NeoPixel Strip with Dead NeoPixels

EL Wire/Tape/Panels, LEDs, pixels and strips, LCDs and TFTs, etc products from Adafruit

Moderators: adafruit_support_bill, adafruit

Please be positive and constructive with your questions and comments.
cardcomm
 
Posts: 5
Joined: Sun Jun 30, 2013 10:06 pm

Re: NeoPixel Strip with Dead NeoPixels

Post by cardcomm »

Does this product (#1612, http://www.adafruit.com/product/1612) use the WS2812B variant?


Those look pretty cool. I may have to order some, darn it! :)

User avatar
adafruit_support_bill
 
Posts: 88088
Joined: Sat Feb 07, 2009 10:11 am

Re: NeoPixel Strip with Dead NeoPixels

Post by adafruit_support_bill »

Yes. It uses the WS2812B.

brueglju
 
Posts: 7
Joined: Mon Jun 13, 2011 10:06 am

Re: NeoPixel Strip with Dead NeoPixels

Post by brueglju »

Hello all,
same happened to me TWICE last week: two dead Neopixel Stripes with 144 RGBs (!)
I thought I've done everything right: Lab-power supply, resistor in the data line, and a 1000µF condensor
and I'm still killing my Neopixels: First WS2812B is dead.

I'm not sure whether just one big capacitor is enough:
Big capacitors handles low frequency ripple and mains noise and major output load changes.
Small capacitors handle noise and fast transients

I assume that the problem is the fast transients - what means a big and a small cap should be connected in parallel.
Maybe a Zener diode would be also useful.

Adafruit, could you please investigate this and come up with the component suggestions? Maybe you can add an already built circuit board to your store.

Thanks a million !

Juergen

User avatar
adafruit_support_bill
 
Posts: 88088
Joined: Sat Feb 07, 2009 10:11 am

Re: NeoPixel Strip with Dead NeoPixels

Post by adafruit_support_bill »

Hi Juergen, Can you tell us a bit more about your system? A photo would be helpful too.
What processor are you using to control the Neopixels? What is the distance between the processor and the first pixel? What other components are in your system?

We do keep our Best Practices page up-to-date, and will update it with new information as we get it. http://learn.adafruit.com/adafruit-neop ... -practices

brueglju
 
Posts: 7
Joined: Mon Jun 13, 2011 10:06 am

Re: NeoPixel Strip with Dead NeoPixels

Post by brueglju »

my system looks like this:

Arduino Mega 2560, USB-powered, signal cable lenght 2 ft, dia 0.25mm², resistor already in stripe; power cable 0.50mm², 1 ft, Lab power supply 5.03V (20A); 1000µF 50V cap.
The stripe is normally fully inserted in the aluminum profile.
Neopixel RGB strip (144 LEDs)
Neopixel RGB strip (144 LEDs)
IMG_3251NeoPixel setup.jpg (244.09 KiB) Viewed 537 times
Here I have 'injected' the signal line right after the first LED
signal 'injected' after the first LED
signal 'injected' after the first LED
IMG_3250NeoPixel 1_LED ueberbrueckt.jpg (17.4 KiB) Viewed 537 times
Works ! - means I have blown the first LED :(
Q: does the first WS2812 blow because of a surge on the signal line (Arduino is powered off my PC) or because of a voltage spike?
If voltage spike: why only the first LED ??

Observation: both times the stripes froze while running the sample sketch.
When disconnected they did not light up anymore.

Hope that helps :)
Juergen

brueglju
 
Posts: 7
Joined: Mon Jun 13, 2011 10:06 am

Re: NeoPixel Strip with Dead NeoPixels

Post by brueglju »

Here is a suggestion how a voltage swing suppression could look like
(Thanks to Andrew from Toastedcircuits.com)
Fuse depends on stripe length (144 LEDs = 10A)
Fuse depends on stripe length (144 LEDs = 10A)
Voltage Swing Suppression.jpg (29.61 KiB) Viewed 522 times
Would this work?

User avatar
adafruit_support_bill
 
Posts: 88088
Joined: Sat Feb 07, 2009 10:11 am

Re: NeoPixel Strip with Dead NeoPixels

Post by adafruit_support_bill »

Q: does the first WS2812 blow because of a surge on the signal line (Arduino is powered off my PC) or because of a voltage spike?
If voltage spike: why only the first LED ??
We are not entirely sure. We have not been able to reliably reproduce the problem here. Some users have reported success using the capacitor on the power supply. Others have reported success using the resistor on the signal line, so at this time we are recommending both. The cautions about making connections before power-up and connecting the ground first are just good practice overall.
Observation: both times the stripes froze while running the sample sketch.
When disconnected they did not light up anymore.
That is an interesting data point. Most users have reported that the problem occurs at startup or when re-connecting. I have not heard of issues with freezing while running.
Here is a suggestion how a voltage swing suppression could look like
(Thanks to Andrew from Toastedcircuits.com)

Would this work?
I think it is worth testing. If you could give that circuit a try (injecting the signal into the second pixel) and let us know how it works for you. I will forward your observations to the rest of the engineering team and see if we can identify the root cause of the problem.

Thank you for your cooperation in this matter :)

brueglju
 
Posts: 7
Joined: Mon Jun 13, 2011 10:06 am

Re: NeoPixel Strip with Dead NeoPixels

Post by brueglju »

What confuses me is that if it is a voltage problem I'd expect all LEDs to blow out. If only the first LED is dead it points towards a problem with the data line.
That would explain why the series resistor in the data line helps other users.

Do you have information on the capacitance of the WS2812 ?

Andrew suggests three possible approaches:
a) A fast-response diode from the data lines to ground, reversed. This will ensure that if the data line wants to go below zero (i.e. the GND line rises due to current issues) it can't fall too far.

b) A low value series resistor and then a zener (in parallel -- i.e. to ground) on the data and clock inputs.

c) A voltage divider to drop the data line voltage to about 3.5v...

Could you please discuss these solutions?
Thanks a million !

User avatar
adafruit_support_bill
 
Posts: 88088
Joined: Sat Feb 07, 2009 10:11 am

Re: NeoPixel Strip with Dead NeoPixels

Post by adafruit_support_bill »

Thanks for the input. I'll review these with the team :)

User avatar
Kramy
 
Posts: 1
Joined: Sun Jul 30, 2017 5:17 am

Re: NeoPixel Strip with Dead NeoPixels

Post by Kramy »

Sorry to be a thread BANNED, but was there a verdict on what to wire in and how?

I got an RGB LED strip, turned it on and ran a sketch... had some neat colours happening, and then the first LED turned off. It wouldn't come on again until I left it off for a while. Now it's alive again, but I don't want to risk killing it, so I'm looking into what parts I need to wire in before it to keep it going.

User avatar
adafruit_support_bill
 
Posts: 88088
Joined: Sat Feb 07, 2009 10:11 am

Re: NeoPixel Strip with Dead NeoPixels

Post by adafruit_support_bill »

Our current recommendations are in the Best Practices page: http://learn.adafruit.com/adafruit-neop ... -practices
The capacitor is strongly recommended when powering with a plug-in power supply to suppress any start-up spikes.
The resistor is recommended when there is a long run (substantially more than 8") between the processor and the first pixel.
Note that the resistor needs to be mounted as close to the first pixel as possible. Many people put it on the processor end, or somewhere in the middle where it doesn't do any good.

Locked
Please be positive and constructive with your questions and comments.

Return to “Glowy things (LCD, LED, TFT, EL) purchased at Adafruit”