Mass Produce?

Please excuse my electronics ignorance, I am a software developer that is just starting to play with electronics.

As I understand it, Arduino is a development environment for rapid prototyping or one-off projects.

If I build a device using an Arduino, what is the process to get it produced on a PCB?

Thank you

Arduino is a development board, there is no facility to get a PCB develop!
There is http://fritzing.org/ that could help you develop a PCB that i know of

Anyone have idea?

Design PCBs yourself (using Eagle from www.cadsoft.de for example), or have someone do it for you (me for example), then send out to have the PCBs made up from places like this

http://iteadstudio.com/store/index.php?cPath=19_20

and assemble them.
How's your soldering?

For creating your own designs, you will find that the majority of the community uses the free version of EagleCAD (http://www.cadsoftusa.com/). You can design your circuit, then move to the board layout and create your board. Export the Gerber files and send it off to the fabrication house. Some folks here also use KiCAD (http://kicad.sourceforge.net/wiki/Main_Page), and others, as Patgadget mentioned, will use Fritzing (http://fritzing.org/).

There are several options to get your custom design realized. Personally, I use two different fabrication houses for my designs: Itead Studio (www.iteadstudio.com) and BatchPCB (www.batchpcb.com).

Afterwards, the soldering is up to you. There are places that will do a full run for you, fabricate the PCB and populate/solder components, however I have no experience in that area as I've always done my own.

The largest "production" run I've done so far was 300 tiny boards for a light show project. They each had 6 components on them, double sided. There were 20 controllers with that, each one with 15 components, single side. The tiny boards were panelized 4x3, so I placed the components for the one side that has the most, reflowed them, then hand soldered the single components on the other side (an RGB LED). Used a Dremel tool to cut them apart. For the controllers, I placed everything, then reflowed them. Easy peasy, Japaneasy. Took me about two weeks of working (since I only did it after I got home from the office for a few hours every evening), but the satisfaction that I did do it all from beginning till end is, well, priceless. Proud moment when I flipped the switches on the controllers and everything came to life.

An extremely vague question that will get just as many different answers as users posting. Everyone gets to market a bit differently.

I am a software engineer, so my journey through the electronics world will probably be similar to yours. Fortunately, I have the privilege of working with some extremely knowledgable double E's.

What I typically use Arduino for is quick-and-dirty proof of concept. I usually code directly on an AVR for serious projects. So your understanding of what Arduino is, I would say is accurate.

First thing that I started with was some solderless breadboards - http://www.ebay.com/itm/390339573267?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1439.l2649#ht_1198wt_967

Get a bunch of those, add a nice variety of resistors, LEDs, etc.. Learn circuitry and theory, move on to ICs (555s Op-Amps, etc..)

Once you have a circuit that you like, solder it to some perfboard. perfboard: Search Result | eBay

The entire time, keep an up to date schematic of your design. It helps with learning theory, and for posting to forums for help. No doubt you will let the magic smoke out of a few components and will need to know why.

Download Eagle. Better yet, if you know someone knowledgeable with a PCB CAD solution, take their advice. (You'll ask for their their help later, no doubt) Or, if you are not planning (or interested in) on learning PCB CAD - have them create your PCB layout for you.

Depending on your PCB design, the number of layers, complexity, etc.. And how often you plan to do this sort of thing, and what kind of physical area you have available to you. You may want to create your PCB at your home, or have it shipped out to a proto-board house. But - before you pay any cash to make your PCB, I strongly suggest posting your files to a forum to have them looked over. Sparkfun.com is a great place for that.

When it comes time to send your board layout out. I would ask around in your local area to see what everyone there is using. If you, or anyone you know work for a manufacturing company that designs PCBs for their products, ask them who they use. They will usually have good reasons on why.

Depending on what you are doing, you may not ever venture further than the perfboard step. I have seen some cool gadgets running entirely on perfboard.

Keep everyone posted on the awesomeness that you create. Have fun and good luck!

Lots of useful info.

Thanks to everyone for the advise.

There’s mass-production and… mass-production.

I have worked for an assembly-firm which I unfortunately only will recommend to “enemies”.

If the number of boards are a few hundred I’d probably assemble them myself. Set up costs will be quite high and workers may need time to get to know your product which may result in a lot of failures.

But with a few thousand… the speed/skills and price may be interesting. Before choosing a firm I would visit the working environment though, have a good look around and ask a lot of questions.

When I started working for example, people were eating while working, had no protective clothing and only 2 out of 30 workers had heard of the term ESD. I did learn to assemble better at that place, but although I still think I just know a little bit about electronics I also immediately was one of the workers with most knowledge of electronics. Important stuff like calibrating equipment was only done once every two months…

I’ve got no experience with other assembly-firms though and hope this was the worst place ever.

