I'm trying to build out a little gadget on a breadboard that, if I could get it working, I'd want to see if I could build out as a PCB and sell on my website.
That lead me to this article and this quote:
"I’ll warn you that obtaining certifications isn’t cheap and most products will cost at least $10k to $30k to certify...[FCC Certification] is required for all electrical products sold in the USA. Products that don’t purposefully radiate electromagnetic energy (i.e. no wireless functions) are classified as non-radiators. On the other hand, wireless products purposefully transmit electromagnetic energy and are classified as intentional radiators. It is much more expensive to obtain FCC certification for an intentional radiator."
So, help me out here.
Let's say I build my basic little gadget (which will basically consist of a microcontroller, an LED, a buzzer, a proximity sensor, and a transmitter/receiver) and want to start selling it. Is that quote valid in my case, or is there something I'm missing there? I've been planning on using these dime-a-dozen 433 MHz transmitters/receivers - would I really need to spend 30k on certifications before I could start to sell these on my website?
elveax:
Let's say I build my basic little gadget (which will basically consist of a microcontroller, an LED, a buzzer, a proximity sensor, and a transmitter/receiver) and want to start selling it. Is that quote valid in my case, or is there something I'm missing there? I've been planning on using these dime-a-dozen 433 MHz transmitters/receivers - would I really need to spend 30k on certifications before I could start to sell these on my website?
The rule is: any electronic device with a circuit operating at 400Khz or above must be certified by a licensed lab before it can be sold commercially. Moreover, anytime you change anything, including software, on your device, it must be recertfied. Your complete and operation device is what is certified, not the components.
If you device is AC powered, it must also be certified by a U/L certified lab, or CSA certified. The use of a wall wart means the wall wart means it must be the certified part.
The interesting takeaway for me here is that unintentional radiators (above 400 KHz but without any known RF producers) certification is closer to 1k, and that if you use FCC certified transmitters, you can run your device thru the FCC as an unintentional radiator. Basically, you're saying the transmitter is already good to go, and you want them to check to make sure the rest of the device is good to go, too.
The interesting takeaway for me here is that unintentional radiators (above 400 KHz but without any known RF producers) certification is closer to 1k, and that if you use FCC certified transmitters, you can run your device thru the FCC as an unintentional radiator. Basically, you're saying the transmitter is already good to go, and you want them to check to make sure the rest of the device is good to go, too.
Thanks everyone!
No, it's the final assembly that is tested in operation, and you have to supply the operating instructions and other than DC power supplies, any other equipment needed for the test, such as keyboards, displays, etc.
Paul_KD7HB:
There is no such thing as an FCC certified transmitter sub-assembly. IF there was, as soon as you add it to something then it is no longer certified.
Paul
Help me out here - I'm quoting from the article from Sparkfun here:
"Some transmitters come with an FCC ID on them. Does using these devices exempt me from further testing to achieve FCC authorization?
Not entirely. While it does lower the bar from the costly Certification process, you are still responsible for ensuring that your product does not emit other radio frequencies. In short, you must test it as though it were an unintentional radiator."
So, if I had a project that used, say, this transmitter (and no other):
Then I'd be able to certify as an unintentional radiator, yes? I wouldn't have to ask them to recertify the XBee I bought from Digi-Key.
In England we're allowed to self-certify to CE standard based on an assessment , which, for RF stuff, means using pre-approved modules. Unless the overall product is ghastly and interferes with others, it's unlikely you'd have any problems.
I don't know how I can explain this any simpler than you cannot FCC certify a product without making it operational and being tested in a certified lab. If you have stuff in the box that is not tested, then I suspect the lab would not even accept it for testing.
We have enough RF noise generated by imported devices that are marked FCC certified that I can hear the noise level rise when I point my very low noise 144MHz antenna toward Portland, OR, 100 miles away.
You can do what you want, but investigate the fines being paid by businesses that import non-certified electronic devices.
Also, keep in mind that if you are using the Arduino IDE and any of its libraries to build your s/w inside your device, then for sure there are open source licensing terms that you must conform to once you begin to sell or distribute your device.
Depending on the Arduino libraries used, it may even mandate that you fully disclose ALL your source code to anyone obtaining and using your device.
Even if you use only libraries that are very permissive and do not require any disclosure of your code, such as BSD, you are still subject to LGPL licensing rules since the bulk of the Arduino IDE bundled libraries, including the arduino core library are licensed LGPL.
This means that if you choose to keep your source code private, you must still provide a way for your users to be able to update the LGPL portions of the code inside your device.
They must be able to rebuild your s/w image and update the device with newer or modified versions of any LGPL code inside your device.
This is a very tricky & sticky area for embedded platforms, and with Arduino it is particularly difficult to comply with if not impossible.
