What are the maximum number of pixels people are driving with one Photon (or

Archive FastLED Archive
,
1

What are the maximum number of pixels people are driving with one Photon (or other board) without data corruption?

I’m using the following and getting a flicker after about 300 pixels and complete color mush after about 600 pixels:

  • One Photon board
  • Level shifter
  • 9 strips of 66 pixel/m APA102 each approximately 1.4m in length
  • Strips are connected with 4-pin IP67 waterproof screw-on connectors - about 200m (edit: I mean 200mm!! Yikes.) distance between strips (signal degradation possible?)
  • Injecting power with one Meanwell LPV-60-5 at every junction point (9 power supplies rated at 60 watts nominal each, or about 45 usable watts)

This seems like it should be TOTALLY doable. Am I missing something? Have people had success with similar topologies? (1 MCU w power injected along up a long set of strips up to 1,000 pixels)

Thanks in advance.

2

I have never used apa102s specifically. “about 200m distance between strips” - Yeah. Amazing that you are getting a whole 600m before things start to break down. Have you tried reducing the data rate? https://github.com/FastLED/FastLED/wiki/SPI-Hardware-or-Bit-banging see also https://plus.google.com/102282558639672545743/posts/eqQ78D87Zra

3

Oh geez sorry 200mm

4

LOL, that makes a little difference!

5

Yea, 199k+ millimeters difference! Ha.

But thanks for that link. Maybe I do need to dial in the data transmission rate even for this relatively short length of LEDs.

6

some apa102 strips need the speed aggressively dialed back the longer they are - this has something to do with the strips themselves, and not the apa102, as i have a board with 768 of them that i can drive at 24mhz with no problems.

7

Yeah, definitely try reducing the data rate.

8

Thanks. Will do. I have some plans coming up to drive 12 separate lengths of up to 1,200 pixels each! So hopefully I can get that dialed in appropriately. (1 photon or other per 1,200 pixel section)

9

Boom. 12MHz still showed flicker, but an improvement. 6MHz is a dream. …wondering if I still should be expecting this though. I’m wondering if my level shifter is crummy. 3.3v logic would produce similar effects I believe (at least it would with the WS2812b that I’m used to - even at snails pace data rates)

10

No, this particular problem doesn’t seem to be related to level shifting - it’s something about the way that the strips are wired/printed.

11

Got it. I’m wondering if the connector system I spec’d out is interfering. There’s a decent distance (not 200m, but 200mm/8" or so) between each strip, and there are 8 of these junctions between the 9 strips. Each one is likely adding some noise. I’m guessing that’s the culprit.

12

I’ve had this problem on a single 5m roll with 300 LEDs even with injecting power at both ends.

13

Interesting. Do you think it has to do with manufacturing defects or is it just inherent in trying to address that many ICs sequentially? Regardless, I don’t think I’ll have any issues with the reduced data rate for my project. (…until I eventually want it back!)

14

Again - it is something specific to the strips. It isn’t the IC’s or the number directly - I have a board with 768 apa102’s that have no issues running at 24mhz and higher. Something with the strips (voltage drop? Rf interference? Line noise? Aliens?) causes the signal to degrade the higher the clock rate, the fewer LEDs it takes to show off the degradation.

15

…and my 8 junctions is likely artificially injecting extra Aliens.

16

I have same problem. I can only run at 4 MHz. But… since I’m running 4 5m strips, I’m going to try using teensy’s 4 SPI ports. One for each strip. I’m hoping this gets me back into the 12 MHz game.