LED array out

[shackscs]

Some time ago I built a series of arrays that fit inside of an old Solaris shell.

There are 12 leds per array, wired negative to positive. Some the arrays have gone out.

I am guessing some of the leds have gone out.

My question is how do I troubleshoot this issue without having to unsolder each LED to test to see if it works?

I do have a voltage meter to assist if need be.

I have also tested the drivers and each one is working fine.

 

[rwb500]

use a piece of wire to bridge each LED on the broken strings. if bridging one diode lights the rest up, that diode has failed open. You could also use two AA batteries to light up the diodes one by one.

i would say it is unlikely that multiple LED's have failed. you may have some short circuit issues, likely involving the heatsink.

 

[oreo57]

Some time ago I built a series of arrays that fit inside of an old Solaris shell.

There are 12 leds per array, wired negative to positive. Some the arrays have gone out.

I am guessing some of the leds have gone out.

My question is how do I troubleshoot this issue without having to unsolder each LED to test to see if it works?

I do have a voltage meter to assist if need be.

I have also tested the drivers and each one is working fine.

A digital VOM will usually dimly light 1-3W LEDs if set to ohms or "diode"
All you need is to use the correct polarity..
As also mentioned a battery will do the same.. A single Lithium button cell will work and light them, usually quite brightly..

using the VOM and testing from say plus to heatsink and then minus to heatsink (using the correct probe polarity) will show if you have any shorts to the sink..

 

[shackscs]

Thanks,

I tried the battery option and got nothing.

What's the easiest way to troubleshoot a short?

 

[oreo57]

Thanks,

I tried the battery option and got nothing.

What's the easiest way to troubleshoot a short?

2 AA's may or may not work.. Lithiums are over 3v(more on the line of 4 for the bigger ones) and above most diode V(f)'s (except big multichips of course)
you can test 3 in series w/ a 9V.......

 

[theatrus]

Battery across each LED? Got the polarity correct?

Do you have a DMM?

 

[perkint]

If using constant current drivers, don't power up your system and then try shorting over each LED

A constant current driver will keep ramping up the voltage as far as it can, until it gets to its rated current. If the circuit is broken and powered up, this eans the driver will be trying to pump as much out as it an. Make the circuit while this is happening (eg by bridging the shot LED) and you run the risk of blowijg the others! Never make the connections between a constant current driver and LEDs whilst the power is on!

As has been said, use a multimeter or batteries

Tim

 

[rwb500]

If using constant current drivers, don't power up your system and then try shorting over each LED

A constant current driver will keep ramping up the voltage as far as it can, until it gets to its rated current. If the circuit is broken and powered up, this eans the driver will be trying to pump as much out as it an. Make the circuit while this is happening (eg by bridging the shot LED) and you run the risk of blowijg the others! Never make the connections between a constant current driver and LEDs whilst the power is on!

As has been said, use a multimeter or batteries

Tim

I thought it went without saying to bridge them while the driver was turned off, and then turn it on.

the easiest way to check for a short is to inspect your soldering joints. A good solder should not be anywhere close to shorting on the heatsink or the wrong pad of a diode. anything within 1-2MM of shorting should be re-done anyway.

 

[perkint]

I thought it went without saying to bridge them while the driver was turned off, and then turn it on.

LOL! This is the internet! Nothing goes without saying! You may be confident of the people in the thread knowing what they are doing, but there will be someone googling, trying to work out why their off the shelf unit has packed up who finds the thread...



Tim