|
| If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. |
|
|||||||
|
|
Thread Tools | Display Modes |
|
|
|
#1
April 18th 04, 02:45 PM
|
|||
|
|||
Eric Jones wrote an nice description of leds as nav and strobe lights he
http://www.periheliondesign.com/down...ing%20LEDs.pdf the rear and side strobes are fairly easy do with about 5 ea 5 watt luxeons i used 6 in my tail strobe. the required strobe intensity in the forward direction is a bit tougher to accomplish with LEDs. still possible. i think for my tail strobe i will uses 50% duty cycle. not as "flashy" as a quick pulse, but actually easier see at a distance. the human eye averages for 200 ms so for quick pulses its the total light energy in the pulse that matters. stobes cant make long pulses but LEDs can, so long pulses are the right choice with leds. circuit will be like this IRF2804 --------- | | |----parallel LEDs----Resistor----+12V | LM555 |--------| | | |-----gnd --------- the 2804 is WAY overkill. it can switch 100s of amps. but its only 3 bucks, so why not use the best. i ordered a handful and when i get back form snf i will give the circuit a try. -Jeff 
|
|
#3
April 26th 04, 12:36 AM
|
|||
|
|||
|
|
|
#4
April 26th 04, 05:18 PM
|
|||
|
|||
|
|
|
#5
April 27th 04, 02:48 PM
|
|||
|
|||
|
Hi Guys...
Havent followed this thread carefully...been away for awhile...so forgive me if this has been covered already... There is one disadvantage to using mostly the forward voltage drop of a LED to "span" the battery voltage ...as opposed to say a big resistor taking a good fraction of the voltage drop... An LED's voltage drop/resistance is pretty temp dependent.... And here is what can happen....the LED gets warm.....resistance drops....more power flows through....LED gets warmer still...still more power flows....warmer,more power....can easily become a runaway situation...leading to burnout or even worse a fire somewhere in the electrical cicuit.. Now, this isnt much of a problem in very low powered LEDS, but in high powered ones that get warm to touch just sitting on the bench, its something you need to worry about.... If you have a circuit in mind...you need to make it and THEN run it in absolute hottest environment at the absolute highest voltage it will ever see (ie the 14 or whatever volts is MAX from the alternator rather than the nominal 12 from the battery)....then if that works and current limits werent exceeded back the power flow down say 10 to 20 percent just to be safe.... What the max temp of a "lexan lighting housing" at the tip of a wing in August on a no wind day sitting on a black asphalt runway at high noon in Texas? I can easily imagine it being 60 degrees or more than the temp you tested it in on the bench.....so just because it didnt suffer thermal runaway on the bench doesnt tell you much about how it"ll behave in the plane.... Another related problem is if you have multiple parallel LED/resistor paths running from the same voltage source....at room temp you may test the setup and each path is taking the same power....but then there is that pesky temp/resistance property of the LEDs...if they are different enough, when your setup is run in a higher temp or poor cooling conditions....one path may become a significantly lower resistance path than the other (the resistors can cause it as well)...and all of a sudden you have way more power flowing through one path than was intended to....and you have another thermal runaway situation... Which brings up fuses....lets say you 3 parallel LED resistor paths, running off of one wire pair that runs out the wing tips.... You design it so each branch takes a bit under 5 amps, so you have that wire pair fused at 15 amps.... If you suffer something like one dead branch in combo with a thermal runaway of another branch (or even perhaps just a thermal runaway of one branch)....you can have a situation where one branch is WAY exceeding its design limits...yet the total power isnt enough to trip the fuse back in the cockpit... So, EACH branch should be fused seperately in addition to the main overall wire fuse... Now, if youve designed and tested your circuit worth a hoot the chances of one branch going amook are pretty low, so something like "semi permanent" circuit board fuses soldered into each branch probably make more sense and would provide better reliability than removeable fuses... Another thing....precision resistors, temp stable resistors, and even resistors with a significant negative temp coefficient ARENT that expensive compared to other expensense for an airplane/high powered LED project....so the might be worth investigating, especially since your only talking a handful of them and they could significantly ease your engineering problems as well as improve the reliability of the setup... Just my 10 cents.... take care Blll
|
|
#6
April 28th 04, 05:35 PM
|
|||
|
|||
|
With a properly biased LED stack and resistor there is no thermal
