Need some LED help...
Need some LED help...
I noticed a couple of weeks ago that one of the six green LED's on the port side had gone dim. Not out, but dim. A couple days ago, I flipped them on for someone to see the lights, and now I have two that are completely OUT!! I don't know enough about LED's to begin trouble shooting, other than to check the solder connections. I thought maybe they had broken from vibration from taxiing, but they appear to still be solid connections.
I started my first condition inspection today to get ready for Tuesday, and I pulled the wingtip to start figuring this out...
Any suggestions on where to start?
I started my first condition inspection today to get ready for Tuesday, and I pulled the wingtip to start figuring this out...
Any suggestions on where to start?
Bad news is it looks like current is passing through the burnt out ones. Just for fun, first check if there is any voltage on your aluminum mounting frame. There is a chance a few of them got shorted. If not, looks like you might of cooked them. How did you glue them to the aluminum? They need a special heat conducting adhesive because those puppies do get hot when not hooked to a heat sink. My best guess is you melted one
Mike Bullock
http://www.rvplane.com
RV-7 | Superior IO-360 | Whirlwind 200RV
Garmin GTN650 | GRT Dual Sport SX EFIS
Status: FLYING!
http://www.rvplane.com
RV-7 | Superior IO-360 | Whirlwind 200RV
Garmin GTN650 | GRT Dual Sport SX EFIS
Status: FLYING!
Chad,
I can't see how the circuit is laid out... cant make out the +- signs.
Can you explain how they are wired and what do you have driving em?
What model LED's are they? Some must be isollated and some don't have to be...
The bad news is like Mike sez, those are gonna be bad and there won't be no reviving em. Might as well get some ordered....We may be able to make em last longer though depending on what you have and how their wired.
I can't see how the circuit is laid out... cant make out the +- signs.
Can you explain how they are wired and what do you have driving em?
What model LED's are they? Some must be isollated and some don't have to be...
The bad news is like Mike sez, those are gonna be bad and there won't be no reviving em. Might as well get some ordered....We may be able to make em last longer though depending on what you have and how their wired.
Brantel (Brian Chesteen),
RV12-IS, #121606, N912BC - Building Now!
RV10, #41942, N????? - Project Sold
RV-7/TU, #72823, N159SB - SOLD
RV12-IS, #121606, N912BC - Building Now!
RV10, #41942, N????? - Project Sold
RV-7/TU, #72823, N159SB - SOLD
Hmmm...interesting. These are Jeff Bordelon lights...I did not make these. He used the special glue to mount them (silver looking stuff).
I'll take some closer detailed pics of them in the morning, but its weird that these have worked just fine for a year and now out of the blue,two of them quit.
They are just wired to the switch...
I'll take some closer detailed pics of them in the morning, but its weird that these have worked just fine for a year and now out of the blue,two of them quit.
They are just wired to the switch...
Chad,
I zoomed in and found what I needed on the LEDs.
Looks like you have two parallel runs of three LEDs in series. You have one shorted LED on each side of the circuit.
That means that whatever your drivers rating is, you are driving each group of three at half that things rating.
What MA is your driver you are using and what make and model are the LED's?
I zoomed in and found what I needed on the LEDs.
Looks like you have two parallel runs of three LEDs in series. You have one shorted LED on each side of the circuit.
That means that whatever your drivers rating is, you are driving each group of three at half that things rating.
What MA is your driver you are using and what make and model are the LED's?
Brantel (Brian Chesteen),
RV12-IS, #121606, N912BC - Building Now!
RV10, #41942, N????? - Project Sold
RV-7/TU, #72823, N159SB - SOLD
RV12-IS, #121606, N912BC - Building Now!
RV10, #41942, N????? - Project Sold
RV-7/TU, #72823, N159SB - SOLD
Hi,
I don't know a lot about specific Lumileds products, but I have more than a passing familiarity with WBG LED products. That particular package looks like Lumileds Star packge which is an excellent product, but the chassis is not neutral for that product, each package should be isolated electrically from others. If I understand the circuit correctly, it looks like you have two parallel sets of 3 LEDs in series. I don't see any current limiting resistors or other mechanism to insure you get an even split of current between the two sets of three. In the on-state, the impedance of the devices vary and the current voltage curve is awful steep, so you cannot be sure of even current splitting without some active means - that can lead to current hogging that may exceed the devices limits. Perhaps the power supply you have is a.c. That would be a novel way to accomplish an even current balance if you have a fairly high frequnecy drive, though as I recall, Lumileds does not encourage operating their devices in reverse blocking. I think that there may be some good news since in my experience, when power LEDs fail, they fail open. I suspect the devices are ok, but the circuit (including the mounting plate) isn't doing everything you want.
my 2c, and congratulations on nearing completion on your excellent project!
