FCC Certification / EMI Compliance Question

[yanir]

Is is necessary to get testing done on a non RF electronics design? I am designing a board with a freescale i.mx233 and it can reach 433Mhz. I will follow EMI reduction best practices but I don't have the equipment to test for such things. (I do plan on getting a spectrum analyzer for some basic testing and noise reduction). Is anyone familiar with the rules in the US? Does this product need to be certified and if so to what standard?

Thanks

[Neilm]

I don't know too much about the US requirements, I believe that the requirements are only that they don't cause interference.

If you are planning to sell the equipment into the EU, then the appropriate technical file would be needed to show compliance to the relevant standards.

Neil

[yanir]

Doing some digging on the ol' internet I found that I should be exempt! Here are the reasons:

(FCC) Part 15 covers devices, including digital devices, which use radio-frequency energy and may be intentional or unintentional radiators. Certain devices are exempted, including:

Digital devices used exclusively as industrial, commercial or medical test equipment
Digital devices used exclusively in an appliance, e.g. dishwasher, air conditioner, etc.
Digital devices having a power consumption not exceeding 6 nW
Operation of an exempt device may be stopped by the FCC if the device is found to cause harmful interference. Operation may be resumed only after the condition causing the interference has been corrected.

I am trying to dig up the actual FCC documents to confirm (this was quoted from: http://www.i-spec.com/EMC/usa.html

But so far so good.

[qno]

Start here

http://www.tmworld.com/electronics-blogs/the-emc-blog/4378141/Pre-Compliance-Testing-for-Radiated-Emissions


the this

http://www.tmworld.com/electronics-blogs/emc-emi-rfi-esd/4378152/An-EMC-Troubleshooting-Kit--Part-1a-Emissions-

etc

[AndyC_772]

As you've identified, there are different standards which apply to different types of equipment. Establishing which - if any - apply to your product is an important first stage in the process. A compliance testing lab can usually help with this.

Chances are your product will need to conform to FCC emissions regulations. On the plus side, that's a lot easier and cheaper than proving compliance with the relevant EU regulations, which are similar on the emissions side but also include tests for immunity to rf and to ESD.

[bfritz]

Actually you don't give enough information to determine if the device must be certified or verified.

I think you'll find the following overview by the FCC quite helpful.

http://transition.fcc.gov/Bureaus/Engineering_Technology/Documents/bulletins/oet62/oet62rev.pdf

Note that if the device connects to the power line through a power supply, that conformance to conducted emissions is likely required as well.

In the US, if you will be selling your device, the easiest and surest route to compliance is to contact a testing lab, and they will give you guidance on which specifications are required by law.

Intertek and UL both provide testing.  I believe both also allow for you to fill out a questionaire on your device, and they will then give you guidance on which specifications you must meet.  Of course they will be happy to give you a quote on the amount to perform this testing as well.

I've found that for a small company, Intertek seems to be easier to deal with, and generally quicker.

http://www.intertek.com/

http://www.ul.com/

[Neilm]

Chances are your product will need to conform to FCC emissions regulations. On the plus side, that's a lot easier and cheaper than proving compliance with the relevant EU regulations, which are similar on the emissions side but also include tests for immunity to rf and to ESD.

True, but usually I have more problems getting things to pass emissions tests than immunity tests. Generally because we have bought in a mains PSU module from a 3rd party. This will be the cheapest they do (cause we have to reduce the BOM cost to save money). This will pass its relevant emissions tests but when assembled into our product the result is a fail. I then have to reduce the noise in my circuit knowing that it is not the main culprit.

That was exactly what happened on my last project. The only comfort was knowing that none of the readings moved by more than 4 digits when exposed to 10V/m RF fields.

[yanir]

Thanks everyone for the usefull information. I wll review the links. We have relationships with UL and Intertek for our other products (appliances). I already have a call into our rep and am waiting to hear back. I'll let you know how it goes.

[tld]

Hey,

Although it won't get you certified, I just wanted to toss in the idea of using a computer-controllable radio for debugging or prototyping.  Personally I'm using an icom pcr 1000 (deprecated, other models available).  It's no spectrum analyzer, but if you're worried about specific frequencies, you'll at least be able to have a look.

I suppose you could say that a radio is to a spectrum analyzer, what a multimeter is to an oscilloscope... Sort of.

(simplified to the point of being wrong, but you get what I mean... )

Terje

[yanir]

Thanks for the tip, I'll check it out.

[AndyC_772]

Generally because we have bought in a mains PSU module from a 3rd party. This will be the cheapest they do (cause we have to reduce the BOM cost to save money). This will pass its relevant emissions tests but when assembled into our product the result is a fail. I then have to reduce the noise in my circuit knowing that it is not the main culprit.

It's something I'm always trying to drum into the heads of PSU suppliers. It's not enough to just scrape through an emissions test on a resistive load at maximum power - in order to be usable in a commericial product, a PSU needs to pass with margin at all loads.

I count myself lucky that generally I have the luxury of choosing quality supplies from reputable manufacturers - but even then, they're a mixed bag at best.

The biggest EMC surprise of last year was discovering just how quiet my laptop's PSU is. I expected it to be a nightmare, but it's easily compliant. Kudos to FSP for getting this right.

[yanir]

I too am using an of the shelf psu in this design. It's something that I took for granted until reading your post, thanks. I  will be more selective now.

[Neilm]

The cause of the problem looked like it was a filter on the OUTPUT of the mains PSU. As far as I could see, there was switching noise from the 5V SMPS on my board. Noise was being conducted up the the main supply rail and caused the output filter on the mains PSU to ring. It was of course very dependent on the module, some resistances, how well the wires were crimped and for all I know the phase of the moon.

A small PI filter on the output of the module fixed it and went home that day with that nice feeling you get when you have solved a REALLY weird problem.

[AndyC_772]

I too am using an of the shelf psu in this design. It's something that I took for granted until reading your post, thanks. I  will be more selective now.

The trouble is, without lab testing, it's nigh on impossible to know just how much noise will come off your chosen PSU + product combination. As Neil has illustrated, it's the system as a whole which can give rise to spurious emissions, not just one component or the other in isolation.

The best advice I can offer is to get at least two, and preferably more, power supplies which will functionally work with your product, and take them all with you to the EMC lab. Do a quick scan with each one of them, and if your PSU emissions (generally a broad 'hump' in the plot between about 30-100 MHz) are anywhere near the limit, pick a different supply.

It's also well worth feeding back any EMC problems you have with a PSU to the supplier. Let them know why you can't use their product - it's the only way they'll know that they have to improve. I've generally found manufacturers more receptive to this feedback than I might have expected, and on more than one occasion the design of a supply has been modified as a result.