13.8V and showed an almost linear decrease in input current. This is contrary to Tikidriver's findings, so there might be a 3rd combination of driver/enclosure out there.

What does your reflector housing look like when you un-screw if from the outer housing?
 
Component Design

First off, I would like to thank you all for your research and for sharing it. This is a great thread and I am excited to see how this all goes! I am building a Bushcaddy L164 and plan to fly it in the Alaska Bush. The RV forums have been stellar in providing info for my build.

Let me get to the point of the post. I am looking for a set of LED landing lights, that will mount externally. More than likely they will mount to the upper-outboard portion of the strut (high wing airplane). The Baja Squadrons look like a great solution, but CREE just came out with a new LED that produces 15,500 Lumens! (their specs)

http://www.cree.com/LED-Components-and-Modules/Products/XLamp/Arrays-NonDirectional/XLamp-CXA2590

News article about them here

http://online.wsj.com/article/PR-CO-20140212-910737.html

Available from Mouser Elecronics

http://www.mouser.com/ProductDetail/Cree-Inc/CXA2590-0000-000R00Y430F/?qs=oowD9IqsgIJeP06MH4f16Q==

I am considering trying to design my own set. Mouser also sells some reflectors. Does anyone have any ideas on how these things could mount in an external housing? or know of an existing housing that I could modify? I wont have any lack of cooling air as the housing will be out in the slipstream. Designing a driver doesn't seem too difficult and there is a whole lot if information out there on how to do it.

It is dark up here for a good portion of the year and the more light I would have available the better. Seems like there are tons of good options available!

Keep Up the good work!

Justin
 
How fast are you building that Bushcaddy? When did the XM-L come out vs. where we are now? I am all for building up your own rig, but if time is on your side you might get a retail version soon enough. Maybe the Baja's could be your first set of lights, with external mount you can upgrade pretty easily.
 
What does your reflector housing look like when you un-screw if from the outer housing?

Tiki - my reflector housing looks like the Left Hand one in the photos shown in your Post #134. No ground connection via coil but rather solely through threads on the body.

I note your observation about the lights coming on around 3V. I've found ours comes on at around ~4V. Unfortunately there's a very fast ramp up between the "is it on?" state and the "oh nuts, I shouldn't have looked right at those emitters" state, thanks to a not-so-fine voltage control on my power supply. Certainly by 5V one doesn't want to be spending much time looking at the emitters. By 6V they're at full brightness.
 
How fast are you building that Bushcaddy? When did the XM-L come out vs. where we are now? I am all for building up your own rig, but if time is on your side you might get a retail version soon enough. Maybe the Baja's could be your first set of lights, with external mount you can upgrade pretty easily.

Not a bad idea. Maybe I can get ahold of the guys at Squadron and see if they will sell me just the housing to use. Or see if they have plans to use the newest LED. The news article talks about using them in search lights, so they have to be bright.

The plane is coming together quickly. Now if I can just get the time to actually build it instead of work. The wings are on and flight controls rigged, engine comes out of the shop soon and the panel is almost done. Probably about 45 days of solid work left. Not sure how long thats going to take though.

The new Cree lights came out just last week. Im sure we will see some companies putting them in a set of lights soon. At $80 per LED, I'm Sure they wont be cheap! At least most of the reflectors are cheap.

Justin
 
Back
Top