What is the maximum signal wire length between the Microcontroller board and the TM1809 strip that you folks would feel comfortable with? I am not referring to voltage drop, I am referring to the signal integrity of the TM1809 data signal. (Assume my power wires are sufficiently large that voltage drop is not a problem). I have run WS2801 CLK+Data 30+ft before with no problem (I think I used twisted pairs of a Cat5e cable with Clk on Blue, GND on Blue/Wht, Data on Green, GND on Green/Wht). With TM1809, however, I have seen weird flickering with a length of 10-20ft. Makes me wonder if using a twisted pair would make a difference. (I am assuming the TM1809 is identical to the WS2811, the WS2812 and the WS2812b in this case). The microcontroller in my case is a Teensy3.0 or Teensy3.1 with NO level shifter IC. Thx in advance for any comments!
Btw, I am not using a series resistor between the Microcontroller and the TM1809… I am wondering if I should? 100ohm? I remember seeing PJRC Paul’s discussion on this and it seemed more black magic than science… I suppose I should use an oscilloscope to verify the signl integrity of the dataline on both ends…
I would suggest you use an ‘off-the-shelf’ RS485 differential driver tx/rx pair at each end of the cable, then you can drive 100’s of meters with very little issue. Plus it will save you having to common the 0v up between your Teensey and your LEDs and this will go nicely down your twisted pair. And you protect the Teensy with the driver chip.
One end can be 3.3v logic to match the Teensy and the other end can be 5v logic to match the LEDs. Then you get a cable driver and level converter all in one go. The chips are dirt cheap.
And a third point is that you can add several Rx off the single driver chip to run lots of LED strips in sync.
I have found tm1809 strips to be a bit more susceptible to noise - twisted pair helps a lot, I’ve also been able to help by level shifting.