FCC To Investigate Raised RF Noise Floor

If you stand outside on a clear night, can you see the Milky Way? If you live too close to a conurbation the chances are all you’ll see are a few of the brighter stars, the full picture is only seen by those who live in isolated places. The problem is light pollution, scattered light from street lighting and other sources hiding the stars.

The view of the Milky Way is a good analogy for the state of the radio spectrum. If you turn on a radio receiver and tune to a spot between stations, you’ll find a huge amount more noise in areas of human habitation than you will if you do the same thing in the middle of the countryside. The RF noise emitted by a significant amount of cheaper modern electronics is blanketing the airwaves and is in danger of rendering some frequencies unusable.

Can these logos really be trusted? By Moppet65535 (Own work) [CC BY-SA 3.0], via Wikimedia Commons
Can these logos really be trusted? By Moppet65535 (Own work) [CC BY-SA 3.0], via Wikimedia Commons
If you have ever designed a piece of electronics to comply with regulations for sale you might now point out that the requirements for RF interference imposed by codes from the FCC, CE mark etc. are very stringent, and therefore this should not be a significant problem. The unfortunate truth is though that a huge amount of equipment is finding its way into the hands of consumers which may bear an FCC logo or a CE mark but which has plainly had its bill-of-materials cost cut to the point at which its compliance with those rules is only notional. Next to the computer on which this is being written for example is a digital TV box from a well-known online retailer which has all the appropriate marks, but blankets tens of megahertz of spectrum with RF when it is in operation. It’s not faulty but badly designed, and if you pause to imagine hundreds or thousands of such devices across your city you may begin to see the scale of the problem.

This situation has prompted the FCC Technological Advisory Council to investigate any changes to the radio noise floor to determine the scale of the problem. To this end they have posted a public notice (PDF) in which they have invited interested parties to respond with any evidence they may have.

We hope that quantifying the scale of the RF noise problem will result in some action to reduce its ill-effects. It is also to be hoped though that the response will not be an ever-tighter set of regulations but greater enforcement of those that already exist. It has become too easy to make, import, or sell equipment made with scant regard to RF emissions, and simply making the requirements tougher for those designers who make the effort to comply will not change anything.

This is the first time we’ve raised the problem of the ever-rising radio noise floor here at Hackaday. We have covered a possible solution though, if stray RF is really getting to you perhaps you’d like to move to the National Radio Quiet Zone.

[via Southgate amateur radio news]

90 thoughts on “FCC To Investigate Raised RF Noise Floor

  1. do they even do any kind of random test on all the garbage being sold with FCC/CE markings? Can you report products violating regulations somewhere? and has there ever been a case of FCC fining retailers/importers for selling garbage?

      1. Ha! Answers to all your questions? Surely you jest! It’s the Chinese stupid!

        The FCC site is geared to self help. If you are an engineer and can define the problem then you already know what the problem is and you don’t need the FCC. On the other hand if you are not an engineer you probably don’t have a clue what the FCC site is all about. Catch 22?
        The truth is this is an epidemic that is far broader than most would imagine. To bring a product into the American or European markets you must run a number of very expensive tests and when you have jumped through all the hoops the manufacturer can self certify compliance with FCC regs. Same for safety regs and UL or CE
        If you fake the test and there is problem the FCC won’t be bothered but the aggrieved party can sue you ass off.
        It is different with the Chinese. They just self certify without running any of the tests and therefore sell the product even cheaper. If the Chinese get caught there is nothing that can be done about it because you will never get the Chinese company into court. Same for safety issues which can kill you but you still can’t get the Chinese into court even if they kill someone. It is a very lopsided issue. The Chinese counterfeit FCC, CE mark, UL logos and the FCC or FTC won’t do a damned thing about.

    1. I heard a very interesting story from John Stossel about what he called lawyer disease when it came to his discussions with HRC and how their ideology leads them to believe that regulations are best. My own observations have lead me to believe that this is a government wide issue… regulations to fix problems and more regulations when it didn’t fix it. There is no government accountability and nobody enforces (except for the EPA who breaks the above mold) because it pisses off donors or because the regulation was based in reality.

