I fault myself for not educating myself better. Last year was so easy with the GECEs and some short runs of 3001s. I'll make it through this season then start the reading. I used to know this stuff way back when, but didn't use it very often, so the knowledge has faded. The weather was crap today, so I was able to take my time and make some good null pixel runs to replace the others. I should know the verdict tomorrow.
Hopefully the new null pixels take care of things for this year.
In regards to the wire that comes on the pixels, at least the ones I've seen, are a kind of ribbon cable. Ribbon cable impedance can get complex, as the measured impedance relies on relative positions of the wires, primarily ground and signal, in the cable. If that changes, so can the characteristic impedance of the cable. The same factors mentioned previously apply as well.
The cable that comes with the pixels may not have the wires arranged in a way that's ideal. While that can be worked around by wiring the feed ribbion differently, if you're going to do that, you're so much better off using a cable with better and more consistent properties. Ribbon cable just isn't intended for distances over a few feet. Even once you've got the impedance dealt with, they don't have great noise immunity properties.
Does the propeller have the option to decrease the slew rate of its output signals? If so, that might have a beneficial effect if the issue is loss of signal integrity due to transmission line effects.
I'm not aware of an option to do that on the prop. In this specific case, it's really the 74HCT541 that you need to deal with, which has a rise time spec somewhere between 10 and 12ns @ 5V supply. Planning for the worst case, that'll give you 100MHz to deal with, which leaves takes cable length to 5ft. Even a pretty slow 74 series digital chip will have a 20ns rise time and get you to 10ft. Either way, a 20ft cable puts you in the area of needing to choose the right resistor. Most of the pixel controllers I've seen allow for the necessary resistors to deal with transmission line effects, it's just a question of getting the right value.
Not sure there's much avoiding these effects. If all it takes in many cases is the right resistor value and a reasonable cable, may as well document that.
That makes sense about the ribbon cable. I hadn't considered the positioning of the signals in the conductors. I've always used 3 wire pixels with data always in the middle. I guess that stuff is ok where the distances are short pixel to pixel. But probably not the best lead wire type.
Personally I'll stick with my 2 pair control cable since I have the magic formula all figured out.
--Jon Chuchla--
Sent from my iPhone using Tapatalk
