Wifi antenna opinions? Bi-Quad vs cantenna as dish feed or alone?

[()blivion]

OK, been building antennas for extending my WiFi even farther recently. And in doing so I came up with an interesting quandary.

Which in your opinion would be better/best for LONG range Wifi?

1: A Home made Bi-Quad built perfectly.
2: A Home made Bi-Quad built slightly wrong.
3: A Home made cantenna built perfectly.
4: A Home made cantenna built slightly wrong.
5: All the above as a feed for a large parabolic dish.
6: A Home made phased Loop Yagi Built perfectly.
7: A Home made phased Loop Yagi Built slightly wrong .
8: A manufactured "Cookie cutter" grid parabolic. ("cookie cutter" as in, they are all "stamped out" and look the same)

(Edit: My opinion and questions on the matter)
Obviously the dish with a built right feed is the best choice for the longest range possible. But better with which feed is best? And when compared, which out preforms what when built right vs being built wrong? I don't exactly have a 2.4Ghz analysis lab at my disposal to do the tests myself. (well, short from the dongle RSI values and careful observations of network integrity) And radio is finicky at best when it comes to home made IMO. So I like to take in to account the best and worst case scenarios and get every ones opinions.

Any input on this would be helpful for more than just me.

 

[KJ6EAD]

Grid parabolic 24dBi $46.00.

**broken link removed**

 

[()blivion]

Edit: Added the grid parabolic to the list. Making this post obsolete.

 

[KJ6EAD]

I guessed that you were trying to snipe free WiFi from your local McDonald's, Starbucks or public library within a few miles. Apparently you're looking to expand the state of the art and set a new world record.

 

[()blivion]

LOL, There is no local WiFi hot spot that I could get reasonably good signal from. It's 1~3 miles of thick foliage and other obstructions 360 degrees around my house. Id never get through, unless I had a 200 foot tall tower with a dish on top. Plus I have wired high speed DSL anyway.

I'm just doing research. Pondering the possibility's. Looking for optimizations. That sort of thing.

 

[davenn]

The gridpac dish that KJ6EAD linked to is an excellent choice
cant get much more optimised than that. They are used commercially and also by many of us amateur radio operators for
voice and TV transmissions on 2.4 GHz. They are even available for 5.8 GHz where for a similar sized antenna the gain is ~ 30dB

looking at your initial list the best choice would be a good cantenna in front of an appropriate dish. The bi-quad is likely to have too much gain and result in the dish being under illuminated.

Dave

 

[trash]

I've found both good and bad results Dave. Each case has basically been more frequency related.
These are some of the 3.5GHz stuff I've done for unwired.
http://home-and-garden.webshots.com/photo/1209322665048941318fXFsgZ
While this works well for 3.5GHz, the exact same setup for 2.4GHz does not work well.
http://home-and-garden.webshots.com/album/183973395WcCdiw?start=0

Gridpacks and long loop yagis are the winners for 2.4GHz.
I have a 2 x 50 element phased loop yagis and they're good enough to get a signal to the moon and back
I used to use them to snoop on their neighbour's video senders around the neighbourhood.
The gridpack on the tower works well for ATV and Wifi.

I find that most cantennas I've been asked to inspect and test are poorly build and not much consideration was put into their design or construction.

 

[()blivion]

Some good input so far. Nice work by Trash. (I especially liked the Geiger counters)

WALL OF TEXT MODE GO!!!!!

I don't completely dislike the cookie cutter grid parabolic dishes that are out there in the wild. Though I do dislike how they are fed signal. I prefer the radio device to be as close to the focus point as possible. That way there is no cable carrying RF to worry about at all, even if they do come with the right cable. This is the main reason I use USB dongles.

As for phased loop yagis, I find them quite interesting but most I have seen on the web were built crappy. Yours looks like one of the better ones. However infrequency of good samples makes it appear like they are hard to build altogether, not that it's a problem for *ME* since I can "solder wood to plastic". (though having to align 50 yagi elements does seem like a pain). Lastly... you can't feed a dish with one of these, lol.

