NeoPixel Ring, no blue leds working

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.
Locked
firmwareguy
 
Posts: 3
Joined: Tue Feb 25, 2014 10:52 pm

NeoPixel Ring, no blue leds working

Post by firmwareguy »

I picked up a 16 section NeoPixel Ring intending to experiment with using it to display both the flow rate and relative temperature of a liquid heating/cooling circulator.

Before going further, I hooked it up to an Arduino Uno and ran strandtest. The test works fine insofar as the red and green leds but fails to display any blue either in isolation or in combination. Is this a common failure mode or common user error? Is there any solution?

FWIW. the ring is powered by a battery pack supplying 4.8VDC with 1000uF cap across the +/- and 360Ohms on the data input lead. The Uno is powered from a USB connection. The ground from the battery pack and from the Uno are tied together on a proto-board.

Also. Since I had an 8 section NeoPixel stick handy, I connected it into the same circuit, ran the modified strandtest (for 8 pixels) and it displays all colors.

Thanks for any suggestions to either recover the blue or avoid whatever error I made in the future.

User avatar
adafruit_support_mike
 
Posts: 67446
Joined: Thu Feb 11, 2010 2:51 pm

Re: NeoPixel Ring, no blue leds working

Post by adafruit_support_mike »

firmwareguy wrote:Is this a common failure mode or common user error?
No, that's flat-out bizarre..

It seems highly unlikely that an entire ring of pixels would all have dead blue LEDs, though the law of large numbers sits ready to slap me upside the head any time I say something like that. I'd be more inclined to think there's a problem with the first pixel, and that it's only passing alog 2/3 of the data it receives. That's marginally less improbable, at least.

Either way, you have a ring that doesn't work. Let's get you a replacement.

Send a note to [email protected] with a link to this thread and the folks there will take care of it.

Since that makes the existing ring fair game for experimentation, try connecting the data line to the second pixel's input pin. See if that makes the remaining pixels work as they should.

firmwareguy
 
Posts: 3
Joined: Tue Feb 25, 2014 10:52 pm

Re: NeoPixel Ring, no blue leds working

Post by firmwareguy »

Thanks for the response. I will give the folks at support an email and as soon as I can I'll experiment further. I haven't had much time to read about the support chip (WS2812) but the traces on the ring make it appear that the only connection between "pixels" is the data/clock and power. If that is the case, I suppose that a dinged 2812 could be failing to respond to the blue part of the command and therefore not regenerate it going forward. The back (non chip) side of the ring is heavily lacquered, so I'll have to figure out by trial and error which is the power and which is the data/clock. I'll post further when I do.

Thanks again.

User avatar
adafruit_support_mike
 
Posts: 67446
Joined: Thu Feb 11, 2010 2:51 pm

Re: NeoPixel Ring, no blue leds working

Post by adafruit_support_mike »

You're right about the connections: each WS2182 chip has a 24-bit data register that it uses to control the LEDs. Each chip copies the first 24 bits it sees on DI to its control buffer, then passes everything else to the next chip.

There's only a single data signal, and the red, green, and blue values are interleaved as they travel along the strip. That's why having a single color drop out would be so unusual.. it would require synchronized failure of several chips in exactly the same (unusual) way.

User avatar
grayconstruct
 
Posts: 12
Joined: Fri Feb 28, 2014 6:20 pm

Re: NeoPixel Ring, no blue leds working

Post by grayconstruct »

Try putting a meter on your power line while running the test. Make sure you have enough power and it is not browning out when lots of LEDs are on.
I suggest this because on typical battery powered blinkies, the blue is usually the first to stop lighting up when the battery gets low. The blue LED requires more voltage (typically 3.2v) than the red (typically 2.1v) to light.
The guide says you can run a short length of pixels as low as 3.7v lithium cell but if it drops too low you will have issues.

Make sure the Arduino is not brownout resetting if it shares the power supply.

Try Changing the program to think there is only one or two NeoPixels (so very little power draw) and see if that works any differently.

User avatar
step180
 
Posts: 1
Joined: Sat Jul 26, 2014 1:32 am

Re: NeoPixel Ring, no blue leds working

Post by step180 »

I recently encountered the same behavior with a Neopixel stick. Green and Red worked fine but no blue. I purchased two so I immediately tried the other one in the same circuit and again encountered the same behavior. Given that the results were the same with two different sticks (and having never encountered any problems with Adafruit products before) I had to question my circuit and parts (wiring, capacitor, etc.) rather than the sticks.

A voltage test revealed that only 2.3V was actually arriving at the sticks which was clearly too low for the blues.

The cause? Honestly, I'm still not completely certain (though I do have a suspicion...mentioned below). What I can say with a significant degree of certainty is that the eventual solution was to remove and then reinsert the 1000uf capacitor into the original circuit while it was powered by the battery.

The overall troubleshooting process I went through: (1) I tested the voltage entering the stick and discovered only 2.3V. (2) I then tested the voltage exiting the 5V regulator I had between a 9V battery and the 1000uf capacitor (5V was exiting the regulator. Steps (1) & (2) indicated that the problem was likely the capacitor. (3) Next, I disconnected the battery, removed the cap, and tried different ones. After reattaching the battery I once again encountered the same behavior (green, red, but no blue). (4) I disconnected the battery, rebuilt the circuit five+ times using alternative capacitors, voltage regulators, wires, etc. Reattached the battery, and still no blue. (5) At this point, I decided to assemble the circuit without the capacitor (what's the worst that could happen, a brownout, an extremely unlikely but nonetheless deadly surge to the pixel stick ending the lives of even the reds and greens, etc.... either way, those were risks I was willing to take). Surprisingly, this experiment immediately solved the problem. No capacitor immediately equaled perfectly functioning blue leds (and for those that might now be wondering, the cap polarity was properly aligned during each of the previous steps). One might (as I did) also begin to question whether the 8+ capacitors I tried during stages 1 & 2 might have all been bad. Nope, they each worked fine in other test circuits. (6) After the discovery that the blues worked without a capacitor in the circuit I just had to try adding the original capacitor back into circuit while it was connected to the battery. The addition of the capacitor had no ill effect (the blues were still working, etc.).

This bout of troubleshooting and the eventual solution makes me wonder whether there might be some aspect, element, etc., to a few WS2182 chips that could at times prevent or "block" a capacitor from fully charging. I must defer to those who know more than me about the internals of the WS2812. But I am genuinely curious. Might there be something in the internal design, layout, etc. of the WS2812 that might at times prevent a capacitor from fully charging when, e.g., 5V, is applied to the fully assembled circuit? I would appreciate reading what others think.

As a final note: I have been able to recreate the "no blues" behavior at least a half dozen times by simply assembling the circuit and then applying battery power. And, I have likewise solved it a half dozen by simply removing and reinserting the cap into the powered circuit (I've done this with different breadboards, etc.). Interestingly enough, all it takes to make the odd behavior go away is just one round of removing and reinserting the cap. After it's been removed and reinserted once, the battery can be disconnect and reconnected as many times as one wants and every time the blues will continue to work perfectly. The "no blues" phenomenon seems to only occur when one completely assembles the suggested circuit and then applies battery power, and then powers the Arduino. In the end, who knows, maybe I happen to be the lucky receipt of the only two neopixel sticks that will, in just the right set of circumstances, block a cap located between the +/- on the stick from charging. Either way, given that I bought the sticks solely for the sake of experimenting, etc., my workaround/solution leaves me completely satisfied with the product (neopixels are absolutely awesome leds). But I'm certainly curious about what others might think in regards to whether there's anything in the layout and design of the WS2812 chip that might at times prevent the cap from charging.

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

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