Lessons Learned from Antenna Experiments

[Just Checking]

[Ed. Note: Moved to its own thread to increase visibility]

I spent several months researching and trying different antenna configurations for my RT-AC66 routers. I went to radio sites and posted on their antenna design sections. I bugged people on this site till they were sick of me.

The things I found out were:
1. You can make significant improvements to the 2.4GHz band in terms of longer range signal strength and quality (both sending and receiving) by using relatively cheap high gain (9dBm) antennas. I liked the TP-Link 8dBm antennas until I found some cheap ($11 for 3 antennas) on Amazon which had metal ferules that didn't slip or break and which maintained orientation when set off-vertical. Without changing the output power of the radio on the router I was able to get stable WiFi signals downlink and uplink at more than 20 meters away through 4 plaster walls filled with romex wiring, copper water pipes, cast Iron waste lines, and quite a bit of low voltage wiring. I am able to get signal out through a metal sheathing and reasonable signal strength 50 meters outside the building (enough to stream video).

2. There is a thing called "rotational polarization" which means that when a signal is transmitted through a solid object (like a wall), or bounced off a surface, the signal orientation can change such that an antenna oriented in one direction (vertical for example) cannot pick up the signal as effectively as two oriented in different directions. (This is a major oversimplification so people who have the background and experience to understand all that is behind this statement are free to offer clarification if you believe it contributes to the discussion.) I found that having two omni-directional antennas oriented 90 degrees from each other (example one antenna vertical and one antenna horizontal; or one antenna at -45 degrees from vertical and another antenna at +45 degrees from vertical) allowed me to have the strongest signal at the receiver of the router WiFi and the client. It also helped to get around the multi-floor concern of a omni-directional antenna oriented vertically.

3. I found no commercially available antennas which would improve the signal strength of both the 2.4GHz band and the 5GHz band simultaneously.

4. I found no commercially available antenna which would improve the signal strength or quality of the 5GHz band better than the stock antenna on the RT-AC66 that came with the router.

5. On the RT-AC66 router, having all three antennas oriented vertically seemed to cause "interference" and decreased sensitivity. I eventually ended up with an outside antenna (2.4GHz high gain antenna) oriented vertically, the middle antenna (Stock RT-AC66 antenna) oriented at a 45 degree angle away from the vertical antenna, and the third antenna (2.4GHz high gain antenna) oriented horizontally (45 degrees away from the center antenna). The 2.4GHz high gain antennas I use are 38cm (15 inches) long. I didn't notice the same signal interference when I was just using the stock antennas. Most observers will look at the antenna configuration and think it is "ugly", "weird", "asymmetrical", "ungainly", or "unattractive". You will have to make the judgement yourself of "function" or "form".

6. If you want to concentrate on the the 5GHz Band, use two of the stock RT-AC66 antennas with the center antenna being a 2.4GHz high gain antenna. The smaller stock antennas should be oriented at 45 degree angle from the center vertical antenna.

7. It is possible to make a 5GHz antenna array that is more sensitive and better signal strength than the stock antennas. My final antenna was very large with multiple smaller antennas oriented and separated in the array. The gains I got from this were not remarkable. It was not worth it IMO so I went back to the stock antennas for the 5GHz band.

8. The actual signal strength/sensitivity varied from antenna post to post on the RT-AC-66 routers that I tried. I believe that this is because of the different wire length from the actual radio on the PCB in the router to the antenna stub. There may be other reasons that an electrical engineer with radio background can discern. All I know is that on the two RT-AC66 routers that I have, the right antenna (looked at from the back side) has the strongest signal.

9. Changing antennas and configurations will not increase your maximum data transmission rate over what you achieve when you put the transmitter and receiver close together. What the antennas and configuration will do is increase the distance that you can separate the transmitter and receiver while still maintaining a high data transmission rate. True data transmission has error corrections and connection delays. The better the signal quality at a distance (Higher signal to noise ratio), the higher the data transmission rate that can be maintained.

10. Surprisingly, changing the orientation of the transmitter and receiver in the same plane (moving the client device to a different position on the same floor of the home wrt the orientation of the router) does make a measurable difference, even with omni-directional antennas. The difference is enough to cause connect or disconnect issues at far distances. The better antennas extended this distance in the 2.4GHz band such that I never had disconnect problems but the 5GHz is still problematical at distances over 20 meters from the transmitter or where there are significant signal attenuators (brick/concrete/stone walls, metal sheathing, Large Appliances, or other RF transmitters).

