Black Lives Matter - Action and Equality. ... Adafruit is open and shipping.
0

Testing signal cables for DotStar
Moderators: adafruit_support_bill, adafruit

Please be positive and constructive with your questions and comments.

Testing signal cables for DotStar

by tjwood on Thu Feb 09, 2017 1:44 pm

I'm wiring up a DotStar panel where a 10 pixel test strip works fine, but a 255-disc and 8x32 panel both get weird artifacts after some number of pixels (and stop working all together far enough down the strip). The connection I have hooking them up is most likely the issue (some jumpers + 2 female 4-wire JST connectors soldered together to make a F-F converter), but I was wondering for the future if there is a easy/reasonable way to test the signal quality across a connection like this for the needs of a DotStar connection (instead of just resoldering/making a whole new cable until it works). Square wave generator + scope?

tjwood
 
Posts: 21
Joined: Fri Nov 29, 2013 3:58 pm

Re: Testing signal cables for DotStar

by adafruit_support_mike on Fri Feb 10, 2017 1:39 am

I doubt it's an input issue.

The PWM controller inside each pixel works as a buffer for the rest of the strip. It consumes the first 24 bits of data it sees coming in, then copies all the rest to its output pin. That eliminates a lot of long-wire transmission problems.

From your decription, you're probably seeing the effects of power loss along the strip. Each pixel removes up to 60mA from the 5v rail (80mA for RGBW pixels), and that produces a slight voltage drop. When you add up the slight voltage drops from a few dozen pixels, you start getting values large enough to cause problems.

If you have a scope, connect it between the 5v and GND rails near the point where the pixels start misbehaving. Have the scope calculate the average voltage, but also measure the peak-to-peak noise.

adafruit_support_mike
 
Posts: 61124
Joined: Thu Feb 11, 2010 2:51 pm

Re: Testing signal cables for DotStar

by tjwood on Fri Feb 10, 2017 2:00 am

Ah, thanks! I rerouted my power (which is also going over small-ish jumpers) closer to the DC barrel jack on my testing breadboard and now more of the matrix behaves. So, better power management seems spot on. I'll work that side of my test setup some more to see if I can get the whole matrix working (possibly at low intensity).

tjwood
 
Posts: 21
Joined: Fri Nov 29, 2013 3:58 pm

Please be positive and constructive with your questions and comments.