LED Indicators Keep Failing
Discussion
The OS rear LED indicators on my A8 have started failing every few months. The first time I got some more LEDs, desoldered the failed ones from the board and put new ones on, sorted. A couple of months on and probably half had failed so I replaced them as well. The failure light has just come on again.
I can soon dismantle the unit and put a few more on but I'd rather solve the cause than keep soldering new LEDs on.
What's a likely cause of the repeated failure and how can I track it down with nothing more than a multimeter?
I can soon dismantle the unit and put a few more on but I'd rather solve the cause than keep soldering new LEDs on.
What's a likely cause of the repeated failure and how can I track it down with nothing more than a multimeter?
Is voltage/current regulation done in the lamp ?
Its normally overcurrent that kills a LED, so either the voltage is too high or the current limiting device is not right.
Depends on the circuit design. If it had been 15 years ago, it would probably be just a series resistor, now it might be a LED driver IC.
Thing is, a LED has 0.7 of a volt dropped across it regardless of the current so something needs to drop the voltage and limit the current when fed from 12V. I bet its a small IC on the circuit board thats allowing too much current (either that or the wrong LED's)
Its normally overcurrent that kills a LED, so either the voltage is too high or the current limiting device is not right.
Depends on the circuit design. If it had been 15 years ago, it would probably be just a series resistor, now it might be a LED driver IC.
Thing is, a LED has 0.7 of a volt dropped across it regardless of the current so something needs to drop the voltage and limit the current when fed from 12V. I bet its a small IC on the circuit board thats allowing too much current (either that or the wrong LED's)
Is it an OEM or aftermarket unit?
Most probably LEDs are getting a tension which is out of range, you may try to check with the multimeter the tension on the LED on the working unit and compare with the failing one.
The above if my understanding is correct and you have one unit only with this problem.
Most probably LEDs are getting a tension which is out of range, you may try to check with the multimeter the tension on the LED on the working unit and compare with the failing one.
The above if my understanding is correct and you have one unit only with this problem.
Is it the replacement LEDs which have failed on the second occasion, or more of the old ones.
If it's more old ones, I'd just change all of LEDs for new ones and it should last as long as the unit did originally.
If it's the new LEDs which failed this time there may be something wrong with the drive circuit causing it to put too much voltage across them. If not sure how complex that is, but if it's a fairly basic arrangement with LEDs on series with a resistor and a DC drive, you will be able to measure the voltage with a multimeter and check against the LED datasheet. If the drive circuit is more complex, using a high frequency pulsed drive (to control brightness and synthesise "fading" as it switches on and off), you would need something better than a meter to diagnose unfortunately.
Another possibility would be that the LEDs are getting too hot. Is there a heat sink which has come detached or is very dirty?
If it's more old ones, I'd just change all of LEDs for new ones and it should last as long as the unit did originally.
If it's the new LEDs which failed this time there may be something wrong with the drive circuit causing it to put too much voltage across them. If not sure how complex that is, but if it's a fairly basic arrangement with LEDs on series with a resistor and a DC drive, you will be able to measure the voltage with a multimeter and check against the LED datasheet. If the drive circuit is more complex, using a high frequency pulsed drive (to control brightness and synthesise "fading" as it switches on and off), you would need something better than a meter to diagnose unfortunately.
Another possibility would be that the LEDs are getting too hot. Is there a heat sink which has come detached or is very dirty?
I *think* there's five resistors wired in parallel and the LEDs are also parallel wired. The indicators are the top two boards.
What I think happens is the + is the top wire, it runs along a track to the resistor bank then on to the parallel LEDs.
Until I pull it again I won't know which ones have failed.
My test rig is a stripped down USB cable that I temporarily attach to the board and then plug it into an Apple charger. Everything is fine for a few months when I put it back on the car.
What I think happens is the + is the top wire, it runs along a track to the resistor bank then on to the parallel LEDs.
Until I pull it again I won't know which ones have failed.
My test rig is a stripped down USB cable that I temporarily attach to the board and then plug it into an Apple charger. Everything is fine for a few months when I put it back on the car.