11. Antenna configuration is going to depend on the router you use. If you have a router with one antenna, you are pretty much stuck with improving the 2.4GHz band with, at best, no change to the 5GHz band. If you have a 2 antenna router, you can put a 2.4GHz high gain antenna on one post, and a 5GHz antenna on the other. Three, four, or Six Antenna configurations (if you can change out the antennas) are going to be a trial and error configuration like I did. It just takes time to do the testing.

I see that some of the new routers like the Asus RT-AC87 have different antennas than the RT-AC66. I'd like to try those to determine if they have a better design for 5GHz than the stock RT-AC66 antennas. Currently, I get 50% of the signal strength on the 5GHz band that I get on the 2.4GHz band and that is with the power setting of 140mW for 5GHz vs. 110mW for the 2.4GHz band.

You may ask "How did he do these measurements?". I did not use any other test chamber (other than my own home). I did not use specialized signal processing equipment (I don't have this at my home). I used two RT-AC66 routers with one being the transmitter and the other being the receiver. I used the "scan feature" available in the Asus-Merlin F.W. and InSSider software on a HP Elitebook 8460p laptop with an Intel 7260 WiFi 2X2 card to measure signal strength. I put different antennas in different numbers, and orientations on one router in different configurations and tested signal strength. I connected a computer to each router so I could measure both the sending and receiving signal strength. I left the routers, and PC's in standard positions and orientations so I could reproduce the results. Finally, I tested "real world" results by sending a "standard" 3GB MP4 video file back and forth between PC's on the routers to determine if there were changes in the wireless transmission rates with different configurations. I did measurements at two standardized distances. The first at 2 meters distance where the transmitter and receiver were on the same plane (same height off the floor) with no obstructions in between. The second where the transmitter and receiver were approximately 20 meters apart on the same plane but with 4 walls and various pieces of furniture between them. I also did some testing where the transmitter and receiver were in different planes (on different floors of the building) and where the transmitter and receiver were oriented at different angle to one another (moving the receiver to a different position 90 degrees from the face of the transmitting router.)

 

[Just Checking]

I have to say that much of my router antenna testing was prompted and inspired by THiggins articles which tested different wireless radios on different routers. Thank you Mr. Higgins. I attempted to see if I could introduce some of the same scientific approach and rigour to antenna configuration that he puts into his testing without the use of test equipment. I also wanted to use "real world" situations where I am actually sending and receiving files I would normally do in locations that are encountered in most homes or business spaces.

I have to admit that most of the changes I made caused only subtle differences in signal strength. The absolute biggest bang for the buck still comes down to a high gain antenna for the 2.4GHz band but usually at the sacrifice of signal strength in the 5GHz band.

If anyone can recommend a (relatively) small antenna, or array, that will improve the 5GHz signal quality, please post it here. I have to admit that I am stumped by that one.

 

[stevech]

One antenna good for both 2.4 and 5GHz, and with lots of gain in 5GHz?
Here's who I buy from, professionally. But I've not expected the above.
http://www.l-com.com/wireless-antenna?cmp=LM1

 

[azazel1024]

Hmmm, my experience has been different.

With my AC1750, I found higher 5GHz speeds with 5dBi dual band antennas over the 3dBi that shipped with it, about 10-15% higher absolute max. At range speeds were from 5-20% higher. 2.4GHz speeds were not higher at max, but at range they were 5-30% faster. The 5dBi antennas were pulled from a pair of WDR3600's (so TP-Link antennas). Inssider at least reports higher 2.4 and 5GHz signal strengths at all tested locations (5 test locations) between 1-5dBm.

With my N600 router I found using a pair of 7dBi antennas from superpowersupply (stock where the 5dBi antennas I sacraficed for my AC1750) showed lower 5GHz signal strength at the 4 test locations I used, higher 2.4GHz signal strength, but 2.4GHz showed a 5-10% increase in performance at all but the closet location (no change there) and 5GHz showed a 10-20% increase in performance at all but the closest location (no change there). So on that, just note, InSSIDer claimed closer 5GHz signal strength at most of the close locations. At the furthest location it showed a small gain, but PERFORMANCE on the other hand and link rates (which I mostly don't care about, but I was checking) WERE higher at all but the closet location (which was same room, about 8ft away).

Spent a lot of hours between the two testing.

