I recently spun a board with an atmega328-mmh. I'm trying to program it with the ISP. I modified the avrdude.conf file so that the device signature would match that of atmega328-mmh (and not the default atmel part on the uno board). The arduino IDE claims that everything programmed fine.
First thing I'm trying to do is just blink an LED -- it's not functioning (the code is fine/works on other arduino boards). I'm wondering how the arduino ide handles different chip packages. For example my led is on pin 11 of the 328-mmh, which corresponds to PB1. PB1 on the 32 pin MLF part is pin 13. Does the IDE/compiler translate the I/O (e.g. Arduino pin 9) to PB1 or does it need to know something about the physical pin assignments on the atmel part? In other words can I even program the 328-mmh by faking an Uno board or are all the I/O assignments going to be botched? Thanks!
Pins_arduino.h is used for any '328 port to software naming.
The software does not care about the physical package - the die inside the part is the same.
So PB5 will map to what the software thinks is D13, no matter what actual physical pin it goes to - 19 on a 28-pin DIP, 17 on a 32-pin TQFP or MLF, 15 on a 28-pin MLF.
Thanks a lot for the reply (and for deleting the duplicate post)! Turns out that I apparently had a cold solder joint (that package is pretty small for hand soldering). I now successfully blink an LED though I observe that delay(n) results in a delay that's actually 16n. So delay(10) results in a delay that's 160 mS. I'm using the same resonator that's found on the uno (murata CSTCE16M0V53-R0). Some more debugging is probably needed, but it would seem that the arduino is operating at 1 MHz and not 16 MHz -- or at least that's an obvious theory. Are there any particular reasons that this behavior would be exhibited? I can't see how any timing registers would have changed. Thanks!
doov:
Thanks a lot for the reply (and for deleting the duplicate post)! Turns out that I apparently had a cold solder joint (that package is pretty small for hand soldering). I now successfully blink an LED though I observe that delay(n) results in a delay that's actually 16n. So delay(10) results in a delay that's 160 mS. I'm using the same resonator that's found on the uno (murata CSTCE16M0V53-R0). Some more debugging is probably needed, but it would seem that the arduino is operating at 1 MHz and not 16 MHz -- or at least that's an obvious theory. Are there any particular reasons that this behavior would be exhibited? I can't see how any timing registers would have changed. Thanks!
What clock source did you select when you set the fuse bits?