I also agree that the average cantenna is built to a crappy standard. Most of the time I see people measure the probe, THEN install it in the can. Or not account for the little overhang of the connecter. Or underestimate how important the measurements really are. Such as, measuring from the BOTTOM of the can instead of the INSIDE BOTTOM when placing the probe. But then again, some things are also completely forgivable since they are not often in the instructions from the start. For example, something I have another thread for is RF cable substitution. Most guides just say, "use 'this' connecter on your antenna, then hook it to your WiFi device". I think to myself, "That's Great!!! But your forgot to mention that there is really strict criterion for 2.4Ghz feed cables". I guess they all assume that a person *BUILDING AN ANTENNA FROM OLD TIN CANS* is going to spend lots of money on good lengths of high quality RF coax. Which as I said before, is why I use USB dongles as close to the antenna as possible.

Now (and this is more of a personal thing) but I UTTERLY DESPISE using a connecter on the can it's self. I prefer the smallest break into the can possible. I personally punch a small hole in the right spot with a thumbtack. Then feed in a unmeasured amount of either enamel coated magnet wire, or some of the core of the RF cable. Then I solder the shield of the cable directly to the can all the way around (it's tricky if you have no soldering skills). Then test for shorts with a meter. If all is good, the small length of RF cable coming off the antenna then goes directly to the dongle and is soldered in. After the outside is in and secure. Then I get out a thick piece of card stock. Measure and cut a *PRECISE* guide or "feeler gauge", and use that to trim down the internal antenna/probe. And it's also good to check that the probe is the right distance from the bottom of the can again. When all is cut, then I rinse and repeat measure, straighten, measure, straighten until I'm satisfied it's in tolerance.

Another thing I deal with is that the the bottom of most cans are not very flat at all. So I usually cut out the bottom and solder on a very flat piece of tin sheeting. This also allows you to work from the shorter side when measuring the probes distance from the back of the guide. And I think it's more accurate positioning of the probe in the end. I would do something like this to remove the ridges in the sides, but that's much more difficult and it would just be better getting a length of pipe or tube instead of a tin can from the start.

And finally, Most people use tape to join more than one can. This is great, and for most every one is definitely preferred. But I tend to solder the lips together. I dislike any gaps in my wave guide walls. But this is WAY over board and completely unnecessary to be sure. I doubt it even does anything honestly.

WALL OF TEXT MODE, OFF.

-Oblivion

 

[KJ6EAD]

I have a 2 x 50 element phased loop yagis and they're good enough to get a signal to the moon and back

Show off! Don't you find the internet a little laggy with the moon bounce?

 

[trash]

I will post a picture of the 2GHz loop yagis because I don't have one on that page.

I've had longer latency on a double satellite hop GSM mobile phone system I used to manage.
The audio did a 150,000km round trip for two handsets side by side and the delay sure sounded like a lot more

 

[()blivion]

I edited and expanded the first post to include your suggestions. Since I did ask for them it would be a shame to dismiss them.

 

[trash]

A picture of my 2.4GHz loop yagis as promised.
http://home-and-garden.webshots.com/photo/2883559600048941318ARDKjo

 

[()blivion]

Holly crap that thing is huge. Now I understand when you say that it can get signal to the moon. You got plans to go with that monster maybe?

 

[trash]

It's actually a reasonably small antenna. It's very light and easy to mount and steer. A clear advantage over a dish.
I've used it for Wifi and snooping on neighbours 2.4GHz video senders etc.

The disadvantage is that the elements are fragile. While they can handle birds sitting on them etc, bumping it against the tower while lifting it into position
can bend or break the elements.

These thin wire elements also tend to make the antenna more narrow in it's frequency response, so the driven elements are log periodic.
Other smaller loop yagis like this one; https://cheatthezone.tripod.com/wireless/loopyagi.html I prefer to make using brass strips.
They give a much broader frequency response and brass is very easy to work with.
It's disadvantages are that it's a soft metal and easily deformed. It's also heavier that other metals.
So while brass is not too good for big antennas, it's great for small and medium size high gain antennas.

Brass corrodes much quicker than other metals too. So it helps to coat an antenna with something. I've always considered paint, but never tried it.

 

[()blivion]

Brass corrodes much quicker than other metals too.

That sounds like galvanic corrosion to me. A really neat way I learned to stop that from happening is to electrically attach a more chemically active metal to your protected device if you plan on having it out in the weather. Something like magnesium, zinc, or aluminum works well. You can also put a small DC bias voltage with respect to ground on it. All this is called "Cathodic protection" I guess.

Edit: Removed oxymoronic statement.

 

[Sceadwian]

Brass electrical connectors are almost universally Nickle plated for corrosion resistance (and wear resistance as well) You could probably use paint but you'd have to clean it well and use a good metal primer.

 

[chicowoodhill]

Interesting thread. Learning much.