I am sure a lot has to do with AP location, structure layout and construction, radio/radio chains, firmware, etc, etc, etc.

The orientation of my antennas is straight up and down on the N600, just like I was using the 5dBi antennas (I played with orientation a little and found only decreases in performance). On the AC1750 they are canted to the side about 15 degrees or so as I need to cover both my basement where it is, as well as the room directly above the router. Overall this provides the best performance in all locations, not just averaged (IE it is faster in the basement than straight up and down, same with in the room above). Changing individual antennas around found only drops in performance.

I didn't try mixing antennas.

 

[ChristineBCW]

Was there any mixing-matching antennas on the receiver unit (such a PCI card with detachables)?

 

[TonyH]

Hi,
I am glad to see there are some who learn things by doing vs. only who wants quick fix answers.
Knowledge gained by theory and experience is real thing which will
last life time. Without both it is useless.

 

[Just Checking]

One antenna good for both 2.4 and 5GHz, and with lots of gain in 5GHz?
Here's who I buy from, professionally. But I've not expected the above.
http://www.l-com.com/wireless-antenna?cmp=LM1

Thanks for posting the website link. Someone on a ham radio site recommended it to me. I didn't expect my post to be a sticky or to be referenced when I wrote it so I didn't include any charts or links.

I did purchase a 5.4GHz omni-directional antenna (among others) to test. Here's the link:
http://www.l-com.com/wireless-antenna-54-ghz-12-dbi-omnidirectional-wireless-lan-antenna

I didn't get any better results than with the stock RT-AC66 antenna. More research and a long time later, I came to realize that, since My routers were on channel 149 and 161, I should have been using a 5.8GHz antenna like this one to more closely match the frequency put out by the router radio on those bands:
http://www.l-com.com/wireless-antenna-58-ghz-12-dbi-professional-omnidirectional-antenna

A key learning that I didn't mention in my previous post is that the 5GHz band is much more sensitive to antenna matching than the 2.4GHz band. All you radio operators will be saying "well, DUH!!!". Hindsight is always 20:20.

If your routers are broadcasting on the 36 to 48 channel bands, use the 5.4GHz antennas. If your routers are broadcasting on the 149 to 165 channel bands, use the 5.8GHz antennas. The most precise match would be channel 161 or 165 (if your FW allows that setting) and the 5.8GHz antenna.

I believe that many of the discrepancies in peoples observations with even the same router with different antennas is due to the fact that their routers are broacasting on different channels that another tester.

Another lesson to learn is that you should not set your router to "Auto" channel selection. That feature is to allow your router to sense interference signals from other devices and attempt to move to a clearer band to give maximum transmission rates. What ends up happening is that you get poorer signal quality from band hopping and antenna mismatch. I do a site survey to find competing radios and decide if I want to attempt to avoid those. I can only detect one other 5GHz radio in my area other than my own. I set my router radios at 149 and 161 and crank the power output to max for a day which causes the competing router to kick down to the lower band. I then reset my router radios to the power output I feel provides the longest life (I try to keep the router temps below 60 Celcius) and good reliable connections to the clients.

I tried a big antenna like this one:
http://www.l-com.com/wireless-antenna-24-49-58-ghz-dual-feed-90-degree-sector-panel-antenna
It was expensive, very obtrusive, and only had a narrow broadcast angularity (90 degrees). Yes, it did improve the signal quality within the cone of the array. I needed 180 degree coverage at a minimum for my application.

As I alluded to in my post, I ended up combining a couple 90 degree antennas to make something like this:
http://www.l-com.com/wireless-anten...ector-array-w-4-single-fed-90-degree-antennas
This one is almost US$1000. Mine was about US$400 (I returned the antennas). It did provide increased signal strength, but did not give significantly higher data transmission rates at a distance over the stock antennas which were included with the routers. I decided to abandon that approach because A.) The cost; B.) The obtrusiveness of the antennas and wiring; C.) Objections from tenants in my building who thought they were being irradiated and going to get cancer (says something about the ignorance of the general population).

If I had a large campus of buildings that I wanted to create a outdoor WiFi Hotspot for, I would use an antenna like this one and set up my router radios on the 161 channel:
http://www.l-com.com/wireless-anten...array-w-4-individually-fed-90-degree-antennas

I saw the following antennas on some of the newest AC routers and I will purchase a couple of these to see if they make any difference. The specs say they work for the 2.4 to 5.9GHz frequencies.
http://www.l-com.com/wireless-anten...and-rubber-duck-antenna-rp-sma-plug-connector

Unfortunately, most of the antenna suppliers are not like Amazon.com. It is difficult to purchase stuff just to try it out one time. I ended up having a whole bunch of antennas just lying around collecting dust. I try to pawn them off on family, friends (cause enemies will not take them), and anyone else who even mildly expresses an interest but it is hard to justify others using stuff that I believe is just a placebo.