Gary C said:
Is voltage/current regulation done in the lamp ?
Its normally overcurrent that kills a LED, so either the voltage is too high or the current limiting device is not right.
Depends on the circuit design. If it had been 15 years ago, it would probably be just a series resistor, now it might be a LED driver IC.
Thing is, a LED has 0.7 of a volt dropped across it regardless of the current so something needs to drop the voltage and limit the current when fed from 12V. I bet its a small IC on the circuit board thats allowing too much current (either that or the wrong LED's)
I've got LEDs that show up as 2.6v on my meter. Might be a good idea to test replcament LED on a meter to see what th voltage (diode test) is and then set the circuit to let them get the correct voltage on them. Its normally overcurrent that kills a LED, so either the voltage is too high or the current limiting device is not right.
Depends on the circuit design. If it had been 15 years ago, it would probably be just a series resistor, now it might be a LED driver IC.
Thing is, a LED has 0.7 of a volt dropped across it regardless of the current so something needs to drop the voltage and limit the current when fed from 12V. I bet its a small IC on the circuit board thats allowing too much current (either that or the wrong LED's)
For simple LED circuit ,the idea is to limit the current to less than the max( I'd go further and say 50%) using the formula
for the series resistor r = (car volts ( I'd say 14.4) - forward volts of LED )/ 50%of the ideal max LED current. THEN work out the power needs of the resistor, from W( power rating of resistor ) = Amps ( flowing through circuit and LED) times the voltage across the resistor( = 14.4-forward drop of LED).
e.g. IF diode has a drop of 0.7v( when tested on meter, with diode lit) then FORWARD DROP is 0.7.
14.4-.7 = 13.7.
For a LED with a current of 10mA, this would need a resistor of9nearest value for 10 mA= 1200 ohms, but I'd start off with 2000 ohms .
WinstonWolf said:
What's a likely cause of the repeated failure and how can I track it down with nothing more than a multimeter?
Overheating, probably caused by excessive current. To know how to fix it you would need to understand how the LEDs are being driven and in particular how the current is controlled. It might be a simple series resistance, in which case a bigger resistance might be all you need.[quote=Who me ?]
I've got LEDs that show up as 2.6v on my meter. Might be a good idea to test replcament LED on a meter to see what th voltage (diode test) is and then set the circuit to let them get the correct voltage on them.
For simple LED circuit ,the idea is to limit the current to less than the max( I'd go further and say 50%) using the formula
for the series resistor r = (car volts ( I'd say 14.4) - forward volts of LED )/ 50%of the ideal max LED current. THEN work out the power needs of the resistor, from W( power rating of resistor ) = Amps ( flowing through circuit and LED) times the voltage across the resistor( = 14.4-forward drop of LED).
e.g. IF diode has a drop of 0.7v( when tested on meter, with diode lit) then FORWARD DROP is 0.7.
14.4-.7 = 13.7.
For a LED with a current of 10mA, this would need a resistor of9nearest value for 10 mA= 1200 ohms, but I'd start off with 2000 ohms .
[/quote]
Normal silicon diode drop is about 0.6v or so. LEDs will be in the 2.2 -2.6v region.
Where did you get your replacements? There are plenty of crap cheap ones that won’t last available online.
Get replacements from a reputable supplier - decent brand such as kingbright and check the spec on the supplier website. Not all LEDs are created equally!
I've got LEDs that show up as 2.6v on my meter. Might be a good idea to test replcament LED on a meter to see what th voltage (diode test) is and then set the circuit to let them get the correct voltage on them.
For simple LED circuit ,the idea is to limit the current to less than the max( I'd go further and say 50%) using the formula
for the series resistor r = (car volts ( I'd say 14.4) - forward volts of LED )/ 50%of the ideal max LED current. THEN work out the power needs of the resistor, from W( power rating of resistor ) = Amps ( flowing through circuit and LED) times the voltage across the resistor( = 14.4-forward drop of LED).
e.g. IF diode has a drop of 0.7v( when tested on meter, with diode lit) then FORWARD DROP is 0.7.
14.4-.7 = 13.7.