Perhaps I should clarify my intentions.

I am looking to create a few hundred boards. The cost of the Arduino is ~30.00 board.

My question is, if I create the proof of concept/prototype on the Arduino, can I have the Arduino design mapped to a PCB at a lower cost per unit?

Ex.

For an Arduino sketch that controls a servo, can you have the Arduino servo library and the sketch written as a firmware that is loaded onto a PCB along with the MC and have that produced at a low cost?

The final product should not need many of the Arduino components such as the USB interface, etc.

can I have the Arduino design mapped to a PCB at a lower cost per unit?

Yes, you can make a board without most of the standard Arduino components, have a search for “standalone arduino”.

As for the programming, if you are trying to be cheap you’ll use SMD chips and for you they will be difficult to program off board. Hardware wise the cheapest method would be to allow 6 pads on the PCB for ISP and use a $20 programmer with some pogo pins to make the connection while you download.

For an Arduino sketch that controls a servo, can you have the Arduino servo library and the sketch written

Whatever program you have working on the Arduino will run the same on your new board, technically it doesn’t matter what libraries you use. What does matter is the license for the code you use, that has to allow commercial use. I’m pretty sure all Arduino libraries would allow redistribution like this but I guess you’d have to check.


Rob

If I am understanding this correctly then you are saying that I can use the Arduino to offload the compiled binary program, license permitting, and then create a cheaper "standalone Arduino" that I can load up with my binary program as firmware?

Yes, I think you can.

Buying components in such numbers will decrease costs a lot.

A good PCB is probably key to succes though, the more difficult it is to assemble a board, the higher the costs. Using Through Hole Technology components is probably easiest to assemble, but cost since a human normally has to place the components. SMD-components come in many varieties, can.... also be quite easy to assemble, but using extremely small components or... a lot of components cramped on a small space can make it a lot more difficult.

Disadvantage of SMD may also be the use of rather expensive stencils to apply solder (reflow oven) or glue (solder wave) to keep the components where you want them.

In an assembly firm a lot of the component-placement is done automated and soldering often isn't done by hand, but with a reflow-oven or solderwave. A bad PCB-design... usually results in solderbridges which have to be corrected. That takes man hours and... heating a component more then once can easily result in damage.

With good designs I often only needed to check for misplacement/solder-bridges, flash and electronically test a board, which all was done in 2-3 minutes. With bad designs just correcting a board could already take 5-8 times as much time.

When working with such numbers finding and having the right components often was a problem for clients. In a first run you'll probably make sure everything's present at the moment you send all materials to an assembly firm. Quite often I've seen clients who wanted a second/third batch though and who spent loads and loads of time finding all components again.
At the moment atmega-chips is available better as a year back, at the time I've seen several projects being delayed by 3-4 months though.

You could design it yourself or like Crossroads offered hire someone to create a PCB using your proof of concept.
Everything you can do by yourself I would do by myself to save costs, but you'll ofcourse have to ask yourself how good you can do it.

When it comes to licensing and firmware, Arduino is Open source, just like most 3rd party libraries and you can use'm as you see fit. But do check what you want to use to be sure. Loads&loads&loads of people already made/sold arduino-compatible boards to use as developmentboard or as endresult.

I am going to build my Arduino device now and then pursue the avenues highlighted in this thread.

Thanks again to everyone for the helpful info.

You should go and do more market research.
For one thing, which "Arduino"? $30 is getting to be high for an UNO.

License.... it's Open Source. You not knowing that speaks loudly.

Well, it's your shirt. Buy some hot sauce to go with the salt.

GoForSmoke:
You should go and do more market research.
For one thing, which "Arduino"? $30 is getting to be high for an UNO.

Just bought an UNO locally for $37.09 so how is $30.00 high?

GoForSmoke:
License.... it's Open Source. You not knowing that speaks loudly.

There are several Open Source licenses with certain restrictions on each. You not knowing that speaks loudly.

GoForSmoke:
Well, it's your shirt. Buy some hot sauce to go with the salt.

Thanks for the constructive input... oh wait....

Thanks to everyone else.

GoForSmoke has a good point: market research. If you plan to sell something at 300 quantity, you don't want them to sit in your garage for years to sell off. Things happen too fast (life cycle of electronics with accelerated metabolism) in electronics world and in that amount of time your project could have already been dead. I am not a software guy but if you ask me, I'm maybe 70% software 30% hardware. I've never designed a board before I came to know arduino. I remember placing my first batch of 100 phi-1 shield PCBs, a bit anxious since I didn't know how the arduino users respond to such shield. I was lucky I made a popular board and a fame for myself (in the order of epsilon). If you can easily reach your potential users, do so.

