[TriEmbed] ATMEGA 1284p "Mighty Mini"

Alex Davis alexd at matrixwide.com
Thu Apr 9 15:05:21 CDT 2015


I wasn't aware of the Mega 2564. I took a quick look, and the main
immediate drawback is it's 3.3v only - not sure if it's 5v tolerant.
More and more that's less of an issue, but there are still 5v parts out
there.

Otherwise it looks interesting.

I think the main motivation behind the Mighty Mini is to get the
crystal, ISCP and Serial headers in about the same space as the DIP
chip. I've done breadboard ATMEGA1284s, and a problem area is the
crystal pins in proximity to UART1 pins. You have to take care to not
introduce crosstalk.

My feeling is if you want big RAM (like > 32 KB) you probably also want
more power. In that case, PJRC's Teensy 3.1 is really good for only $19.
You get 64 KB RAM, native 32-bit, DMA, much higher SPI speeds, much
faster CPU, etc... It's a step up in complexity, but Paul (PJRC) does an
excellent job of hiding it with his very good libraries.

If someone wants to pursue 2564 and enough people are interested, I can
at least offer fabrication help. It isn't easy working with QFN-size
parts, but it's doable.

On Thu, Apr 9, 2015, at 03:22 PM, Jon Wolfe wrote:
> I love the Mega 1284. But the 1284 is also the biggest 8 bit AVR chip 
> that comes in a DIP form factor.  I run them at 20Mhz in breadboards all 
> the time with just a xtal and loading caps, and I even have a home-made 
> JTAG-breadboard adapter for them, inspired by this 
> http://quinndunki.com/blondihacks/?p=550
> 
> If it's already going the SMD route, I wonder if the Mighty Mini would 
> work with the new Mega 2564. It's sort of the upgrade to the 1284, and 
> so far it doesn't come in DIP form. It's got an insane 32kb of sram, and 
> 256kb of flash. The 2564 also is marketed as a device for RF 
> applications, I think at least a variant of it has built-in Zigbee, or 
> hardware to talk to a Zigbee device. I'm also not sure if they make 5V 
> version of the 2564, they all seem to run at 3.3V, with a max clock of 
> 16Mhz.
> 
> The other potential chip would be the 2560 used in the Arduino Mega. 
> It's got the flash, but only 8kb of sram. The 2560 does have an external 
> sram interface though, so you could have potentially 64kb (or more if 
> you "bank switch", there are shields for the Arduino Mega that do 
> this).The AVR parallel memory interface also uses around 18 I/O lines 
> too, so there's that. It's hard to add too much to an AVR based board 
> these days without hitting the point where a Cortex-M based design would 
> be a better choice.
> 
> There are seveal XMega's that could potentially work too, but I don't 
> know nearly as much about xmegas.
> 
> 
> On 2015-04-09 13:19, Alex Davis wrote:
> > Martin,
> > 
> > there's nothing on-board which should cause an issue with power-saving.
> > There is no FTDI chip, a only 5-pin header to attach to an external
> > USB-to-serial adapter. Some people might prefer an 8 MHz crystal vs. 16
> > MHz, but you can always set the fuse bits to use the internal RC
> > oscillator and get 8 MHz. There's an amber LED on the SCK line with a 
> > 1K
> > resistor to ground. That would only matter if you left it high and went
> > to sleep. You can easily remove it with a soldering iron if you want.
> > 
> > I will be bringing my board to the next meeting. It's a cool board - my
> > thanks to Jack Christensen for sharing it on OSHPark. I can talk about
> > the aspect of soldering the SMT parts. It's actually much easier than
> > through-hole once you do a few.
> > 
> > Message: 1
> > Date: Wed, 8 Apr 2015 13:47:38 -0400
> > From: Martin Brooke <martin.brooke at duke.edu>
> > To: TriEmbed Discussion <triembed at triembed.org>
> > Subject: Re: [TriEmbed] ATMEGA 1284p "Mighty Mini"
> > Message-ID:
> > 	<CALjzumpMQzbMS2C3eeuooMNHbX=wKNgErmYmFHL7mk=BaMi=8Q at mail.gmail.com>
> > Content-Type: text/plain; charset="utf-8"
> > 
> > Does it have anything on it that would interfere with Low power modes?
> > 
> > Sound like no regulator what about  other chips?
> > 
> > I would love a bare bones board that supported more RAM and Program but
> > could be put to sleep.
> > 
> > Martin
> 
> -- 
> Jon Wolfe
> Anibit Technology LLC.
> 
> https://anibit.com


-- 
"The theater of noise is proof of our potential."
|\ |  (¯  \/ |¯\  |V| |\ ¯|¯ |¯) | \/ | | | |¯\ (¯   /¯  /\ |V|
|-||_ (_  /\ |_/ @| | |-| |  | \ | /\ |^| | |_/ (_ . \_  \/ | |

You won't find me on Facebook.




More information about the TriEmbed mailing list