Atmega 2560

[ivars211]

I will start by saying thanks to Felix for this wonderful library and module. I did my research and if nothing has changed Felix said that there will be no support for atmega2560. Which is shame, as I need more I/O pins then atmega1284p can offer. I really like the wireless programming feature Moteino offers. So the question is, how difficult would it be to implement dual optiboot with the wireless programming feature on the bootloader for the atmega2560? Is it just adjusting few lines or is it basically redoing the whole thing?

Any input would be much appreciated. 

[Felix]

As much as I would like to take the challenge, it boils down to economics and the fact is these chips are $13 in 100x QTY so the final board would cost something like $45 (no radio). It would require an exhaustive amount of testing and I just can't stretch that far at the moment. Also the upfront investment is huge in man hours and prototyping and production costs. Without knowing I will be able to sell these quite fast it really makes no sense do invest all that and then wait for months while one or two of these are sold and I got thousands of dollars worth of parts in my inventory. The 1284 was a good compromise between awesome features (vs 328p) and low cost (in the AVR realm). Still some people think everything should cost $5 because they can get it from china for $5, I'm not addressing to that crowd.
Also the massive amount of IO would make the board much much larger that I'd want it to be for the purpose of a "Moteino" (the philosophy was to make it as small as possible while retaining as many features as possible), and omitting some of that would be a waste.
Back to economics, it looks like you can get a fully fledged ArduinoMEGA for $45 so I'm afraid I'd be competing with much bigger houses that can make a much better margin while keeping low prices  (stocking parts in much larger quantities). So the price/feature ratio turns somewhere between the 1284p and the 2560. In that respect I think it would make more sense to come up with a shield for the radio and use it on the ArduinoMEGA. At that point I believe low power would not be as big of concern and a 3.3V regulator and data voltage buffers could be included in that shield.

I believe optiboot is will only support up to a 1284 and not a 2560 so you might have to rewrite whatever bootloader you're using with the 2560 to support wireless programming. Also you might have to modify the WirelessHEX69 library and wireless programming protocol therein to support the massive flash of the 2560 - the current out of box supported amount is 64kB.

[ivars211]

Thanks for your reply and I fully understand and support you. Totally agree, atmega2560 just won't fit Moteino ideology. Moteino shield is a good idea though. Would fit standard uno/mega
However, I am very interested hacking the code myself to adjust it for atmega2560. But I guess it's not as simple as changing few lines...
If that's the case, I will just go with atmega1284p + MCP23017

[Felix]

Any possibility you could use two MoteinoMEGAs in place of 2560 and have one of them be a slave and talk over Serial1 or i2c to the other?
That would give you a bunch more IO and still take a relatively small area. And it would be dual core hehe