Got a big property to cover? Try this! (Ubiquiti PowerAP-N mini review)

[Mangusta1969]

There have been many posts on this forum about range/distance problems and frustrations with the performance of 2.4GHz wifi devices on large residential properties. Some of the discussions have been theoretical and some have been based on actual performance results.

I thought I would post my initial real world experiences with the Ubiquiti PowerAP-N router/access point on my difficult to cover property. I have posted the following review on another technical forum, but I thought my observations would be of benefit to many forum members here, too:



I just received my new Ubiquiti PowerAP-N access point/router yesterday and got it up on line today. So far I am very impressed with its power, range and throughput, especially when compared to my almost new Cisco/Linksys E4200 wireless router.

Probably because of its much higher output power (28dbm) and increased receive sensitivity from two relatively high gain omni antennas, this Ubiquiti unit has approximately 3-4 times the range and 3-4 times the throughput of my E4200 on a large in-city property of about .8 acres. From the ground floor of an outbuilding that is about 70 feet to the north of the main house, the AP-N is providing 4-5Mbps of sustained throughput in all 22 rooms and three floors of coverage for an old Victorian mansion, penetrating a large number of exterior and interior walls and floors. By contrast, the E4200's 2.4GHz radio covered about 8 rooms and two floors, with much reduced speeds (.5-1.5Mbps) in some of those 8 rooms. My ISP source is a nominal 6Mbps DSL source from ATT.net. The best hard-wired performance I have ever received from my .att ISP has been 5.2Mbps from my Zyxel P660 DSL modem, with 4.9-5.1Mbps speeds being more typical.

In walking around the property with my Mac laptop today, the AP-N is covering almost all of the .8 acres with a good signal and two-way communications; there were two outdoor spots that had poor performance of around 1Mbps, probably due to interference from neighboring networks. My Mac's IStumbler application picked up 15 other neighborhood networks as I walked around the property, checking signal strengths and network performance at various interior, porch and outdoor locations. Some of these neighboring networks had periodically been giving my E4200 some performance fits when they were very active. So far, the AP-N's much greater signal strength appears to overwhelm these neighboring networks.

According to the IStumbler application on my Mac laptop, the AP-N's received signal strength inside the house is typically about -45 to -50dbm, with a -82 to -87dbm noise floor. My Mac's Airport signal strength meter is always showing 5 bars of received signal strength as I walk the interior of the 8500 +/- square foot house. The neighboring networks have about a -72dbm strength, with a similar noise floor of -82 to -87dbm, as measured the IStumbler application. The AP-N appears to have a 20+ dbm signal strength advantage over the E4200 on the transmit side and the two 6 dbi gain antennas also seem to help with reception of weaker wifi signals from my laptop, computer and two networked DVD players. My Mac's Airport signal strength meter is always showing 5 bars of received signal strength, as I walk around the house and exterior grounds. Probably due to the increased receive sensitivity of the two 6dbi gain antennas, my local noise floor has increased from -92dbi to around -85dbi; the increased receive sensitivity and transmit power seems to more than compensate for the increased noise floor.

If the AP-N proves reliable and its throughput performance holds up over time, it will have solved all of my large property wifi signal problems. It is a little tricky to setup compared to my previous wifi routers, but so far, I am very impressed. I will provide some more feedback as I gain experience with this unit.

 

[stevech]

Don't overlook that WiFi is a bi-directional medium.
You should use antenna gain to improve coverage.

Don't use high power in the access point/WiFi router - because that is one-way, it doesn't deal with the weak signal from the client devices. The High Power Amplifier for a rock band is the analogy.

A 12dBi gain omnidirectional (horizontally) antenna on the access point or WiFi router is the best thing for improving the range bi-directionally. This kind of antenna is about 4 ft. long and needs to be collcoated to avoid a coax cable's losses.

