Does 5 GHz Disrupt Sleep?

I don’t have a lot of research to back this up. I did find a few links that claimed 5 GHz (or any GHz) WiFi would do various bad things from lowering your IQ to mutating your mitochondria; but nothing I’d call definitive, and a lot of it was whacko prattle.

Yet what has happened did happen.

So you will get a “story”.

I bought a new WiFi router for our “inside” network. It does both 2.5 GHz and 5 GHz. Things came up nicely. Never thought to connect that buy to anything health involved that was coincidental.

Then the spouse started having some health issues and I started having disrupted sleep. I just figured it was the Old Man Thing. I’d wake up after about 4 hours, visit The Little Room, and often be unable to return to sleep. About the same time I started getting what seemed like peripheral neuropathy. I’d wake up with my feet tingling or a bit of burning sensation.

So I checked my blood sugar. Normal. Cut alcohol intake to zero (for months now). No change. On down the list of possibles… made sure B vitamins were taken, etc. Over time it seemed like my heart rate was a bit irregular, but it was only a feeling. The Spouse had “hear palpitations” and some various medicines were tried. They would come and go.

My interior router was not always used, only turned on when needed for something most of the time; but sometimes left on for days depending on what I was doing, or forgot to do ;-)

Then one day I’d noticed that the spouse seemed to not have irregularities when the router was off. I rearranged things so that it wasn’t really needed and left it off for a week or two. She was better. OK, off it goes.

Then about 2 weeks after that, the AT&T router died. The replacement does both 2.5 GHz and 5 GHz. It runs 24 x 7. After a few days, spouse is having palpitations again, I’m sleeping “4 hours and pee” with feet tingles sometimes waking me up, and we’re both just feeling stressed and tired. I went from a “couple of cups a day” to “coffee or tea about every 2 hours” and still felt tired and slow all day. I posted about cheaper fuels for making coffee. ;-) Spousal insomnia returned; remember she is of the “bed at 9 or falls asleep at the TV” group – so her up reading at 11 pm because she can’t sleep is bizarre.

So I decided to run an experiment. Back when I’d shut off the interior WiFi router, I had gone through the house and made sure we had no LED bulbs running. (The excess blue from them is known to cause insomnia for her). Only one was in the garage, so OK. At the same time, any CFL bulbs were swapped out (just to assure no emissions from the switching power supply or short arc as spark gap transmitter). That and the router off had helped the spouse sleep better. Now I decided to shut off the AT&T Router nightly.

That was about 5 days ago.

Spouse was immediately sleeping better. No longer “waking up tired” and no waking up at the wrong time. I’d not really noticed, but my tendency to discomforts had lessened. (Part of why I didn’t notice was that I’d go to bed about 1 AM and get awakened at 5 AM to turn the WiFi back on so the spouse could check email & watch TV…) Well, after a few nights of that it was pretty clear she was doing better and I needed a better network option.

Looking at the router, it has instructions for ‘advanced settings’. Basically point your browser at it, and a password. Turns out it has a nice control panel where you can shut off either the 2.5 GHz of the 5 GHz radio. 2 nights ago I shut off the 5 GHz radio. Now I can’t say if it was just me imagining things, or something more, but at the moment the radio was turned off, I felt a kind of jittery tension just drain away. I’d not been expecting it, nor even really noticed the feeling being there until it left. It was just a feeling of relief.

Since then, my coffee consumption is drifted back toward more normal. I no longer feel the need to rush from cup to cup. The spouse reports sleeping comfortably through the night. My feet have not tingled in days; and last night I slept through the night, comfortably. Both of us have reported a deeper more restful more normal nights sleep.

The Environment

The telco entry and WiFi routers are in the room next to our bedroom. It is about 10 feet from the routers to our bed. The wall between is simple wood studs and drywall (i.e. not much interaction with radio waves). We have about 50 to 100 feet to any neighbor’s possible EMF sources so plenty of IR^2 loss from them. There are power lines over the back of the lot about 25 feet away.

We are about 10 miles from an airport (so radar sweeps overhead) and immersed in the WiFi / Cell Tower 5 G “electronic smog” of Silicon Valley – but in a part that is not yet full of 5 G towers “near”.


As I don’t have any equipment that needs / uses 5 GHz I’m leaving those radios turned off. For now, the Telco router returns to being turned on 24 x 7 but as 2.5 GHz only. I am NOT going to be running any further “blind tests” with turning on the 5 GHz and seeing what happens; as I’m hoping both the spouse (and me) can get some decent sleep for a few months…

The number of “accidental experiments” was high enough and the results consistent enough to lead me to conclude that 5 GHz is highly likely at least a sleep disruptor, and likely causes other neurological responses as well. At 5 GHz the wavelength ought to be about 60 mm, or about 2 1/3 inches. Just about right to couple into many conductive structures that would fit in a human head (or body).

I will be much more diligent about observing any changes of “EMF Environment” when any sleep disruptions or other effects show up.

It will take a few weeks of this regime to assure all the bad things are gone. Perhaps partly due to the intermittent nature of my interior router use, or perhaps for other reasons, the negative health observables have been sporadic. Irregular heartbeat coming, then going, at various intervals often separated by many days. Now we will see if it is “just gone”, by waiting a few weeks. Similarly for other supposed manifestations.

I will instead look for any literature with a bit more authority in it that might shed some light.

There is a very real possibility this is all just wishful thinking mixed with placebo effect / hypochondriac ideation. That can take a while to figure out. But I don’t think it is. We’ve got 2 people, lack of expectation that this was the answer (until late in the process), and onset the second time before reading the router and seeing it did 5 GHz. Also the spouse was blissfully unaware of the technical aspects throughout. At least until I tried to explain my thinking 2 days ago ;-)

As the use of 5 GHz in home WiFi is just becoming significant, most folks will still be running older 2.5 GHz gear. Similarly, 5G cell service (a different but related thing) is not yet common in most areas. To the extent there IS an issue with very high frequency microwaves, we will be at the leading edge of this as it rolls out globally. My suggestion would be to drag your feet on it in your location.

A Couple Of Random Readings

Not particularly recommended, just things I ran into that might or might not have bearing. One I ran into stated that the move to 5 GHz had been accompanied by an allowance for much higher transmit power, but I don’t have the specifics.

Joel M. Moskowitz, Ph.D.
Center for Family and Community Health
School of Public Health
University of California, Berkeley

Monday, August 7, 2017

5G Wireless Technology: Millimeter Wave Health Effects

The emergence of 5G, fifth-generation telecommunications networks, has been in the news lately because the wireless industry has been pushing controversial legislation at the state level to expedite the deployment of this technology. The legislation would block the rights of local governments and their citizens to control the installation of cellular antennas in the public “right-of-way.” Cell antennas may be installed on public utility poles every 10-20 houses in urban areas. According to the industry, as many as 50,000 new cell sites will be required in California alone.

Although many major cities and newspapers have opposed this legislation, the potential health risks from the proliferation of new cellular antenna sites have been ignored. These cell antennas will expose the population to new sources of radio frequency radiation including MMWs.

5G will employ low- (0.6 GHz – 3.7 GHz), mid- (3.7 – 24 GHz), and high-band frequencies (24 GHz and higher). In the U.S., the Federal Communications Commission (FCC) has allocated “low-band” spectrum at 0.6 GHz (e.g., 600 MHz), “mid-band” spectrum in the 3.5 GHz range, and 11 GHz of “high-band” frequencies including licensed spectrum from 27.5-28.35 GHz and 37-40 GHz, as well as unlicensed spectrum from 64-71 GHz which is open to all wireless equipment manufacturers.