 

[L&LD]

Could you elaborate on your point 5 in the first post?

Maybe a picture or diagram to show exactly how you set the antenna up? I think I know what you mean, but it would be great to know for certain.

Great post btw.

 

[Just Checking]

For those of you out there who will ask "Why not try something other than a omni-directional antenna"; here are a few others that I tried.

Flat plate type
http://www.l-com.com/wireless-anten...losure-antenna-with-rp-sma-plug-pigtail-cable
I thought this was going to do it for me. I was wrong. It was too directional - even though the specs say it is wide angle. The one I tried had something like a 60 degree angle of sensitivity with severe signal fall off outside of that. Vertical angularity was about the same. If you have an array of 12 of these in a spherical shape (Make a "modern" sculpture hanging from the ceiling in the hallway?) you might get the coverage you want.

Dome type
http://www.l-com.com/wireless-anten...y-ceiling-antenna-11in-rp-sma-plug-connectors
If you are willing to mount these on the ceiling in every room and also in the hallways, you can create excellent WiFi coverage. You have to have either an AP to go along with the antennas. That's exactly what these are for in a commercial application. I don't have false ceilings with wire chases in any of the buildings for which I support networks so they didn't work for me. Individually, these antennas broadcast a pretty uniform hemisphere of signal. Perfect for a single square or round room coverage. I would use multiple antennas for longer rectangular rooms or, if I wanted very high signal strength in a cluttered room that was more than 10 meters in any one dimension.

Yagi type
http://www.l-com.com/wireless-antenna-58-ghz-165-dbi-yagi-antenna-10in-n-female-connector
What can I say, I actually made one myself out of a pringles can and a cheap antenna I took apart. It does work. I was able to get a signal that I could do web surfing from a unsecured college buildng more than 200 meters away. It would be ok for setting up a wireless connection between an outbuilding (detached garage) and your primary building (home) if you wanted to do low bandwidth applications. Don't expect to be transferring files at 100+Mbps. I did stream music over a connection from 50 meters away. It really helps if the receiver is a wide parabolic antenna to capture the signal. Even a 30 degree cone spreads out quickly and signal strength drops as the square of the distance. This is what you need as a receiver:
http://www.l-com.com/wireless-antenna-49-58-ghz-27-dbi-lightweight-die-cast-grid-antenna

Parabolic Dish Antenna
http://www.l-com.com/wireless-antenna-51-ghz-to-58-ghz-31-dbi-broadband-parabolic-dish-antenna
Again, I made one of these out of a discarded parabolic dish from a satellite TV receiver and an antenna from a router. It worked for sending out low angularity signals. Good for transmissions between two separate buildings, terrible for any application where clients are spread out.

In other posts, I have read of neighbors who want to share their internet connection between their homes. My experience tells me to "buy a long Cat6 ethernet cable and string it between". If you have to use wireless, this is what to use. If you use tuned parabolic reflectors on both sides like these
http://www.l-com.com/wireless-antenna-5-ghz-dual-polarity-solid-parabolic-dish-antennas
You might get 150Mbps - maybe. Don't expect much if it is snowing or raining heavy.

Anyway. Have fun trying all this yourself. In my buildings I have the fastest wireless networks of any that I have tested but I now am in the process of stringing a lot more CAT6 ethernet cables than I originally thought. I bought some managed switches with POE and I am learning how to do port agglomeration. This wireless stuff was a great concept but if you really have a lot of data to transfer, wire is the only way to go.

 

[Just Checking]

Was there any mixing-matching antennas on the receiver unit (such a PCI card with detachables)?

To do my testing, I kept the "stock" router antennas as a constant to test the effect of changing the transmitter antennas.

When I got into doing download/upload transmission rates I did a few of the "best" antenna configurations on both the sending and receiving routers.

That was to verify that there was nothing different about the radios in each of the routers. I was suprized at how close the two RT-AC66 routers performed. Both of mine were RT-AC66R units that I bought "refurbished" from Newegg.com. I bought them several months apart.

