Router with high gain antennae VS router + repeater

[fizikz]

Which is preferable? One good router with 1 or 2 high gain (5dBi, 7dBi, etc) antennae or the same router with 2dBi antennae and another router in repeater mode?

Also, on a 2 antenna router, is it ok to have one 2dBi and one, say, 5dBi antenna?

I would like to increase the wireless coverage a little in my house and I'm wondering about the best way to do it.

 

[thiggins]

The Best Way To Get Whole House Wireless Coverage



If you upgrade antennas, use two of the same gain.

 

[fizikz]

Thanks for the article. I had previously wondered about the impact on throughput when using a wireless repeater. Given that running an ethernet cable is not practical, and there is no cable in the house for MoCA, and the powerline option requires the purchase of more (somewhat expensive) equipment, it seems to me that getting 2 higher gain antennas would be cheaper and work better for my needs.

Perhaps a 7-9 dBi antenna should be good. I have seen some 12 dBi ones advertized, but I don't want it too directional. The router is in the corner of the building in the basement, and needs to provide coverage in the shape of one eighth of a sphere.

 

[stevech]

12dBi gain with an omni-directional.. means you'll have a 3 ft long expensive antenna. It'll have a flat-doughnut coverage pattern - OK for a 1 story home. One that's truly 12dBi (there are a lot of phony antenna gain claims) - will cost much more than a power-line IP connection - these are $75 or so a pair.

 

[fizikz]

I have seen several websites and local buy/sell sites selling high gain antennas in the 5-12 dBi range for around $10 and I wondered how to tell if they're genuine, and if they would make a difference. Ideally I would want to try before buying something.

To test what a directional 9 dBi antenna might do, I tried making this parabolic reflector: http://www.freeantennas.com/projects/template2/index.html

As far as I can tell, it made absolutely no difference. I used InSSIDer on a laptop to measure the signal strength in one corner (on the same floor) of the house, once with the reflector mounted on the router antenna aimed towards that point, and again with the reflector pointed in the opposite direction. In both cases, InSSIDer reported the same signal strength.

 

[stevech]

I do RF engineering by profession... home made antennas rarely work well. Antenna design is an art, and tiny dimensional changes, at this 2.4GHz frequency, make big changes in antenna performance gain.

Antennas can be directional or the horizontal plane or vertical plane. A gain omni is directional on vertical and not horizontal. Vice versa, an antenna like a yagi (sort of like a TV antenna) is directional mostly on the horizontal plane. Sector antennas, like you see on cell towers, exist for 2.4GHz and are directional on both H and V.

Parabolics are normally used outdoors. A "real" one for 2.4GHz is about 12-24 in. in diameter. The little 4 or 6 in. ones you see for WiFi are silly/phony like a pet rock, to an engineer like me.

The vendor below has a large variety of antennas. I really encourage you to buy, don't make, and don't buy one of these fake consumer antennas. A proper antenna comes with a gain patter plot chart, as you'll see on this web site.
Also, don't bother with an omni with less than 6dBi of gain - it's insignificant.

http://www.l-com.com/home.aspx
and http://www.l-com.com/category.aspx?id=2073

A good directional antenna for indoor use is a patch/panel such as in the above.

 

[fizikz]

Thanks for sharing your expertise. I find it interesting that the rubber duck 9dBi antenna http://www.l-com.com/item.aspx?id=22067 has more vertical coverage than the 7dBi one http://www.l-com.com/item.aspx?id=22063. The only problem is that I'm not sure this vendor ships to Canada, since their shipping calculator only takes US ZIP codes.

I have also found 5, 8, and 9dBi antennas made by (branded as) TP-Link, Asus, D-link etc that have radiation pattern plots or specifications and they tend to be similar or cheaper in price. Are they reliable?

Eg. http://www.tp-link.com/en/products/details/?model=TL-ANT2408CL#spec

 

[stevech]

Here's why 2, 3, 4, 5 dB difference in antenna gain has only a tiny affect on data rate and performance...

The through-the-air attenuation (line of sight), at 2.4GHz, is many tens of dB for common distances in WiFi, and more yet, with walls/floors in the path. Commonly 60dB or more. So improving antenna gain by 2 or 3 dB is small in comparison to the path loss.

If the received signal is either too weak for reliable data transfers, or is too weak for the desired data rate, it takes a lot of antenna gain to make up the shortcoming.

Higher transmitter power from the WiFi router DOES NOT cure the problem of a weak transmitted signal FROM CLIENT TO ROUTER.

The enclosed spreadsheet for path loss might be helpful. Note the path loss formula in the spreadsheet.
Do it twice: once for to-client and again for from-client direction, using the transmitter power and antenna gain for each end device. Typical 802.11g/n transmitter power is 30mW RMS. This can be 100mW with the same hardware, but only at the lowest data rates (this is not well understood in consumer-land).

 

[stevech]

I have also found 5, 8, and 9dBi antennas made by (branded as) TP-Link, Asus, D-link etc that have radiation pattern plots or specifications and they tend to be similar or cheaper in price. Are they reliable?

(Opinion-alert) Most consumer-branded antennas, I think, come from some mom-and-pop shop in China so it's a gamble. If you buy from a distributor for MAXRAD or Mobile-Mark or some such old school US vendor, though they may buy from China, Inc., I think they choose the fab shops carefully. It's curious that these kinds of antennas are just rubber and cheap wire - a truly good antenna, patch/panel/yagi, has lots of RF engineering and anechoic chamber tests in R&D costs and that raises the price. As ever, you usually get what you pay for, if you're an informed buyer.

