Extend a WiFi network originating with an omnidirectional antenna to a garage?

Folks:

In setting up our new suburban paradise I would like to extend WiFi coverage to a garage that is a separate building. The signal must get through about 75′ of air and about five walls. The base station is an ActionTec router supplied by Verizon FiOS (i.e., an omnidirectional antenna). Measured with a Samsung Note 3 there is either 0 or 1 bar of WiFi strength in the location where I want WiFi. Since it is almost working between two omnidirectional antennae (ActionTec and phone) I am thinking that what would work is a repeater box with a directional antenna pointing at the ActionTec. Has anyone tried this? The locations have separate electric feeds from the pole so it would not be possible to use a powerline-based system.

Thanks in advance for any help!

17 thoughts on “Extend a WiFi network originating with an omnidirectional antenna to a garage?

  1. It might even work without the directional antenna. In my setups, I’ve found that the repeaters often get 2-3 bars where phones are down to 0-1, presumably because they’re bigger and not battery powered.

    Maybe get one with the option to attach a special antenna, but try with the included basic one first?

  2. I replaced the FiOS router with a modern D-Link router and instantly got much much better coverage in our house. And then set up a DLink repeater on the second floor to cover the corners of the house that were marginal.

    I had to call Verizon and get them to switch the ethernet at the FiOS box in our basement to come out the ethernet port rather than the cable. Side effect is that their cable box doesn’t get internet and so can’t do some things like tv guide or something, but we returned the cable box anyway a while ago.

  3. If you’re willing to take the time, you could probably setup a couple of Raspberry Pis to create a dedicated point-to-point wireless link.

  4. I’ve had lousy luck with repeaters. Part of the problem is that they’re on the same channel that you’re trying to repeat.

    You can try adding an inexpensive router in bridge mode. We have an expensive (when I bought it 6 years ago) Airport Extreme router in bridge mode. In front of it we have a TP-Link AC router because the TP-Link has 802.11ac for devices that support it, broadcasts simultaneously on the 5GHz and 2.4GHz bands and supports prioritizing by port so it can give VOIP priority if needed. The TP-Link was about $60 or $70. Devices from ASUS, NetGear, Linksys are 2 or 3 times that.

    Or you can try ethernet-over-powerline. There’s even devices (from TP-Link and others) that put out a WiFi signal. That’s how we get WiFi in the kitchen which would otherwise be blocked by the chimney.

    I have an unused Cantenna. I’ll mail it to you if you want it. You need to be able to attach it in place of a regular antenna. The TP-Link, for example, has 3 external antennas.

  5. You may want to try Powerline networking. Sometimes works better than Wifi (sometimes not). The gizmos are cheap ($40 for a pair) and you will know very quickly whether it works for you or not. If not, just return them. You can plug in a router at the garage end so that you can broadcast wifi within that room.

    http://www.amazon.com/TP-LINK-TL-PA4010KIT-Powerline-Adapter-Starter/dp/B00AWRUICG/ref=cm_cr_pr_product_top

    If you print out this template:

    http://www.aegisarts.com/wifi/ez12.pdf

    you can turn a omnidirectional antenna into a directional.

    Instructions here:

    http://www.freeantennas.com/projects/template2/

    Again you have little to lose by trying it.

  6. In short, there’s a chance it might work as described (somewhat distant omni to well aimed directional antenna to go from 0-1 bars to “3+ bars”), but the reliability may be more marginal than you’d like without changing other things too.

    The “WiFi Strength” bars are both based on a logarithmic scale (often something like dBm received signal strength), and then made further non-linear for display (ie, “5 bars” is deliberately much better than 5 times as good as “1 bar”). In numbers 5 bars is typically something better than -65dBm, and 1 bars typically down under -85dBm — so at least a a 20dB difference, but the real numbers may be even further apart. The exact mapping depends on the devices, and I have no idea what Samsung decided to do, but “0-1” bars probably means something around -87dBm to -90dBm. Typically “3 bars” will be the lowest suitable signal for reasonable stable performance, and (depending on receive sensitivity/error correction) maybe around -75dBm.

    Antenna gain is measured in dB. The tiny inside-device omni antennas might be something like 3dB gain. A good (say 1-foot square panel) might be 23dB gain, assuming it’s carefully aligned to the specific signal origin (the gain in unaligned directions is worse, and often dramatically worse in between the lobes of gain). A maximum difference of 20dB, which might be enough with careful directional antenna alignment to give you a stable useful signal. But it’d need to be well mounted to avoid it being bumped/drifting out of alignment.

