Re: [avr-gcc-list] very strange behaviour...

[Erik Christiansen]

On Wed, Jul 27, 2005 at 03:26:28PM -0700, Reza Naima wrote:
> I've got a bizzare problem.  I'm trying to send out a fast waveform
> which should result in a 4Mhz signal (F_CPU = 16Mhz, 4 single cycle
> instruction per waveform, hence 16/4=4Mhz).  However, it's a lot
> closer to 1Mhz when I look at it on the scope.

When I generated slowed-down read/write strobe timing, using nops as you
are doing, each gave 62.5 nS at 16 MHz, as expected. On the basis of the
posted code, it looks a little like the assumed clock source is not
being used, doesn't it?

What do the fuse setting look like?

Erik