    2. Guys…


      Do some reading besides about Arduinos.


      Spread spectrum devices and IoT is going to make this all a lot worse. What’s going to help are the smaller channels now being crammed into existing bandwidth, ie CAT-1 or CAT-M cell radios, etc.

      There was a reason why the CTIA absolutely did not want general purpose “data” being sent in RF bands below 900 MHz in the US. The compromise there was that the data had to be related to control signals.

    3. Your reading skills are lacking. From the article “It is also to be hoped though that the response will not be an ever-tighter set of regulations but greater enforcement of those that already exist.”

  2. Remembering unit that were sent to labs for CE/FCC (and other) certifications, I can say that most marking are jokes. They are prepared, run special firmwares. Much like VW ECU unit in fact…

    1. Not sure what you are talking about.

      They have to run special firmwares to test the radios so that the quasi-peak emissions can be tested on each RF channel.

      Regarding other compliance, if your end product is depended upon for safety then you would have had to comply with one of the myriad of firmware safety documentation, software development practices and testing, everything must be documented including the software development tools, any updates made, testing methods, practices, FMEA, etc.

      I get it, that most people here do not have to comply with such requirements, but that doesn’t mean that they don’t exist and or are worthless.

      1. I read that x86 clocks are typically set to randomly jitter in firmware to spread the noise across channels avoiding peaks during testing and that as delivered the firmware is not jittered because it can cause instability.

    2. The fines for knowingly breaking these rules are quite high. However, the problem is the amount of checks being done is so low. And proving that it was knowingly is very hard.

      You can, if you want, put a CE logo on your product without any testing if you want. It is your risk. There is no regulation preventing you from doing this. You just say “this product is according to the rules”. If you do no tests, but it also complies with the rules, you have done nothing wrong.

      Unlike for example “UL”, which you actually have to do work for, and actually proves a whole lot more before you get this label.

      1. Even UL ratings can be fake. Ebay and Amazon Marketplace are littered with with fake UL Listed products. Especially in the laptop power supply category. Many of those products have fake UL logos and are downright dangerous yet no one ever does anything about it. Most consumers are still not even made aware that they are likely buying counterfeits.

    3. Another big problem is that financial penalties may only work locally. A company in an overseas low-cost production area may make a product that does not comply. When fines are levied, do they pay? Or do they just declare bankruptcy, sell off their production line to “another” company, and have it start right back up? Differences in accountability (for patents, compliance fines, and legal proceedings) is a huge problem in global commerce. The fact that the USA’s legal system is not the same as in other countries makes our legislation difficult or impossible to enforce. So even a well-written, good piece of legislation can be meaningless. Of course, most of our laws are not good (since they are often bundled with other special-interest laws) or well-written (since they are not authored by technical experts, or are authored by experts with ulterior motives).

  3. Good luck with this FCC when it comes to doing something about it where the problem is caused, in China.

    Don’t get confused between the CE mark and the China Export mark:

    The CE mark has a specified minimum size and dimensions when displaying it.
    The China Export mark is different.
    So to answer the article question, yes the CE mark can be trusted, it’s not on the product :)

    Make of it what you will for the purposes of the China Export mark.

    1. Hogwash. There’s nothing trustworthy about the CE mark. With the wrong dimensions you may believe it’s shady. With the right dimensions you’re never sure.

      I have a lovely very expensive Peltor headset with the China Export mark on it. Printed on the arm to save space, the letters wouldn’t fit otherwise. I have a horrible electrocute-o-adapter with a defeated ground and nothing but little spring contact to carry the supposed 20A. It has a perfectly rendered CE mark on it.

      The mark is meaningless without enforcement.

    2. The “China Export mark” is fictitious. It’s a poorly faked CE mark, and anyone with half an ounce of common sense realizes that.

      Plenty of cheap Chinese stuff has faked testing agency marks. The FCC mark, for example, should include an file code, which you can look up on their website:

      …as should the UL mark:

      The cheap Chinese stuff won’t have these, or the file code will come back to a different product model or a different manufacturer.

