Questions about FCC Certification

[deweyusa]

Hi

I have located a customer who is very excited about the possibility of incorporating wireless into his products. I have tested some prototypes that operate with cheap transmitter-receiver pairs running in the unlicensed 433.92 MHz ISM band.

I am running into some questions around the law when it comes to selling something similar:

1) I'm fairly sure I cannot sell my product using the transmitters I mentioned above, as they aren't FCC certified. Can I buy a pre-certified transmitter (such as those sold by **broken link removed** or **broken link removed**), and interface to it with a microcontroller with USB PC interface, putting everything in the same box, as long as I don't modify the RF section, and still have the overall device maintain the FCC Part 15 certification the radio came with? (I include the PC part because I read something in Part 15 that calls some PC parts unintended RF radiators in some circumstances, and don't understand what part they play in the ability to get something certified).

2) If the answer to #1 is no, are there any other options for me when it comes to purchasing transmitters I can then couple with a microcontroller of my own choosing, or maybe those that come pre-certified with a microcontroller I can further program, for use in a product for commercial sale that don't require me to go through the relatively expensive process of FCC certification?

Thank you for any help!

-Dewey

 

[Sceadwian]

Yes, but keep in mind the transmitter/receiver AND the antenna have to be certified together. You can't combine an aftermarket module with a custom antenna for which the device did not pass it's original FCC certification. This includes altering the transmission line. So pretty much the only modules you can blanket certify are those with built in antennas, this does not however cause the device which operates it to be Part 15 certified and something it outputs could void the units overall part 15 compliance outside of the RF module itself.

All you could legally do with a module with a built in antenna is certify the the module has FCC part 15 certification, NOT your entire device.

 

[deweyusa]

Hi again

Thanks for your reply. I don't quite understand what you mean. I would like to avoid having to certify anything myself. Are you saying it's okay for me to purchase a pre-certified transmitter and receivers and add a microcontroller of my own without having to get the whole thing re-certified?

Thanks again
-Dewey

 

[Sceadwian]

I thought I was pretty clear. If you purchase a certified module, ONLY THE MODULE is certified not the device you place it into. You can not alter the board including the antenna that comes with it or it voids certification, without certifying the entire unit the unit could in theory cause the certified device to fail it's compliance. There is no way around these technicalities, other that to overlook them as unimportant as they are.

If you build a certified module without altering it into something you make you can certify that the RF front end is FCC Part15 compliant, nothing more.

 

[deweyusa]

Yes, I did understand the points you made - a) that once the antenna is changed, certification is lost, and b) that only the RF section that is pre-certified retains any previous certification when added to other circuitry. So I guess a follow-on question embedded in the first that I am trying to answer is: When I include a pre-certified transmitter in a product, and the circuitry I add to it doesn't generate any RF, and I don't alter the pre-certified RF portion, does the law require me to re-certify the entire unit afterward as a new device, or is it enough to say that the RF portion was already certified, and therefore the entire bundle is in compliance with Part 15?

Thanks for your help

-Dewey

 

[Sceadwian]

If you want the entire device to be certified yes regardless of what you consider the rest of your device the FCC can not classify it until tested. You can not legally rubberstamp the whole unit from the RF sections certification, period no if and sor buts.

 

[deweyusa]

Thanks for your second reply. So if I understand you correctly, if I put my device in one box, leaving the RF device in its original box, and connect them with a wire so I can control the RF section with my device, the FCC still requires me to certify the both as a new device to be in compliance?

 

[Sceadwian]

Yes, that is correct. The certification itself is generally speaking not required unless the customer requires it.

 

[cachehiker]

There is the law and then there is the enforcement.

As an intentional radiator, you should stay within the conditions necessary for the transmitter and receiver to maintain their FCC certification with your initial design.

As an unintentional radiator, a digital device operating above 9KHz (the rest of your design) is technically required to have FCC certification but there are a lot of companies operating in the grey area like we do. They eventually get the certifications as the product is refined and use these certifications as evidence that later, similar products are likely to be in compliance. This seems to satisfy most customers and keep the FCC out of their hair.

Bottom line: Don't panic.

Develop your product employing good design techniques to minimize radiated and conducted emissions. When the design is stable and ready to start shipping in quantity, get it tested. If it passes all is fine and dandy. If it fails by a little bit, fix it with version 2, discontinue version 1 when version 2 is ready, and get version 2 certified. If it fails miserably, rethink shipping anything beyond prototypes of version 1.

Once you know what you need to do to be in compliance, start integrating things a bit more, use the certification of the prior product to demonstrate to the customer that the new product is likely to be in compliance with FCC requirements, and get the testing underway as the design matures.

Certain customers though, are likely to demand you do more.

 

[Sceadwian]

As an unintentional radiator, a digital device operating above 9KHz (the rest of your design) is technically required to have FCC certification but there are a lot of companies operating in the grey area like we do. They eventually get the certifications as the product is refined and use these certifications as evidence that later, similar products are likely to be in compliance. This seems to satisfy most customers and keep the FCC out of their hair.

Can you reference the section of FCC rules that this information is found under, I'd like more information if possible? Are you aware of any exceptions?

