Simple Wi-Fi Yagi (2014) | Hacker News Reader

Simple Wi-Fi Yagi (2014)

ham-radio.com

211 pointscyanoacry67 comments

This makes the antenna more directional. It increases the range in one direction while decreasing it in every other direction, so I'm not sure how well this would work for most people without relocating the router/AP.

That being said, it's a clever bit of engineering.

Given that it's only somewhat directional, this would actually be useful for the common case that the AP sits in a corner of a one-story house. No sense sending all that RF energy outside, so direct it inside.

> the common case that the AP sits in a corner of a one-story house

I was going to make it the centerpiece of the living room but my girlfriend vetoed that.

When you upgrade from consumer-grade equipment into dedicated APs, they start looking more like something you can put into the center of your living room without being ashamed of it.

The Unifi APs for example look like oversized smoke detectors. They have obnoxious blue LEDs like seemingly all modern equipment, but you can turn them off and then it emits no light at all, and fits quite well into the living room when mounted on the ceiling.

Mikrotik has been moving in this direction as well. This sucker [1], while unfortunately 2.4 GHz only, is an inconspicuous 3.5" white circle you can stick on your ceiling.

[1] https://routerboard.com/RBcAPL-2nD

My current (rental) place has the router and AP in one corner of the apartment, so signal strength at the other end is terrible. Is there a better (cheap but effective) option than this?

You can try adding a parabolic (or circular) reflector: http://www.freeantennas.com/projects/template/

As with the original link, the antenna is directional (and since it's a reflector, probably even more directional), so you lose out on flexibility. It'll be quite good if you don't plan on moving the other point, though.

Better is to get a WDS bridge (wireless repeater). A high gain antenna (higher than in article) won't work well in the "long house" scenario because two clients may not be able to hear each other and may thus collide. (RTS/CTS is supposed to fix this but it relies on clients choosing to use it.) Unfortunately with WDS, the bridge must be the same brand as the AP (which itself must support WDS), so you might be stuck purchasing two new pieces of hardware.

But if the clients are all near each other, you have some money to spend (~$100), and don't mind being stuck with one band (2.4 GHz or 5 GHz), I'd recommend one of these:

https://routerboard.com/RBSXTG-2HnD

https://routerboard.com/RBSXTG-5HPnD-SAr2

https://routerboard.com/RBSXTG-5HPacD-SA

They're really meant for outdoor installations but they work well for indoors with a cluster of clients far from the AP.

I bought a WiFi extender for this and works perfectly. I don't know why I didn't do this earlier. Mine is a Xiaomi one for $9.

This would be great if you lived on a boat, then you could have a wifi router on the shore/marina with signal directed at your boat.

What is the legality of this? I have a vague suspicion that operating a device like this would violate some FCC regulations.

FCC regulations (afaik) only apply to the power output of the device, modifying the antenna like this shouldn't change the power output.

FCC regulates EIRP, which is output+antenna gain. For this to be strictly legal you would need to adjust the power down until the EIRP is within the legal limit. Actually to be strictly legal I think it would need to be type excepted by the FCC.

That being said if you're using this in your home/office on a legal channel it would be really difficult to even tell from outside you were using it and almost certainly not worth the FCCs time to even try to watch for it. It's not that much extra gain over the stock antenna.

To be fair, the FCC regulates both EIRP (4 W / 36 dBm) and max transmit power (1 W / 30 dBm). E.g. you can't just have a 34 dBm transmitter hooked up to a 2 dBi dipole.

If it's like here in Australia, they do consider EIRP. So a directional antenna would increase the EIRP, but with just 100mw tx at the output, I doubt EIRP would be over the limit (4W for 2.4Ghz ISM in Australia).

FCC regulations apply to the device with the supplied antenna in place.

The antenna is not supposed to be replaceable, hence the reverse SMA connector on most modems.

Modifying the antenna in any way invalidates the type approval.

Good intuitions. FCC regulations restrict EIRP[0], which is essentially "what is the signal strength in the strongest direction". It's possible that increasing directionality of your antenna puts you over the EIRP limit (given how routers already try to get _really_ close to it to improve performance).

[0] https://en.wikipedia.org/wiki/Equivalent_isotropically_radia...

Yep. Specifically you'd need to adjust your TX power down to 27 dBm (from a max of 30 dBm) for an AP, or 29 dBm for a station. http://www.air802.com/fcc-rules-and-regulations.html gives a nice summary.

Basically APs on 2.4 GHz are only allowed 36 dBm (4 W) EIRP, no matter what antenna shape. This is usually not a problem though, since stations tend to have weaker transmitters anyway, but will still benefit from the AP's enhanced antenna gain.

OK, so if by the time the signal exits my property it has dropped below mandated limits, there is no harm done right?

If I aim my antenna such that I contain the illegal power region within my property, will there be an unintended consequence?

