AVR-Chat

As I was having a lot of trouble getting PonyProg to
talk to my board, it turned out that the problem was
being caused by the "Reset IC ZM33064" that equips the
Olimex header ( http://www.olimex.com/dev/index.html )

This IC was holding the reset line so the parallel
prog could not pull it down... 

Once I replaced it with a more conventional RC circuit
it became wonderful.

On Wed, 06 Jul 2005 14:02:47 -0000 you wrote:

>As I was having a lot of trouble getting PonyProg to
>talk to my board, it turned out that the problem was
>being caused by the "Reset IC ZM33064" that equips the
>Olimex header ( http://www.olimex.com/dev/index.html )
>
>This IC was holding the reset line so the parallel
>prog could not pull it down... 
>
>Once I replaced it with a more conventional RC circuit
>it became wonderful.

I had considerable problems using ponyprog which mostly disappeared when I used the CVAVR
programming software. The rest disappeared when I switched to using an STK500 for programming
instead of a simple dongle. The Olimex mega128 boards always program ok now.


--
Ralph Hilton
http://www.ralphhilton.org
C-Meter: http://www.cmeter.org
FZAOINT http://www.fzaoint.net

>
>I had considerable problems using ponyprog which mostly disappeared 
>when I used the CVAVR
>programming software. The rest disappeared when I switched to using 
>an STK500 for programming
>instead of a simple dongle. The Olimex mega128 boards always program ok now.


There is a fun little gotcha there, for the unaware..

When the AVR programmers let loose between program and verify, the 
code actually runs for a moment.
This can cause some really interesting things, when you interrupt it 
mid-process to go verify.
I had a case where it was wiping out data in an external storage device.
I had a line in the code to nuke it, which I would comment out when I 
didn't want that to happen, but between stopping emulation (Jtag ice) 
and programming, the ice would let the target run just long enough to 
start the "erase all" and POOF went my data.. This caused some 
anxious minutes till I worked out what was happening.

Now, that app has a short delay before the write happens, about 1s, 
which is long enough that the jtag reasserts control before the time expires.