LED RGB IN-P32TATRGB is used in an array of 30 LEDs.
The image below includes the circuit for half the total amount of LEDs.
The signals LED+REDSTRGA, LED+GRNSTRGA & LED+BLUSTRGA, at the bottom left corner of the image, connect to the drain of NCH MOSFETs (not shown for simplicity)
The current-limiting resistor is set to allow 10mA of current for each Red, Green and Blue elements.
It has been observed a significant number of failures of the red element, and a handful of failures on each green and blue.
A failure is defined as the LED element (either red, green or blue) does not turn on at all, and it remains in the off state regardless of the driving condition. Every test conducted confirmed the LED is properly driven and bias, and I am beginning to suspect the failure is related to the component itself.
Has Digikey received reports of failures as described above?
Can Digikey perform failure analysis on site, or should it be forward to the manufacturer?
We have purchased 9000 pcs already, and are scheduled to purchase an additional 100K pcs for a build late in the year. Before the purchase of 100K pcs I need to come to an understanding of the failure mechanism.
Could you recommend what steps to take next?
Regards,
Oscar.
Greetings,
The typical protocol for formal failure analysis requests is to return a sampling of failed product to the manufacturer via the channel through which the product was purchased. Assuming that to be DK in your case, a next step would be to contact customer service, explain the situation, and request an RMA for failure analysis. Be advised however that this process can take a while and require jumping through numerous hoops; formal F/A is costly, and the typical result points to user/application error so there’s justifiable reluctance to accept such requests absent a degree of qualification.
In the meantime, I’d suggest taking a critical look at things presuming the cause to be electrical overstress of some kind. I recall a customer some time ago that was seeing bridge rectifiers fail in quantity, and the cause turned out to be a static discharge event related to an assembly/test process. It can happen…
Also, it wouldn’t hurt to put some of the offending devices under your own microscope and see what you can see. Might offer clues, might not, but it’s a fast & cheap option with little downside.
Hi @oscar.medina ,
Thank you for your inquiry. Are the LEDs not working when independently tested from the board? If verified indeed LED failure, agreeing with rick_1976 then most likely electrical overstress of some kind.
A few questions to troubleshoot:
- Is your ground a true ground? Does it have static, noise, or voltage spikes? This is more common with floating grounds.
- Is the voltage supply free from static, noise, and voltage spikes? An analysis of the 5v supply will help determine if this is the cause.
- Are there any inductive loads on either the 5v supply, or main voltage supply feeding the 5v supply? If there are noisy loads on the same supply, the 5v supply might not be filtering out the voltage spikes.
-As rick_1976 mentioned, wouldn’t hurt to make sure the handling and assembly is following proper ESD protocol.
A few other topics regarding static, if this is the problem:
Establishing an ESD Safe Workspace
Filter for Static discharge in audio equipment
What is an Ionizer?