      1. Interesting you should say that.
        Since I’ve had the very same China Export feedback from exporters in China when questioned about their CE markings.
        Obviously it’s there to confuse, but I wouldn’t go as far as you have in saying it’s fictitious when it’s coming from the mouths of those using it for the purpose I linked to.

  4. There is a huge problem with rf noise in Morgan city la from a tv service and electrical of Morgan city I tracked down a bunch of it using sdr# I be leave a cheap rtl-sdr and a gps. I had it all mapped out across the hole city it was A lot of driving and the fcc did come but last I heard was they won’t fine them because it would do no good if they do not have the man power are resources to fix it now why fine them they won’t be able to fix it. I wish I could get out of speeding tickets that way I’m sorry officer I need to buy some food for my 5 kids so don’t give me a speeding are seatbelt ticket I can’t pay it wtf

      1. Without regular periodic measurements of each point, the data you have collected is effectively worthless. Nearly all RF noise is random in nature, and a single sample point tells you nothing. That sample could have been taken when one or more of untold hundreds of potential RF sources were emitting noise onto the sampling frequency (potentially including sources such as your vehicle, your power system, your computer, your equipment, your other RF emitting devices, and multiple harmonic and IF products that could be created or influenced by such sources).

        The drawback to any noise floor management is properly accounting for all variables that could be influencing your frequencies of interest over time, and properly sampling those frequencies of interest in such a way that the data you retrieve is both meaningful and accurate (either one of those, without the other, is worthless on its own under most circumstances). Tracking and measuring RF noise is a tricky business, and while some of the data from a tool such as the rig you used might end up having a positive correlation with future data from a reliable source, a great deal of it will be meaningless by virtue of everything you’re not accounting for (or that you are unable to account for).

        1. the fcc agent confirmed what we tracked down with there own investigation. the ham radio club that i was in is who reported it. so for a group of licencend and trained ppl to be rong i dout it. the noise is still there today if u get out of morgan city it goes away. covering the entire city 24/7 only when a storm comes and they louse electricity the noise is there its not random at all we used every thing from a yagi looking for noise also used a rf doplar tracker on top of my truck it was borrowed but still the same they did nothing and that was over 2 years ago

          1. i spent alot of time i could of spent with friends and family trying to help a club and a city just to find out the fcc will do nothing to help when they are the one who shood enforece the law who do we turn to now im still pissed sorry if its long winded bad grammer and spelling but i feel better i hate rf noise

      1. automatic i shood still have all the recordings you can even open it up set a freq and bandwith and it will show the hot spots wile doing the recrdings id we saw on the spectrum jump up we slowed down to get more data

      2. i guess to better answer the question i drove every street and cross street it seemed 100 times with my computer recording the spectrum then the computer compiled the data then i put what freq and how wide the signal im tracking and it autimaticly plot it to google earth so i can see the hot spots

  5. The problem would be historic data. I would think those that would be concerned with the noise floor for things like calibrations or cell sites, etc. would be taking snapshots of the local noise floor before taking noise measurements. But as for the average Joe or Jane, I doubt them having any historical data to reference. I guess a good time to break out the wide band spectrum analyzer would be during a mass power outage. But then you’d still have what are probably the biggest contributors: cell phones. I am sure there well be a lot of people trying to help after reading this, which will only harm the situation, because Jimmy the hacker is going to plug his Chinese SDR dongle in and use it to try to capture the noise floor, rather than an laboratory grade Rohde & Schwarz… Multiply that times thousands, and well….

    1. According to the article most of the issues come from cheap crap that is skirting if not out right ignoring regulations. I imagine a good portion of the products are ebay specials type deals. Now if this is true then cell phone should stay in there own bands. They tend to be far better built and don’t fall into the cheap crap category. More importantly apple and Samsung have a lot more to loose than a no name ebay seller if they broke regulations.

      Although like what you saying, it’s impossible at this point or to verify this since the cats out of the bag so to speak. Now I need to get a good spectrum analyser and put it on a battery backup. I wonder how much nose those crappy apc battery back ups makes……

      It would be a really cool case study if some one has the equipment. After all even what I wrote is all gross speculation and ultimately meaning less with out data.

  6. Square waves have unlimited harmonics, and in our digital age they are produced at so many frequencies. Sometimes I wonder how the old radios are able to receive any useful signal at all.