 

[Mangusta1969]

Stevech,

I think the basic PowerAP-N performance results and effective ranges that I posted speak for themselves. The unit has 1000 mw of transmit power (compared to the 50-80 mw transmit power of most other wireless routers) and it has also has twin 5-6dBi receive antennas to help with the reception of weak client signals at greater range. I think the reception "rule of thumb" is that a 3dBI gain in received signal will double the effective range.

Your point is well taken, for those trying to cover even bigger structures than mine (10,000 plus square feet) and larger acreages, higher gain receive antennas would be advisable. The PowerAP-N has detachable antennas with SMA connectors, so even high gain antennas could be fitted without much difficulty.

My new router is continuing to operate without any glitches, disconnects, or reboots. That certainly was not the case with my Cisco/Linksys E4200 or my older Netgear RangeMax. With the E4200, I had to disable the 5GHz RF radio, otherwise it would slowly lose range and throughput and required frequent reboots.

 

[Internetraver]

I would also add that the Power APN is less than $100.....

 

[stevech]

Stevech,

I think the basic PowerAP-N performance results and effective ranges that I posted speak for themselves. The unit has 1000 mw of transmit power (compared to the 50-80 mw transmit power of most other wireless routers) and it has also has twin 5-6dBi receive antennas to help with the reception of weak client signals at greater range. I think the reception "rule of thumb" is that a 3dBI gain in received signal will double the effective range.

Your point is well taken, for those trying to cover even bigger structures than mine (10,000 plus square feet) and larger acreages, higher gain receive antennas would be advisable. The PowerAP-N has detachable antennas with SMA connectors, so even high gain antennas could be fitted without much difficulty.

My new router is continuing to operate without any glitches, disconnects, or reboots. That certainly was not the case with my Cisco/Linksys E4200 or my older Netgear RangeMax. With the E4200, I had to disable the 5GHz RF radio, otherwise it would slowly lose range and throughput and required frequent reboots.

But to be meaningful, in the context of bi-directional WiFi, tell us about the client device(s) used to assess the PowerAP-N. Are they low power as are iPads and smart phones, or are they more laptops or desktops, usually with higher power, better antenna?

The WiFi AP/router could have 100 watts (illegally) but if the client device has the typical 10 to 30mW (0.03 watt) there is, er, a bit of an unbalanced situation.

Each 3dB of antenna gain or reduced path loss improves the received signal strength by 3dB. So a 3dB improvement in antenna gain helps, but 3dB is a tiny percentage of the path loss which in residential, is 80dB or so. The antennas with 10-15 dBi gain do make a difference, e.g., the 4 ft. long "poles" for 2.4GHz that have 12dBi and are 360 degree omni on the horizontal and about 10 degrees on the vertial. Or the "patch" antennas that are often 14dBi or so by forming a cone shaped beam about 60 degrees. I think the internal antennas of most 11n routers are about 0 to 2dBi gain. The typical small rod antenna on routers is about 2-3dBi gain. A bit longer one with 5 or 6dBi isn't very much compared to the path loss/link budget.

Let's not forget too that the increased antenna gain also increases the interference! This is a real problem for those 12dBi omnis when elevated/outside. The desired signal is improved but the amount of interference goes up sharply. It can be a net sum loss in some settings.

 

[Mangusta1969]

I was simply comparing apples to apples with regard to client devices and router wireless performance, let's not confuse the issue by commenting on theoretically reduced ranges with smaller WIFI devices, like WIFI capable cell phones.

My observations were intended to characterize the wireless performance of the Ubiquiti PowerAP-N with computer clients ONLY. I also compared the performance of the Ubiquiti router with the supposedly good 2.4GHz range (as reviewed on this forum) of the Cisco E4200. In my testing, the PowerAP-N has been a far superior performer WRT the Cisco in terms of effective range, throughput and stable wireless performance.

