ArduinoGuru wrote:What's involved with being considered a cellphone manufacturer?
A lot of communication with the FCC, a lot of paperwork, and a lot of money. Adafruit had to be on a pretty solid financial footing before we could seriously consider it, and the company has been in the black pretty much since day #1.
The first step is to find an FCC rep in your area and talk to them. They're good people, they know the rules inside out, and they're much happier to answer questions before you do something than to come in as enforcers after you do it.
ArduinoGuru wrote:What's a small company's best bet for selling OSHW?
Don't think of it as selling products.. by the definition of OSHW, anyone willing to invest the resources can sell the same product, and there's always someone with deeper pockets.
Our actual business revolves around services.. in order of relative importance: education, accesibility, and provisioning. The hardware ties into all of those, but the value lies in the intangibles.
Our first goal is to help people learn technology, primarily electronics hardware and code. These forums and the project tutorials in the Learning System are our main tools for doing that. In a given week I'll explain everything from how to measure voltage with a multimeter to transmission line theory, depending on how much background information the person asking questions has coming in. We see those as equally important if they help someone make the next step in understanding how to make stuff. The hardware tie-in is that we build the projects with our own boards (or more accurately, we sell the boards we use to build our own projects), and after a while we know the designs by heart.
Our second goal is to make the hardware of 2018 accessible to people who don't have a $100k pick-and-place line. Most of our breakouts carry devices that only come in QFN packages, which are beyond the hand-soldering skills of even experienced hackers. We provide the service of putting in a form most hackers can use. Again, over time, we learn the hardware inside out, and get good at teaching people how to use it.
The third service, provisioning, means we do all the legwork of sourcing parts, having PCBs made, and bringing all the pieces together. We have a lot of different suppliers, and can find/procure stuff that most people would never hear about otherwise. We played a strong part in making NeoPixels (WS28xx RGB LEDs with integrated PWM driver) a normal thing among Makers. That also includes things like paying for time in an EMF lab so we can FCC certify FONAs so the person who wants to make a remote sensor doesn't have to.
Whatever you want to build and sell, look at it from the perspective of someone who has your schematic and BOM. Even if you didn't publish them, reverse-engineering them isn't hard. Then ask yourself what you have that they don't, but which would be valuable to them. That's what you can build a business around.
ArduinoGuru wrote:The FCC rules don't make sense to me for things like cellular shields and modules. If, for example, a cellular shield is EMC-tested with an Arduino Uno as the host, how could that cover the case of using the shield with an RPi or Beaglebone?
Imagine putting the board in an RF lab with all the appropriate equipment plugged into an Arduino, then moving it to an adapter that connects it to a RasPi or BBB. What effect does that have on anything you can measure in an RF lab?
ArduinoGuru wrote:It's like you say about changing the batteries. Would a cellular shield be considered a "kit" and therefore be exempt from testing?
No, the cellular shield will always be an intentional emitter that requires certification (for cellular, hoo boy will it require certifcation). If you don't pay to have it certified yourself, you'll have to buy a certified module from someone else and follow the conditions of their certification to the letter.
The Arduino, RasPi, and BBB fall broadly under the heading of "fancy voltage regulator".
To the extent that an Arduino, RasPi, and BBB operate at frequencies that can generate RF signals at frequences and power levels the FCC cares about, the makers of those boards have had to get them certified. You'll find the markings on the PCB.
If you put the Arduino and cellular shield in an RF lab and simply plug them together.. no signals from the Arduino telling the cellular module what to do.. does it change the RF emission profile of either device? If not, there's nothing to certify.
If you send the cellular shield a set of control signals from the Arduino and measure the shield's RF output, then swap connections and send exactly the same set of signals from a RasPi or BBB, does the cellular shield's output change in any way? If not, again, there's nothing to certify.
In general, swapping the data connection from an Arduino to a data connection to a RasPi will have no more effect on the cellular shield's RF emission profile than swapping the Arduino's LM117 5v regulator for an LM7805 5v regulator, or changing the teal soldermask to yellow.
The first question to ask is always, "what's emitting signals the FCC might care about?" Then for any change, ask, "what effect will this change have on the measurable emissions of the thing the FCC cares about?" More generally, you don't just set a device on a table and certify it. You connect it to power, you send it control signals, you transmit messages with it, and so on. Those operating conditions are part of your certification.. "when connected to 5v power at pins X and Y" and so on. Those conditions are part of your certification.. the device is certified to operate under those conditions and no others. The report from the RF lab shows that the device complies with FCC regs when operating under those conditions.
If a change external to the certified device doesn't take the device outside its certified operating conditions, the FCC doesn't care.