using OctoWS2811 to control custom AC Power circuit

Status
Not open for further replies.

stephanschulz

Well-known member
Hi Forum.

I am using the OctoWs2811 library on a teensy 3.1 with the Octo Adaptor to control 8 line of each 36 x WS2811 RGB drivers. I do not have LED strips but rather use the RGB-Driver IC.
I am sending either 255 or 0 for each RGB channel per LED-driver to turn 24VAC devices on or off.

Everything works but not all RGB channels (the attached 3 device) come on consistently. I also attached a AC led to make sure my TRIAC to provide power. The LED work consistently.
I know people will say now, well it's your AC circuit or AC device, but if i send an all on signal to all s2811 chips then they all come on without delay.

I also tried to attach a zero crossing circuit to only send led.show() on a zero crossing, which did not help either.

Setting the library to WS2811_400kHz turns all circuits on right a way but i can't control the s2811 after that.
Nonetheless it feels like that maybe something can be done on the firmware side to fix my problem.

Those AC device that do not react right away do come on after a few ms, and it feels like they come on with the same delay. it is very rhythmic?


ws2811.jpg

thanks,
stephan.
 
I see C1 and R3 connected to IC1's VDD pin. This may be overly obvious, but have to tried also adding a 0.1 uF or 0.01 uF ceramic capacitor directly between the VDD and GND pins?
 
Thanks for your input.
I just tried adding a 0.1uF between the GND and VDD pins of the WS2811 IC but i still see the same delay.

Here is a short video of two "RGB pixels" being turned on one after the other.
Pleas only look at the top left 3 device row (since the others are broken). The top 3 devices get their control signal from the same WS2811 IC and you can see how some turn on with a slight delay.

https://www.dropbox.com/s/x3vzdb7beuge3w1/this is with 120 Resistors and zero crossing.mp4?dl=0

It is strange, because that delay seems to always have the same timing no matter which of the pixels is delayed.
?
 
Last edited:
My design already has LEDs on the AC side of the optocouplers and TRIAC. See LED7.
These LEDs do come on reliably and do not display the delay i see with the AC devices, even though they both get the same AC switched power.
I know this would mean my AC devices are the reason for the delay and they might (they are very cheap devices). BUT it's not the same device that always this places the delay, and all device respond right away when all "RGB LEDs" are set to on.

Someone else said maybe the TRIAC's trigger current is not high enough. I tried differed values for RES R11 and also tried a lower voltage and current TRIAC. But still the same behaviour.

?
 
Oh... I guess I was thinking in terms of a Teensy-to-WS2811 issue. Sounds like that part's working fine.

Maybe the triac is actually not on 100% of the time, and neither is the LED. Visually you can't see the brief off times, but perhaps your AC equipment is sensitive to occasionally missing part or all of one AC cycle?

The WS2811 uses 400 Hz PWM. Even with 255, there's still a substantial off time 400 times per second. Maybe that's playing havoc with your AC waveform after the triac?

Probably need to use an oscilloscope and a good (safe) 10X probe to measure the AC waveform to see what you're actually getting.
 
i followed your advice and connected an oscilloscope to one of the 3 AC outputs which are controlled by the WS2811.
the AC curve looks find to me, considering that i am not monitoring the zero crossing.

here a quick video (removed video), the scope is connected to the channel that feeds AC to the far right device.

Screen Shot 2016-05-26 at 12.09.26 PM.jpg
 
Last edited:
Status
Not open for further replies.
Back
Top