 

[fizikz]

Given the large path losses, I wonder if a higher gain antenna makes any real world difference at all. I'm thinking it might be good to get some antenna from a store with a good return policy to see first hand what difference it makes in my specific case. Reviews of high gain antennas are all over the place, from great results to no difference. The trouble is, it can be impossible to tell whether it's because the product is no good, or whether the reviewer did not set it up properly.

If the rubber duck omnidirectional antennas are just rubber and wire, I would hope it should be easy to make and that they should be more or less the same regardless of brand or manufacturer. It reminds me of this link I found on yet another home-made solution: http://www.instructables.com/id/WIFI-Antenna-Hack!/?ALLSTEPS

I don't think the "engineering" there is going to be very precise. In the end I suppose I'll just need to actually try some antenna(s) and see how it turns out.

 

[stevech]

yeah, that's why I say that you need an antenna type/size that is 10dBi or more better to make a real difference.

 

[sfx2000]

(Opinion-alert) Most consumer-branded antennas, I think, come from some mom-and-pop shop in China so it's a gamble. If you buy from a distributor for MAXRAD or Mobile-Mark or some such old school US vendor, though they may buy from China, Inc., I think they choose the fab shops carefully. It's curious that these kinds of antennas are just rubber and cheap wire - a truly good antenna, patch/panel/yagi, has lots of RF engineering and anechoic chamber tests in R&D costs and that raises the price. As ever, you usually get what you pay for, if you're an informed buyer.

This is an area where Stevech and I strongly agree... most add-on antenna's are not properly matched to the RF front-end's on the AP (or client) so the value is somewhat limited.

FWIW - I've stared at enough Smith Charts to know this, even though I don't consider myself an RF guy (I'm mostly MAC <not apple> experienced, and most recently in the application layer for multiple access technologies).

 

[stevech]

And I believe that most consumer antennas are marketed with far more gain (in dBi) than is true. In antennas, size matters, even at 2.4GHz.

 

[lifespeed]

use a wireless AP with proper antennae design

I recommend the use of a proper wireless access point, ceiling mounted, with properly-designed directional antennae. The radiation pattern is optimized for this location, each of two antennae firing in opposite directions with generous coverage to the sides.

I use the low-power EAP9550, located in the center of a single-story house with excellent results.

Their latest and greatest EAP350 is high-power and gigabit ethernet interface. While this will only increase speeds/range in the downstream direction, for those applications that are primarily downloading there is a modest benefit.

I use my WLAN for VoIP phone calls from Android smartphone, so asymmetric speeds/range are not of much benefit.

Lastly, these are Power Over Ethernet devices, making installation no more difficult than running a single ethernet cable. If you can't be bothered to pull an ethernet cable you deserve what you get for wifi coverage

 

[stevech]

Wish it was also a smoke detector!

 

[lifespeed]

Wish it was also a smoke detector!

You've got to admit it is a nice form factor for an ideal location. Not to mention above-average hardware and firmware performance.

I think Engenius and Draytek are perhaps two of the best kept networking secrets, at least in the USA.

 

[stevech]

I used one Engenius WiFi outdoor bridge. It worked well and was low cost on newegg.

They are owned by a huge chip vendor in Asia who has a large WiFi market share.

 

[stevech]

yes, the laws of physics tell us the attenuation versus frequency versus distance, in line of sight conditions. Then add attenuation from walls, floors, and some other factors. Yes, it's normally 60-100dB, depending on many things.

So diddling with 2 or 3dB difference in antenna gain makes little difference in contrast to the attenuation in the RF propagation path.

Also, it's just 3dB from a 30mW WiFi transmitter to 60mW, and on to 100mW is not much either.
If that increased power is just on one end of the link (say, the WiFi router), the other end (laptop/handheld) is left with a great disadvantage called an imbalanced link (you hear me well but I can't hear you at all or just barely!)

As said above, the cure for poor coverage is an access point in the weak signal area.

 

[tipstir]

To me their not worth it. I know I have two RE1000V1 around here in the areas where the signal is weak. I use an app called WiFi Analyzer for Android free one it sees all 3 APN (access point networks) test each one to see how well they do and how far the main overlap SSID doing. Cisco idea is to have RE1000V1 anywhere and they're not cheap at $80 a pop! In call they keep the signal steady.

 

[TonyH]

Thanks for the article. I had previously wondered about the impact on throughput when using a wireless repeater. Given that running an ethernet cable is not practical, and there is no cable in the house for MoCA, and the powerline option requires the purchase of more (somewhat expensive) equipment, it seems to me that getting 2 higher gain antennas would be cheaper and work better for my needs.

Perhaps a 7-9 dBi antenna should be good. I have seen some 12 dBi ones advertized, but I don't want it too directional. The router is in the corner of the building in the basement, and needs to provide coverage in the shape of one eighth of a sphere.

Hi,
Gain and coverage, it is complex issue. The higher gain, the pointed radiation pattern(vert. & horiz. plane) Also radiation angle moves up and down depending on the gain. If you have a wide area to cover, thought about
some thing like Ubiquiti PowerAP N? It can put out upto 1000mW. You know that on paper to increase signal by 3db you have to double the power output.