Prior to widespread deployment, major cell phone carriers are experimenting with new technologies that employ “high-band” frequencies in communities across the country. The “high-band” frequencies largely consist of millimeter waves (MMWs), a type of electromagnetic radiation with wavelengths of one to ten millimeters and frequencies ranging from 30 to 300 GHz (or billions of cycles per second).

by Dr. Edward Group DC, NP, DACBN, DCBCN, DABFM
Last Updated on October 2, 2015

1. Contributes to the Development of Insomnia

Have you ever felt more awake after using Wi-Fi or even struggled to sleep through the night? Reports of these phenomena have been frequent and even prompted a study in 2007 that evaluated low-frequency modulation from cell phones and its impact on sleep. Participants were exposed to the electromagnetic signals from real phones or no signal from fake phones. Those exposed to the electromagnetic radiation had a significantly more difficult time falling asleep and changes in brainwave patterns were observed.

It’s been suggested that sleeping near a phone, in a home with Wi-Fi, or in an apartment building with many Wi-Fi signals can create chronic sleep problems as the constant bombardment of Wi-Fi pollution interferes with falling asleep and sleep patterns. For many, sleep deprivation is just the start for larger problems. The development of depression and hypertension have also been linked to inadequate sleep.
8. Provokes Cardiac Stress

If you think your heart races when surrounded by wireless networks or 3G or LTE cell phones, it may not be in your head. A study involving 69 subjects reported that many of them experienced a real physical response to electromagnetic frequencies. Exactly what was the physical response? Increased heart rate — similar to the heart rate of an individual under stress.

In any case, while this is all interesting, it wasn’t this WiFi information that caught my eye so much as the reason that the girls thought to conduct this experiment in the first place. The article stated that all of the girls noticed that when they slept near their mobile phones when the phones were turned on, their sleep was restless and difficult.

This got me to thinking. My mobile phone is always next to my head on my bedside table. I’ve had a mobile phone since at least 1996, which is about when my insomnia started. What if my insomnia is not caused by my inability to relax, but by the mobile phone signals?

I tried my own experiment. I began turning off my mobile phone at night. I also turned off our household WiFi. It isn’t in the same room with me, but it is just through the wall, not six feet from my brain, and WiFi waves travel through walls.

Immediately my sleep improved. I’ve had a few nights where I awakened, but I fall instantly back asleep without the little buzzing brain going on. No more song worms. No more spinning thoughts. No more waking so much my kidneys think it’s time to turn on the bladder. No more tossing and turning and finally falling asleep at dawn, only to have to rise a half an hour later exhausted and worn.

I’ve gone for almost three months sleeping through the night. Nothing else in my living situation has changed to explain the improvement in my sleep. Even when I wasn’t having insomnia episodes before, my sleep was always restless. Now I actually rest.

So basically it isn’t just me; and other folks have observed similar things. Either there’s a lot of us having the same delusion, or there’s some “there there”.

Searches on “WiFi Insomnia” find a lot (much of it cell phone too). A search on “WiFi mitochondria” finds articles asserting it damages mitochondrial DNA leading to long term damage to cellular energy production. I’ve not sorted any of those articles yet so no links yet. An interesting idea, but needs a lot more before I’m signed up to the notion my WiFi is damaging all future generations though mtDNA destruction… Then there’s just the fact much of this is about 2.5 GHz and we seem OK with that, but not 5 GHz, so what the heck does that mean?

Gives a hodge podge of quasi-scholarly stuff and whacko fringe Luddites. So lots of time consumed per unit information gained.

This will be a low priority “mostly observational” thing for me, going forward. To the extent we’ve identified and got a fix, I don’t really care much anymore. To the extent 5G and 5 GHz become ubiquitous in classrooms, hotels, roadside dining, neighborhoods, airports, and everywhere there is humanity, it will become much more important to me… And, to the extent it is self delusional imaginings, well, my tinfoil hat looks pretty! ;-)

Subscribe to feed


About E.M.Smith

A technical managerial sort interested in things from Stonehenge to computer science. My present "hot buttons' are the mythology of Climate Change and ancient metrology; but things change...
This entry was posted in Biology Biochem, Human Interest, Tech Bits and tagged , , , . Bookmark the permalink.

59 Responses to Does 5 GHz Disrupt Sleep?

  1. jim2 says:

    So, instead of a mosquito net around the bed, maybe an EMF net is required?

  2. Larry Ledwick says:

    Interesting observations.
    Little story from many years ago. I had just moved into a small apartment and was playing with a CB radio to see if I could monitor local CB traffic. I had the mag mount antenna sitting on a cookie sheet on the bed and was sitting just a few feet away from it on the corner of the bed. After tuning around a bit I decided to check the SWR (standing wave ratio) which required some short key down broadcasts on an unused channel to do the tests. Moments after I did those tests, I felt suddenly feverish and “antsy”. I am sure it was my body reacting to the very high RF levels just a couple feet from the antenna. At 27 mHz one body length is about 1/8 wave just long enough to be an acceptable antenna.

    I currently usually leave my cell phone in the next room next to the computer but I occasionally have similar experiences, I never noticed if those were on nights when I brought the cell phone into the bedroom (like when I was on call). I might also see if it makes any difference if I shut down my computers completely. I often have 3 of them on in the next room all night as I don’t shut them down all the way. With modern processors operating in the 2-3 gHz range it would not be unreasonable for them to broadcast some related harmonics, although I turn off their wi-fi systems since I don’t use it.

    It would be interesting to see if there is a low cost gHz signal strength meter available, looks like the ones that have a high enough bandwidth run just short of $200.

  3. E.M.Smith says:

    Interesting that there is a market for such an EMF net. I think I’d go for lower cost though… Like an aluminum mesh. Need about 1/2 inch weave is all. Wonder if anyone makes aluminum chicken wire? ;-)

    Unfortunately, our bedroom is on a direct line between the WiFi router and the Living Room TV, so too much “shielding” in the middle will be an issue.

    I might look at putting an extension on the telco drop (looks like an RJ-45 I think, but that or RJ-11, so standard wires & sockets, and I already have conduit through the walls…) and relocate the routher to the L.R. Then I put shielding on the wall to the bedroom… That would limit signal in my office, but I could easily put a back haul Ethernet and run wired.

    Might be easier to just get a non-WiFi telco presence and then put a low power WiFi where I want it… That’s what I had before. DSL to the office, Hub, eithernet to “places”, and a WiFi router I could turn on when desired.

    Well, “We’ll see”. This test has a while to run before I’m going to start worry over the 2.5 GHz.


    I know for fact PCs radiate LOTS of RF. Set an AM radio next to one and just try to get a strong local station through the static… Network gear is leaky too. Historically even video screens (one spook technique was to listen to your CRT paint the screen and just repaint it and see what you saw). I may test the Raspberry Pi vs PC at some point. With all the traces Damn Small and not much power anyway, it can’t be radiating very much !

    I think a trip to the hardware store to price various metal screens might be in order. A cylinder 8 feet tall and a bit under 3 feet diameter would be about 8 x 9 feet. Would make an interesting ‘step in’ test rig. Just add ground wire… IIRC you need a mesh size under 1/2 the wavelength to block. So a 2 inch wavelength would need a 1 inch or smaller mesh. I think that’s even smaller than chicken wire… Maybe window screen… but most of that has gone plastic now. Wonder if anyone sells brass mesh in big sizes… or if aluminized mylar is conductive enough to work… They used a conductive paint inside the Mac for EMF screening… Wonder what it costs / can…

    Oh, Wait! Carbon Fiber is all the rage for making stuff and carbon conducts! Fine weave too. Just ground it. If desired, dunk in a metal salts bath and plate on a few atoms of copper or chrome…

    But then we’re back at how to test it again…

  4. John F. Hultquist says:

    Post gives me the thought that “bleeding-edge technology” is for other people, not me.