By just replacing my Cisco E4200 wireless router (which was operating only on the 2.4GHz band) with the Ubiquiti PowerAP-N wireless router in the exact same location, my effective range and performance throughput was substantially increased with my same 6 wireless computer clients. While the E4200 had difficulty in penetrating 3-4 house walls and operating at a distance of 100-160 feet from my client devices, the PowerAP-N has had no problem in penetrating 8-10 house walls and operating at distances of 200-300+ feet. Depending on interference from neighboring WIFI networks, the E4200 typically provided .5-1.5Mbps throughput, while the PowerAP-N typically delivers 4-5Mbps performance in the same locations/to the same clients.

Here are my client devices for all tests and routers:

Apple Macbook Pro laptop 2.4GHz CPU; 802.11 a/b/g/n Broadcom-based
Airport Extreme wireless card
Apple Airbook laptop 2.13GHz CPU; 802.11 a/b/g/n Broadcom-based
Airport Extreme wireless card
Apple Imac 2.4GHz CPU; 802.11 a/b/g/n Broadcom-based
Airport Extreme wireless card
LG BD590 networked DVD player (2.4GHz only)
LG BD690 networked DVD player (2.4/5GHz capable; operating on 2.4GHz
only
6 year old HP Laptop (wired connection only, as this unit does not support
(WPA2 security)
2 year old Toshiba laptop (unknown wireless chipset; operating on WPA2
security)

While Stevech's comments on increased sensitivity antennas potentially causing more noise floor interference are accurate, the increased receive sensitivity of the PowerAP-N in my location has not been a problem, despite around 15 neighboring WIFI networks. As previously reported, my Cisco E4200 appeared to have periodic interference problems, as evidenced by fluctuating throughput performance measurements. The PowerAP-N delivers very solid wireless performance of 4-5Mbps, depending only on the distance of my client device from the wireless router and the number of walls/floors the signal is required to penetrate for each location. My DSL ISP's maximum throughput speed is typically 5-5.2Mbps, when measured with a wired Macbook Pro, connecting directly to my Zyxel P660 DSL modem.

Perhaps others can comment on observed WIFI ranges and throughput with WIFI capable cell phones and the PowerAP-N...

 

[stevech]

As previously reported, my Cisco E4200 appeared to have periodic interference problems, as evidenced by fluctuating throughput performance measurements..

The performance, if this means throughput/speed, can degrade without interference simply due to the client OR router end changing air link bit rates (modulation order). This can happen quickly, several times a second, according to error rates and signal quality assessment that are vendor-specific (not given by the IEEE standard).

Often, when there's significant interference (i.e., competition for air time under CSMA/CA), you'll see an increase in average packet latency if you ping your own router (like 192.168.1.1 or whatever), from the client for a long time and do the math (or use a pingplotter).

 

[hawkeye62]

I have a Ubiquiti PowerAP N that I use as an AP located in a central location of my 3,400 square foot house. I have tested several other routers as AP in the same location and only one comes close to the PowerAP N's performance as measured by signal strength and throughput. (Engenius ECB3500.) The toughest location is in my computer room which is about 60 feet away through 5 interior walls including a tile wall shower and an oak desk and an oak cabinet door. NetSurveyor shows a signal strength of 56-60%, while most of the other routers showed a signal strength of mid to low 40%.

I just did a test using my Asus TF101 Android tablet. In the living room which is only about 20 feet from the PowerAP N and only an oak cabinet door as an obstruction, the TF101 wireless shows 4 bars solid on a 4 bar scale. TF101 reports a link speed of 26 Mbps and signal strength as excellent. Speedtest.net shows a download throughput of 22 Mbps. In the difficult location, the TF101 wireless shows 3-4 bars on the 4 bar scale. TF101 reports a link speed of 19 Mbps and signal strength as good to excellent. Speedtest.net shows a download throughput of 13 Mbps.

I call that performance outstanding!

Some of the other APs I tried include Asus RT-N13, Buffalo WRZ-HP-G300NH, Zyxel WAP3205 and the Dlink DIR-655.

