Shop OBEX P1 Docs P2 Docs Learn Events
Through-Hole "Smart" RGB LEDs (WS2811-compatible) - Page 3 — Parallax Forums

Through-Hole "Smart" RGB LEDs (WS2811-compatible)

13»

Comments

  • varnonvarnon Posts: 184
    edited 2014-09-26 12:18
    Do you have the start method inside or outside of the repeat loop? It would be easy to accidentally put the start method in the repeat loop. I don't know exactly what that would cause, but it seems like things would get weird after a few repeats.
  • NWCCTVNWCCTV Posts: 3,629
    edited 2014-09-26 20:36
    I have played around with the code and have come up with some weird results. Any how, getting the LED's to change color in this way is really not my goal. My ultimate goal is to use 6-8 of these LED's on a strip. I want each LED to cycle through all of the possible colors repeatedly. I would prefer that they are random so each LED is a different color throughout the cycle. Does this make sense and can anyone help me out? I have standard RGB LED's but they would require a resistor and changing voltages on each line which I do not want to mess with.
  • John A. ZoidbergJohn A. Zoidberg Posts: 514
    edited 2014-10-05 08:51
    I just got these thru-hole little PL-9823 smart LEDs (seems a bit bulkier than the standard RGB ones). The problem with it is, the LED does not cooperate with the majority of the 8-bit microcontrollers (notably PICs) I have in my drawers due to the very strict timing. In the end, all I manage to do is to send the bit-zeroes through the SPI, but it couldn't get the bit-ones to the LED without the thing flashing random colors at random brightness.

    I read it elsewhere that these LEDs have the capability to "Reshape" the incoming pulses so that it is compatible to the next chained LED. I've poked the probes of the "Data Out" pin on the LED and I got the bit-zero correct, but on any kind of timing for the bit-one, it gets filtered back into a bit-zero.

    I'll test this with a faster microcontroller such as Prop later.
  • JonnyMacJonnyMac Posts: 9,105
    edited 2014-10-05 11:47
    You realize that "all the possibilities" is a 24-bit value (16,777,216 colors), right? Even at 1ms per color, it would take a very long time to cycle through them all.

    Many of us are willing to help -- you may need to narrow and clarify your spec.


    NWCCTV wrote: »
    I have played around with the code and have come up with some weird results. Any how, getting the LED's to change color in this way is really not my goal. My ultimate goal is to use 6-8 of these LED's on a strip. I want each LED to cycle through all of the possible colors repeatedly. I would prefer that they are random so each LED is a different color throughout the cycle. Does this make sense and can anyone help me out? I have standard RGB LED's but they would require a resistor and changing voltages on each line which I do not want to mess with.
  • David BetzDavid Betz Posts: 14,516
    edited 2014-10-05 14:05
    I have 10 of these http://www.adafruit.com/products/1837 and I'm trying to get four to work on my ActivityBoard. I'm pretty sure I have them wired up correctly but when I power them on they all glow blue without any signal being fed to them and if I pass them a signal using my version of JonnyMac's PASM driver nothing happens. The continue to glow blue. Is there some additional trick to getting these individual LEDs to work? The picture on the AdaFruit site seems to just show them wired to +5, GND, and the DIN and DOUT lines chained together. That's how I wired them on the ActivityBoard's breadboard area. Any suggestions as to why this didn't work?

    Thanks,
    David
  • RobotWorkshopRobotWorkshop Posts: 2,307
    edited 2014-10-05 14:51
    David Betz wrote: »
    I have 10 of these http://www.adafruit.com/products/1837 and I'm trying to get four to work on my ActivityBoard. I'm pretty sure I have them wired up correctly but when I power them on they all glow blue without any signal being fed to them and if I pass them a signal using my version of JonnyMac's PASM driver nothing happens. The continue to glow blue. Is there some additional trick to getting these individual LEDs to work? The picture on the AdaFruit site seems to just show them wired to +5, GND, and the DIN and DOUT lines chained together. That's how I wired them on the ActivityBoard's breadboard area. Any suggestions as to why this didn't work?