  5. E.M.Smith says:

    $5 / sq ft copper shielding mesh:

    NASA shielding guide PDF:

    Big roll off in shielding effectiveness with frequency. Only 20 to 30 dB at 5-10 GHz.

    They also have a nearly transparent stainless steel mesh for windows…

    Golly… I wonder if “The Mesh Company” makes wire mesh? ;-)

    Looks like getting materials is just a matter of money.

  6. Larry Ledwick says:

    Yes I think you can buy light duty aluminum chicken wire at places like Home Depot.
    Kitchen aluminum foil is also useful for such testing if you want to set up a quick temporary RF cage structure.

    I live on the 3rd floor and end apartment of a 3 story apartment complex, so my only close adjacent source of wifi would be my next door neighbor to the south. Our floors are cast concrete decks so good shielding below as far as RF is concerned. If I turn on wifi on the lap top, I can see about a dozen wifi networks from other tenants in nearby buildings, but due to inverse square law they would be very low signal strength, and not be much of an issue unless one of them was running a signal booster.

    I am not particularly worried about it as in serious health concerns but it is an interesting idea to check regarding low grade impacts like quality of sleep.

  7. philjourdan says:

    I would say you did the experiment well. Unfortunately for me, I just no longer sleep well. I will sleep soundly for 3 hours, and then off and on until I get up. And that was before 5ghz.

    So could be. Instead of getting up to turn it on for the wife, you may want to invest $5 in a timer.

  8. ossqss says:

    Interesting stuff. I run an AC router and do use the 5 GHz side. Throughput on that side provides a 200+ Mbps (much higher if configured for a local media server etc) connection to the net in testing vs. about 100 on the low side 2.4 GHz. Range on the 5 is significantly shorter than the 2.4 side. For the record, many hotspots can use the 5 Ghz space. I use it on my phone and most folks don’t even see it.

    I will shut it down and see what happens as an experiment ;-)

  9. jim2 says:

    To test, put a cell phone inside the barrier and call it.

  10. Larry Ledwick says:

    Well the down side of that is that cell phones dial up their power output to get a signal, so it will try to ping the cell towers and crank up the output power to max to try to get a response.
    a) short battery life
    b) might make the test not very meaningful

    In my case I leave it on the desk next to the computer but can just barely hear the ringer in the bedroom, so signal strength at that distance will be only a small fraction of what is if it is on the night stand beside the bed 2 ft from my head.

  11. I’ve been collecting stuff to work at 5.8GHz for the EMDrive-equivalent (but hopefully better) idea. For a cheap sensor that tells you roughly how strong the signal is (supposed to run to 6.5GHz but I haven’t yet tested it to that) and around $10 there’s . Cheap and cheerful, but has an audio output to find hotspots. The antenna is probably a bit long for 5.8GHz though.

    If you want to actually measure it there’s which is an AD8318 on a board (around $16), for which I wrote a quick program on a PicAxe 14M2 in order to display the measurement on an LCD display. Given that each version of PIC has a manual of around 150 pages and it’s a simple loop program with a lot of waiting, the compiled BASIC version of PicAxe is good enough for the job. That means somebody else has set up all the registers to useful values and I don’t have to read all the manual. The range of measurement is from -56dB to around 4dB with increased inaccuracy at the limits but still near enough for jazz. If people want the program I’ll put it up. There’s also an audio output from the program that updates the pitch around 50 times/second. Spare output pin so may as well use it, and I’ll be using this to tune the antennae. Total cost of the boards maybe $30, but it does mean you can put a number on the measured power and hear a 20ms blip in the power it’s measuring. Since the program occupies around 10% of the available space, it should be easy to add more functions once they are needed.

    To get a waterfall display of the spectrum there’s the HackRF project (search on Great Scott HackRF one). I also got one of those and it does work nicely on Linux, with a somewhat steep learning-curve if you’re not already into ham radio (I wasn’t). This has a limit of 6GHz and some change. This costs around $160 or so, but as a software-defined radio it can transmit and receive across a very wide range and using pretty-well any format you want. Rather than buying the dedicated $200 device that Larry was looking at, maybe more useful to get the HackRF and be able to do very much more with it. Other software defined radios are available, but most don’t have the capabilities to reach 6GHz.

    Note that the experiments I’ll be doing will be very-well shielded and at relatively low-power anyway. 600mW transmitters are cheap and probably enough to get a measurable result.

    The new router here does do 5GHz or so, but I haven’t yet measured the frequency or power. Still getting all the kit together to do that…. I haven’t noticed any problems with sleep yet, but AFAIK it would in any case not run the 5GHz signal much unless there was something there to receive it. Also I don’t sleep close to the router and there are some thick walls (half a yard of rammed earth bricks) between me and it.

    There have been some experiments showing that people are a bit smarter when exposed to the 2.4GHz phone radiation. Sorry, I didn’t bookmark that. The wavelength of that is around 12.5cm in air, so it may depend on head-size as to whether you get a complete standing-wave inside your head. The 5.8GHz will have more resonance modes available and so seems likely to have more effect. Your head is basically a dielectric waveguide, where shape and size will be important. Tests on safety of the 5.8GHz frequency were probably done on lab rats, and of course their heads are a lot smaller and won’t sustain any standing waves inside.

    Like Phil, I think the correlations are good enough to not really need more experimentation, and the problems seem severe enough to want to avoid self-experimentation here anyway. AFAIK the guy who worked on the original radio phones for BT ended up with a brain-tumour.

    Seems like tinfoil hats may be coming into fashion, given the roll-out of 5G.

  12. interzonkomizar says:

    Hi Chief. Thanks for the heads up notice on 5 gigahertz problem. Last year I got the dual-frequency Wi-Fi router from the telephone company but only use the 2.5 gigahertz for my tablet. I haven’t noticed any health problems but would like to turn off the 5 gigahertz. Do you have a link or app recommendation that takes step by step turn off procedure. Thanks

    Sandy, minister of future.

  13. Sandy MCCLINTOCK says:

    The 5Ghz signal is too weak to work in the bedroom in my place so the 2GHz is the only option there.
    In the other parts of the house, where sleep should not happen, the 5GHz signal is strong and makes a big difference to browsing speed etc. So it’s all about Location Location Location ;) You can have the best of both worlds.

  14. E.M.Smith says:


    It will vary with the type of router. Some ISPs may even block you entirely from controlling the router. IF you have the same one I have, from AT&T, you just point your browser at it. Such as:

    It takes a minute, then it pops up the usual kind of web page with tabs at the top. I clicked around and one of them was “home network” (3rd over) with a sub-tab of “Wi-Fi” (at this point it prompts for the password). Then there’s a “drop down’ near the top of the page that lets you choose the 2.5 or the 5 GHz radio, and below that an on / off toggle.

    That’s pretty much it. Save the config and then go to “status” (first tab) and see that the 2.5 is running and the 5 is off.

  15. ossqss says:

    You can always call the providers tech support and they can remotely make the changes you need. You will need a password or an admin account to do anything in your ISP provided router/modem for certain. Typically putting a base line address like (typical router default) into your browser and will get you to the login page. You can ID the exact path through looking at your WiFi connection path and see what the gateway is on a mobile or PC device. Many providers use default passwords, so consider changing it if so. Just be careful if you mess around with settings in there.

