Estimate max current for LEDs from various broken lights?

[RoGeorge]

Most LEDs from broken LED lights are still good.  Some are quite powerful, thus tempting to reuse in other projects (not to repair the original lightbulb).  They usually already have their own radiator, which is a big plus.

How to find the maximum current?  Maybe to measure the LED's (or radiator) temperature while increasing the current?  What would be a safe max temperature for 24/7 use and DC (not pulsed) power?

Any other ideas for how to estimate the max current faster than just letting the LEDs to heat their own radiator?

[S. Petrukhin]

Only the datasheet can tell you about this. Rated current for LEDs in very large ranges from a few mA to several amps. It should also be taken into account that the lamps use a mixed circuit of inclusions in series and in parallel, which has a direct effect on the total current, which is a generalization of individual currents.

[RoGeorge]

Indeed, they are combination of serial and parallel connected LEDs, and many times some of them are dead, so they must be either bridged or cut out of the circuit.  Also, they are almost never marked so no chance for a datasheet.  Some of them don't even have the rest of the lightbulb any more, so no way to deduce the max current from the specs typed on the light's socket.

Assuming I cut the traces to measure only one LED, are there any chances to find a good max current from something other than temperature, for example by looking at the I-V curve, or maybe at something else?

[BravoV]

As olde saying ... in order to make money, sometimes you have to spend more money. 

Similar to your case here, if you insist, and don't like to assume, buy another new identical light, tear it down and measure the current. 

[ealex]

hello

you can usually infer the operation conditions from the driver schematic.

usually the driver is ok - i've  mostly seen failed leds.

if it's a current mode driver you can usually check the schematic, or you can bypass the dead led(s) and check the current that way

i don't think the remaining leds will be in their prime - maybe you can bin them by checking how much light they are still emitting, and keep the best ones to fix other units

[S. Petrukhin]

hello

you can usually infer the operation conditions from the driver schematic.

usually the driver is ok - i've  mostly seen failed leds.

if it's a current mode driver you can usually check the schematic, or you can bypass the dead led(s) and check the current that way

i don't think the remaining leds will be in their prime - maybe you can bin them by checking how much light they are still emitting, and keep the best ones to fix other units

It is very easy to check the current that the designers decided to apply to the LEDs from the driver: close it with an ammeter. 

[Ground_Loop]

While measuring current through and voltage across the LED with a reasonable resistor in series with the power supply, slowly increase power supply voltage until voltage across the LED levels off. At that point the current observed will be your answer.

[Ian.M]

You can set up to use LED Vf to measure the die temperature.  See: https://www.electronicdesign.com/markets/lighting/article/21775687/use-forward-voltage-drop-to-measure-junction-temperature

Pick the max die temperature you are happy with, calibrate the LED for Vf vs temperature, then ramp up the current (with brief pulsed interruptions to measure the Vf at the test current you calibrated for temperature at) till the die reaches your chosen max. temperature then back off the current till the temperature just starts to drop.

[RoGeorge]

You can set up to use LED Vf to measure the die temperature.

That's a very good idea, thank you!