    1. In our digital age squarewaves exist only in your imagination. They are all trapezoidal because they all have rise times, no matter how square they appear to be.

      The difference between then and now is that the rise times are faster now.

          1. I assume that’s from a “mathematical” square wave. In the real world there are no real square waves, because a voltage can’t change from high to low instantly. So the sides of the “square” are ramps. This limits their harmonics. Still more harmonics than you might like, but not quite as bad as in theory. A theoretical square wave is an infinite series of sines.

          2. @Miroslav your statement about a mathematical square wave in the frequency domain is accurate. Your assumption that real world digital square waves look like mathematical square waves is not.

    2. You have to consider the difference between a simulation and reality (rise/fall timings and non-idealities like dumping RC parasitics). Moreover, nowadays certified (or well engineered) devices are shielded.

  7. Most hams have been aware of this problem for years. The noise floor just outside of my home often reads full scale on my radios, making hearing all but the strongest signals impossible. It’s a bit like shining car headlights into someone’s face from 5 feet away while expecting them to be able to read an unlit sign 50 feet away.

      1. But it shouldn’t be. Legacy modes can be much more reliable under adverse circumstances. Proper enforcement of existing rules would allow those modes to remain robust and useful. Those who deprecate the past simply because something novel is commanding their attention do so at their own peril.

        1. “Legacy modes can be much more reliable under adverse circumstances.”

          Obviously “adverse circumstances” does not include high RF noise floor. I’d say you’re probably referring to natural and man-made disasters, causing the collapse of infrastructure required for modern modes of communication.

          The irony being that during such a disaster, the RF noise floor would typically drop. So the legacy modes, even if not normally useful, would instantly *become* useful when actually needed.

          And even in disasters, that need appears to be shrinking. From Wikipedia:

          “The largest disaster response by U.S. amateur radio operators was during Hurricane Katrina which first made landfall as a Category 1 hurricane went through Miami, Florida on August 25, 2005, eventually strengthening to Category 5. More than a thousand ham operators from all over the U.S. converged on the Gulf Coast in an effort to provide emergency communications assistance.”

          No doubt HAM operators really helped out in this case, and that’s something to be proud of.

          However, even though land lines and cell phones were down, HAM was *not* the only means of communication. Satellite phones also worked. There just weren’t enough of them on hand. A lesson learnt from and soon rectified. Iridium shipped 5,000 sat phones, mostly to the Gulf states, before hurricanes Gustav and Ike.

          While the HAMs still helped out, now their numbers – and utility – were dwarfed by sat phone users. Give a government or relief official a sat phone, and they can use it almost instantly – without license or significant training as would be required for HAM. They can also communicate sensitive information if needed, since communications are encrypted; whereas HAM operators cannot use encryption.

          1. Is that a fact? Funny, I’ve never seen anyone use a satphone in my entire life, but I see people using ham radio every day. Those same people are practiced, ready, and in place already in case of need of any kind.

            Ham radio has been, is, and will continue to be the very best way of communicating without infrastructure. You seem to forget that satphones still need a functioning phone network on the other end in order to be useful. Nothing against satphones, but they just aren’t a replacement for ham radio. They are a whole different thing altogether, with different uses, functionality, and cost. Apples to oranges.

  8. well! if they are not enforcing the rules then we should be able to modify the routers or even build our own 5 ghz transmitter and as long as we are not drowning or dos’ing any emergency or medical devices the fcc wont bust you

  9. Maybe there is a consumer solution to this. Spread the word about the effects of interference on the quality of wifi, cell signal, speakers, etc and also spread the word on how to easily test for such interference. That way consumers themselves can rate and weed out the poorly designed electronics.

    1. Before they buy them?
      Hardly likely. I think it’s more likely the chinese export mark is getting confused with other markings or the manufacturers cost cutting is causing the problem.

    2. How do you ‘easily’ test for it. RTL-SDRs are pretty crappy. I would never feel confident that I was measuring a real noise signal vs some artifact of the dongle itself. Actual decent spectrum analyzers are far outside the price range of most radio hobbyists who have an actual interest in the problem let alone the general public that mostly doesn’t care.

