Debugging sewable neopixels

General project help for Adafruit customers

Moderators: adafruit_support_bill, adafruit

Please be positive and constructive with your questions and comments.
Locked
User avatar
BackwardsLED
 
Posts: 19
Joined: Sun Aug 05, 2018 4:16 pm

Debugging sewable neopixels

Post by BackwardsLED »

I have a set of sewable neopixels that I've sewn into a strip of cloth. They used to work perfectly, but have started behaving strangely as the cloth is flexed. The data out->in connectors are chained using conductive thread on a backing layer, and the power/ground lines are 22awg stranded wire on the back, with the sheathing stripped and sewn to the neopixel power/ground using more conductive thread. (I think I used a blend of 2-ply and 3-ply thread, but don't recall which is which anymore).
1.jpeg
1.jpeg (142.51 KiB) Viewed 118 times
2.jpeg
2.jpeg (170.95 KiB) Viewed 118 times
They're wired up to a Feather using a a 3-pin JST SM connector that I use to control the LEDs. However, they've recently started misbehaving; when plugged in (when just getting just power/ground and prior to any data signal), some LEDs turn on dim red. Bending around the strip of fabric causes different LEDs to turn on and off in that dim red.
3.jpeg
3.jpeg (112.35 KiB) Viewed 118 times
I can't find any place where the data thread might be shorting to power or ground. How do I debug this? I've gone through my pixels and checked each for low resistance between DIN and VIN, as well as DIN and GND, with multimeter set for 2k Ω of resistance. I'm not spotting anything (but also, maybe I'm looking at the wrong threshold?)

User avatar
adafruit_support_bill
 
Posts: 88037
Joined: Sat Feb 07, 2009 10:11 am

Re: Debugging sewable neopixels

Post by adafruit_support_bill »

I'd suspect a loose connection somewhere. The knots in conductive thread can loosen over time - especially if subjected to much movement. Poking them with a meter-probe just tends to make them more conductive, so that is not a very conclusive test.

User avatar
BackwardsLED
 
Posts: 19
Joined: Sun Aug 05, 2018 4:16 pm

Re: Debugging sewable neopixels

Post by BackwardsLED »

You're most likely correct, but I'm not sure how to test a might-be-loose-when-moving-connection, and the problem is multiplied by 100 or so. Maybe alligator clips to the multimeter and watching the resistance while wiggling things around? Anyway, I'd love help making sure the theory lines up before I start several hours of testing.

Can a loose connection lead to a pixel spontaneously enabling itself? Presumably that would have to be loose on power or ground, since there's no signal coming in over DIN. And given my construction, that must be the knot between the bus wire and the pixel.

Can I confirm that by alligator clipping a pixel to power+ground and checking that strange behavior stops? Also, is there a compound (conductive glue? grease? etc) I can squirt into the knots to increase their conductivity, or am I stuck removing and retying them all?

User avatar
adafruit_support_bill
 
Posts: 88037
Joined: Sat Feb 07, 2009 10:11 am

Re: Debugging sewable neopixels

Post by adafruit_support_bill »

There are a few strategies for tracking down a loose connection. You can use alligator clips, but you might need quite a few if you have many pixels. Another technique is to jam a toothpick into the hole so that it presses the stainless thread hard against the side of the hole to make better contact.

In any case, you should start at the beginning of the strip & work your way down.

Conductive grease is pretty messy stuff. There are some conductive glues, but I have not tried them on an application like this. https://www.amazon.com/Sciplus-Electric ... 000Z9H7ZW/
https://www.amazon.com/Conductive-Tempe ... B08RJ94KLF
You might be able to snug up a loose knot by tying another knot over it with another piece of thread. A drop of clear nail-polish on the knot can help keep it from loosening(Just on the knot itself. Don't let it get between the thread and the plated hole.)

User avatar
BackwardsLED
 
Posts: 19
Joined: Sun Aug 05, 2018 4:16 pm

Re: Debugging sewable neopixels

Post by BackwardsLED »

Thanks, those techniques will be really helpful!

And is my theory plausible that a loose power or ground connection can cause a neopixel to turn itself on? And each different pixel will be an unrelated loose connection?

User avatar
adafruit_support_bill
 
Posts: 88037
Joined: Sat Feb 07, 2009 10:11 am

Re: Debugging sewable neopixels

Post by adafruit_support_bill »

That is plausible. The state of a Neopixel at power-up is technically 'undefined'. They *almost* always power up in the OFF state. But they sometimes come up displaying some color. This is more likely to happen if the power is a bit flaky as it would be with a loose connection.

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

Return to “General Project help”