Please be aware of all shipping deadlines before placing your order - we cannot guarantee orders will arrive before Christmas!

LPD8806 Pixel Anomaly / Power Bleeding (Packaging Issue?)
Moderators: adafruit_support_bill, adafruit

Re: LPD8806 Pixel Anomaly / Power Bleeding (Packaging Issue?)

by cjbaar on Sat Dec 01, 2012 8:03 pm

Based on my experiences with these strips, I don't think it is an issue with the LED. It is an issue with the signals getting too close on either the copper traces or inside of the chip itself. The effect is something to do with the clock coming out of the arduino to the first chip. Since each chip basically takes the signals (data and clock) and passes them down the chain to the next chip, it is somehow limiting the signal or cleaning it up enough that the issue does not persist going down the chain.

I do not have a logic level tester, so I cannot say what the difference is between the signal coming into the first chip and the signal coming out of it. If someone has one (and the free to play with it), I would be interested to see what the difference is. Might help us figure out a more permanent solution.
cjbaar
 
Posts: 73
Joined: Fri Aug 26, 2011 5:58 pm

Re: LPD8806 Pixel Anomaly / Power Bleeding (Packaging Issue?)

by DoctorDidj on Sun Dec 02, 2012 12:24 pm

It seems that what's going on is that the first LED in the strip is picking up induction from the sharp edges of the clock signal, just enough to light the green LED a little. This induction rounds off the edges of the clock (removing the high frequencies) so that the LEDs farther down the strip don't experience the phenomenon.

With my strips, the first LED does not light green when I apply power & ground, but it does light when I apply clock with no data. I'm clocking at high rates, using the SPI library.

Myself, I don't find this to be a problem. The green LED is quite dim, and doesn't interfere with light patterns that I've programmed into the strip. In fact, I use it as a handy monitor, so I can see my strip is getting clock.

If you want a fix, round off the clock edges. A small RC circuit will do this, as shown by other people on this forum who've used either a capacitor or a resistor to solve the problem. (There's probably enough trace resistance and capacitance in the strip that either one would do.)
Food Processor with Ears
User avatar
DoctorDidj
 
Posts: 28
Joined: Fri Apr 27, 2007 9:30 am
Location: Boston