  16. kneel63 says:

    “I think a trip to the hardware store to price various metal screens might be in order.”

    You should be able to get aluminum fly-screen wire pretty cheap. Most is plastic, but they do make aluminum stuff too.

  17. p.g.sharrow says:

    interesting proposal. Our Old Timers maladies might have an RF cause or at least make them worse. I have to take 600 units Ibuprofen to sleep the night, maybe I just need to unplug the router. I will try that and see if it makes any difference. I can operate my computer direct on the modem as it is on Ethernet wire and nearly 80 feet away. turn off the computer as well! No operating cell phones here. Go back to being a technophobe! ;-) …pg

  18. E.M.Smith says:


    I’ll be interested in your results. I’d gradually started taking 2 x aspirin a couple of times a day (rising and before bed). Last 2 days, none… Joints just didn’t complain today… No idea if it’s related, or not. Too small a sample… / duration.

  19. Larry Ledwick says:


    I don’t know about you but aspirin acts a lot like caffeine for me, If I take a couple doses of it with the last too close to bed time I will be wound up and unable to sleep just like as if I had some caffeinated drink.

  20. interzonkomizar says:

    Hi Chief. Thanks for the tip. Need to find username. Mabe thai.

    Sandy, minister of future.

  21. E.M.Smith says:


    You may or may not have a “username”. For the AT&T router, it is just “point at the gateway IP” and then give it the password. (Said password being printed on the side of the router… not exactly thinking about physical security exposures at AT&T…) It is literally the case that anyone who can read the outside of your router can get root on it…

  22. philjourdan says:

    Some of the ISPs have a “special” username – e.g. cusadmin – that you can use to set the stuff like the Wifi Password and SSID name. Simple stuff. But it hides the true admin stuff behind their “admin” account. And almost all of them have it printed on the unit itself. So just a matter of a magnifying glass (hey! I am old!) and careful reading of the labels.

    But whatever you do, if they “reset” your system, all of your changes go away. So best to document them so you can quickly and easily get back in and reset them to your preferences.

    Note for you Comcast users. Look at your SSIDs. See that “Xfinity WiFi”? That is Comcast giving access to your bandwidth to all their paying members. (Part of their wifi anywhere)

    Another reason I do not use them.

  23. E.M.Smith says:


    I’d seen that and wondered. Any idea how isolated it is from the house net? I.e. is it a DMZ type thing? Aside from just some random being able to slow my download of a linux blob with their watching Oprah, there’s a security aspect. It is a LOT easier to break a firewall if you are already logged in v.s. blocked out… (Apply warz vs crack passwd then apply warz)

    FWIW, I’m pretty sure they will not let me turn off the 5 Ghz on “their” Xfinity WiFi service, so added to their cost and onerous contract terms, they are out of the running for me. It’s looking like either continue with AT&T (but internet only) or Earthlink. The low end Earthlink is pretty cheap so not a big risk if I don’t like it, and looks fast enough for 2 x HDTV sets at a time with some left over (provided they deliver what they sell…) Regular ADSL, so I can apply my own WiFi (meaning I don’t have to shut down my internet access when I power off the WiFi… and I can set frequencies and power levels).

    Most of the rest of my choices are either horribly expensive, too slow a service max level, or unavailable in my particular area.

    You would think that, by now, we’d have an internet option available more like a standard phone service or basic cable. Something like “10 Mb/sec, all you can eat, full bandwidth guaranteed, for a fixed monthly $25 and no long term contract, just month to month.” That’s just enough for 2 TVs or one TV and a computer doing interesting things. As ISPs are selling 100 Mb (peak, good luck on that average actual …) services it ought to be near trivial to provision a guaranteed 10 Mb. All I can figure is that they don’t want to demonstrate what is technically possible as it would dramatically undercut their existing exploitative contracts and pricing plans.

  24. philjourdan says:

    @E.M. As I have not been a customer of theirs since it occurred, I can only give the circumspect response. i.e. it appears to be DMZed off as it is set up as a Web authentication (once you connect, you have to authenticate to your Comcast Account). At least in this area, they are getting into cell service and promoting the service – I take it to cut down on their costs of using other cell providers network to get you service.

    Here’s the kicker. While Comcast claims you “have a choice” the choice in reality is you buy your own router. If you have Comcasts, it is going to be on.

    As a matter of policy, I always have a router between my internal network and my ISPs regardless of their policies. So they can play with their equipment all they want. But not touch any of mine regardless. So yes, in effect, I have 3 wifi signals available. The ISP I set up as “guest” for visitors to my place), then one for my “smart” TVs, and the third for my computers and real ‘smart’ devices (computers, tablets, etc.).

  25. E.M.Smith says:

    Similar “level of trust” on my side.

    Telco Router – has “guest like” WiFi. Only direct plug-in is MY DNS server (that is easy to re-set to saved state) and MY router for the rest of the house (and the occasional experimental desktop thing). The WiFi is passworded and locked so not exactly “guest”, but used for “typical” things like the spousal cell phone and my tablet (that both are exposed to the world a lot so not really desired inside the house network…)

    The TVs had been WiFi to the house net, but for now are on the Telco Router since I started trying to test reduced total signals… Not too worried as they are basically “TV only” devices (ie. the stick is not equipped with things like microphone, camera, communications other than WiFi) At some point I’m going to make a real IdiOT Network and move all devices I don’t fully control (like TV Sets) to it, but for now it’s just out here. Now that I’ve learned to config the Telco router I might make that DMZ like subnet for “things”… (Or wait until I”m sure what ISP I have for the next year…)

    Then my router spits out the “House” WiFi. It has a distinct Guest Network feature set up, but not often used. Basically a DMZ between house and Telco. It also usually has direct plug-ins of the ethernet switch for the Pi Cluster / Beowulf in the making; plus some gear on my desktop. Secure laptops and stuff like that connect here. I.e. devices that don’t wander outside much, or at all. My private servers for files and such are here, too. I’m likely going to make a dedicated router only to replace it (no WiFi) so as to avoid the “Router on means WiFi on too” given recent experiences. But for now, using the private side means private side WiFi on as well. (Thus moving the TVs one layer higher… so this can be “off” most of the time).

    Then there’s the “back room” router. For experimenting with things. Older WiFi D-Link that’s slow but stupid (i.e. not enough smarts to be hacked ;-) Various investigation, experimental, and R&D things go in here, plugged or WiFi. As it is rarely on, it’s near impossible to be hacked, and things behind it are very secure, but also more rarely used. Think of it as isolating the Lab so Lab Accidents don’t get the spouse asking what I did to the TV ;-) (Oh, sorry dear, was just playing with a Ping Of Death script and it got away from me and seems to have infected 1/2 the house…)

    But right now the Backroom and House nets are basically down so only one Router WiFi to worry about as I test effects; and that one now with 1/2 the radios turned off…

    Probably about time to start a slow ramp back up of operational bits… I did sleep well again last night, though not quite as good as with routers totally off. I may need to arrange a remote on / off for the power strip. (I’ve got the gadgets, in a box. Little clicker and a power strip like thing that gets turned on and off by it).

    Well describing all that has me feeling guilty. I think I’ll go start working on getting the office and network into more of a production shape…

  26. Steven Fraser says:

    …and I thought there would be a design for a stylish fractal nightcap!

  27. Power Grab says:

    I will have to re-read this page again completely and carefully, but here is (are?) my two cents’ worth:

    There was a time when I would wake up (at the normal time) with my eyes so dry that they refused to stay open unless I held them open enough to get them to water. I didn’t do any specific research for that problem, but did read a comment by a forum poster that EMF is drying.