Regards, Jim

 

[Mangusta1969]

The performance, if this means throughput/speed, can degrade without interference simply due to the client OR router end changing air link bit rates (modulation order). This can happen quickly, several times a second, according to error rates and signal quality assessment that are vendor-specific (not given by the IEEE standard).

Often, when there's significant interference (i.e., competition for air time under CSMA/CA), you'll see an increase in average packet latency if you ping your own router (like 192.168.1.1 or whatever), from the client for a long time and do the math (or use a pingplotter).

Back when I was using the E4200, highly variable pingtest results were one of the reasons that I decided to replace it with the PowerAP-N.

Here is this morning's pingtest result to my PowerAP-N:

--- 192.168.2.1 ping statistics ---
3189 packets transmitted, 3158 packets received, 0% packet loss
round-trip min/avg/max/stddev = 0.748/5.103/492.111/19.662 ms


By contrast, the E4200 typically had 2-4% packet losses and highly variable response times when using pingtest from a 160 foot distance/3-4 walls location. I had always suspected local interference to be a cause for the performance problems, but perhaps the E4200 had some link re-negotiation problems with my Mac laptops that I used for most of my testing. The Ubiquiti router doesn't exhibit any such problems when testing from the identical physical location/distance in my house.

The PowerAP-N operates much better with all of my computer-based clients, with typical ping response times of 2msec. The E4200 was much slower to respond, typically about 10-15msec and had some very long ping response times (1200-2000msec) during long term testing. That's in addition to much higher packet losses with the E4200.

Like Hawkeye62, I remain very impressed with the Ubiquiti router's outstanding wireless performance. Thanks for the input on the android tablet, too.

 

[stevech]

Of course, you can change channels and retry ping tests, if at that moment, you suspect high activity by neighbors' WiFi on or within 3 channel numbers of the number chosen in your router (20MHz mode).

When comparing WiFi router/AP product A and B, same client device, same locations, then the same channel number should be used, and same 20/40MHz mode of course. What you can't control is how active your neighbors' WiFi are at test-time (is the guy next door doing an FTP/YouTube now?). Nor non-WiFi interference causes in 2.4GHz (cordless phones, analog wireless video cameras, baby monitors, etc).

 

[Mangusta1969]

Of course, you can change channels and retry ping tests, if at that moment, you suspect high activity by neighbors' WiFi on or within 3 channel numbers of the number chosen in your router (20MHz mode).

When comparing WiFi router/AP product A and B, same client device, same locations, then the same channel number should be used, and same 20/40MHz mode of course. What you can't control is how active your neighbors' WiFi are at test-time (is the guy next door doing an FTP/YouTube now?). Nor non-WiFi interference causes in 2.4GHz (cordless phones, analog wireless video cameras, baby monitors, etc).

With respect to trying ping and throughput tests on channels 1, 6, and 11, I've already tried that with my E4200, with similar varying results. I also tried the same tests at 0300-0400 in the morning, in order to help eliminate the possibility of neighborhood Netflix and Vudu downloads, as well as big FTP jobs causing WIFI interference. Again, no change, and still somewhat inconsistent performance from the E4200.

Just changing the wireless router from the E4200 to the stronger and more sensitive PowerAP-N and keeping everything else the same (20 MHz bandwidth, same security settings, testing on channels 1, 6, and 11, locations unchanged, clients unchanged) was like a night and day difference.

When I watched the download graphic from www.speedtest.net, the PowerAP-N immediately jumped up to 5Mbps and it stayed there for the whole download test. With the E4200 on the same test, I would see a maximum wireless throughput of 4-5Mbps, but I also saw many "valleys" in the performance graphic as the router and client communications throughput did not remain constant throughout the test. I saw some improvement by testing in the off-hours, but the E4200 almost never maintained a constant throughput with this test. The PowerAP-N's throughput was almost always a constant 5Mbps, regardless of the time of day that I was conducting the test.

