I've been working with controlling RGB LEDs through PWMs on a Duemilanove using Maxuino 009 in Max 5 (latest version). I have the correct version of Firmata and the latest version of the Arduino software available on their website, Arduino 018. I am using an Apple G5 Quad with OS X Server 10.5.8. The first time the Arduino was connected via USB, I uploaded Firmata and all six PWMs on pins 3, 5, 6, 9, 10, and 11 functioned as expected.
The following day, I opened the patch (without restarting anything or unplugging the Arduino) and the red and green channels for the second light (pins 9 & 10) would not work when the pins were set to PWM in the pin modes. I was able to get an output from those pins when they were set to servo mode, but they were 100% unresponsive when set to PWM. Pins 3, 5, 6, and 11 (RGB on the first LED and blue channel on the second) were functioning as expected. The connections to the LEDs were tested independently and functioned as expected.
I restarted Max 5 and the computer with no effect, pins 9 and 10 still not functional. Next I reloaded Firmata on the Duemilanove, problem was still there. Reopened the original, unmodified Maxuino patches from a fresh download, pins 9 and 10 still did not respond in PWM mode. Finally, my frustration got the better of me and I started changing settings back and forth in the maxuino.help, Pin-Modes patch, but only in the sections for pins 0, 1, 2, and *, not in the section where I had specified any of the PWM pins 3, 5, 6, 9, 10, and 11. Pins 9 and 10 suddenly started working again.
The day after that, the same problem returned, but no combination of pin mode adjustments would force 9 and 10 to work again this time. This time I carefully rebuilt my patch from scratch without copying and pasting to rule out a problem with my patch. Pins 9 and 10 still refused to work. I reloaded Firmata, unplugged the Arduino, restarted the Mac, opened my patch in Max 5, and then plugged the Arduino back into the Mac. Pins 9 and 10 now respond when set to PWM mode again, but there is a new problem.
Regardless of which patch is used (mine or the original) or how many times Firmata is reloaded and the Mac is restarted, pin 10 always triggers pin 11 and vice versa. When the duty cycle for pin 10 (green) is raised above 0, pin 11 (blue) also responds even though the duty cycle in the patch still reads 0. The same is true for pin 10 when the duty cycle on pin 11 is raised above 0. Through testing, I have ruled out the wiring between the LEDs and the Duemilanove or the LEDs themselves as a cause. This leaves either the Duemilanove itself or the Mac as the problem.
Anyone have an idea what could be going wrong?