Hi all,
So after we have tested our wonderful project and prototyped it using an Arduino and a shield(s) and are happy with the results.
How does one go about making the project it's own dedicated circuit?
Are there any companies out there that can take the project (Arduino and all components) and turn it into a dedicated circuit board?
Thanks in advance
Sure, guys like me do it for hire all the time.
You can see some of my designs here
http://www.crossroadsfencing.com/BobuinoRev17/
and there are many others that I have not posted as well, '328, 1284P, and '2560 based.
I've had designs go all over the world - US, UK, Netherlands, Germany, Spain, Middle East, South America, Malaysia, Australia, India to name a few. Has been quite interesting 8)
Sometimes just Gerber files & a Parts List, sometimes completed boards.
Good advice above.
Im new to arduino and code but old to PCB layout and schematic design. I would recommend you write a quick design brief. Detailing exactly what your code does and what hardware you would need in order for you product to work.
Further, list any information which is 'set in stone'. For example many people who I work with have already done the mechanical design and have a chosen off the shelf or custom enclosure. Who ever is doing the hardware design would need to know dimension of this for the PCB outline and mounting hole positions.
In your case, it might be simplest to give a description of what the code does and what the product is and then detail (accurately) in a table which arduino (or shield) pins you are using for what. You can expect the final design to be simplified from you current test set up as there will no doubt be some sections of shields not used.
Once the schematic has been completed and signed off by you the PCB layout can begin. At this point you need to choose whether you want to receive the boards pre populated (with components) or without. And how you would like to assembly the PCB.
For simple projects people often want to assembly which normally means through hole components. However when my firm produced PCB's populated in any quantity greater than 10 we always use surface mount components as I have a pic and place machine.
Hope that helps describe the process 
If you have any questions about specific parts then post up and I will do my best to give an informative reply. It might also help if you released your code to the designer, if they are building the board too so that the design can be properly tested and signed off before shipping.
Rob,
Thanks to CrossRoads and Rob
I guess with what you are saying that next question would be, is there any costs savings ?
Let us for discussion say we have created a simple Arduino with Bluetooth shield, that allows a smartphone to pair with the bluetooth and thus send a code which will have Arduino activate a power relay and turn on a household switch.
A very simple project, sure one can use the Arduino with the bluetooth shield, however is would look more professional if it was it's own circuit board that would do it all.
Thanks again
There are indeed several things which you can do to minimise cost.
Reduce the time spent by the PCB designer, this can be achieved with a good brief and as much coherent information as you have. You could sketch out a schematic for example which could be modified by the designer to include circuit protection and other useful hardware features which might not be obvious at first. As a minimum you will need to supply a pin chart. Assume the designer has no knowledge of code and knows nothing about your product.
The enclosure will impact upon the product cost. The smaller you need the PCB the more likely you will need multiple layers or components on both sides of the PCB, this increases build cost more than design cost.
Keep complexity to a minimum. This will save design time and re-design time of something goes wrong or changes. It is very easy for a design to go beyond scope, some of the worlds most profitable products have also been the simplest.
Remember that it is likely your product will need to be externally tested to comply with various region specific regulations before it can be properly sold. There are many who don't do this but it is certainly something you should consider if this is being used with mains electricity and is in a house hold. EMC and immunity and ESD testing is a specific test which would normally be carried out.
You (or the pcb designer) should also plan to do your own testing of the product. You will want to know what the failure mode of a product is. If for example your unit is controlling a heater, something fails, your blue tooth app thinks it has turned the unit off but actually hasn't.. the heater catches fire and wipes out the owners David Bowie record collection... you would feel pretty guilty as a minimum and legally liable in a worse scenario.
Further to my last post.
You should in this case think of also writing a technical brief, giving specific information about the maximum voltage and current which will be switched, the life time expectancy of the unit (probably in total number of actions), the maximum and minimum operating temperatures etcetc.
Minimise design time and complexity to save money.
Rob,
Cost savings really depends on quantity. If you only doing 1 or 2, then the design costs are spread over few units. It you are doing 100, then the design costs (Non Recurring Engineering, NRE) are spread out and adds less to each unit.
So if it cost you say 4 hours to combine the two and develop the parts list, and that cost $240, then that adds $2.40 to the price of each unit.
If it then cost $350 for a manufacturer to create a solder paste stencil and program a pick & place machine, that's another $3.50 per card.
Assembly costs, say $8/card.
Programming, Shipping, Packaging, you have those either way.
Material - whatever it is - your 100 lot will never be as low as someone making thousands.
So you can see how it adds up.
You can do an initial 10-lot, with little markup, to test & use to get the word, and then a larger lot, and those little markups will start adding up, maybe you start saving on material by buying in larger lots, or figuring out lower cost shipping methods, etc.
So, it really depends on what end goal is.
Well said, my reply above was assuming a prototype batch.
IMO a safe route is always one with some pre production prototypes. The greater the quantity the lowers the unit cost.... as long as they work.
Rob,
EDIT - unless its a very simple design, in which case a pre production run really isn't needed.
Arduino basics + bluetooth module can be pretty simple, with just a serial interface to the module. For example