Seems like many provided their valuable opinions. Don't want to regurgitate any of those. Here's my addition: between breadboard and 300 production, you got to add a 10-pcs run, for testing, for firmware tweaking, and for giving them away if you want, for exchange of feedback. I used to order 100 boards at a time, but then 50, and now 10 at a time. It's much less pressure for me to order less :). Plus you can make revisions between orders if you want. If you ask a risk manager, you will be told profit increases with risk. :stuck_out_tongue:
In my case most what I sell are in kits form so there's not much need for assembly.

By the way, does your product not need an enclosure? I have not seen you mention that. Hope it needs not. Otherwise you should design your circuit around an enclosure you have in mind. Can't just do that afterwards.

My stand-alone arduino project: phi-panels

Amazon - official UNO made in Italy - $27.50

That's what I paid, Slick. Do more looking around, assembled clones sell in the low $20's.
You can easily find that in under 1 hour of "market research".
But what did you do?

There are several Open Source licenses with certain restrictions on each. You not knowing that speaks loudly.

Whereas I read the actual one that applies. No, you don't need special dispensation. Neither do the many and varied people marketing clones and compatibles today.

You don't supply enough data for "constructive input" beyond, hey, you need to get your head out of your plans and look before you leap.

GoForSmoke:
Amazon - official UNO made in Italy - $27.50
http://www.amazon.com/Arduino-A000046-UNO-board/dp/B004CG4CN4/ref=sr_1_1?ie=UTF8&qid=1323025351&sr=8-1

$27.50 + shipping is $30.00

$27.50 is $30.00 for that matter. $2.50 is negligible.

GoForSmoke:
That's what I paid, Slick. Do more looking around, assembled clones sell in the low $20's.
You can easily find that in under 1 hour of "market research".
But what did you do?

I never asked about clones. I am sure that they are cheaper (as knockoffs always are) and I appreciate the heads up but you dont need to be rude about it.

GoForSmoke:
There are several Open Source licenses with certain restrictions on each. You not knowing that speaks loudly.

Whereas I read the actual one that applies. No, you don't need special dispensation. Neither do the many and varied people marketing clones and compatibles today.

Hold on here....
I said I would check the licensing...
You commented that if I needed to check the license then I dont know what open source means...
I said that there are multiple levels of open source licensing...
You said that you checked the specific open source license that pertains to Arduino...
which is exactly what I said that I would do in the first place!

So just admit that you were wrong to suggest that the license did not need checking and that I did not understand open source in the first place. Your comment was unwarranted and incorrect... and rude.

GoForSmoke:
You don't supply enough data for "constructive input" beyond, hey, you need to get your head out of your plans and look before you leap.

I am "looking before I Ieap". What do you think this thread is for in the first place?

I have gotten lots of valuable info from other forum members. I have gotten nothing but improper, rude comments from you. Please stay out of my threads if this is all you have to offer.

I often help people in my areas of expertise, Linux systems engineering and web application development, and I dont see the need to make the types of comments that you have made here. If thats your personality, fine but please respect my thread enough to stay out of it and let me converse with those willing to help.

Let's not start a fight, we're talking starting a mass production, point is getting missing info available, there's no production yet.

I do think a lot of good points are made though, spicy ones included.

One thing I hadn't thought of this morning, what... do you want to mass-produce ?
You gave a board driving a servo as example, but that's all.

Thing I'm thinking about is possible requirement of certificates.

Should you, as example, need several hundred boards for an art-project you're making I guess you could do almost everything you like, please... keep it safe though. Should you, on the other hand sell a product that controls AC-equipment as well or otherwise could be dangerous, you very probably need your product to be tested by a safety-lab to get the right certificates before you can sell.

Different countries use different rules. If you need your product to be tested that will unfortunately be a big investment, it may be too expensive to gain back with a few hundred boards.

liudr mentioned enclosures, I find that's the single hardest part of a project, getting the right box at the right price. In fact I've just given up yet again and my current project will be naked.


Rob

That is a good point. I do carry business insurance for general liability and the device that I am controlling is UL listed and NOT sold by my company. I can resell the Wall Wart separately so it is covered by the manufacturer leaving the only liability on me being that of the Arduino board and the Servo. Since both use under 6v operating voltage, the safety risk is very minimal.

As for the packaging, this is something I have not had to deal with yet and I expect it will be difficult. I plan to have a "controller" box that can just be a glorified project box with a few holes for LED display. The smaller box for the servo motor is more of a challenge as it has to fit to a specific form factor for proper mounting. I would like for both boxes to be branded either via a sticker or possibly engraved/raised lettering/logo. I expect any local machine shop with a CNC machine can handle that.

Thanks for the valuable points.