My due is working with “adafruit_Neopixel” samples so i assume the voltages are all ok.
But when i try to use the FastLED 3.0 the program seems to stop at FastLED.show()
I don’t see any error messages. Am i missing something DUE specific ?
void setup()
{
delay(3000);
FastLED.addLeds<CHIPSET, DATA_PIN, COLOR_ORDER>(leds, NUM_LEDS);
pinMode(13, OUTPUT);
digitalWrite(13, HIGH); //turn on the led on the arduino
}
void loop()
{
for( int i = 0; i < NUM_LEDS; i++) {
leds[i]=CRGB(128,128,128);
}
FastLED.show();
digitalWrite(13, LOW); // This never happens. led stays on.
delay(200);
}
Try 1.5.5 or 1.5.6 - they made a major compiler change in 1.5.7 that was causing serious problems with avr - I wonder if it is also causing problems with the due? (Ftr I use a due here with both fastled3.0 and 3.1, but I am still mainly using 1.5.5)
Ok, this is another arduino 1.5.7/1.58 issue. If I build for the due using master on 1.5.5 - it works, if I build for the due using 1.5.8, I get the hang up that you’re seeing.
The good thing is, the hangup does not occur with the FastLED 3.1 branch - which also gets you the ability to live with interrupts as well
I can see about pulling over the due specific changes to the master branch, but I suspect you’ll be happier w/the 3.1 branch anyway (plus, that’s where all the new toys are going to be going for a bit!
I did the test using this small sample on an Udoo Quad (including an Arduino Due), compiled on a Linux Ubuntu 13.10 (64-bits) with arduino-1.5.8 and FastLED 3.1 branch but it seems to stop at FastLED.addLeds(). It works fine when I comment out the code related to FastLED. Do you have any idea ?
No, I don’t - i’ll have to see if ubuntu in my VM recreates the problem that you’re having. I think you’re the second person who has ever mentioned using an Udoo to me (the other sent theirs to me since they weren’t using it to debug some other problems).
Unfortunately, it seems/feels like the mac/windows versions of arduino are better maintained/more consistent/stable.
The strange thing is it stopping after the call to addLeds - the previous issues (with it not coming out of show) had to do with a clock register/timing issue in show and on the due that changed from 1.5.5 to 1.5.8. There isn’t anything in addLeds that should be not returning.