For a LED with a current of 10mA, this would need a resistor of9nearest value for 10 mA= 1200 ohms, but I'd start off with 2000 ohms .
[/quote]
Normal silicon diode drop is about 0.6v or so. LEDs will be in the 2.2 -2.6v region.
Where did you get your replacements? There are plenty of crap cheap ones that won’t last available online.
Get replacements from a reputable supplier - decent brand such as kingbright and check the spec on the supplier website. Not all LEDs are created equally!
SkodaIan said:
Is it the replacement LEDs which have failed on the second occasion, or more of the old ones.
If it's more old ones, I'd just change all of LEDs for new ones and it should last as long as the unit did originally.
If it's the new LEDs which failed this time there may be something wrong with the drive circuit causing it to put too much voltage across them. If not sure how complex that is, but if it's a fairly basic arrangement with LEDs on series with a resistor and a DC drive, you will be able to measure the voltage with a multimeter and check against the LED datasheet. If the drive circuit is more complex, using a high frequency pulsed drive (to control brightness and synthesise "fading" as it switches on and off), you would need something better than a meter to diagnose unfortunately.
Another possibility would be that the LEDs are getting too hot. Is there a heat sink which has come detached or is very dirty?
ThisIf it's more old ones, I'd just change all of LEDs for new ones and it should last as long as the unit did originally.
If it's the new LEDs which failed this time there may be something wrong with the drive circuit causing it to put too much voltage across them. If not sure how complex that is, but if it's a fairly basic arrangement with LEDs on series with a resistor and a DC drive, you will be able to measure the voltage with a multimeter and check against the LED datasheet. If the drive circuit is more complex, using a high frequency pulsed drive (to control brightness and synthesise "fading" as it switches on and off), you would need something better than a meter to diagnose unfortunately.
Another possibility would be that the LEDs are getting too hot. Is there a heat sink which has come detached or is very dirty?
Need to know if the replacement LED's are blowing or some of the originals before guessing a possible cause
WinstonWolf said:
I *think* there's five resistors wired in parallel and the LEDs are also parallel wired. The indicators are the top two boards.
What I think happens is the + is the top wire, it runs along a track to the resistor bank then on to the parallel LEDs.
Until I pull it again I won't know which ones have failed.
My test rig is a stripped down USB cable that I temporarily attach to the board and then plug it into an Apple charger. Everything is fine for a few months when I put it back on the car.
If there are LEDs wired in parallel without current sharing resistors (and that look to be the case from the PIC), then the LEDs have to be reasonably well matched in terms of both forward voltage and Vf temperature coefficient. Replacing just some with random off the shelf parts is almost guaranteed to cause problems with current sharing, with the diodes with lower Vf taking the bulk of the current.What I think happens is the + is the top wire, it runs along a track to the resistor bank then on to the parallel LEDs.
Until I pull it again I won't know which ones have failed.
My test rig is a stripped down USB cable that I temporarily attach to the board and then plug it into an Apple charger. Everything is fine for a few months when I put it back on the car.
The best solution would be to replace all the LEDs, and use branded LEDs from a reputable component distributor rather than bulk packs of LEDs from eBay, the failure rate on the unbranded Chinese parts can be dire especially when run close to their maximum current rating and at elevated temperatures. Are these standard 5mm LEDs?
Paralleling LEDs is a pretty crappy solution in general, since even with decent LEDs you never get perfect current sharing which results in brightness variation. I'd expect this sort of thing from aftermarket Chinese lights but a little surprised if that's an OEM light.
Thanks folks, sounds like I should probably replace them all from a batch which isn't a major problem.
I'm currently using these
https://www.ebay.co.uk/itm/161155991147
Which seem to be popular as the A8 LEDs do seem to fail. The brightness etc is a good match.
Does anyone have any suggestions for a UK source of decent ones?
I'm currently using these
https://www.ebay.co.uk/itm/161155991147
Which seem to be popular as the A8 LEDs do seem to fail. The brightness etc is a good match.
Does anyone have any suggestions for a UK source of decent ones?
Gassing Station | Home Mechanics | Top of Page | What's New | My Stuff