When I used the built-in Airview utility/poor man's spectrum analyzer of the PowerAP-N to look at network activity across the 2.4GHz RF band, I would see more neighboring network activity in the lower part of the band, but there was not a large amount of activity and the signal strengths seemed relatively low to me. In testing for best range and throughput by changing channels and walking around the house and property doing measurements, I actually achieved slightly better results by using channel 1 (although that had appeared to have more activity near it than channels 6 or 11).

Perhaps the PowerAP-N has a better RF front end design than my other two wireless routers and could reject adjacent channel activity better?? In the FM radio world, this would be measured as a better capture ratio of the desired signal to nearby signals at lower or higher frequencies. I can't explain why this would allow the PowerAP-N to ignore other wireless router signals that were operating on the same channel 1 frequency, except for the fact that these neighboring signals from other routers were 22-27dBm lower in strength, which is a huge difference. I suspect there was a similar difference in signal strength in competing signals from much weaker client devices located on neighborhood properties, but I was not able to measure the client signal strength.

Bottom line, I have tested the E4200 and the PowerAP-N extensively with my client devices and large distances/many house walls between the router and the clients, under test conditions as similar as I can make them. There is really no comparison--the PowerAP-N is a much stronger wireless performer for bi-directional WIFI communications at extended ranges. It also provides a more stable throughput level during almost all download tests, unless the E4200 is within about 40 feet of my Macbook Pro laptop. Within 40 feet or less and with one wall or no walls to interfere, then the wireless throughput of the two routers is similar and is equivalent to a wired connection to my DSL modem.

 

[stevech]

...When I watched the download graphic from www.speedtest.net, the PowerAP-N immediately jumped up to 5Mbps and it stayed there for the whole download test.

How did the upload graphic look?

 

[Mangusta1969]

How did the upload graphic look?

Good idea to test and report on! However, while www.speedtest.net provides a nice and very granular graphic of throughput vs time for its download test, it does not provide any such graphic display for its upload test.

My DSL service is spec'ed at .5Mbps upload and 6.0Mbps download. The PowerAP-N delivers those speeds very consistently (well, the download is actually capped at 4.9-5.2Mbps, but that is the same speed that I get when directly wired to my Zyxel P660 DSL modem). I do get a solid .5 or .6Mbps on the upload test with the PowerAP-N.

The E4200 was much less consistent on the download speeds, depending on my location in the large house; I think it was more consistent at .4-.5Mbps on its upload tests, though. Most of my testing was focused on download speeds, since my laptops and the wireless DVD players are generally receiving many more bytes than they are sending back to the router. The bytes sent back to the router are typically just acks for downloaded packets that were just received.

While the PowerAP-N has a graphic utility that shows throughput vs time for both transmit and receive in packets/sec on both the LAN and the Wireless LAN, its display is not very granular, so its not nearly as useful as the speedtest.net download throughput graphic. On its transmit/WLAN side (laptop to router), the PowerAP-N graphic quickly climbs to something slightly above .5Mbps and stays there for the short duration of the upload test.

 

[Mangusta1969]

Longer Term Reliability Report on PowerAP-N

My Ubiquiti PowerAP-N wireless 2.4 GHz router continues to work like a champ!

In almost 2 months of operations with the same 6 wireless and 1 wired clients (see my 6/18/11 post for a description of these clients), I have only rebooted the PowerAP-N one time.

I am not sure if that one reboot was the router's fault or a problem with my Mac Airbook, as I had rebooted the router wirelessly from my Airbook; i manually rebooted the Airbook at the same time. The only symptom that I had prior to the reboot was slow communications between the Airbook and the PowerAP-N. At the time of the reboot, I was in a hurry and needed to get a long document printed, reviewed and sent out via email. I'll be more careful with the next reboot, so the technical fingers can be pointed at the culprit device(s).

In the preceding 4 months of operations with the same clients on 2.4GHz, at the same locations, I typically had to reboot my Cisco E4200 several times per week, as its communications with several clients would slow to a crawl over time or it would not even respond to a client's manual ping request. The only difference than was the old HP client was operating wirelessly on a guest network protected only by WEP security; all other clients were operating wirelessly on the main network with WPA2 security.