    At that time, I had several EMF-emitting devices that were on 24/7 close to the head of my bed. There was the router that I had my laser printer plugged into, and I had my cell phone plugged in to charge right by my head, and a table and/or iPad. When I decided to turn off the “extra” stuff (laser printer, router, and tablet/iPad), I stopped having the dry eyes in the morning. I still keep the cell phone on its charger next to me.

    One time, after reading how disruptive curly light bulbs can be, I decided to avoid using them to the extent possible. At home, I turned off the landlord-supplied curly bulbs and opened the blinds and curtains. I felt like I was on vacation! The level of stress/anxiety that wasn’t something I even particularly noticed, disappeared!

    Here’s another thing that I have found has an impact on how long I sleep and how good the quality is: I have been taking cod liver oil daily since about 2003. At first, I would take it in the morning with breakfast. Then I got to taking it at bedtime. I found that I slept through the night when I had my cod liver oil at bedtime. It also helped with some female stuff and dental stuff . Medical folks get wide-eyed and pay close attention (some even take notes!) when I tell them about the effect on the female stuff. Apparently, there’s nothing in the way of pharmaceuticals that will do what the CLO does.

    In the years before I started taking the CLO at night, I would often wake up around 4:00 a.m. and be unable to get back to sleep. Now, my only sleep-related issue is that I stay up longer than I should, just because I work such long hours and can get nothing done at home unless I stay up late. My sleep is great. I have a nice routine that usually ensures that I get to sleep within 5 or 10 minutes from turning off the light.

    I have a friend who told me 5G is bad stuff. We didn’t have much time to discuss it, but I have been meaning to start looking into it. Thanks, EM, for this article. This page will give me a nice jump-start.

  28. E.M.Smith says:


    The spouse has Seasonal Affective Disorder issues, traced to low Vit-D. We’ve bought a “Lizard Lamp” and she uses it in winter to get her Vit-D up (it emits soft UV). When we were in Florida, often in lots of sun, there were no issues… I now take lunch outside for at least 1/2 hour any non-overcast day and feel better for it.

    At the time Incandescent bulbs were being demonized, I bought a “lifetime supply”. I have articles up about it. The super short form:

    You can still by halogen bulbs and they are color neutral non-EMF bulbs. Get them.

    You can “manufacture” your own 100 W incandescent bulbs via a 3-way lamp. Buy 50-100-150 bulbs and run them on 50 as ambient light in the living room or office. WHEN the 50 element burns out, you still have a nice 100 W element on “the regular” connections (that is, it works in the standard non-3 way socket like a regular 100W bulb). It works fine where you want that 100 W light.

    You can buy “ruggedized” incandescent bulbs for things like shop lights. Often these are already a bit dimmer and running for long life (thick element for rugged behaviour).

    While halogens are a bit more tricky on dimmers, regular incandescents have their life prolonged DRAMATICALLY when dimmed just 5% to 10%. I have dimmers for the regular bulbs in my office, spouses workroom, bathroom, hallway, and bedroom. (Living room is the 3-ways “manufacturing” more bulbs… as is one bedroom and one spousal workroom lamp). You can get “dimmer plugs” where a lamp plugs into it and a control on the end of a wire can be placed in about a 6 foot radius. These even let you dim your 3 way lamps so you can control your rate of “manufacture” of 100 W bulbs. (Too many too fast? Just dim the 50 W by 5% or so… )

    This looks like the ones I use but mine are in white. Got it at OSH Orchard Supply Hardware for one and at Home Depot for another.

    The prohibition on incandescents was for mid-Wattage bulbs. About 60 W to 100W. You can still get 200 W bulbs. A 200 W dimmed to a more golden color and 100 W light output is not very efficient, but will last forever. (At that 50% to 70% dimmed range, life span moves out to decades).

    You can still get 25 W and I’ve seen 40 W too. It is possible to put a Y fitting in a lamp and use 2 x 40 W to make an 80 W. Two more of them in that Y gives a 4 bulb (if gangly) fixture for anywhere from 100 W to 160 W total. I’ve not needed to do this, but did test it with skinny bulbs intended for a chandelier.

    “Decorator” bulbs were exempted (and I think they still are, though have become harder to find). Look for the really big ones intended for bath room light bars and the little flame shaped ones intended for ceiling fans and chandeliers.

    For halogens, when run just a bit off full, dimming improves life. If run a LOT off full power, the halogen gas no longer is restoring the element but is eating it and you can get a big shortening of bulb life. Any Halogen used on a dimmer must be run full power some of the time to regenerate the element.

    Tungsten evaporating from a very hot filament reacts with halogen gas and that compound, when it hits a VERY hot tungsten filament breaks down putting the tungsten back where it came from. If NOT hot enough, the halogen gas doesn’t redeposit that way and the tungsten just evaporates and deposits elsewhere in the bulb where cold… Look for halogen bulbs with a VERY high lumens / Watt and a relatively low lifespan for use on dimmers. These are already being run at the high end of hot, so dimming 15% or 25% isn’t an issue. Be wary of a low efficiency halogen with a 3000 or 4000 hour life on dimmers – it is already running at the lower end of hot… They are good as full on long life bulbs though. The good news is that the cheapest halogens shaped like regular bulbs are made to be maximally efficient and run too hot, so are suited to modest dimming.

    That’s the big lumps.

    FWIW, between still buying Halogens for places where we just want fast full light (kitchen, office one fixture, one living room lamp), and the “manufacture” of 100 W bulbs in the lamps, then the others on dimmers; I’ve not used much of my stored inventory of bulbs at all.

    Mostly we cycle through the halogens in the “full on” fixtures. They are about 1000 hour life? Something like that. 5 hours a day is about 200 days and I think that’s about my average replacement interval. The 3-Way lamps take a long time to burn out the 50 W element, but then the 100 W on a dimmer just doesn’t die very often at all. Bathroom is the biggest user (we have a wall switch dimmer there) and it just never gets turned up to high except on cleaning day or when someone REALLY wants an intense inspection of a spot somewhere. So it’s “every few years” to change the bulb. The one in the bedroom on a dimmer is almost always run dim and has not been changed in years. Running a 100 W at about 20 W (someone sleeping) or 50 W (normal) just doesn’t age it at all. Even the sporadic 80% to 90% on cleaning day doesn’t do much. It is on a wall switch dimmer too, but that one doesn’t seem to quite go all the way to full bright.

    My “curly bulbs” have largely been banned to the Garage and Yard Lights. As have the LEDs. (I went through an efficient bulb fad before it became mandated by law and learned why I didn’t like them… but have a lifetime inventory of them,now too.) So they are slowly being worked off in the yucky places. I do have 3 in a fixture (those flat ’50s kind of flying saucer shaped pull down ones ;-) that mostly just gets turned on / off over the dinning table (and occasionally the 3 way switch in the middle gets turned). I’m not going to put more in it, but it gets used so little those will last a few years / decade+… But why risk breakage and mercury in the dining room if you can just leave it alone…

    Since we found out that the LED blue causes spousal insomnia, they got purged. Then with this EMF causing it to return, my first sweep was to remove the remaining CFLs; so not going to put them back (though the removal didn’t do much – the WiFi off did much more – why bother putting them back?) At this point the IC process is working nicely and I can ignore the “ban” indefinitely, so I’m just going to run with that… Besides, the IC bulbs on dimmers are very mood friendly ;-)

  29. The Orange LiveBox also does 2.4GHz and 5GHz. Since I have nothing here that uses the 5GHz band I thought I’d switch that off. There is an option to turn off the 5GHz in the menu, and when I untick the box it asks me if I’m sure I want to want to switch it off. I say I’m sure and it comes back with 5GHz enabled again. I can’t disable it. I have however turned off the SSID advertisement for the 5GHz side, which should reduce the amount of radiation.