    If you could either move the omni base station closer, or replace its antenna with a slightly more directional antenna (eg, a 60 or 90 degree sector antenna, instead of a 360 degree omni) you’d have much more noise margin (and thus more reliable signal) — ie a slightly directional antenna (maybe 10-12 dB gain compared with say a 3-6dB gain of a basic external omni) linked to a highly directional antenna (another 20-23dB gain), for a total gain in the 30-35dB range. Or add another base station close to the garage (ie fewer walls away), connected either by copper or via “wireless relay”. (Wireless relay will definitely eat into your bandwidth, as it needs to use the same radio bandwidth to receive and then moments later transmit on the packets; if it’s a second base station connected by copper, put it on another frequency to avoid the two base stations interfering with each other!)

    Ewen

  7. First thing to do is bring the signal closer to the garage by moving the access point. Second thing to do would be to try different and higher end hardware. At the end of the day you can’t repeat a weak signal, so you could mess around with directional antennas attached to some consumer grade hardware, Or jump straight to a couple of ubiquiti nanostations, which will certainly work.

  8. You will be happier if you just run a wire and use an access point or another wifi router. It will come to that when you get feed up with the slow speed and unreliable signal.

  9. I have a situation that sounds very similar. Detached garage, about 75 ft from
    main house. Powerline doesn’t work well, so I link the garage using Wifi.
    I used two Netgear AC-66U in bridge mode, and that drives a wired network
    in the garage (security cams, etc). I monitor the house network using Cacti,
    so I have a good idea of the long term stability of the connectivity (generally very
    good, in spite of densely populated Wifi bands around our area). The link
    operates in the 5 ghz band, using up to 80 mhz channels.

    The two routers were placed in “convenient” windows – one in the walk attic
    of the house, the other in a window of the walk-up garage loft.

    The AC-66Us delivered a fairly dramatic improvement in overall link stability
    relative to the previous 2.4 ghz link where I used dd-wrt running on Linksys
    WRT-54G wireless AP. I attribute this to operating in the 5 ghz band. 2.4 ghz
    is very noisy here (we live in the center of town, and have a cable operator with
    a fairly dense outdoor Wifi network).

    I don’t do regular speed tests on the network, but when I commissioned the network,
    it did turn in 50+ mbps in both directions in a quick speed test.

    Both the attic and garage get very hot in the summer, probably up to 98 degF, and
    so far, the routers have tolerated the temperatures. Low temps in winter don’t seem
    to be a problem either, since the boxes generate enough heat to keep them warm.

    Of course, your mileage will vary, but this solution has worked very well for me.

  10. I’ve had great success with 2 Ubiquiti NanoStation loco M5’s in Bridge mode. (http://www.ubnt.com/airmax/nanostationm/)
    They cover a 200′ air gap and I’m consistently getting 200 Mbps. They are powered by Ethernet, so they don’t have to be near power outlets and can be mounted outside. In 3 years, I’ve never had them go down. Great software as well.
    $67 apiece from Amazon.
    Good luck in your new house! Matt

    P.S. Long time lurker, first time commenter. I really enjoy your point of view on many issues, especially because I don’t always agree!

  11. I have nothing to add to Ewen’s theoretical discussion, so I offer the following practical example:

    I rode out the 2009-2010 recession in suburban Detroit by pointing a Hawking HA12W (Hi-Gain 12 dBi Window Antenna) at my apartment clubhouse about 200′ away and connecting it to a Belkin Wireless Gaming Adapter and plugging that into an ordinary WiFi router.

    The challenge is getting good reception when the leaves are out and ambient temperatures are high.

  12. All the proposed solutions appear to be relatively expensive and not very robust. Are there strong impediments to using an Ethernet cable to carry the signal between the two locations?

  13. Not robust perhaps, but in terms of expense, all of the solutions are cheap compared to calling in an electrician, trenching to bury cable, fishing thru walls, installing jacks, etc. You could do a hack job with a premade 100′ cable and save $20 but it would look like hell.

  14. Surely no one here would require an *electrician* to run cat5? If there is only a flat uniform lawn separating the two locations, a bare cable will be impractical. However, if there is a sidewalk to parallel, or a hedgerow to snake through, or any combination of similar structures, wired could be the most robust and convenient solution.

  15. PS if you are going to try powerline, the results can vary widely from receptacle to receptacle (and the 500 Mbps number is a fiction). By trial and error I was able to find a pair of receptacles that gave adequate speed – sometimes receptacles that were only a couple of feet apart gave widely varying results. I think this has a lot to do with which one of the 120V legs the receptacles are connected to.

    In US domestic power service, you have two 120V “hot” lines. 1/2 of the breakers in your house are on one leg and half on the other (and the 240V appliances are connected across both). Theoretically powerline internet will work across the legs but in practice it works better if both receptacles are on the same leg.

Comments are closed.