0

Is it possible that an APA102 led could "die" due to softwar
Moderators: adafruit_support_bill, adafruit

Please be positive and constructive with your questions and comments.

Is it possible that an APA102 led could "die" due to softwar

by KevAQN on Wed Jun 14, 2017 5:05 pm

Looking to purchase a few APA102 strips from adafruit, but in the mean time I have an APA102 strip I got from a friend for testing. I keep running into issues where an LED will randomly die in the strip and I have to cut it out. Is there even a remotely possible chance that software would be to blame for this?

I'm asking this before I buy because I really don't want to waste money on a strip only to screw it up.

KevAQN
 
Posts: 11
Joined: Sun Jun 04, 2017 12:06 am

Re: Is it possible that an APA102 led could "die" due to sof

by adafruit_support_bill on Wed Jun 14, 2017 5:15 pm

I don't know of any way to kill one via software. Are these pixels at the start of the strip, or random pixels in the middle somewhere?

adafruit_support_bill
 
Posts: 61093
Joined: Sat Feb 07, 2009 10:11 am

Re: Is it possible that an APA102 led could "die" due to sof

by KevAQN on Wed Jun 14, 2017 5:29 pm

adafruit_support_bill wrote:I don't know of any way to kill one via software. Are these pixels at the start of the strip, or random pixels in the middle somewhere?


They happen to be random pixels in the middle of the strip. Some times it's near the end of the strip too. Surprisingly it happens to the 30/m and 60/m test strips I was given, but not the 144/m test strip. The LED portion doesn't burn out either, occasionally I can get the LED (and anything past them) to light up 255,255,255 at max brightness. Something with the controller inside the LED seems to be messing up?

Edit: To give more of an idea on what I'm testing... I have a RPi hooked up with an 74AHCT125 connected to it's own 5v power supply, then I have a 5V 15A power supply for just the LED strips. I hard wired the 144/m and then connect the other 60/m strip to the end of it. I wrote my own software (C++) and have only been testing at brightness settings at or below half the max. Occasionally I will set it to max for a few settings. There appears to be no rhyme or reason for anything to be dying like they are? Could it just be a set of faulty APA102 LED's I'm working with? (I'd really really hate to order a bunch here and then come to find out I get the same results and are suddenly out of a large sum of money.)

KevAQN
 
Posts: 11
Joined: Sun Jun 04, 2017 12:06 am

Re: Is it possible that an APA102 led could "die" due to sof

by adafruit_support_bill on Wed Jun 14, 2017 6:35 pm

If it is happening in the middle of the strip, I'd suspect possibly power supply spikes and/or bad chips. Are the pixels mostly dying at power up, or after they have been running for a while?

adafruit_support_bill
 
Posts: 61093
Joined: Sat Feb 07, 2009 10:11 am

Re: Is it possible that an APA102 led could "die" due to sof

by KevAQN on Wed Jun 14, 2017 7:23 pm

adafruit_support_bill wrote:If it is happening in the middle of the strip, I'd suspect possibly power supply spikes and/or bad chips. Are the pixels mostly dying at power up, or after they have been running for a while?



I can't really tell, I do have reason to think most of it is at "power up" or when I change the brightness levels. But I have had 1-2 that just seem to go at random. The problem is most of the time I don't notice it dead until 15 seconds after due to having the strip laid out. I don't have a 6.3v capacitor attached like the guide here on adafruit would recommend since I'm planning on scaling this up to 35m worth of strips. I just don't know what kind of capacitor I should be putting on a power supply of 5V 40A. Would this be due to a lack of a capacitor? Or could the power supply be faulty?

KevAQN
 
Posts: 11
Joined: Sun Jun 04, 2017 12:06 am

Re: Is it possible that an APA102 led could "die" due to sof

by adafruit_support_bill on Wed Jun 14, 2017 7:37 pm

Some power supplies will put out a spike of high voltage at power-up before the regulator has a chance to stabilize. This can be enough to cause damage to a pixel. This type of problem usually causes pixels to fail at power up. But sometimes a 'wounded' pixel may limp along for a while before actually dying.

Adding a large capacitor between 5v and GND at the power supply will help to absorb any power-up spike.

adafruit_support_bill
 
Posts: 61093
Joined: Sat Feb 07, 2009 10:11 am

Re: Is it possible that an APA102 led could "die" due to sof

by KevAQN on Wed Jun 14, 2017 8:06 pm

adafruit_support_bill wrote:Some power supplies will put out a spike of high voltage at power-up before the regulator has a chance to stabilize. This can be enough to cause damage to a pixel. This type of problem usually causes pixels to fail at power up. But sometimes a 'wounded' pixel may limp along for a while before actually dying.

Adding a large capacitor between 5v and GND at the power supply will help to absorb any power-up spike.



Thank you for the help, I will give a capacitor a shot and see how it goes. Would you have any recommendations in regards to a capacitor for a 5V 40A or even a 60A PSU?

