Some small suggestions.
Looking at the board layout above, it looks like you snipped out the reset switch. This is OK although it eliminates handy resets – although a piece of wire from the reset pin would do as well – sometimes a useful thing, to restart the program aside from download.

You might want to think about putting the sixth cable pin and reset cap back in (plenty of room under there if you’re going to surface mount – since it would facilitate going to the Diecimila bootloader, which is a little faster on some (Windows) machines. As much as I like Adaboot, it doesn’t kill all the wait states on certain Windows machines. For those who are affected (or afflicted) Diecimila does end up saving a lot of time over a long development process, and would involve only minor change in your board. It also is going to facilitate lack of the reset switch, since having to either hook up a switch or mess with a piece of wire on each program download seems like a fuss.

Another thing to think about is the addition of a Ground and 5V pins near reset and RX TX pins, to facilitate people who wish to use our P4 RS232 programming adapter or something similar. There seems to be a diehard group of RS232 fans out there who like that option. Also it’s cheaper than the cable. We are proposing the pin order of the P4/RBBB as a standard (RST, G, 5V, RX, TX) but suit yourselves (the P4 is easily remapped for other configs). Of course I know adding a pin here, a pin there, you’re up to the Queen Mary in no time.

It is also true that the P4 (or similar) could easily be remapped to work with the cable headers.

The one meg resonator load resistor turns out to be unnecessary and you can eliminate that entirely – it’s built into the atmega circuitry.

On another note, I still have 30 or so of the Rev A boards that you’re such big fans of, and know whether another 100 possibly are squirreled away, send me some email if you want them.

Paul Badger moderndevice.com

I am a proponent of moving the reset pulse onto the programmer. Either with a cap (like this design from sparkfun — http://www.sparkfun.com/commerce/product_info.php?products_id=8551), or in firmware (like my own design (http://www.dorkbotpdx.org/blog/feurig/benito_7_the_next_big_thing). My experience has been that if you do want the reset you will want it on the outside of the enclosure and not on the board. So in this board it is brought out twice once next to the pin and once on the programming header. The new pinouts were not as easy to get into the board as I thought given the profile and the need to have the clearances required by our local board fab (www.sunstone.com) but the rev I am going to roll with incorporates your proposal in stead of maintaining pin compatibility with the rbba.

So I do hope you are serious about this change. Thank you for the solid and constructive feedback.

Don.