I have an array of 8 SMD LEDs driven by a 74LS244 with each one having a 100Ω resistor in series.
The '244 pulls down, then the resistor, then the LED to +5V.
Two of the LEDs are now very dim (but still lit) and the others are OK.
The resistors read correct.
Checking the voltages. the outputs of the '244 are correctly low for all LEDs
The voltage at the resistor/LED connections read higher for the dodgy LEDs. This implies higher current (more voltage across the resistor) and lower voltage across the LED so, to me, this looks like the LED is faulty but it seems an unusual fault.

The LEDS are 2.54mm and placed right next to each other and it is most unlikely that I will get matching colour (blue BTW) and brightness so I would need to replace all eight. I just want some reassurance that this is a possible way that an LED could fail beofre I start the challenging task of replacing them all.

 

The typical failure mode for high quality LEDs is reduced light output. I read it in an HP Optoelectronics Manual. That book is a gold mine of information.

The cheap stuff from China seems to fail with bad wire bonds. I've had several that stop working when they warm up. So far, they start working after they've cooled down.

 

The maximum current a 74LS244 buffer can sink is 24mA. You are using a 100 ohm resistors in series with the LEDs. The voltage drop across a blue LED can be anything from less than a volt to 4.5V depending on the type. What is the spec for the ones you are using? If the voltage spec is less than 3.6V, you are overloading the buffers.
Keith

 

I don't know any spec for the LEDs and I didn't design the driver circuit. I'm just repairing it.

 

I don't know any spec for the LEDs

What does the voltage measure across a good LED?

 

It seems I measured wrong.
Working LED: '244 output 0.25V, voltage at diode/resistor junction 1.84V, so 1.59V across the resistor and 3.16V across the LED.
Dim LED: '244 output 0.15V, voltage at diode/resistor junction 0.38V, so 0.23V across the resistor and 4.62V across the LED.

 

so 1.59V across the resistor and 3.16V across the LED.

Yeah 3.16v typical of a blue LED. So 15.9ma current draw, no problem there.
Steve G.

 

I am having similar failure issues with outside lighting for a restaurant. 48 inch LED tubes with 4 series groups of LEDs and they are failing to light. Each series group of 24 LEDs gets the same voltage, a bit over 76 volts. The failure mode starts with flashing, then goes to blinking and then not lighting at all. My analysis is that one or more LEDs in each string has a higher voltage to light it, instead of 3.2?? volts it is up to 4.++ volts, and so the string does not light. These are SMD style LEDs and it does not seem possible to unsolder them to repair a string, since they must have been glued in place prior to reflow soldering.
My point being that requiring a higher voltage seems to be a common failure mode for cheap Chinese LEDs. Failure in less than a year. The really strange part is that identical tubes inside the building. with far more hours of use, have had no failures at all. The outside lights are protected from the weather and show no signs of moisture damage at all. So what is happening?

 

I believe my ones are 15+ years old so they don't owe me anything.