I understand the don't panic note I was just quoting what little of the law I'm aware of, haven't exactly sat down and read through all the FCC rules that are applicable, so I'm being very careful about what I say is okay.

 

[cachehiker]

Part 15, Section 31, Paragraph 6:

(6) Digital devices authorized by verification, Declaration of Conformity, or for which an application for equipment authorization is filed on or after May 1, 1994, and intentional and other unintentional radiators for which verification is obtained, or for which an application for equipment authorization is filed on or after June 1, 1995 are to be measured for compliance using the following procedure excluding section 5.7, section 9 and section 14: American National Standards Institute (ANSI) C63.4–1992, entitled ‘‘Methods of Measurement of Radio-Noise Emissions from Low-Voltage Electrical and Electronic Equipment in the Range of 9 kHz to 40 GHz,’’ published by the Institute of Electrical and Electronic Engineers, Inc. on July 17, 1992 as document number SH15180. This incorporation by reference was approved by the Director of the Federal Register in accordance with 5 U.S.C. 552(a) and 1 CFR part 51. The Commission encourages the use of this procedure for testing digital devices, intentional radiators, and other unintentional radiators as soon as practical.

My verbiage might be a bit dated as part 15 received a major rewrite/update in 2002 (shown above) and it has been somebody else's job since then. I only inspect boards here and there for glaring design blunders that will lead (or have led) to excess conducted or radiated emissions. My favorite is supply tracks around a board's perimeter that resemble (and behave like) UHF loop antennas.

If you've stayed within the conditions necessary for the transmitter and receiver to maintain their FCC certification, you don't need to formally register your product with the FCC, you can simply declare its compliance with a decal saying it has been found to comply with FCC Part 15 yada yada after it passes testing.

It also refers to testing as soon as practical. Different manufacturers and different customers interpret this different ways. We have a few that insist on testing before accepting delivery. Most will accept evidence that similar products have been found to be in compliance and having something from every major project category tested at some point.

 

[deweyusa]

Wow, that's really good news cachehiker. Thanks for your reply. Sorry I haven't responded sooner, I kind of did give up some hope there and hadn't checked back in a while.

I had no idea I had the option to get my product certified or not (that IS what you're saying, isn't it?) Also, I should be clear - I think the words "certified" and "verified" are being used interchangably here? In the end, what you're saying is that I can wait on testing at an FCC approved lab, do my own, and put a decal on the product saying "this device passes FCC part 15 rules"? Hot damn. I was under the assumption I HAD to get my product tested by a 3rd party FCC-sanctioned lab.

So, it sounds like you also agree that even if I buy a "pre-certified" transmitter module and antenna, when I couple it with my device, I need to make sure the new combination of circuitry meets the requirements of the law? (In other words, its not enough to just say "the transmitter and antenna passed, so no further testing is required"?)

That leads me to a final question: for a small-time operation that would rather not (translation: can't) go out and buy a $10K spectrum analyzer, can I do the testing myself with simpler tools and be relatively sure I'm within the bounds of the law? Say if I were to use a field strength meter or something to know I was close, at least for the radiated power requirement, then perhaps I could attenuate my circuit's performance slightly and guarantee I'm within the bounds of the law? There have to be some tricks of the trade that will let me avoid the normally expensive lab tools.

Thanks so much for your professional help! I really appreciate getting your advice, since you do this kind of thing regularly!

 

[cachehiker]

As long as you have documented evidence from an accredited lab, you can give your customers a Declaration of Conformity and attach a decal indicating compliance for unintentional radiators. Evidence from a accredited lab is verification of compliance. If you're confident that changes to the original design (like software or minor bits of hardware) will not increase emissions, you can often use verification of a prior version as justification for issuing a Declaration of Conformity for an updated version.

Actual certification of an intentional radiator involves a formal application and evidence of compliance be submitted to the FCC. If you introduce changes to the transmitter and antenna, you're legally required to do this. The fines for non-compliance are stiff for both unintentional and intentional radiators but the chances of getting caught are low if you've done a good job and are shipping low enough quantities. Just don't ignore the process and assume it will all work out. If you accidentally start shipping something that takes down somebody's wi-fi, people will complain and the FCC will start investigating.

You can verify the conducted emissions cheaper than radiated emissions. We have about $15K worth of gear including a line impedance stabilization network, spectrum analyzer, cabling and rails to do the former but our test environment isn't really up to using it for verification purposes. Our problems are generally related to radiated emissions anyway so we leave the detailed work to the lab. IIRC, a day in the lab costs us between $1000 and $2500 depending on the amount and type of testing we're trying to accomplish.

We primarily use the equipment for troubleshooting and development only, that is, when the lazy ... in charge of it gets off his fat ...

When we run into problems with radiated emissions, we use various homebuilt antennas, primarily a homebrew biconical, with the spectrum analyzer to make comparative measurements and then resubmit an updated design to the lab when we think we're there. This is a tedious affair given all the ambient noise we have to filter out but we're usually concentrating on a single frequency and possibly a harmonic and know how much we need to lower it.