    Wire mesh as shielding doesn’t work so well at these frequencies because of the skin effect. It could be improved by using Silver plating on the mesh – skin depth is around 0.6 micron at 5GHz so you don’t need more than a couple of microns to get 3 skin depths and almost all the energy flowing in the low-resistance Silver.

    Given that Orange give me little control (and of course that my changes can be reversed by them at any time) I’m thinking of putting the router in a cage and attaching a 2.4G wireless access point.

    More details on power levels measurable once the round tuit arrives and I’ve got some confidence in the accuracy.

  30. E.M.Smith says:

    @Steven Frasier:

    No need for a fractal night cap, a simple aluminum hemisphere is fine ;-)


    Folks, including doctors, often forget that every human being is metabolically unique. Aspirin does nearly nothing to me. It does very slightly reduce pains, and seems to drop fevers a little, but mostly it is just near nothing. Certainly no metabolic “uplift”.

    Yet give me a tranquilizer to try to get me to sleep and I’ll wake up (for whatever mild one that they gave me in the hospital as a ‘pre-anesthetic’. The reality is that as a night person I was reading at about 10 pm and they wanted me asleep so tried to slip me a mickey of sorts. I was on the verge of sleepy enough and a few more boring pages would have gone under. BUT, detecting the drug effect, MY body issued a call to action ( i.e. experienced at drinking the body reacts…) and along with a bit of adrenaline stepped up the ‘burn it up” process. I was wide awake with increasing heart rate for about an hour… And informed them to please “Just leave me alone.”.

    The spouse goes to hospital admissions if given a particular class of antibiotics that are fine for me. But I can’t eat garbonzo beans…

    We are killing off minority genotypes with blanket drugging and inoculations. A good friend of the spouse was put in a wheel chair (and no kids…) from a reaction to the polio vaccine. Her genotype now removed from the pool…


    So does Orange not have an option of “hard wire only”?

    Maybe you could find the antenna and put a nice resistance wire to ground… About 50 ohms ought to be workable…

    In General:

    Having gone back to 2.4 Ghz on all the time, I’m having a few issues return a little bit. I’m thinking it may be a matter of degree between the two frequency ranges; or maybe the 5 Ghz higher power allowance. “Going Forward” I’m going to put the thing on a clicker and only turn it on when we want internet / TV access. Or, I suppose, I could just turn off both radios in it and treat it like a router box, using MY WiFi gear more remote from the bedroom and have it on a clicker… Mine also has a power setting control… Maybe I need to see if the telco one has a power setting +/- dB?

    Yesterday I turned it off mid-day and had a very nice nap ;-) I’m tempted to make that a “regular thing” … “I’m not being a lazy bum, dear; it’s A Scientific Experiment of Great Importance! Now let me get back to that nap…” ;-)

  31. E.M.Smith says:

    Well, that was easier than I thought…

    So my Netgear interior WiFi / Router has a little button on top. Hold it down for 2 seconds, it toggles the WiFi on / off.

    I’ve now got my whole office cluster / private side up and running with NO WiFi active. That interior router then plugs into the AT&T / Telco box where using their browser configuration page, I’ve shut off both radios.

    This means I can now run, WiFi free, my entire office cluster. Only the TV sets are SOL as they are WiFi only. I’m OK with that, as I’m the only one here during the days (as chief Doggy Doorman and Bowser Butler…)

    It does mean not using the laptop from the living room sofa, but that’s probably a good thing too.

    So with this set-up, I’ll have to turn WiFi on when the spousal unit gets home and / or I want to watch The Tube (is it still The Tube when it’s an LCD screen?).

    I figure I can likely get a good 12 hours a day, at least, of zero local WiFi signal. (I may need to go through some of the various devices and inform them not to “do” WiFi as well – like the Pi M3 that has a WiFi built in, and then pull the dongles from some others, but I think that’s pretty thin gruel…)

    So now my “Big Question” is just “leave the TVs on the Telco router that needs a browser interface to toggle it; OR, put them on the Netgear with the button on top… ” I think I’ll leave that one for later. For now I can just glory in the fact that I have WIFi free networking end to end running, and a very easy way to toggle either WiFi access point on / off at my desktop. Yay… I think…

    At this point I’m just going to leave both routers up and running and not give them power transients nor bounce the telco line anymore, and enjoy the functionality…

  32. E.M.Smith says:

    Found the “power” setting. I took it down by half. Then half again. It is now at 25% and the Mac still shows “Full bars” in the far end of the house… I may try 12% later tonight…

  33. Larry Ledwick says:

    Depending on your physical room layouts, you might also be able to set up a small corner reflector so that all the power gets sent in the direction of the far end of the house and use the lowest possible power setting.

    If the needed RF path does not include the bedroom areas you might be able to effectively exclude them from useful beam width.

  34. E.M.Smith says:

    Unfortunately, the WiFi router is at one end about 2 feet from the wall, the TV at the other, bed in between… I’d need a 2 foot widening to 4 foot sharp edged beam to get any benefit.

    As I have 1 inch pvc pass throughs in each wall, easier would be to run 50 foot of cat-5 and just plug the TV in to the router… (just scoped it out 3 hours ago… TV has an RJ-45…) so would only need the WiFi when wanting a Roku only channel, which is maybe 20% of usage…

    IF it goes down to 12%, and I think it will, I’m pretty sure that’s close enough to zero to be fine.

  35. Sabretoothed says:

    That could be the mechanism?

    I put my router in a galvanized rubbish bin (Faraday cage) and have a timer for it to switch off at midnight till 7am (when sleeping).

    Maybe more vitamin C is needed around wifi?

  36. Sabretoothed says:

    Another thing, when I use the ‘turnoff’ button on the wifi or do it via the software and put a microwave meter next to the wifi router, it doesn’t actually turn it off! It stops the signal showing on the laptop, but the router keeps emitting no matter what, so best to turn it off on the wall when not in use

  37. Sabretoothed says:

    Interesting about the UV lights. I think there was a place in Chicago a club, where they ran those UVA party lights all the time, and nobody there had a sick day for like 20 years or something :P

  38. E.M.Smith says:

    So looks like 5 GHz IS more power allowed and the USA is fairly high:

    From my read of it, effective radiated power in mW. Band a is 5 GHz while b is 2.4 GHz.

    Country   Band     Channel               mW
    US 	  a 	   36, 40, 44, 48       200
                       52, 56, 60, 64      1000
                       149, 153, 157, 161  4000
    US 	  b 	   1-11 	       1000

    So 2.4 GHz is 1 W and 5 GHz ranges from 200 mW to 4 W.

    So by shutting off 5 GHz and then cutting the 2.4 GHz to 1/4 we’ve got 250 mW max and I might be able to get it down to 100 mW with some work. Compared with 4 transmitters at 1 W, 4 W max, 1 W, 4 W max (or 10 W total) cutting that down to one at 100 mW is a 100 : 1 reduction…

  39. E.M.Smith says:


    Well, I’m a bit concerned as I’m at 20% power setting yet still 100% bars on the laptop… so I’m still not sue there’s been any real instrumental power reduction. We’ll find out after the spouse has gone to bed and I can play more with power levels…

  40. ossqss says:

    So, a couple questions EM. Do you wear metal framed glasses?