  10. “We hope that quantifying the scale of the RF noise problem will result in some action to reduce its ill-effects. ”
    Really? HaD wants all the items people on HaD use to be banned by the FCC? What are the plans for afterwards then? I’m not sure gawker is hiring..

  11. I blame HaD and all the dang kids with their hot-rod Tesla coils! At least you have not shown the tremendous explosion from mixing non-fat milk with sweet cream.

    Why, back in my day you had to make your own bandwidth with a hand cranked spectrum spreader! And stand back if that thing comes loose after it is wound up. Now they don’t talk about them in schools and I can’t even find one on the internet!

  12. Seems like this would be a good use for RTL Power.

    You could get a few hundred people across the US to use an inexpensive RTL-SDR tuner to dump a bunch of data to CSV files and then post process.

    They don’t need to be all that accurate as long as you devise a way to factor out the inaccuracy. Perhaps place the dongle inside some RF-absorptive foam for an hour and record the data. Then, remove the foam and record another hour. Should be able to measure the difference and use that to calibrate the rest of the data.

    Wouldn’t cost a whole lot to get people to do this for a year so you could see day/night and seasonal changes. A sample every 5 seconds doesn’t exactly make a huge amount of data. Send out a cheap flash drive to each person volunteering to do it and some stickers that say they participated. You could probably get a bunch of hams to do it for free. I would.

    1. I don’t think you’ll ever be able to characterize a cheap SDR dongle good enough to develop any useful information out of it. It’s connected to a computer of some sort via USB and computers themselves are crazy noisy on their own. This is something you need experts and $100,000 spec/an’s (like the one sitting on my bench) to work out. Not citizen scientists.

      1. First, whoops I accidentally reported your comment. I didn’t mean to, but the report button is right where every other website has their Reply button.

        Second, there are tons of great things being done with RTL-SDR’s. You might not be able to achieve the same accuracy level as a $100K spec/an, but if you could no one would sell $100K spec/ans. But, a large sample size reduces the margin of error. I still think you can get useful data from such a survey.

        You should be able to account for internal spurious signals in the SDR by characterizing it in an anechoic chamber of RF absorptive foam, then again outside the foam. Signals which didn’t drop by the attenuation of the foam are likely internally generated.

        Then, you could take the entire collection system to two different geographic locations. Signals which are constant at both locations are likely generated by the collection system (SDR, PC, Power Supply, etc).

        Additionally, by sampling with and without the foam, in the foam’s calibrated range you can characterize the measured signal strengths to attempt to introduce absolute (as in, not relative) accuracy. If the foam reduces signals by 30dB at a given frequency, then whatever the change in levels recorded by the device, it can be calibrated to 30dB.

    2. FCC: “Odd, there’s suddenly a spike in the EM noise you normally see from all those unshielded TV dongles.”
      FCC: “Oh well, back to the booze and hookers”

    1. The noise floor is a measurement of any unwanted rf radiation at a given frequency and bandwidth. There are parts of the total noise that can be calculated such as Johnson-Nyquist noise. Other types like atmospheric noise (QRN) and rf leaking switching power supplies (QRM) need to be measured.

  13. Gassed out sodium lights (on and off) at night make a lot of noise. They don’t replace them on a predictive failure mode anymore. They want you to report each one, one at a time by phone or website. One light poops all over the AM and HF end of things for blocks. Will switchers in LED lights be clean? LED traffic lights make quite a bit of noise on AM radio. There are apartments level with the lights, right at the corner. Better hope they have internet.

    Some of the late night talk on AM radio gets into the subject of Dirty Power, where the noise on mains and in the air is highly questioned as to it’s effect on human and animal health.

  14. Being able to detect the problem in any meaningful way is an issue.

    We need cheap and available spectrum analysers. If for some reason mobile devices started to have the capability in them as standard (like gps and accelerometers) it might get somewhere.

    Would be great to be able to pull up RF ‘pollution’ maps.

    There were mainstream articles around at Christmas about decorative tree lights possibly affecting Wi-Fi. So it isn’t entirely a techy corner issue.

