Re: [avrdude-dev] programming 128

[Erick Castillo]

On Thu, 15 May 2003, Theodore A. Roth wrote:

>
>
> On Thu, 15 May 2003, Erick   Castillo wrote:
>
> :) The bits read all zeroes because of some strange electrical problem I'm
> :) having. For some reason, when I start the board up at 5 volts, the
> :) controller doesn't respond nicely... in fact, it does nothing. If I step
> :) up the voltage, I notice that the controller responds to 3 volts and runs
> :) the loaded program just fine. I've got the atmega128 which is rated to
> :) operate between 4.5-5.5V so it should work fine with 5V at boot up. Any
> :) ideas why the controller only works when initiated with 3.5-3.7V?
>
> When Atmel (other manufacturers too probably) makes a certain device,
> say mega128, they have various speed grades and voltage grades. All
> come off the same wafer. The only difference between a 5v part and at
> 3.3v (LV) is how the set the secret internal mode fuses of the device
> (and whether it passes tests in that mode). The mode fuses are
> accuessible to anyone if you know the secret hand-shake. Usually done
> with either a special SPI command or a special hi-v programming
> operation (take lines x, y, z to 12V, send these commands, etc).
>
> It's possible (although not very likely) that you have a device that
> has some bogus fuse mode.
>
> Has the device ever worked?
>
> Ted Roth
>
>

The device has indeed worked. In fact, it's running now. But in order to
get it running, what i have to do is use 3.7 V to start up and then raise
the voltage up to 5V so the other chips on the board get powered. Finally,
I reset and everything works fine. It's just the boot up process that is
acting strangely.

I check the fuses when it's running normally and sure enough, FF is what I
get which is exactly what I write to them.

--Erick