Re: [avr-gcc-list] ATMega128, fuse bits, program optimisation, several questions

[Joerg Wunsch]

Torsten Mohr <address@hidden> wrote:

> I only programed the ext. fuse bits to 0xff.  Is that ok?

Probably yes, at least for a start.  The remainder is upon your
requirements, after all, that's why they are fuse bits at all. ;-)

> Regarding the interpretation of the fuse bits i'm a bit unsure, is
> there a general rule about what is active / not-active?

Bits burnt as 0 are active.  That sounds illogical at first, but is
basically due to the way ROMs are manufactured: the unprogrammed bit
value reads out as logic 1.  If you've got a traditional EPROM, and
UV-erase it, all bytes read out as 0xff.  Likewise, all unprogrammed
AVR fuses read out as 1s.

> One bit is called M103C, so i'd guess if it is TRUE (1) this means
> "yes, active" so to say in compatibility mode.

If it is programmed, i.e. 0.  That's the default how the device ships
(as you already noticed), and usually really the first to turn off.

> What about WDTON?

You only need to program that fuse if you are concerned that the
device might jam up during startup code.  Otherwise, it's pretty fine
to enable the watchdog right after startup, at the beginning of
main().

> To switch off the internal watchdog, is it sufficient to do this at
> startup:

Unless you programmed WDTON, the watchdog is disabled by default.

> To my understanding i also don't need to attach an external
> crystal or change any clock options at startup as the part will
> run from internal RC with 1 MHz.  Is that correct?

Using the fuse settings you've posted, yes, this is correct.  The
ATmega128 ships with the fuses set to run off the internal 1 MHz RC
oscillator (calibrated at 1 MHz, btw.).

-- 
cheers, J"org               .-.-.   --... ...--   -.. .  DL8DTL

http://www.sax.de/~joerg/                        NIC: JW11-RIPE
Never trust an operating system you don't have sources for. ;-)