Moving a project from the Arduino to ?????

[dlglaz]

Ok, so I get my project working ont the Arduino, what is next if you want to move it to a sellable product? I assume I get some schematic software, get it all laid out, and find someone to produce it for me? I have to upload the programming somehow, burn it to a chip I guess, and do some soddering.... thoughts?

[jersagfast]

I think the next step would be to transfer your project on a breadboard and get it working stand alone. From there you will need to design a PCB and order parts from suppliers. As far as burning the chips, Ladyada has a great tutorial here:http://www.ladyada.net/library/arduino/arduinoisp.html and a great stand alone programming kit here: http://www.adafruit.com/products/462. Happy building!

[cstratton]

A lot may depend on the volume in which you envision making them.

While you can certainly program chips in a ZIF-socket based programmer, or even buy them pre-programmed, it may actually be preferable to design your board so that you can program the chip installed in the product, in a rig that also does some at least semi-automated testing. You could do that either by bringing out enough signals to headers, or using "pogo pins" (available here iirc) in the fixture pushing against empty holes in the board.

You may also want to think about designing to ease assembly, including the big decision between through-hole components that are great for hobbyists building kits, vs. surface mount which semi-automated assembly services can deal with using pick-and-place machines and solder paste.

[dlglaz]

Still a little confused, so I have a project, it works all the way to field tests. I have a schematic that shows how I have the Arduino and the wiring setup with the components. I need to make it all a lot smaller, fit in a case of some type and be a lot cheaper than buying an Arduino for each unit.

[neutron spin]

Still a little confused, so I have a project, it works all the way to field tests. I have a schematic that shows how I have the Arduino and the wiring setup with the components. I need to make it all a lot smaller, fit in a case of some type and be a lot cheaper than buying an Arduino for each unit.

Ok here is what i would do in your case. First acquaint yourself with a good CAD program for designing PC's...Eagle or whatever. You re-create the schematic using the CAD. Then verify everything is per your circuit design and generate PCB files form the program to have the boards fabricated. Sparkfun or others offer a small prototype service for small numbers of boards. Cross your fingers and hope you did everything correctly and go from there. If you lack experience in doing this perhaps you could contract it out to someone who can do this. Not knowing your engineering experience it is hard to give such advice.

If this is going to be a mass produced project then that is ythe only way to accomplish this. You can substitute surface mount equivalent parts in the design to make the footprint smaller if you wish. But a prototype will need to be developed first.

[abqlewis]

You have a project that works, but is based on an eval board that does a number of things that you hopefully don't need in your project. If your project uses the USB, it probably doesn't need the VIN, or the power sensing/switching circuitry. If it doesn't use the USB, you can gut the FTDI or 8U2 circuit. All of this will help save money and space. Even if you use the USB for debug output, it's probably best to move that off the repeated project board (where you have to have duplicate parts for each board), to a separate, single FTDI cable or board. Even if you use the complete Arduino schematic, it will be tough (availability and $$$) to get the exact same components. So no matter what you have working, it will be cheaper and easier to prototype your actual deliverable circuit on a breadboard. Also, custom circuit boards don't always work the first time. Mistakes can be made in schematic capture (you), or manufacturing (PCB house). If you've already breadboarded the circuit, it can help debug those problems too.

[jersagfast]

This might help you out a bit if you are not familiar with running Arduino projects on a breadboard. http://thecustomgeek.com/2011/12/07/get-on-board/

[Jack52]

Still a little confused, so I have a project, it works all the way to field tests. I have a schematic that shows how I have the Arduino and the wiring setup with the components. I need to make it all a lot smaller, fit in a case of some type and be a lot cheaper than buying an Arduino for each unit.

You probably don't need a lot of the support chips/hardware if the project doesn't need to communicate with a PC. You will need to decide first how the project will be powered, which determines if you need an on-board voltage regulator and its support parts. Then it is a simple matter to place the chip on the pcb layout and connect the needed parts to it.

PCB layout is not for beginner and many of the CAD programs like Eagle have a steep learning curve. You might want to hire a consultant on your first project to get it going.... or who knows, there might be a volunteer from this forum to help you get laid out on the pcb.

regards, Jack

[jersagfast]

Also, these stickers are a great reference when starting to breadboard. Wish I had them when I started..
You can buy them here:
http://www.adafruit.com/products/554

[kscharf]

I've been down this route, at least partway. I've prototyped a design using Arduino and shields and then drew up a rough schematic showing the connections between the processor and off board parts. We had a draftsman available who re-drew my rough hand drawn schematic on a CAD program to design circuit boards to the size and shape required to fit the mechanicals. What I had to capture was transfering the Arduino digital and analog pin numbers to the actual AVR port/pin numbers.

[jersagfast]

This might help. - With the correct file this time.

[westfw]

so I have a project, it works all the way to field tests. I have a schematic that shows how I have the Arduino and the wiring setup with the components. I need to make it all a lot smaller, fit in a case of some type and be a lot cheaper than buying an Arduino for each unit.

So you need to design, or have designed, a new board (and possibly packaging) that omits all the Arduino components that you don't need, includes all the additional components you DO need, provides some mechanism for loading the firmware socket for chip, ISP connector, etc), meets your desires for size, and is "manufacturable" by one or more definitions (that will depend on target price, target number of units, etc.) Then you need to find a way to get the actual manufacturing to occur.

You can learn some or all of those things and do them yourself, or you can hire consultants and/or design companies to do it for you (at significant costs/etc.) These are significantly difficult tasks; it's pretty much a myth that a sufficiently clever design is profitable without all this non-design stuff happening as well.

A start would be to look at how the "dangerous prototype" folks have their stuff built by Seeedstudios.

[jersagfast]

These are significantly difficult tasks; it's pretty much a myth that a sufficiently clever design is profitable without all this non-design stuff happening as well.

I could not agree more with that statement. It's exciting and a lot of fun, but it's some serious hard work.

I wish the best for you!