next topic »

[Reply #75 on:]

I've started playing with the Mega & NKC's Mega-Shield, and it is amazingly fun to develop without worrying about memory, IO lines, or requiring SoftSerial.  

My hat is off to the designers, developers and testers.  Designing a small footprint board with a relatively large MCU while retaining shield inter-op is an amazing feat!  

There are a few minor shortcoming that I've read or noticed:

• As mentioned above, only 6 of the external interrupts are available (the MCU has .
• Due to the 1280s pin-out, the 4 "new" external interrupts are on dual-purpose pins. 2 are on the TX/RX pins for Uart#1 (pins 18-19), and 2 are on the I2C pins (pins 20-21).
• While most pins are the same as an Arduino, the I2C pins have moved (from 4-5 to 20-21).  Any Arduino shield, or project, using I2C will require re-work for the MEGA.

I agree with the posts above that the Arduino team did a fantastic job getting this to market so quickly and smoothly (a few early pictures and rumors is no big deal).

I have two Megas & sheilds here, and I'm starting a project on one of them.  If anyone has any tests or questions let me know.

Cheers.

[Reply #76 on:]

Maybe a converter pass through shield would be useful to swap the MEGA I2C pins to the original pins, maybe it could have some dip switches to let the user switch between the original annoying pin spaces and a version with the correct pin spacing as well.

I think I might wait a while for the MEGA, I'm hoping seeedstudio will do a version with mini-usb, and the white power plug adapter (and hopefully all the Pins)

[Reply #77 on:]

While most pins are the same as an Arduino, the I2C pins have moved (from 4-5 to 20-21).  Any Arduino shield, or project, using I2C will require re-work for the MEGA.

SPI has also moved.

-j

[Reply #78 on:]

There are some other changes that are more subtle.
The timers associated with specific PWM pins have changed, Interrupt 0 and 1 are on pins 20 and 21, Input capture for timer1 is not available, and pins 14 through 19 are no longer shared with the analog pins

[Reply #79 on:]

thinking of the future: sure that after 'mega' comes 'hyper', then 'ultra', but what's after that?
 

[Reply #80 on:]

Obviously, after mega comes giga, tera, peta, ...

[Reply #81 on:]

Don't forget about the Arduino ÜBER.....

[Reply #82 on:]

Naaaa, i think it's just going to be calle Arduino 7

:-)

[Reply #83 on:]

There are some other changes that are more subtle.

All of which makes me wonder why the bullet just wasn't bitten (lol) and the shield gap fixed--instead of leaving people thinking shields would be compatible out of the box.

--Phil.

[Reply #84 on:]

@hailey
you are thinking too technical. i'm more into comic style.

[Reply #85 on:]

All of which makes me wonder why the bullet just wasn't bitten (lol) and the shield gap fixed--instead of leaving people thinking shields would be compatible out of the box.

Bah, fixing SPI and TWI shields is a software problem; you just have to exchange SW protocol support for HW protocol support.  "Fixing" the gap would require a PCB spin for all the shields; MUCH harder.   Sort of.  (I am reminded of some other HW problems that I have had to fix in software...)

[Reply #86 on:]

So why didn`t they go the seeduino route and have a second set of sockets with .1 spacing.

That way if they made that the standard you could use both old and new standards of shields.

          

[Reply #87 on:]

Bah, fixing SPI and TWI shields is a software problem; you just have to exchange SW protocol support for HW protocol support.

Are you volunteering?    

Has anyone checked the 1280 to see if the SPI and TWI hardware is driven the same way (register names, etc) so that code for a mega8 SPI shield would work software-wise on the 1280, if the pins were just connected?  I just haven't had time to check this yet.

-j

[Reply #88 on:]

Hmm looks like someone has already started on sotware TWI/I²C:
http://www.arduino.cc/cgi-bin/yabb2/YaBB.pl?num=1235009256/7#7

-j

[Reply #89 on:]

looks very nice