LED’s and Continuity testing

AN LED or diode will conduct (and light in the case of the LED) in one direction and not the other. Best test for an LED is a 1.5V-3V battery. 1.5 will light a red,3V will light a white.

Dale H

Well I learned something. I would have thought it was because the LED uses arcing to light and therefor doesn't  have continuity all the way thru to the other lead like an incandescent bulb. So the only way to tell if an LED works or not is if it works ..or not?
Makes sense it's really hard to look at one and see if it's bad.

David

To test a diode, you apply a fixed current from a current source and measure the voltage drop across the diode.  This is what the "diode" setting on a volt/ohm meter does.  Unfortunately, if you have diodes that take higher voltages, like white LEDs or series-connected non-white LEDs, the meter may not have enough voltage behind the current source to get a reading.

Does your meter have a "diode" setting?

I know of no way to look at an LED and tell if it's good.  The exception is if you smoke it with really high current you can sometimes see the effect if it's a clear plastic model.

My Fluke 77 doesn't handle white LED's.  I test them with a 1K resistor and my bench supply at 12V DC. 

Also remember that these LEDS are most likely wired in series.  Probably 4 or 5 wired in series so each lamp gets about 2.2V (assuming nominal 13.8V battery).  And speculate there is probably a simple constant current source circuit to maintain steady brightness regardless of auto battery supply fluctuations (within reason of course).

  If you took the bar apart you could test each LED until the bad one is found, assuming the constant current circuit had not failed.

Relatively easy. You will need the specs on the LED's you plan on using. They have voltage and current limits which are unforgiving. Exceed a limit and the LED burns out like a fuse. A LED is simply a diode, a rectifier which happens to give off light when it is forward biased and is conducting current. Your problem is one of voltage and current. Too much of either and teh diode burns out like a fuse. A typical led strip lights forward biases at something over 0.5 volts, and can usually tolerate about 10-30 milliamperes (MA) of current. If we These are rated in milliwatts (MW) of heat dissapation. A typical flexible led strip  (the kind which is about 0.2 inch diameter and about 0.3 inch long) can dissapate about 10 MW, which means I have to adjust current and voltage according to Ohm's Law, E=IR, voltage equals curent times resistence, and P=IE, power equals current times voltage. 10 MW means I can have 10 MA and 1 volt across the diode, 5 MA and 2 V, or any combination which does not exceed any of the rated limits. So, how do you get this? Use a stepdown transformer as working with the power line directly is VERY VERY DANGEROUS. Any wall wart style will do. Radio Shack and several other electronics stores will have these at a reasonable cost. Look for one with about a 6 volt DC output. With the example of the 1 volt drop and 10 MA current, you should be able to connect 6 of these in series (anode to cathode, anode to cathode, just like batteries in series) and connect them directly across the 6 VDC supply and they all should light. If it isn't bright enough (as I am being quite conservative in my estimates), you might be able to remove one LED from the string to up the current (light output) without exceeding the current limits. Use a meter capable of measuring MA to monitor the current drawn by the string. By using a transformer to lower the voltage you use as a power source, this means you can parallel a rechargeable battery downstream of the transformer to supply power to keep the LED Light Bars  when the power line fails. This is how a lot of emergency lights in buildings work these days. The low voatage power supply lights the lights and keeps the battery charged and if the power line fails, the battery keeps the lights lit. You will need to do more research on how to add the backup battery since there is not room here to give much in the way of detail here to charge the battery and keep it in good health (overcharging is the cause of early failure of standby batteries). Using a combination of strings in series and parallel up to the current capacity of the power supply will determine your end result. I am only giving a basic outline here since I am not legally going to allow myself to be obligated in the event you do something stupid which gets you killed working with the power line drectly. Again I say, DO NOT WORK WITH THE POWER LINE DIRECTLY! ALWAYS use a stepdown transformer to protect yourself from becoming a crispy critter and NEVER assume a circuit is dead until you verify it is really deenergized. You can DIE and there is no coming back from death as we know it.

Thanks to everyone for your input. My meter is pretty old school analog about 20 years old so I did not see a diode setting (but not sure what the symbols would be for that). I tried the cotinunity test both ways on the LED terminals (old and new working light) and no needle movement. The meter had a new 9 volt battry but it sound as tough that may not have had enough power to make the test work??

Out of curiosity, I will try to dis-assemble the sealed unit to see what may have caused the issue. (BTW the dealer wanted $220 for a replacement 3rd brake light, Good grief!).

To complicate the issue, the fuse was also blown. Perhaps an internal short in the old light could have caused that??

( If you own a GM Suburban be forewarned there are 3 separate fuse panels and the 3rd brake light is fused in a different fuse block than the regular stop lights. That tid bit took me 2 hours to figure out!!)

If doing a lot of these,I would make up a test board with a CL2-N3 chip. That way only 20ma would go to the LED regardless of voltage input. This assumes testing an LED on a circuit in which all power is off and secured.

Dale H

I don't get so many bad LED's that I need a custom fixture.  A quick dive into the resistor drawers and pick up a suitable value  and it's done.

If you're getting enough bad LED's that you need a custom test fixture, you need to deal with the reason, not the testing.

Bingo. The max voltage your meter could apply to the module whether in resistance-mode or diode-mode (if it has it) is 9V.  Since diodes are the principal load inside, the resistance (measured indirectly by voltage divided by current) changes with voltage and drops dramatically as you approach 12V and the LEDs start to turn on.  That is, I think you were not applying enough test-voltage.

Even when you get inside and attempt to measure individual LEDs, it may be tricky. It appears you have a "needle" analog meter presumably the kind with selectable resistance ranges such as Rx1, Rx10, etc.  Some analog meters put out a hefty test-current when in the Rx1 range which might pop a traditional 20-mA LED.

I still say you can simply probe around with a 12V power supply and a resistor of 500 ohms or more and try to light individual LED's in the array. 

Well I have to agree with john. All you need is a 9 volt battery and a 500 ohm or better resistor. I keep a set of jumpers around with alligator clips on both ends. A set of red ones and a set of black ones. stick a resistor in line and go to the hot side of the LED and the + side of the battery a black jumper and go to the common side of the LED and the - side of the battery.... simple. I do this anyway when I make my own LED's to use in my engines and buildings . No use putting something in that doesn't work.

David