    2. Do you have any embedded RF reflective stuff in you?

    Just sayin, I have some plates in me that do have some reflective RF nuances .

    I feel cell calls when not using my BT headset sometimes. Just some thoughts.

    Cell ranges do vary some however.

    It is late and I probably should not be commenting now, but butt, I will anyhow ;-)

  41. EM – useful data on maximum allowed power, given that the 5GHz lower channels are limited to 200mW here in France and the auto setting is giving me channel 132 (so maybe 1000mW and somewhat higher than the 200mW legally allowed for channel 36). Just changed to channel 36, given that I can’t disable it. Data on actual power level emitted when I’ve got it.

    Given that part of my bandwidth is open to everyone (and I can’t change that even though as you noted it is a security loophole) I suspect that whatever I’m allowed to do will not stop the wifi being active on both 2.4 and 5G, though it might well lie to me and tell me they are not active. There are no external aerials to remove, either. There are no visible screws on the box, so getting at those (probably fractal) internal aerials may leave traces that I’ve opened the box and may break some of the clips since I don’t know where they are. As such, it’s looking like a shielded box may be the only real option for shutting off the 5G band. Maybe the same for you. Ferrite shielding works better than Faraday shields at these frequencies, since the dissipation factor of ferrite means that the energy is dissipated as heat rather than being reflected.

    Useful data on actual frequencies at .

  42. Sabretoothed says:

    Could be one of these, it’s too close to a biological important one. I think there are Iron particles inside the pineal gland?

  43. E.M.Smith says:


    No glasses other than the occasional (very occasional, like every other day mice type on power brick occasional) readers.

    I wear no metals other than a small belt buckle (and it is gone when I’m sleeping)

    No metal in me with the exception of some gold caps on molars and 2? remaining amalgam fillings.

    No metal in the spouse either, and she has nearly perfect teeth with no metal in them.


    My box claims to shut off and we feel better then too, so I’m willing to believe it (for now).

    Yeah, that power chart is nice. Do hit the link to verify France (it lists each country since some vary and I didn’t check France).

    As of last night I had 5 GHz shut off, and 2.4 GHz set to 20% power, so 200 mW. Everything worked fine. I briefly dropped it to 100 mW (while the spouse was watching TV) and she said the TV was having trouble loading some channel on the Roku so I immediately swapped it back… but some times that happens anyway without low power, so I need to wait until she’s off somewhere and try again on the reaching for lowest power…


    It’s all over but in clumps.

    some gets in up your nose too, but it is angular:

    Oh, and they don’t know what it does, but it is significant, they think…

    Although the exact role of magnetite nanocrystals on human cerebral physiology has yet to be determined, we suspect that it plays a significant role in the nervous system.

    Then there’s the bit where it increases in Alzheimers

    but it isn’t clear if it is causal or symptomatic or coincidental… since lots of strange stuff forms in Alzheimers brains I’m willing to bet it is either a natural process locked on by disease or a response to the disease process.

  44. p.g.sharrow says:

    5G report;
    after 3 nights of router and computer turned OFF, Sleep quality seems to be better, specially in the morning. Previously I HAD to get up, just could not stand to be in bed any longer, hurt too much to continue lying down. The last 3 mornings I could luxuriate, get real rest in bed, in the morning. 8-9:am instead of 5-6:am. I will continue the experiment…pg

  45. E.M.Smith says:


    Nice reporting!

    BTW, the low power / no router regimen makes my dogs unhappy…

    See, they spend the night in a large “crate” (that’s what they were trained to by the prior owners) and instead of me waking up at 5 AM, I slept in to 8 AM…
    so they had to wait for the crate escape ;-)

  46. EM – I thought I should get some information on signal strength here, so using that $10 RF sniffer around the house I found an unexpected hot spot or two in the bedroom. That sniffer won’t tell me the frequency (I need the HackRF connected to a Pi to get a mobile scanner that tells me frequency and power) but putting a sheet of metal behind the Orange router did reduce the amplitude by a lot. No calibration on the cheap sniffer though – adjust the gain so you have 2 out of the 5 LEDs lit in a low-signal area, and see (and hear) the changes in amplitude as you move around. Still, looks like a useful bit of kit for the tinfoil-hat brigade.

    With pg seeing improved sleep, it could be that some of the complaints about smart electricity meters could be justified. Somewhat serendipitous that your router died and thus you got the correlation between health and the 5G signal.

    If the Roku only works on wifi, but you have cable to the room, maybe a small wifi plug-in point in the room would be useful, especially if it was linked to the TV power so it was only on when the TV was in use.

  47. E.M.Smith says:

    There are 3 TVs with a Roku. The one in the office is rarely used (it is usually the monitor on the Odroid XU4). Basically just when doing “long things” like updates and / or downloads and / or moving TB from disk to disk on the private side cluster… where I’m posting this from the R. Pi headend… So, for example, I might turn it on for a news update if I’m at this station much longer. Otherwise, it is off or being a computer monitor. Not much sense in trying to set up a local WiFi for it… but I will do an experiment on lower bound of power possible (when spouse is not using the L.R. TV…) I’m thinking maybe 1 mW for a 3 foot jump would be enough ;-) (Probably ought to just measure with the laptop where the 200 mW gets too weak, then use inverse square and back calculate… but… experimenting is so much more fun ;-)

    The bedroom has a nice TV that’s my main one when the spouse is home and I want to watch something other than her preference. Given that, it will be run when the minium power to reach the living room is in use anyway. Not much can be done there…

    Then the LR TV is a 2-fer… Has a Roku Stick, but also the Smart-TV has an RJ-45. It isn’t all that smart and has far fewer channels than the Roku AND is slow to tune and boot new channels so a dinky processor in it.. BUT it does get Netflix, Amazon Prime, and a couple of the other “biggies” that make up the bulk of all things of interest, so one option might be just to string a 50 foot cable to it and run the Smart TV for most things instead of the Roku. Another is to set up a R. Pi with a WiFi dongle as a minimal low power hot-spot near the TV and use both Roku and “Smart”-TV – but that will take learning if any dongles let you adjust power, and how to do it. A project I don’t need right now…

    But frankly, it looks like 200 mW works fine for everything AND everyone is doing fine / sleeping fine. So I think just shutting off 5 GHz and turning 2.4 Ghz down to 200 mW is more than enough. (At some later point I may try an A / B of 5 and 2.4 both at 200 mW and see if it is the 5 GHz or the 4 W that was the issue… but not for now.)

    My suspicion is that it’s just a total radiated power thing. Going from 1000 mW on 2.4 with the old router, to adding my private side (where we first noticed it) at 5 W total ( 1000 mW at 2.4 and 4000 mW at 5 GHz) then having it show up again with the replacement Telco router adding another 4 W of 5 Ghz. Taking that 10 W total (plus the minor reply bits from the various WiFi clients) and all of it going back and forth through the bedroom, well, I think having 10 W microwave bath 24 x 7 except when shopping is probably “not a good thing”… Then cutting that to 200 mW ( so 10,000 / 200 = 50:1 reduction or 2% strength, 98% gone) is pretty much all it takes. That’s where I am now and it seems “good enough”.

    As I understand our particular “Smart Meter” (mounted just the other side of the wall from the telco router – an accidental coincidence…) it only transmits in bursts when queried. Probably would benefit from an investigation, though. The other good news is that the wall between it and me is stucco with chicken wire in it and my window frames are aluminum with metal screens. I’m not too worried about outside getting in. More about being inside a mesh reflector and transmitting 10 W with me as the main absorber inside… Though the roof is radio transparent as it’s all wood, drywall, and asphalt shingles…so some can escape after a few bounces ;-)

    I really was surprised at the 4 W signal allowed on 5 GHz. Frankly, even the 1 W. I’d figured it would be some hundreds of mW given the needed (very short) ranges involved. Guess they designed it for the McMansions of 10,000 ft sq on a few acres… Heck, I think some Hams have worked DX to the other side of the world on 10 W (when very lucky… and at much longer wavelengths).