I have a 12' parabolic dish (an old sat TV antenna) I've been playing with out in the isolated backcountry to experiment with mountain-to-mountain 2.4 GHz wifi connections. Having fun, learning a lot, not fully legal (although I have a ham ticket) but there's no other 2.4 GHz activity within miles so I don't feel too guilty. It can easily do over 40 miles with pretty good signal.

I use an off-the-shelf patch antenna as a feed. I've tried a homemade bi-quad but it didn't perform as well as the patch. Perhaps it wasn't constructed perfectly.

All that said, does anyone have recommendations about what feed might do better for my dish? I'd like to keep experimenting but would like to start with some good ideas. Thanks in advance.

 

[Sceadwian]

That loop yagi absolutely massive! How much gain does it have compared to a cantenna (which is much smaller) Cantenna's are typically 12-15DB.

The world record is currently held with nothing more than a well placed cantenna fed from a 12 foot satellite dish, and that's lines of site 125 miles.

 

[()blivion]

The world record is currently held with nothing more than a well placed cantenna fed from a 12 foot satellite dish, and that's lines of site 125 miles.

Yup, That was the main reason for the thread. And they used an "over the counter" WiFi device unmodified if I'm not mistaken. I think they were going to try and out do there record, but I don't know any more info than that. The idea they had has room for improvement IMO.

Profesor Trevor Marshall on his website uses a Bi-quad for his dish. He is also using an odd oblong shaped dish, so there is a chance that Bi-quads are still non ideal for a large hemispherical dish. But I don't know, which is another reason for the thread.

I have a 12' parabolic dish (an old sat TV antenna) I've been playing with out in the isolated backcountry to experiment with mountain-to-mountain 2.4 GHz wifi connections. Having fun, learning a lot, not fully legal (although I have a ham ticket) but there's no other 2.4 GHz activity within miles so I don't feel too guilty. It can easily do over 40 miles with pretty good signal.

I have been trying to get my hands on a big dish for cheep/free since I started this thread. Sadly no dice. It seems that in my area all the meth addicts have scraped them. Or people expect to get what they originally paid for them. I may look into making my own.

I too have lived in, and possibly plan on once again living in "the isolated backcountry". Can't stand that the best internet I can get in that area is 256k and like 50$ a month. I was really hoping to some how make a solar powered long range multy point to point connection of some kind to the nearest real town for this reason. It is around ~37 miles away through ungodly amounts of mountains, and both towns are in valleys. If I could figure this out I might be able to achieve ~10meg for a reasonable price. But... then I woke up from that pipe dream, and there are apparently laws against putting things of national forest land :/

If your still on chicowoodhill, could you post what kind of signal your getting exactly? something like inSSIDer plots and graphs. Or the manufacturers signal level numbers? Hell... I'll take windows signal bars if that's all you got.

-()b

 

[Sceadwian]

Yup, That was the main reason for the thread. And they used an "over the counter" WiFi device unmodified if I'm not mistaken. I think they were going to try and out do there record, but I don't know any more info than that. The idea they had has room for improvement IMO.

It's stock hardware, though the software has to be tweaked (just a registry entry setting) and it's not technically wi-fi compliant as the time of flight at that distance is actually greater than some of the timeout delay tolerances in the protocol specifications but that's semantics really.

I'm sure the key to their success is very simple... Minimize insertion loss! Any DIY'er trying this at home really doesn't appreciate how important the connectors and impedance matching is with extremely weak signals, you could with a few bad connections loss 3+db of signal, which would be a 100% drop in pickup range, you can easily loss another 3db with an impedance mismatch... So the record goes from 150 miles halfed twice to 37.5 miles.... Right at where just about any joe can do this if you have line of site.

I may look into making my own.

I do not recommend this at all, the specific curvature is incredibly important you'd need very large scale fabrication experience to do it accurately. A hobbyist could make a dish 2-3 times the size of a commercial dish and still end up with a loss of signal in comparison. The large the dish the more subtle the curves and the more errors present themselves.

Keep in mind, at that low of a signal range, rain will destroy your signal, a bird will cause blips, fog willl cause a dip as well. You'll only get that kind of incredible range vs power if the environment is spot on willing to allow it to happen.

If you really want to do a longer range link you have to delve into a HAM license and the hardware to get it to work right.

Solar powered point to point repeaters might be okay as you can position them closer to each other, but the cost to set the network up and maintain it is probably going to cost you in the end more than a commercial service.