Joe S>
I don't know a lot about specific Lumileds products, but I have more than a passing familiarity with WBG LED products. That particular package looks like Lumileds Star packge which is an excellent product, but the chassis is not neutral for that product, each package should be isolated electrically from others. If I understand the circuit correctly, it looks like you have two parallel sets of 3 LEDs in series. I don't see any current limiting resistors or other mechanism to insure you get an even split of current between the two sets of three. In the on-state, the impedance of the devices vary and the current voltage curve is awful steep, so you cannot be sure of even current splitting without some active means - that can lead to current hogging that may exceed the devices limits. Perhaps the power supply you have is a.c. That would be a novel way to accomplish an even current balance if you have a fairly high frequnecy drive, though as I recall, Lumileds does not encourage operating their devices in reverse blocking. I think that there may be some good news since in my experience, when power LEDs fail, they fail open. I suspect the devices are ok, but the circuit (including the mounting plate) isn't doing everything you want.
my 2c, and congratulations on nearing completion on your excellent project!
Joe S>
I think that as Brian was indicating, some quality time with a multimeter on the installation (removed from the plane) is the best way to procede. If you confirm that the mounting plate is floating at some potential relative to any part of the circuit, then we're on track about electrical isolation complicating things and some installation mods are needed. If there is no potential, then you probably do have some dead chips and could replace them. Actually in this case, I hope I am wrong and you just got two bogus chips because it would be a ton easier and cheaper to just replace two bad ones. I think the stars are like $6 or so.
I did bring up balancing the current splitting, while I do not see a method to do that in the pic., I don't know all about the power supply, so there may be a clever methodology in play that I don't see. If there is no provision for that, it may not be strictly necessary - and I guess that is one of the criteria for including anything in an airplane - then a simple and easy way to accomplish that would be to install a low value resistor in each of the parallel paths. Unfortunately, I am not that familiar with power resiters, so I can't give a really competent reccomendation, but usually you use something like a 1/4 to 1 ohm resistor that can handle the current you have. These resistors are not the familiar little wasp shaped guys, they are honkers, sometimes with heat sinks needed, but they are not terribly expensive.
best of luck,
Joe S>
I did bring up balancing the current splitting, while I do not see a method to do that in the pic., I don't know all about the power supply, so there may be a clever methodology in play that I don't see. If there is no provision for that, it may not be strictly necessary - and I guess that is one of the criteria for including anything in an airplane - then a simple and easy way to accomplish that would be to install a low value resistor in each of the parallel paths. Unfortunately, I am not that familiar with power resiters, so I can't give a really competent reccomendation, but usually you use something like a 1/4 to 1 ohm resistor that can handle the current you have. These resistors are not the familiar little wasp shaped guys, they are honkers, sometimes with heat sinks needed, but they are not terribly expensive.
best of luck,
Joe S>
If there is no driver and no current limiting resistor, that is what killed em. Too much current! If you replace em they will fail again.
You need to add a resistor to the power leads feeding em. Cheap and it works!!!
You need to add a resistor to the power leads feeding em. Cheap and it works!!!
Brantel (Brian Chesteen),
RV12-IS, #121606, N912BC - Building Now!
RV10, #41942, N????? - Project Sold
RV-7/TU, #72823, N159SB - SOLD
RV12-IS, #121606, N912BC - Building Now!
RV10, #41942, N????? - Project Sold
RV-7/TU, #72823, N159SB - SOLD
Hey Chad.
Sounds like you found it - the glob trumps all of the other cards.
If you could post the spec.'s of the power supply you have and the LEDs, that should help figuring the sizing the resistors needed.
If you have a true constant current power supply, then this app. is a bit different from the simpler "put your 12-ish volts across a string of LEDs, expect a 3.whatever voltage drop across each in the string and size the current limiting reisister to match the deisred current under the remaining voltage".
With a constant current power supply that you split to two branches, you could use a lower value resistor since you are really trying to flatten the I-V curve I mentioned in an earlier post.
BTW - did you see any voltage on the mounting plate relative to any of the terminals? I am suspicious since the two dead LEDs are the only ones directly connected to the (+) power supply lead. If an isolation issue remains, then current limiting resistors may not completely address this failure mode. Sorry to pile on more things to wonder about.
Joe S>
Sounds like you found it - the glob trumps all of the other cards.
If you could post the spec.'s of the power supply you have and the LEDs, that should help figuring the sizing the resistors needed.
If you have a true constant current power supply, then this app. is a bit different from the simpler "put your 12-ish volts across a string of LEDs, expect a 3.whatever voltage drop across each in the string and size the current limiting reisister to match the deisred current under the remaining voltage".
With a constant current power supply that you split to two branches, you could use a lower value resistor since you are really trying to flatten the I-V curve I mentioned in an earlier post.