To be clear, I'm referring to a house not an apartment so I don't have any concerns about upstairs or downstairs neighbours.

I think FCC laws technically apply within your property as well. Hypothetically, what happens if your non-compliant device blocks a visitor on your property from calling 911?

I think this falls under the pragmatic category of "you're only breaking the law if you get caught."

Oh of course, and there's actually a surprising number of FCC rules you'll never get caught breaking. GMRS license violation is another.

No, that is incorrect.

The domain of the FCC extends to all property of the United States of America, be it public or private.

The reason this is important is that it prevents private corporations from interfering with public air space (as discussed previously on HN [1]), and prevents landlords from interfering with the rights of tenants to legally place outdoor antennas (such as satellite dishes).

[1] https://news.ycombinator.com/item?id=8406022

RF energy doesn't quite work like this. If your transmitter has a high EIRP (either due to antenna directionality or high TX power), then its signal is going to travel a long way, even after the signal level has dropped to that of a lower-EIRP transmitter. That is simply the inverse square law at work.

You could place sufficiently absorptive materials that this isn't so (e.g. enclose your house in a Faraday cage). But as hydrogen18 points out you're still limited by the FCC.

He's not operating under FCC part 15. He's using FCC part 97 instead, with the appropriate license.

His rules: 1500 watts PEP, unlimited EIRP, channels 1 to 6 only, no encryption

That's better than a normal microwave oven.

Note that licensed ham radio operators are allowed to operate at (much) higher power on the WiFi channels that happen to overlap the 13cm ham band.

http://w5vwp.com/wifihams.shtml

(Of course then you are subject to all the Part 97 restrictions like no encryption, no commercial use, etc)

Right? Why wouldn't the router companies include this somehow in their designs if it actually improved performance?

Because it's illegal, is why.

It's not illegal per se, if you adjust transmission power accordingly. Router manufacturers generally don't use yagi because it's directional, requiring aiming, unlike omnidirectional dipole antenna. Modern wifi stations have much more sophisticated ways to increase range than passive antenna designs.

Dipoles are directional too. There are nulls directly aligned with the antenna (above and below the router in this case).

No doubt this accounts for some of the "my WiFi doesn't work upstairs" stories.

No, because no-one wants a home AP that only works if it's pointed at you.

You can buy directional APs if you want, but they're usually used for outdoor point-to-point links, or as sector antennas. e.g. I just installed one of these today: https://routerboard.com/RBSXTG-2HnD in a place that needed a signal far from where the AP was. Usually patch antennas (like this product) are preferred over Yagis for microwave links due to their smaller footprint, and capability of dual polarization in a single package.

Better is if your indoor AP can transceive in more of a "pancake" shape, like this: https://routerboard.com/RBOmniTikU-5HnD Note that's almost the same gain as this home-grown Yagi (7.5 dBi vs. 8.5 dBi) but with a more useful directionality.

It is illegal, according to another comment:

https://news.ycombinator.com/item?id=13459834

As I said in that comment, it's perfectly legal (and useful) so long as you turn down your TX power a bit.

Antenna gain is a function of directionality and vice versa. Patch antennas are also good because they have often very low "backward gain" but at the same time they are not very directional forward, so they are good to be placed in a corner of a room or near a wall.

It's only illegal if you get net gain out of it[1]; if you turn down the power so that the maximum power transmitted in any direction is the same as before, then it's legal. You can get a small performance improvement by using directional antennas in this way as you reduce the received strength of off-axis signals which can reduce both multipath interference as well as interference from other stations.

It's not all that useful indoors though as 2.4 and 5GHz signals bounce off of walls really well so you lose much the directionality at the very first wall it hits.

1: Actually there is a total ERP requirement, but most routers with external antennas are already close to that requirement, so more than about 3dB gain will often put you over.

I remember reading that unless you amplify the radio above 5W or something, there's no way you're above the regulated limit.

4 W (36 dBm), and the 8.5 dBi gain of this homemade Yagi puts you over that if you're operating at 1 W (30 dBm) max transmit power (as some home routers do).

You're not allowed to mix and match antennas on mass produced Part15 devices, regardless of the ERP. That never seems to stop anyone though.

Citation? Plenty of commercial part 15 devices don't even come with antennas, requiring you to "mix and match", unless I misunderstood what you meant by that.

http://www.ecfr.gov/cgi-bin/text-idx?SID=c2dc3c7e7691816d126...

The FCC has stated that the RP-SMA connector (the connector used on nearly all wifi routers that have detachable antennas) shouldn't be allowed on part 15 devices, but also that it will delay enforcement of this indefinitely[1]. So even the FCC is fairly "meh" on enforcing this unless actual interference is reported.

1: http://hraunfoss.fcc.gov/edocs_public/attachmatch/DA-00-2225...

So this makes the router sends a stronger signal, farther. But how does this help if my laptop still has the same wifi adapter and can't send a strong signal back?