For example, if you use the IDE to build your code, the IDE has no way of building images where some of the code is binary objects and others are source code. So when using only the IDE it is impossible to comply with the LGPL licensing terms if any of the code is not open source.
You will to carefully study the LGPL licensing requirements
And if you are trying to sell a device that is not fully open source you will need to figure out some way of building & providing tools to allow full compliance with LGPL since as of right now, images built with the Arduino IDE depend on LGPL code but the IDE does not support the ability to build images in a way that complies with the LGPL licensing requirements if not all the code is open source.
elveax:
I'm trying to build out a little gadget on a breadboard that, if I could get it working, I'd want to see if I could build out as a PCB and sell on my website.
Is it legal to use 433Mhz for 'gadgets' in the US ?
I thought the (licence exempt) ISM band in the US was 915Mhz ?
A. There's no need to be rude. If you don't want to assist, you don't have to answer. I'd rather have no answer than a rude answer.
B. If I understood everything perfectly, I wouldn't be on here asking questions.
C. But, anyways, I think we're talking past each other.
You're insisting that all products have to be certified by the FCC after they've been assembled, and that's true. I don't dispute that.
But, what Sparkfun is saying (and what I'm saying) is that whether I have to certify the product as an 'intentional radiator' (12k or more) or as an 'unintentional radiator' (1k to 1.5k) depends on whether or not the wireless module on the device already has an FCC ID.
So, for instance, an electronic, fully assembled, where the only radio is an XBee (which has an FCC ID) could be certified as an unintentional radiator. If it was determined that the electronic was outputting RF waves outside of what the XBee puts out, I'd fail and have to resubmit as an intentional radiator. Sound accurate?
I don't see how it even could be a different way. I can't testify as to what is going on inside an XBee, in terms of firmware or circuitry, and the company that makes it isn't going to give it to me so I can get through an FCC screening. The only reason anyone can use an XBee in their product at all is because they've already gotten it through the FCC. The FCC only has to screen the assembled product to see what it's doing outside of what the XBee does.
srnet:
Is it legal to use 433Mhz for 'gadgets' in the US ?
I thought the (licence exempt) ISM band in the US was 915Mhz ?
That I don't know - I know lots of Arduino projects use them, and I was submitting them here as an example. That band might be out of the question as far as the FCC is concerned.
elveax:
I don't see how it even could be a different way. I can't testify as to what is going on inside an XBee, in terms of firmware or circuitry, and the company that makes it isn't going to give it to me so I can get through an FCC screening. The only reason anyone can use an XBee in their product at all is because they've already gotten it through the FCC. The FCC only has to screen the assembled product to see what it's doing outside of what the XBee does.
Keep in mind that the s/w that uses a module such as your s/w can dramatically affect the emissions.
And that is why a device as a whole must be tested in "normal" operation vs testing individual components.
Recently there have been some issues with the open source router code DD-WRT.
Some manufactures have recently stopped supporting it, or even kept certain h/w information secret as an attempt to block it use.
The main reason being that the open source code could potentially turn up the transmitter power beyond what is legally allowed and/or also turn on usage of incorrect rf bands for the country the device is being used in.
The FCC complained and threatened to hold the manufactures liable, so some of the manufactures responded by attempting to disable its use.
bperrybap:
Keep in mind that the s/w that uses a module such as your s/w can dramatically affect the emissions.
And that is why a device as a whole must be tested in "normal" operation vs testing individual components.
Recently there have been some issues with the open source router code DD-WRT.
Some manufactures have recently stopped supporting it, or even kept certain h/w information secret as an attempt to block it use.
The main reason being that the open source code could potentially turn up the transmitter power beyond what is legally allowed and/or also turn on usage of incorrect rf bands for the country the device is being used in.
The FCC complained and threatened to hold the manufactures liable, so some of the manufactures responded by attempting to disable its use.
--- bill
Sure, that makes perfect sense. Like I was saying to Paul, I understand that the device would have be tested in full, as an assembled unit. The sticking point in what we were talking about is whether or not I could submit a device that transmits over RF as an unintentional radiator if the radio device already had an FCC ID.
elveax:
Sure, that makes perfect sense. Like I was saying to Paul, I understand that the device would have be tested in full, as an assembled unit. The sticking point in what we were talking about is whether or not I could submit a device that transmits over RF as an unintentional radiator if the radio device already had an FCC ID.
And I did do a Google search - that's how I came across the Sparkfun article I shared earlier. That Sparkfun article is quite a bit more detailed than the one you're sharing here, and the one you're sharing here doesn't address the question about the effect of using FCC approved transmitters in new devices.
So, in short, if you didn't know the answer to the question, you didn't have to answer. It was a perfectly reasonable question to ask on this forum. Some people here do have experience getting FCC approval for devices they built, starting with an Arduino.
I appreciate everyone's help, I think I have learned everything I needed to know at this point.