I did do some verification of signal strength using an HP Elitebook 8460p laptop and the InSSider software to get measurements of dBm. The position of the antennas in the laptop do make a difference in the signal strength measured so I had to be very careful about orientation of the laptop screen which houses the laptop antennas.

I also did an evaluation with a couple of AC adapters which plug into the USB port. That is a whole other story.

 

[Just Checking]

Could you elaborate on your point 5 in the first post?

Maybe a picture or diagram to show exactly how you set the antenna up? I think I know what you mean, but it would be great to know for certain.

Great post btw.

l
l / ___

A very crude representation of the antenna configuration. If I remember, I'll post a picture when I'm able.

 

[Just Checking]

Hmmm, my experience has been different.

With my AC1750, I found higher 5GHz speeds with 5dBi dual band antennas over the 3dBi that shipped with it, about 10-15% higher absolute max. At range speeds were from 5-20% higher. 2.4GHz speeds were not higher at max, but at range they were 5-30% faster. The 5dBi antennas were pulled from a pair of WDR3600's (so TP-Link antennas). Inssider at least reports higher 2.4 and 5GHz signal strengths at all tested locations (5 test locations) between 1-5dBm.

With my N600 router I found using a pair of 7dBi antennas from superpowersupply (stock where the 5dBi antennas I sacraficed for my AC1750) showed lower 5GHz signal strength at the 4 test locations I used, higher 2.4GHz signal strength, but 2.4GHz showed a 5-10% increase in performance at all but the closet location (no change there) and 5GHz showed a 10-20% increase in performance at all but the closest location (no change there). So on that, just note, InSSIDer claimed closer 5GHz signal strength at most of the close locations. At the furthest location it showed a small gain, but PERFORMANCE on the other hand and link rates (which I mostly don't care about, but I was checking) WERE higher at all but the closet location (which was same room, about 8ft away).

Spent a lot of hours between the two testing.

I am sure a lot has to do with AP location, structure layout and construction, radio/radio chains, firmware, etc, etc, etc.

The orientation of my antennas is straight up and down on the N600, just like I was using the 5dBi antennas (I played with orientation a little and found only decreases in performance). On the AC1750 they are canted to the side about 15 degrees or so as I need to cover both my basement where it is, as well as the room directly above the router. Overall this provides the best performance in all locations, not just averaged (IE it is faster in the basement than straight up and down, same with in the room above). Changing individual antennas around found only drops in performance.

I didn't try mixing antennas.

With stock antennas, I didn't see a lot of interference between them when I had them vertically oriented. When I started using the 38cm antennas is when I started seeing a reduction in signal strength. Especially on the 5GHz frequency band.

Again, radio people will say "DUH" since they are always concerned about interference in their antennas and do quite a few things to minimize or correct for it. They have the ability to "tune" their transmitters and receivers to compensate for changes in the antennas also. I don't have the ability to do any of that with the router radios.

I did not do a lot of testing in the vertical dimension. The building where I was doing testing is an apartment building. The best I could do was test signal strength in the hallway on different floors. My tenants who utilize the wireless broadcast signal are happy with it but they are not computer savy power users (If they can get their devices to connect and stream media content without hangups, they are happy). I am the one who does massive file transfers between clients and servers and I am the most concerned about maximizing my transfer speeds.

In order to try to maximize the signal strength and quality of both the 2.4GHz and 5GHz bands, I had to use different antennas.

 

[L&LD]

I was at a customer location when I read the post with the 'crude' diagram.

They have an RT-AC68U with the 376.47 Beta 1 firmware installed and I tested using my laptop with the Intel AC7260 wifi card.

With the antennae all vertical, the connection was strong but seemed noisy, as it would take websites several seconds to load (intermittently, not always).

With the antennae in a W formation as Asus recommends the connection was still noticeably laggy (compared to my RT-N66U, for example) but greatly improved over the all vertical orientation.

With the antennae in orientation shown in post 11 of this thread, the lag was gone and almost unbelievably, an Ookla Speedtest showed an improvement of almost 10% but most telling was how smooth the graph was. No dips, no dancing. Just shot up to full speed and kept it there for the duration of the transfer.


I'm a believer in this antennae setup!


Thank you again for this great thread.

 

[azazel1024]

Ahh, yes, with antennas that big I can see that being an issue if they are as close together as antennas typically seem to be on routers. AFAIK, the bigger the antenna, generally the more distance between them you want/need.