KevAQN
 
Posts: 11
Joined: Sun Jun 04, 2017 12:06 am

Re: Is it possible that an APA102 led could "die" due to sof

by adafruit_support_bill on Wed Jun 14, 2017 8:21 pm

Different power supplies may exhibit the problem to different degrees, so it is hard to say without looking at the startup behavior with an oscilloscope. But in theory, a higher capacity supply has the potential to put out a more damaging spike. And, within reason, too much capacitance is better than to little.

For a power supply of that size, I'd probably start with about 10K uF. If you have access to an oscilloscope, I'd see how it performs through several dozen power-up cycles at least.

adafruit_support_bill
 
Posts: 61093
Joined: Sat Feb 07, 2009 10:11 am

Re: Is it possible that an APA102 led could "die" due to sof

by KevAQN on Thu Jun 15, 2017 11:53 am

adafruit_support_bill wrote:Different power supplies may exhibit the problem to different degrees, so it is hard to say without looking at the startup behavior with an oscilloscope. But in theory, a higher capacity supply has the potential to put out a more damaging spike. And, within reason, too much capacitance is better than to little.

For a power supply of that size, I'd probably start with about 10K uF. If you have access to an oscilloscope, I'd see how it performs through several dozen power-up cycles at least.


Sadly I don't have acceess to oscilloscope, and I don't think I'm about to go and buy one. I assume voltage might not matter that much on the capacitor? Since I'm looking at picking up a PSU with 3 outputs I guess I'll likely be getting a 6.3v 3300uF capacitor. Thanks again for the help :)

KevAQN
 
Posts: 11
Joined: Sun Jun 04, 2017 12:06 am

Re: Is it possible that an APA102 led could "die" due to sof

by adafruit_support_bill on Thu Jun 15, 2017 12:20 pm

6.3v should be fine. A higher voltage rating is OK too. But things start getting bulkier as either capacitance or voltage ratings increase.

adafruit_support_bill
 
Posts: 61093
Joined: Sat Feb 07, 2009 10:11 am

Re: Is it possible that an APA102 led could "die" due to sof

by KevAQN on Sat Jul 22, 2017 12:11 pm

So, I got a capacitor, got a power supply from Meanwell and went back to testing. Almost instantly I was able to cause an LED to die in an APA102 strip. I set the brightness to max white, then set the brightness to 0 to turn them off and about 30 lights stayed on (out of 140ish.) Then when I adjusted the brightness back to something else I was able to notice a dead pixel. This is making no sense... Does anyone know why this might be happening?


Really looking to make this a large scale project, but I will refuse to go forward if I can work past this stupid issue.

Edit: I'm able to replicate it finally, by setting them to white on max brightness, then setting the brightness to 0, and then setting it back to the max brightness. This appears to kill the LED's somehow, but this should not be possible?

KevAQN
 
Posts: 11
Joined: Sun Jun 04, 2017 12:06 am

Re: Is it possible that an APA102 led could "die" due to sof

by adafruit_support_bill on Sat Jul 22, 2017 12:38 pm

Is this a DotStar strip purchased from Adafruit? We have not been seeing many issues with the DotStars. I don't have any hard statistics on it, but in general they seem to be less susceptible to power glitches than the Neopixels.

adafruit_support_bill
 
Posts: 61093
Joined: Sat Feb 07, 2009 10:11 am

Re: Is it possible that an APA102 led could "die" due to sof

by KevAQN on Sat Jul 22, 2017 12:53 pm

adafruit_support_bill wrote:Is this a DotStar strip purchased from Adafruit? We have not been seeing many issues with the DotStars. I don't have any hard statistics on it, but in general they seem to be less susceptible to power glitches than the Neopixels.



They aren't DotStar ones, I've been using cheap ones to try and test to work out the possible issues. I'm really cautious to spend $100 for testing to get a 5m strip only to have this same issue happen. (Note my end project is looking at about 34m of these strips.)

In the case I have this happen to a DotStar strip, what would end up happening? Would I be able to get a replacement?

KevAQN
 
Posts: 11
Joined: Sun Jun 04, 2017 12:06 am

Re: Is it possible that an APA102 led could "die" due to sof

by adafruit_support_bill on Sat Jul 22, 2017 1:11 pm

Yes. In a case like this with one of our DotStar strips, we would authorize replacement.

adafruit_support_bill
 
Posts: 61093
Joined: Sat Feb 07, 2009 10:11 am

Re: Is it possible that an APA102 led could "die" due to sof

by KevAQN on Sat Jul 22, 2017 1:16 pm

adafruit_support_bill wrote:Yes. In a case like this with one of our DotStar strips, we would authorize replacement.



Thank you, I will be ordering some from you :)

KevAQN
 
Posts: 11
Joined: Sun Jun 04, 2017 12:06 am

Please be positive and constructive with your questions and comments.