I have shamelessly copied the new official Arduino UNO photo (from here: http://arduino.cc/en/Main/ArduinoBoardUno ) and added numerous "annotations" of the changes that I found "interesting." (click though to flicker to see the annotations)
Good observations there. I found the crystal / resonator combo interesting: the USB AVR gets a crystal, while the main MCU only gets a resonator. Sounds like USB timing issues are more critical than keeping the main MCU at an accurate clock rate.
The decision to run the USB AVR at 16MHz just like the main MCU is also interesting. Looking at the schematic it's almost like two Arduinos munged together!
Jon
Nice work.
Is there a schematic avalible yet in pdf? I would like to see if the auto voltage selection circuit has had any improvement or component upgrade.
Lefty
Is there a schematic avalible yet in pdf?
Yes; it's in the expected standard place.
The auto-voltage select looks "approximately" the same; I'm pretty sure that the recent discussions about its problems happened WAY too late to show up in the UNO design.
Hi, nice work!
I'm pretty sure that the recent discussions about its problems happened WAY too late to show up in the UNO design.
What discussions? Is there a thread about autoselect problems?
Yes; it's in the expected standard place.
Thanks, found it. Looks like the uno now has a higher output current capacity at 3.3 volts then the 50ma limit from the FTDI 3.3vdc output on older standard Arduino boards.
That may provide a problem for commerical shield manufactures. If they design their shields to utilize more then 50ma of 3.3volt power then they will not work safely with the older standard Arduino boards?
I guess that's not a new potential problem as the Seeeduino boards have used dedicated 3.3vdc regulators on several of their boards for awhile.
Lefty
Is there a thread about autoselect problems?
http://www.arduino.cc/cgi-bin/yabb2/YaBB.pl?num=1284034373
The "communication processor" would give a lot of new options. However to be compatible - and be "boarduino friendly" - it is not wise to let let him do anything else but the FTDI chip. Is this really a lower cost option?
WRT to power supply selection. I had hoped they would give up that too complex auto-selection (or do it right, or add a battery option with boost regulator, or a processor 3volts option, or...)
My understanding of the design is
- They do not like jumpers
- They want to have a higher parts count on it, maybe to push SMT (?) and to emphasize more Arduino specifics (in contrast to ANY AVR board)
I personally still love the Diecimila (with 328 chip)
Is the Atmega in the picture put in the wrong way? Or am I missing something? 
You might be right 
But it looks much better fitted in the wrong way ;D
But it looks much better fitted in the wrong way
Only a detail too...
Is the Atmega in the picture put in the wrong way? Or am I missing something?
Uh, yup :o You didn't miss anything, in fact you caught something I looked at a dozen times and missed.
Is the Atmega in the picture put in the wrong way? Or am I missing something?
Did you notice how cleverly they did not mark pin 1 - on either side of the board. It looks like the silk screen IC outline (which would also provide orientation information) may be missing as well.
Don
Only a detail too...
Not really, as a platform for (also) beginners in electronics, one might have a hard time wondering why their board isn't functioning.
Did you notice how cleverly they did not mark pin 1 - on either side of the board. It looks like the silk screen IC outline (which would also provide orientation information) may be missing as well.
Indeed. I saw it because I was trying to find the second crystal, which is not in de Uno as westfw pointed out nicely in his "annotations". I was wondering why the resonator was on pin 23 & 24
Replacing the crystal with a resonator (which is less accurate) appears to be a step backwards.