    Hello David,

    When I first wired mine up I forgot to add the jumper from the breadboard to the +5V supply. When that happened a couple of the LEDs lit up blue. It must have been getting some power from the DIN line. Can you double check the supply lines? These also seemed to work ok when powered from the 3.3V supply.

    Robert
  • David BetzDavid Betz Posts: 14,516
    edited 2014-10-05 17:34
    Hello David,

    When I first wired mine up I forgot to add the jumper from the breadboard to the +5V supply. When that happened a couple of the LEDs lit up blue. It must have been getting some power from the DIN line. Can you double check the supply lines? These also seemed to work ok when powered from the 3.3V supply.

    Robert
    I checked all of the wiring and it all looks correct to me. Someone on the AdaFruit forum is trying to help me. I'll let you know if I find out anything. I did notice that someone else reported a problem where their LED glowed blue and they said they "fixed" it by switching to a different batch of LEDs.
  • lanternfishlanternfish Posts: 366
    edited 2014-10-05 18:05
    Tubular wrote: »
    The rumour is the next version of these leds will support dmx protocol, natively. At some point in the future I guess they'll have their own IP stack and internet connectivity.

    You can get the 5V DMX ones here here or the 12V DMX ones here
  • TubularTubular Posts: 4,702
    edited 2014-10-05 19:01
  • John A. ZoidbergJohn A. Zoidberg Posts: 514
    edited 2014-10-10 02:04
    Using JonnyMac's code for the Prop, I managed to get something out of the PL9823 (clones of WS2812) LEDs some hours ago. Only difference is it's RGB and not GRB unlike WS2812, and that's straightforward to handle.

    But one issue on that LED - I can't seem to make the exact "White" without the red color 'drowning' the other color and so on.

    Some colors are almost hard to mix - especially getting a pure "yellow" is totally impossible. I only get a yellow-ochre (brownish-yellow) tint from this.

    I can get purple, turqouise and crimson really easy. But not for the other colors.

    Is it a defect on the LEDs, or are all the programmable LEDs have that funny behaviour?
  • NWCCTVNWCCTV Posts: 3,629
    edited 2014-10-10 17:03
    You realize that "all the possibilities" is a 24-bit value (16,777,216 colors), right? Even at 1ms per color, it would take a very long time to cycle through them all.

    Many of us are willing to help -- you may need to narrow and clarify your spec.
    Sorry about the confusion. I think about a half dozen of each color would work. Such as Blue. dark blue, light blue and so on.
  • NWCCTVNWCCTV Posts: 3,629
    edited 2014-10-12 09:35
    I'm using the new version of the driver from the OBEX and I still believe the RGB vs GRB issue exists.
    Ron, Thanks for the fix. I ran in to the same issue and your modified code fixed it. Now I just need to replace all my original jm_ws2812 files and all should be good.
  • xanaduxanadu Posts: 3,347
    edited 2014-10-12 10:11
    Using JonnyMac's code for the Prop, I managed to get something out of the PL9823 (clones of WS2812) LEDs some hours ago. Only difference is it's RGB and not GRB unlike WS2812, and that's straightforward to handle.

    But one issue on that LED - I can't seem to make the exact "White" without the red color 'drowning' the other color and so on.

    Some colors are almost hard to mix - especially getting a pure "yellow" is totally impossible. I only get a yellow-ochre (brownish-yellow) tint from this.

    I can get purple, turqouise and crimson really easy. But not for the other colors.

    Is it a defect on the LEDs, or are all the programmable LEDs have that funny behaviour?

    If you diffuse it, or look at it from further away you could get better results. If you're looking at the WS LED up close with no diffusing lens you're looking at three separate LEDs and they will act just like three separate LEDs.
Sign In or Register to comment.