I am a little confused, what kind of setup are you trying to use this in? Point to point? Are you trying to cover a single story apartment? Two story? Multi-story?

It sounds like a multi-story apartment building, in which case, kind of duh on big omni's not working well. The bigger the omni, the more compressed the vertical coverage. You can compensate some by doing similar to what you have by angling the antennas, but it reduces peak speed at medium distances as the signal gain is going to be such that some of the spatial streams are going to be exceedingly weak and likely only have one strong spatial stream.

Also, (unless of course you know the gain figures for each band), a dual band antenna is generally only listing the 2.4GHz gain, the 5GHz gain is generally 2-3dBi higher (IE 5GHz band antennas are roughly half the size of a 2.4GHz band antenna for the same gain). So a dual band "5dBi atenna" is actually going to have more compressed vertical coverage on the 5GHz band than it does on the 2.4GHz band, exacerbating the issues of coverage in a multi-story dwelling.

In my case I am only concerned about coverage on a single story for my one AP, or in the case of my router, 1.25 stories (IE the basement and just 2 rooms immediately above the router), so higher gain (though nothing insane like a 12dBi+) antennas make sense, operating on different channels of course.

I haven't tried it, but anecdotally, I have heard lots and lots of issues about using very high gain omni-directional antennas, except in very edge cases.

As for high gain, one thing to keep in mind too, most likely a pre-manufacturered (so long as it was made right) high gain antenna (IE Yagi, parabolic, panel) is probably going to work better than one you make yourself. One of the big reasons why is (again, if made right) the resitance and wave length size are going to be tuned much better than you are likely to do yourself just making one from a pringles can (or other materials). Means both likely higher gain, but also cleaner signal.

I can't speak to high gains other than that as I haven't really played with them, but I can get >60Mbps off some 5dBi omni's at 50 meters distance to my laptop.

Also in my experience with rain (haven't tried it in the winter time), standing in the doorway to my shed about 30-35 meters from the antennas on my outdoor AP (the one with 5dBi omni's, regular old consumer router (TP-Link 841n)) in a rather heavy down pour on my phone I can get around 20Mbps down. That is compared to an effectively identical number with no rain at the same distance (it maxes at about 32Mbps when resonably close to one of my APs). I didn't happen to have my tablet or laptop to test an actual high throughput connection. I also haven't tried it on 5GHz (outdoor AP is 2.4GHz only).

I do know a number of people who have used 14dBi panel antennas on all in one wireless bridges who easily setup line of sight wireless bridges at >100Mbps at well over a quarter mile. I would imagine that the longer the link, the more impact it'll be by weather, but realistically, there just isn't all that much water in the way, even in a down pour or blizzard over modest distances (dozens to hundreds of meters) and water isn't that great an absorber of the radio frequencies we are talking about (peak is 22GHz, not that near 5.2-5.9 or 2.4GHz). An entire human bodies worth of water is only good to attenuate about 4-6dBm, which means only absorbing about 2/3-3/4 of the RF energy passing through us. That sounds like a lot, but that is a difference of roughly halving the range (or less) impact if a person is standing between wifi links. Over a few miles in a down pour I can see there being that much water in the way (yes I know diffraction might be an issue, not just absorbtion), dozens to hundreds of meters not really.

 

[stevech]

Fiddling with vertical rod (dipole) antenna orientation is the wrong focus. The changes this can make are almost insignificant compared to the end to end attenuations from walls/floors plus free-space (line of sight) attenuation numbers. The antenna orientation tweaks might change things by 3 or so dB. Perhaps 6dB if you reorient by 90 degrees.
But the path from A to B will encounter 50, 60, 70 dB of attenuation as above. So 3dB or so is tiny compared to the total path loss. It's like a tiny nudge on the volume control.

A good system needs lots of excess received signal strength (fade margin). In WiFi, which uses adaptive modulation - adapting speeds to current signal strength and error rates. But due to multi-path, moving either antenna a fraction of, or a few wavelengths, changes the average (over time) signal strength considerably.

Because of this, what you need is to decide on the minimally acceptable 802.11 burst bit rate (Mbps). Then lookup the receiver's required signal strength to achieve that rate. Take this and arrange path length, path obstructions, antenna type/gain, to get that signal strength plus, say, 10dB of margin (the more, the better).