runaway problem. I've never seen or heard of this happening to anyone. The slope of the diodes once they go past the knee in combination with the series resistor more than makes up for a small negative temp coefficient of the diodes. Maybe what you're saying is that you need some minimum resistance as a percentage of the total drop to guarantee stability, and this I would agree with. And I totally agree, prototype one leg of the circuit, and run it over voltage and temp (hit your corners) before you build it into your wing. Fuses- I'd still go with the standard central fuse/breaker panel. Since its a non-inductive load, you should be able to use a fuse/breaker without much over rating. I'm a big believer in keeping things simple. But often times a simple final product requires complex analysis and testing. The most complex implementations often reflect design with less than well understood inputs. (BllFs6) wrote in message ... Hi Guys... Havent followed this thread carefully...been away for awhile...so forgive me if this has been covered already... There is one disadvantage to using mostly the forward voltage drop of a LED to "span" the battery voltage ...as opposed to say a big resistor taking a good fraction of the voltage drop... An LED's voltage drop/resistance is pretty temp dependent.... And here is what can happen....the LED gets warm.....resistance drops....more power flows through....LED gets warmer still...still more power flows....warmer,more power....can easily become a runaway situation...leading to burnout or even worse a fire somewhere in the electrical cicuit.. Now, this isnt much of a problem in very low powered LEDS, but in high powered ones that get warm to touch just sitting on the bench, its something you need to worry about.... If you have a circuit in mind...you need to make it and THEN run it in absolute hottest environment at the absolute highest voltage it will ever see (ie the 14 or whatever volts is MAX from the alternator rather than the nominal 12 from the battery)....then if that works and current limits werent exceeded back the power flow down say 10 to 20 percent just to be safe.... What the max temp of a "lexan lighting housing" at the tip of a wing in August on a no wind day sitting on a black asphalt runway at high noon in Texas? I can easily imagine it being 60 degrees or more than the temp you tested it in on the bench.....so just because it didnt suffer thermal runaway on the bench doesnt tell you much about how it"ll behave in the plane.... Another related problem is if you have multiple parallel LED/resistor paths running from the same voltage source....at room temp you may test the setup and each path is taking the same power....but then there is that pesky temp/resistance property of the LEDs...if they are different enough, when your setup is run in a higher temp or poor cooling conditions....one path may become a significantly lower resistance path than the other (the resistors can cause it as well)...and all of a sudden you have way more power flowing through one path than was intended to....and you have another thermal runaway situation... Which brings up fuses....lets say you 3 parallel LED resistor paths, running off of one wire pair that runs out the wing tips.... You design it so each branch takes a bit under 5 amps, so you have that wire pair fused at 15 amps.... If you suffer something like one dead branch in combo with a thermal runaway of another branch (or even perhaps just a thermal runaway of one branch)....you can have a situation where one branch is WAY exceeding its design limits...yet the total power isnt enough to trip the fuse back in the cockpit... So, EACH branch should be fused seperately in addition to the main overall wire fuse... Now, if youve designed and tested your circuit worth a hoot the chances of one branch going amook are pretty low, so something like "semi permanent" circuit board fuses soldered into each branch probably make more sense and would provide better reliability than removeable fuses... Another thing....precision resistors, temp stable resistors, and even resistors with a significant negative temp coefficient ARENT that expensive compared to other expensense for an airplane/high powered LED project....so the might be worth investigating, especially since your only talking a handful of them and they could significantly ease your engineering problems as well as improve the reliability of the setup... Just my 10 cents.... take care Blll
|
|
#7
April 20th 04, 01:13 AM
|
|||
|
|||
|
Jeff,
Who did you order the 2804's from? Thanks Dean "Jeff Peterson" wrote in message om... Eric Jones wrote an nice description of leds as nav and strobe lights he http://www.periheliondesign.com/down...ing%20LEDs.pdf the rear and side strobes are fairly easy do with about 5 ea 5 watt luxeons i used 6 in my tail strobe. the required strobe intensity in the forward direction is a bit tougher to accomplish with LEDs. still possible. i think for my tail strobe i will uses 50% duty cycle. not as "flashy" as a quick pulse, but actually easier see at a distance. the human eye averages for 200 ms so for quick pulses its the total light energy in the pulse that matters. stobes cant make long pulses but LEDs can, so long pulses are the right choice with leds. circuit will be like this IRF2804 --------- | | |----parallel LEDs----Resistor----+12V | LM555 |--------| | | |-----gnd --------- the 2804 is WAY overkill. it can switch 100s of amps. but its only 3 bucks, so why not use the best. i ordered a handful and when i get back form snf i will give the circuit a try. -Jeff
|
|
#8
April 20th 04, 08:03 PM
|
|||
|
|||
|
You can get all the parts from Digi-key for less than $10 +S/H except
those fancy LEDs. Regards "Dean Head" wrote in message ... Jeff, Who did you order the 2804's from? Thanks Dean
|
|
#9
April 24th 04, 03:27 AM
|
|||
|
|||
|
"Dean Head" wrote in message ...
Jeff, Who did you order the 2804's from? Thanks Dean digikey. built the circuit today and will post photos and schematic soon -Jeff
|
|
#10
April 24th 04, 03:06 PM
|
|||
|
|||
|
Jeff,
Anxious to see the results. I am thinking of some pretty slick ways to embed these into my Cozy winglets. I hope to find a way to replace the Xenon strobes completely. Dean "Jeff Peterson" wrote in message om... "Dean Head" wrote in message ... Jeff, Who did you order the 2804's from? Thanks Dean digikey. built the circuit today and will post photos and schematic soon -Jeff
|
| Thread Tools | |
| Display Modes | |
|
|
Hybrid Mode