    1. It’s not just that you need a cheap yet very clean (noise free) wide band spectrum analyzer to begin with, on top of that, you should probably have a high gain wide bandwith antenna and RF amp too to get any data into your device. And then, you have to characterize that system over the full RF band to actually get any useful data points out in the field.
      Honestly, i don’t see anyone pulling that off. Guess there is a good reason why RF lab equippment from R&S (or whoever else is any good in that field) quickly costs between 10K and over 100K$ and usually comes in large “boxes” you don’t really want to move alone or even carry around.

  15. Well this is nice – especially side by side with this http://hackaday.com/2016/04/03/dvb-s-from-a-raspberry-pi-with-no-extra-hardware/ (filters? who need the right?)

    About HF – maybe some HAMs already found how this actually works https://en.wikipedia.org/wiki/HomePlug – biggest HF killer so far

    Besides all that i still think that RF bands should be at lest partially “free” – that “ordinary” people should have frequencies available for their needs – in some parts of Europe there is nearly no possible way to legally fly FPV due to RF band restrictions (433MHz for control is extremely limited, 2.4GHz for video is NO and also noisy as hell, 5.8GHz for analog is extremely limited …). I also think it still somehow better to be able to use HomePlug adapters (which is potentially beneficial to wide group of people), than to listen to HAMs, who are constantly bragging about how these destroyed their hobby (and the same goes to other frequencies – some HAMs are pretty hardcore if you ask them who can and can’t use this or that frequency without their holy document).

    1. I feel I should reply on Evariste’s DVB TX. By “No extra hardware” in that case we mean “No external I/Q modulator chipset”. Not “No filters if you use it in a to-air situation”. Previous SBC digital TV TXs used expensive modulator chipsets, this one doesn’t.

      One of my other activities when not writing for Hackaday is the design and sale of filters for amateur radio. In Evariste’s case I have complete confidence that he is fully aware of the need for filters, and fastidious about their use. He’s provided the Raspberry Pi community with some of their most useful RF sotware.

  16. Having implemented a system with HomePlug AV PLC I’ve come to realize the importance of supply filtering to a point where I’m even considering active filtering.
    Then I realized that most of the people complaining about issues with powerline networks got it back to front: In most cases not the network interface is at fault but the noise on the medium. I get it that overall noise is becoming a more serious issue but it’s not just radiated emissions but also conducted EMI.

  17. If they really cared, they would work to make compliance testing easier and cheaper. Not grant near monopolies to test/certification labs. More carrot, less stick.

    1. Maybe you should change your viewpoint a little when you say that test/certification labs are expensive. There are a few very good reasons why they are expensive to begin with… (highly trained workers, special high precision equippment, accreditations and certifications of the labs so they actually measure to a certain standard of precision and quality,…)
      Of course, if you happen to find a lot of money somewhere on the budget to somehow subsidize EMI/EMC labs so they don’t have to charge the client what they have to charge to stay in the green, that could help. But then again, these people who don’t care about the regulations in the first place to save a few $ would not care to go testing theyr devices even if it was completely free to do so. To meet compliance, you usually need more components and iterations to get to market.
      The only thing that would help is when consumers would actually stop to buy all that cheap electronics crap we have today. Something i really don’t see as long as the current “way of life” and capitalism rules the majority of the earths population.

  18. Military maybe told FCC that the radio hash is getting so bad that their fallback methods of communications are getting iffy. This has been a long time coming. The amount of noise in urban areas is almost silly.

  19. Need to weigh the value of these old modulations on low frequencies against the benefit of the tech that is interfering. I have not turned on a broadcast radio receiver in years other than in the car briefly while cycling inputs between bluetooth and aux in, so see pretty much zero value in the spectrum under 1Ghz other than the car remote control.

  20. The number one culprit is unshielded CAT-5, followed by anything Chinese; Wall Warts, LED Lighting, Etc. In Cities the Traffic system (including cameras) and Computerized Power Meters are big sources. In buildings is modern motor controllers and IT equipment. Solar panels are also becoming a source of noise. Here in Broward County, FL the HF noise floor averages -115 dbm. I’ve heard the Quiet Zone is below -160 dbm. All this RF noise can be good for people.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.