This applies to BOTH directions: client to network access device (wifi router or AP/bridge) where most products don't display that, and as well the to-client direction, where client devices always display something, though it may be marketing-speak bars. If you're really in to this, you need a "sniffer", an independent device to receive and report signal strengths in both directions.

 

[Just Checking]

Ahh, yes, with antennas that big I can see that being an issue if they are as close together as antennas typically seem to be on routers. AFAIK, the bigger the antenna, generally the more distance between them you want/need.

I am a little confused, what kind of setup are you trying to use this in? Point to point? Are you trying to cover a single story apartment? Two story? Multi-story?

It sounds like a multi-story apartment building, in which case, kind of duh on big omni's not working well. The bigger the omni, the more compressed the vertical coverage. You can compensate some by doing similar to what you have by angling the antennas, but it reduces peak speed at medium distances as the signal gain is going to be such that some of the spatial streams are going to be exceedingly weak and likely only have one strong spatial stream.

Also, (unless of course you know the gain figures for each band), a dual band antenna is generally only listing the 2.4GHz gain, the 5GHz gain is generally 2-3dBi higher (IE 5GHz band antennas are roughly half the size of a 2.4GHz band antenna for the same gain). So a dual band "5dBi atenna" is actually going to have more compressed vertical coverage on the 5GHz band than it does on the 2.4GHz band, exacerbating the issues of coverage in a multi-story dwelling.

In my case I am only concerned about coverage on a single story for my one AP, or in the case of my router, 1.25 stories (IE the basement and just 2 rooms immediately above the router), so higher gain (though nothing insane like a 12dBi+) antennas make sense, operating on different channels of course.

I haven't tried it, but anecdotally, I have heard lots and lots of issues about using very high gain omni-directional antennas, except in very edge cases.

As for high gain, one thing to keep in mind too, most likely a pre-manufacturered (so long as it was made right) high gain antenna (IE Yagi, parabolic, panel) is probably going to work better than one you make yourself. One of the big reasons why is (again, if made right) the resitance and wave length size are going to be tuned much better than you are likely to do yourself just making one from a pringles can (or other materials). Means both likely higher gain, but also cleaner signal.

I can't speak to high gains other than that as I haven't really played with them, but I can get >60Mbps off some 5dBi omni's at 50 meters distance to my laptop.

Also in my experience with rain (haven't tried it in the winter time), standing in the doorway to my shed about 30-35 meters from the antennas on my outdoor AP (the one with 5dBi omni's, regular old consumer router (TP-Link 841n)) in a rather heavy down pour on my phone I can get around 20Mbps down. That is compared to an effectively identical number with no rain at the same distance (it maxes at about 32Mbps when resonably close to one of my APs). I didn't happen to have my tablet or laptop to test an actual high throughput connection. I also haven't tried it on 5GHz (outdoor AP is 2.4GHz only).

I do know a number of people who have used 14dBi panel antennas on all in one wireless bridges who easily setup line of sight wireless bridges at >100Mbps at well over a quarter mile. I would imagine that the longer the link, the more impact it'll be by weather, but realistically, there just isn't all that much water in the way, even in a down pour or blizzard over modest distances (dozens to hundreds of meters) and water isn't that great an absorber of the radio frequencies we are talking about (peak is 22GHz, not that near 5.2-5.9 or 2.4GHz). An entire human bodies worth of water is only good to attenuate about 4-6dBm, which means only absorbing about 2/3-3/4 of the RF energy passing through us. That sounds like a lot, but that is a difference of roughly halving the range (or less) impact if a person is standing between wifi links. Over a few miles in a down pour I can see there being that much water in the way (yes I know diffraction might be an issue, not just absorbtion), dozens to hundreds of meters not really.

I have a couple multi-family dwellings (apartments to those of you who are not in that business). I live in one of the units. My goal in doing these tests was to try to get the best signal quality I could to the tenants on all three floors from the minimum number of wireless repeaters or AP's. My buildings are over 110 years old so they didn't have electrical wiring when they were built, let alone low voltage communication lines. I have run some CAT6 ethernet cables but not enough to allow clients (tenants) to connect all their devices up to hard wire. Not to mention all the smart phones and tablets that don't have rj-45 ports anyway. Hence the desire on my part to try to maximize the signal from the routers wireless radios. I also did some tests to see if I could do a wireless communication between buildings that were approximately 50 meters apart (mostly as an experiment).

I agree that purpose built commercial antennas are far more capable than my home made antennas. My home made antennas were A). To reduce cost of testing the systems, B). To see if I could make something myself that worked, and C). For entertainment value.