BTW - did you see any voltage on the mounting plate relative to any of the terminals? I am suspicious since the two dead LEDs are the only ones directly connected to the (+) power supply lead. If an isolation issue remains, then current limiting resistors may not completely address this failure mode. Sorry to pile on more things to wonder about.
Joe S>
Chad,
Go to Radio Shack and get a couple 10 ohm 10 watt power resistors.. One for the red side and one for the green.
Put them in between your switched wire from the panel and the led array. You will need to figure out some way to mount em and they will get pretty warm in use.
Go to Radio Shack and get a couple 10 ohm 10 watt power resistors.. One for the red side and one for the green.
Put them in between your switched wire from the panel and the led array. You will need to figure out some way to mount em and they will get pretty warm in use.
Brantel (Brian Chesteen),
RV12-IS, #121606, N912BC - Building Now!
RV10, #41942, N????? - Project Sold
RV-7/TU, #72823, N159SB - SOLD
RV12-IS, #121606, N912BC - Building Now!
RV10, #41942, N????? - Project Sold
RV-7/TU, #72823, N159SB - SOLD
Potential TMI alert!
Ever read or hear something and only later realize what was said? Another option, perhaps what you are suggesting Brantel, is to forego the constant current power supply and just hook up strings of 3 LEDs to ships power with a current limiting resistor in each string. I have wondered why so many LED installations do use power supplies and then find attendant electrical noise and I have wondered why not use simple passive elements. I always assumed because of the higher power levels, but I think that simpler approach I touched on earlier would work too. You might lose some efficiency with the resistor compared to using a smart power supply, but you save weight and reduce part count, so that’s probably a favorable trade in a plane .
Here is an on-line calculator to do what you want:
http://led.linear1.org/1led.wiz
Assuming you are using the 1 W Luxeon Stars (LXHL-MM1D): http://ledsupply.com/lxhl-mm1d.php
They are rated for 350mA at 3.42V. So plugging three of these puppies into the calculator, you input 350mA and 10.26V (3 X 3.42). That tells you that at 14.4V supplied, you want a resistor of 12 ohms that can dissipate 1.5W. The calculator will gripe that the voltage range is irrationally high, but we are using 3 LEDs in the chain.
Here is a suitable resistor
http://shop.vetcosurplus.com/catalog/pr ... ts_id=8591
to be more conservative, you could step up to a 13 ohm resisor
http://shop.vetcosurplus.com/catalog/pr ... ts_id=8592
Or a larger wattage class resistor as suggested is a great plan - they will get warm in operation.
You can do a similar evaluation of the red guys which probably have different I-V properties.
Ever read or hear something and only later realize what was said? Another option, perhaps what you are suggesting Brantel, is to forego the constant current power supply and just hook up strings of 3 LEDs to ships power with a current limiting resistor in each string. I have wondered why so many LED installations do use power supplies and then find attendant electrical noise and I have wondered why not use simple passive elements. I always assumed because of the higher power levels, but I think that simpler approach I touched on earlier would work too. You might lose some efficiency with the resistor compared to using a smart power supply, but you save weight and reduce part count, so that’s probably a favorable trade in a plane .
Here is an on-line calculator to do what you want:
http://led.linear1.org/1led.wiz
Assuming you are using the 1 W Luxeon Stars (LXHL-MM1D): http://ledsupply.com/lxhl-mm1d.php
They are rated for 350mA at 3.42V. So plugging three of these puppies into the calculator, you input 350mA and 10.26V (3 X 3.42). That tells you that at 14.4V supplied, you want a resistor of 12 ohms that can dissipate 1.5W. The calculator will gripe that the voltage range is irrationally high, but we are using 3 LEDs in the chain.
Here is a suitable resistor
http://shop.vetcosurplus.com/catalog/pr ... ts_id=8591
to be more conservative, you could step up to a 13 ohm resisor
http://shop.vetcosurplus.com/catalog/pr ... ts_id=8592
Or a larger wattage class resistor as suggested is a great plan - they will get warm in operation.
You can do a similar evaluation of the red guys which probably have different I-V properties.
Re: Need some LED help...
Well there's your problem right there! The green LEDs are unhappy about being on the port side instead of the starboard side...cjensen wrote:I noticed a couple of weeks ago that one of the six green LED's on the port side had gone dim.
Buck Wyndham
RV-8 N18XL (Reserved)
Northern Illinois
RV-8 N18XL (Reserved)
Northern Illinois
Re: Need some LED help...
LOL!!! FINALLY someone noticed that!!BuckWynd wrote:Well there's your problem right there! The green LEDs are unhappy about being on the port side instead of the starboard side...cjensen wrote:I noticed a couple of weeks ago that one of the six green LED's on the port side had gone dim.
I did accidentally type that, but left (pun) it alone to see if anyone would catch it!
Good eye Buck!