    So I’m going to run this way “for a while”. Only new experiments will be finding absolute minimal power settings for LR, BR, and office TVs and then adding just a bit for quality… (So a person standing up in the LR doesn’t freeze the TV ’till they walk or sit back down ;-0

  48. p.g.sharrow says:

    If I were to guess, the frequency, power and digital jitters are enough to set up muscular twitching that prevents total rest state. The new 5G systems will make this even worse as they will be everywhere in everything. talking all the time! I’m glad my nearest neighbor is nearly 200 yards away and I can kill all my stuff at night.
    The off part of this experiment will continue for a week and then back on for a week or 10 days. The effect is likely accumulative…pg

  49. E.M.Smith says:

    The Laptop only reports 2 other networks at “full bars”. The other half dozen are all less than full. As I’m showing full bars at 200 mW, it is clearly not possible to measure between 200 mW and 1000 mW with this device. It also says the others are definitely below 200 mW or I’d have full bars on them. This tells me I CAN run under 200 mW (as the laptop connects with less than full bars on public hot spots). So it’s quite likely none of my neighbors are near enough for a strong enough signal to be a bother. Maybe time to do that inverse square calculation and plot a graph… mW at Distance…

    I’m also very glad the house is stucco and relatively RF blocking at bed height… The LR however does have a non-screened picture window in it, so a big open RF entry point. Might be amusing “some day” to set the laptop on a table and walk around it with a cookie sheet to see the direction of those networks… One is an Asian word, so almost certainly the Asian family next to the living room side of the house…

  50. E.M.Smith says:

    Well, a quick test of lowering power had the LR TV fail at both 10% and 15% power (though it didn’t connect at all at 10%, it did connect at 15%,, but only loaded 25% and hung with a forever spin cycle. Then claimed it “could not play that title now”… Raising it back to 20%, all was good.

    So I’m close to the limit at 20% and over it at 15%. That’s 200 mW and 150 mW. Only leaves 160 mW, 170 mW, 180 mW and 190 mW as possible lower power that would still work. Max reduction of 40 mW from the present. I doubt that’s worth the bother. I’ll do a finer search of that space some other time, but I doubt it will make much difference. (Why not now? Spouse came back to the living room ;-)

    Oh, and the laptop reported full bars at 15% but none at 10%, so it seems like it has a very abrupt loss of bars calibration…

  51. E.M.Smith says:

    Netflix (at least on the ROKU) has this nice feature where it will display the bit rate and resolution ( 480, 720, 1080 ) if you ask it.

    Using that, I tried 18% and 19% settings. 18% will only give me 480 with occasional bits of 720, but it does work. 19% mostly gives me 1090, but sometimes slides down to 720. At 20% it’s a solid 1080 all the time. (Same recent test program used as the material).

    There’s a known tendency for throughput to drop at marginal energy levels (more packets need a re-transmission or are just lost, so video streaming drops back to slower baud rates. Also easily seen on YouTube).

    So this tells me that the 20% is in reality my lowest bound I can reliably use, but also that I don’t need to allow more to assure it doesn’t suffer drop-outs. It’s got 20 mW of loss while still giving a picture, just a resolution degraded one. I’m happy to have my buffer be in resolution.

    It also says if you increment up to just functional, you need to either test throughput for the final setting or jump it up a few percent steps above that point for actual full throughput.

  52. Sabretoothed says:
  53. Qualitat says:

    Thanks for posting this information, EMS. It was very helpful for me. Here’s my story.

    About a year ago I moved across the country and in preparation for the move I was working on getting my former house ready to sell, getting stuff packed, cleaning (and no doubt exposing myself to allergens & infectious agents aplenty) and then drove across the country in 3 days. Needless to say when I arrived in my new place I was barely functioning.

    I thought I would bounce back after I got some rest but that never happened. At first I would become wide awake after a few hours of sleep and then be unable to get back to sleep for hours. After a few months that began to abate but I never felt refreshed in the morning, my memory had become very fuzzy and I didn’t feel like I was ever going to get back to my old self.

    After reading this blog I checked into the settings for the new wifi router at the new house and found that 5 gig wifi was indeed supported and enabled and that my old router only supported 2.4 gig. So I turned off the 5 gig support in the wifi and noticed immediate improvements in my mental capabilities. After a couple of months I can confidently state that the 5 gig rf signal has a bad effect on my brain functioning. My wife, on the other hand, reports no issues with it.

    I don’t know what I’m going to do as 5 gig becomes ubiquitous. I’m living in a city now and it seems like 5 gig usage is only going to become more prevalent. I may have to move out to the country to get away from it.

  54. E.M.Smith says:


    Might want to cheack your dates. “After a couple of months” is a bit long when the posting has only been up for May + 5 days each side ;-) It just feels like a couple of months….

    Glad it was helpful to you!

    FWIW, these things usually sort out after a decade or so when enough folks notice, commission are convened, and there is a SHTF moment… So most likely you only need to get a decade of avoidance out of the way.

    If you have ongoing sleep issues, also check out:

    We found that the blue spike in LED bulbs caused the spouse to have insomnia. It resets your biological clock to “morning”, which is not good when it’s the 9 PM lights just before going to bed…

    Replacing all our LED bulbs with incandescent & / or “curly bulb” CFL of 2700 K to 3000 K color temperature “cured” it…

  55. p.g.sharrow says:

    In my experiment with turning off the computer and WiFi at night, I can point to no noticed changes in my sleep at night. Router and computer are about 10 feet from the head of my bed. However my old router that I am presently using is 2.4gig WiFi only. The computer runs both but at low power. Barely one bar on nearby devices…pg

  56. E.M.Smith says:


    It’s pretty clear it is person dependent. The spouse is sensitive to 2.4 GHz while I’m insensitive to it.

    I did find full power x 2 routers was an issue for me, but once 5 GHz was off, I was fine. For the spouse, I had to cut the 2.4 GHz down to minimal power (20%) to just barely let connectivity work. I set it back up to 50% while testing something middle of the day, then forgot to turn it back down. A couple of days later she asked me had I changed anything since she was back to not sleeping well… so an accidental “blind test”.

    For her, the 5 GHz also causes heart rhythm changes. She gets sort of ‘palpitation’ feelings. So it is quite immediately obvious for her. Me not so much. Had gone to the Dr. and tried a couple of drugs for it. Nothing really helped. (some had unacceptable side effects). Shutting off 5 GHz seems to have “cured” it…

    It looks to me like there is definitely some interaction of microwaves with the nerves (which is not surprising as they ARE electrical conductors acting as antennae in an RF field) but where the particular frequency vs. antenna length may make a difference. As we are all different sizes, the different frequencies will couple better or worse to each of us.

    FWIW also: I shifted our 2.4 GHz signal to a different channel and that seemed to help her too. It’s a pretty wide band so moving from one end to the other can shift the “match”.

  57. ossqss says:

    Don’t forget microwave ovens operate at 2.4 GHz also.

  58. Qualitat says:

    Ya, that 2 months statement didn’t seem right when I wrote it so I went back and read it again and then hit post because the math seemed right. On the other hand, I never said that I was back at full clarity. I’m still not but I am still getting better.

Comments are closed.