Regarding degradation in performance in rain or snow, I am just relating my experiences which indicated data transmission slowdown and performance decreases with heavy rain or snow over a 50 meter distance with a parabolic antenna on each end.

I do acknowledge that there is a lot of commercial equipment available which does address the above issues. I also know that, with enough money and time, I can build a wired network and wireless network that would be the envy of any network professional in a fortune 500 company. I don't want to spend the financial resources to enable that. Besides - what else would I have to talk about in these forums?

 

[Just Checking]

Fiddling with vertical rod (dipole) antenna orientation is the wrong focus. The changes this can make are almost insignificant compared to the end to end attenuations from walls/floors plus free-space (line of sight) attenuation numbers. The antenna orientation tweaks might change things by 3 or so dB. Perhaps 6dB if you reorient by 90 degrees.
But the path from A to B will encounter 50, 60, 70 dB of attenuation as above. So 3dB or so is tiny compared to the total path loss. It's like a tiny nudge on the volume control.

A good system needs lots of excess received signal strength (fade margin). In WiFi, which uses adaptive modulation - adapting speeds to current signal strength and error rates. But due to multi-path, moving either antenna a fraction of, or a few wavelengths, changes the average (over time) signal strength considerably.

Because of this, what you need is to decide on the minimally acceptable 802.11 burst bit rate (Mbps). Then lookup the receiver's required signal strength to achieve that rate. Take this and arrange path length, path obstructions, antenna type/gain, to get that signal strength plus, say, 10dB of margin (the more, the better).

This applies to BOTH directions: client to network access device (wifi router or AP/bridge) where most products don't display that, and as well the to-client direction, where client devices always display something, though it may be marketing-speak bars. If you're really in to this, you need a "sniffer", an independent device to receive and report signal strengths in both directions.

Stevech, Thanks for the input. I did say in my posts that the changes I made were often subtle, but measurable. Because I did not use purpose built test equipment, I was often not able to effectively capture the actual gains seen by the clients in terms of the quality of their wireless connections. I can say that, after my antenna experiments, I received far fewer complaints about failed connections, hangups, and slow internet access speeds.

Stevech wrote
"Because of this, what you need is to decide on the minimally acceptable 802.11 burst bit rate (Mbps). Then lookup the receiver's required signal strength to achieve that rate. Take this and arrange path length, path obstructions, antenna type/gain, to get that signal strength plus, say, 10dB of margin (the more, the better)."

This would be calculable if there was one or two receivers in set locations. My application has the receivers moving around with obstructions also movable. Similarly, I have no control over the client devices antenna type, location, minimal acceptable burst bit rate, or frequency. My tenants have smart phones, X-Boxes, Smart HDTV's, wireless hotspots, routers, Laptops, tablets, tower PC's, and other electronic devices that I am not aware of - nor do I want to know. If you can provide me with some software which would allow me to model this system, I would be happy to try it out. However, I'm not sure if it would help me accomplish real world improvements.

I am not a networking professional. I do this as part of a service to my tenants and because I am naturally interested in this stuff. I also have a "day Job" as a research engineer which takes up a lot of my time. Unless the equipment to do a better signal analysis is less than a few US$100's and quite small/portable, I am not sure I could justify such a purchase. I can't really deduct that off as a business expense in my residential rental property business either, so it would be out of my own pocket.

 

[ChristineBCW]

Thanks for the discussions about antennas, too. Are you concerned about bandwidth for a shared WiFi connection? I think I would be. Do tenants have an alternative, like going to the ISP themselves? Or could your ISP offer you some "business class" account that might include cabling for at least every floor?

(Maybe this is too far afield for this one thread's subject.)

 

[misfiringsystem]

Just Checking,

Did you encounter any problems while using single band antennas (2.4 ghz) RT-AC66R, ie, problems or damages to the 5 ghz radio?

I'm interested on increasing the 2.4 ghz range on my N66U, and most, if not all places sell only single band antennas for 2.4 ghz.

I came across several forums that state using a single band antennas is a definite no on a dual band router. Reason, which I don't understand, is something about standing waves destroying the radio. I'm not sure if there's any truth behind that.

I was wondering was there any problems with your 5 ghz radio, or your router at all after testing?

Thanks

 

[ankydu]

I am having a R7000 with stock antennas. I was wondering is the netgear suggested \ | / a better setting or _|_ a better config?