Reciprocity says that path loss is equivalent both ways.

It also makes the antenna more sensitive in the direction its pointing. The elements make the antenna more directional, sort of focusing the energy. This increases the transmit side by concentrating the energy into a smaller area, but the same effect also helps it collect more energy.

The trade-off is worse performance perpendicular to the extra elements.

I'd guess it goes both ways, just like a speaker cone and the cone of an ear.

Antennas are coupled. Transmitting is receiving, and receiving is transmitting.

When you pull on a chain or a rope, it doesn't matter if the other side is "receiving" the pull, or "sending" it. The two sides are physically coupled.

Similarly, antennas are physically coupled to each other. Increase the gain on one side, the other side still benefits. You're physically connecting a "rope" of electrons over the air.

I am wondering if antenna calculations are applicable for all types of routers. If not, how can I adapt it to my own setup?

itll be applicable for any 2.4ghz, 50 ohm radio

Not really. It's only really applicable to any router that uses a dipole antenna.

Pragmatically, this means it's only really an option on routers with external whip antennas. Lots of routers have internal antennas, at which point you're going to have problems attaching the thing. Also, if the internal antennas aren't dipoles, it's not an option at all without significant redesign.

there are a ton of routers with replaceable antennas where the connector is coaxed directly into the front end, like in the article.

i assumed that anyone trying this would have one of those.

> dipole antenna

Erm, looks like a monopole antenna to me. The rest of your post looks correct otherwise.

As an electronics student, are there any resources I could use to learn the basics of RF (antennas, etc)?

1. Find a university course on RF, look up the course textbooks.

In my university we had several courses on "analog electronics" which covered topics like free space path loss and RF concepts (e.g. bandpass filters).

2. Use Google or another search engine (DDG) to find the book in PDF or EPUB version.

3. Enjoy reading the book without forking over hundreds of dollars to a greedy publisher.

balanis

You might start at the ARRL's Antennas page [0] and consider the ARRL Antenna Book [1]. Most of that is aimed at amateur radio in particular but the theory is the same.

There's tons of free and/or open-source software (of varying quality) for modeling antennas and such as well.

[0]: http://www.arrl.org/antennas

[1]: https://www.arrl.org/shop/ARRL-Antenna-Book-23rd-Softcover-E...

And/or find a local ham radio club.

http://www.arrl.org/find-a-club

Take a microwave design class. This would be a second or third course after electrostatic, electrodynamics, then something like microwave theory. At my school we had to petition for the course as a special section. It was great, we built cavity resonators, and final project was an antenna design. All the way from theory, then using modern fea antenna software, and then building it and testing performance. If you have any interest take it. I was not going to but the professor said: "Phil, what's the lower end of the microwave spectrum?", I said "1 GHz", he then asked "How fast is your computer?", I replied "2.8 GHz", He said "I'll see you next semester then?". these days having an understanding of microwave theory will help you no matter your specialization.

if my router has 3 antennas, is there any point to installing it only on one of them? my intuition that's based on exactly no RF experience whatsoever tells me that it'll do more harm than good. (case is weak 5GHz signal in the corner of a house.)

Depends. Some routers have multiple antennas because there are multiple transceivers inside it, and some technologies (i.e. MU-MIMO in 802.11AC) will most likely react strangely having one of their antenna modified. In the end, it would probably make your performance even worse.

theres a two reasons your router probably has 3 antennas:

-beamforming

-split 2.4ghz/5ghz/5ghz-wave2 antennas

if your router is using those three antennas for beamforming, you might trigger some strange behavior if you yagi a single one of them. if run into issues and can turn beamforming off, you should be able to yagi a single one no problem.

if your router pumps each antenna into a separate radio, then you can reliably pump a single radio into that corner of your house w/ a yagi. however, you won't have seamless roaming between the 2.4ghz and 5ghz networks when you move into and out of that corner of your home.

i recommend getting a mesh network, they're the bees knees and are a huge step above the traditional access point + range extenders model

Basic antenna question: I thought with Yagis the elements where connected together, is that not the case? As they seem to be held just by plastic.

No, Yagis have a single driven element like a dipole, and you have longer elements behind the driven element (reflectors), and shorter elements in front (directors I think?).

Interesting, thanks!

Could something like this have any impact on a bluetooth antenna?

Certainly, though the output power is quite a bit less (20dBm for WiFi vs 4 dBm for typical bluetooth).

http://www.tomsguide.com/us/how-to-bluesniper-pt1,review-408...

interestingly, the antenna optimizer (AO) software uses genetic algorithm to find the right the antenna topology, therefore, this antenna is not designed, it's evolved. more details can be found here: https://en.wikipedia.org/wiki/Evolved_antenna

I wonder what this will do for my 20DBi external wifi adapter.. I can already connect to networks over 600 meters away from my 13th floor apartment..

Reply on news.ycombinator.com