Picking a dev platform for a future product

For Adafruit customers who seek help with microcontrollers

Moderators: adafruit_support_bill, adafruit

Please be positive and constructive with your questions and comments.
Locked
User avatar
liveitnerd
 
Posts: 9
Joined: Tue Apr 16, 2013 3:09 am

Picking a dev platform for a future product

Post by liveitnerd »

Hi,

I've been playing with both the Arduino and Raspberry Pi for a little while now and love them both for their own reasons. I've now come to the point where I want to develop a product to go to market with but am not sure what dev platform is going to be easiest to take down this path. I'm looking at all options now, rather than just the 2 I have used so far, and am wondering if any one is able to point me in the right direction.

Needs:-
- IPv4 & IPv6 support
- Web and SNMP server ability
- PoE (can build an add-on for this so not critical)
- Ease of taking the product from a dev board to a finished product.

I have most of the rest of the project mapped out but am stuck on what to use as a board. Any assistance would be greatly appreciated.


Thanks,
LiveItNerd.

User avatar
Franklin97355
 
Posts: 23939
Joined: Mon Apr 21, 2008 2:33 pm

Re: Picking a dev platform for a future product

Post by Franklin97355 »

If you are building your own product don't look at boards look at the processor that will do what you need and design the needed peripherals into the end product.

User avatar
liveitnerd
 
Posts: 9
Joined: Tue Apr 16, 2013 3:09 am

Re: Picking a dev platform for a future product

Post by liveitnerd »

I've been able to build it with the hardware on an Arduino but the lack of support for SNMP & IPv6 has precluded me from continuing with it. The RaspberryPi is perfect if not overkill as I can just load a small linux dev and run everything from that but I'm not sure how easy it is to then turn around and push out a production run.

User avatar
adafruit_support_mike
 
Posts: 67485
Joined: Thu Feb 11, 2010 2:51 pm

Re: Picking a dev platform for a future product

Post by adafruit_support_mike »

Given the list of things you want to do, you'll probably want a card-sized computer more than a microcontroller like an Arduino. A TCP/IP stack for microcontrollers does exist, but the servers that fit into such a small platform are fairly simple.

Make magazine just did a comparison of three major development platforms, and that might give you some ideas: http://blog.makezine.com/2013/04/15/ard ... pberry-pi/

User avatar
liveitnerd
 
Posts: 9
Joined: Tue Apr 16, 2013 3:09 am

Re: Picking a dev platform for a future product

Post by liveitnerd »

Thanks for the link to the article. It was most helpful in getting a decision made. I certainly will find using Linux a bit easier to get this done that using another environment and from the looks of the availability of parts etc I'm leaning towards the BeagleBone. While the dev board is more expensive, the chips are available from Element14 without needing to sign an NDA so I can keep the whole thing nice and open. Now to put my head to the grind stone and read up on the BeagleBone and what I can get it to do.


Thanks for the help!

User avatar
cstratton
 
Posts: 294
Joined: Wed Sep 29, 2010 3:52 pm

Re: Picking a dev platform for a future product

Post by cstratton »

There's a huge jump in complexity between a ATMEGA328p and a board with an ARM SOC and enough external memory to run Linux. From a hobbyist / preview production perspective where you just buy more copies of the same board you used for evaluation that may not be a big deal, but when you start talking about making your own version, the complexity of the pi / beagle type designs is not easy or inexpensive to deal with.

What I think it missing from the range of options for productizable projects are the high performance, fully integrated ARM cortex parts with generous on-chip flash, moderate on-chip RAM, and oodles of peripherals. Just about every chip maker in the game offers a range of them (even atmel and the Arudino folks are giving it a go), and most are available in leaded TQFP packages - you can bring these up on an inexpensive to prototype two layer (or even homemade 1 layer plus jumpers) board, and hand build your prototypes with an iron and a little practice. And unless you need it for data storage, you don't have to add the cost / sourcing troubles of micro SD cards which are pretty much a given for a linux image.

Certainly, there are embedded applications where using an entire pi or beagle is a very cost-effective solution for a lot of sophistication and power in a small package. But when you don't actually need a linux box, it's worth keeping in mind that you are going to have to hit quite large quantities before you can get the cost of your custom respin below that of a volume discount on the existing assembled boards. If you have to have a custom board for special peripherals anyway, and an all-in-one arm microcontroller is up to the task, then putting that on your board may be a lot cheaper than paying to manufacture your board, only to stack it on top of a pi or beagle which you also have to buy.

Ultimately, there's a role for each class of embedded device that is on the market. I'm just sorry to see this huge and very accessable middle category skipped over in the analysis.

Locked
Please be positive and constructive with your questions and comments.

Return to “Microcontrollers”