Anybody want a slightly used E4200 for cheap (still under factory warranty)?

 

[ian2000]

I recently did some research into the best range performance wireless router and being a novice convinced myself, through the numerous reviews and articles including those on SNB, that the E4200 was the state of the art.

Well reading this thread has opened my eyes. What a fantastic experience with the Power AP. I think I might just rush out and buy this plus a Zyxel P660 DSL modem (or is there an even better modem solution - please note, I am with BT on PPPoA).

Anyway folks, many thanks indeed for sharing this info.

Best

Ian

 

[Mangusta1969]

E4200 vs PowerAP-N Effective Range

Long term PowerAP-N range on 2.4GHz with 4-5 Mbps connectivity to my PC and Mac clients is 2-3 times the demonstrated range of my January 2011 build E4200.

I just sold the "retired" E4200 on CraigsList today, too and will probably pick up a spare PowerAP-N while they are still available and under $100 here in the US.

 

[njweb]

Hi Mangusta1969,

How did you configure the firewall rules?
I read that it needs the user to manually set them and be familiar with iptables.
Sample settings and / or instructions would be greatly appreciated.
I am considering potentially giving this a shot (to compare to my E4200 and the Zyxel NBG5715 in terms of experimenting with purely extending the range, but without the use of a range extender).

PS: My E4200 has performed extremely well since March of this year, unlike your experience.
I never need to reboot it. My packet loss tests are fine (think ISP and line issues are likely causes of packet loss - I had some in May which the tech said originated outside my house somewhere at the street level).
My only desire was to test the extra range offered by the Ubiquiti to see if it might be worth looking into.
Although, as I write this, I was just thinking about the downside of the Ubiquiti, for my potential usage:
With my E4200 router and RE1000 range extender I get great signal levels throughout the entire house (4 bars even in furthest corner with an obstructed wifi adapter in my Blu-Ray player which sits in a stereo cabinet on the shelf under a heavy duty A/V receiver with plenty of metal) and also have 5 GHz capability, which the Ubiquiti AP-N lacks.
I was looking into whether I could remove the range extender from the picture by getting the AP-N instead, but it is actually quite unobtrusive, when used without the cord (plugged directly into an outlet) and has those nice B&O-like looks that look at home even in a modern kitchen outlet (compared to the typical range extender).
Additionally, the supposed wireless throughput loss resulting from the range extender is not an issue in my case, since I still get high link speeds throughout the home.

I will think it through some more before deciding.
If I plan to use 5 GHz (might be worth using in the room next to the router) I will stick with the E4200 and RE1000 (since the Zyxel did not do the trick).
If I do not need 5 Ghz, I guess I could always try the AP-N, if I want to deal with the effort to swap devices in and out, test performance and also the added time to pack and return / or advertise / pack and sell the gear I end up choosing not to keep.

 

[dyskinesia]

I also would love some screen shots of the necessary setup.

Considering buying this myself

Thanks

Shawn

Planning to use it as an access point from my wrt54g

 

[Mangusta1969]

Hi Mangusta1969,

How did you configure the firewall rules?
I read that it needs the user to manually set them and be familiar with iptables.
Sample settings and / or instructions would be greatly appreciated.
I am considering potentially giving this a shot (to compare to my E4200 and the Zyxel NBG5715 in terms of experimenting with purely extending the range, but without the use of a range extender).

PS: My E4200 has performed extremely well since March of this year, unlike your experience.
I never need to reboot it....

NJWeb,

Sorry for the late reply, I have been dealing with some re-possessed real estate woes (be careful who you lend to...) Sorry I can't help you on the firewall hardware configuration for the PowerAP-N. I am relying on my built-in Apple software setup, which is less secure than also configuring a firewall in a router. The only thing I can suggest is a visit to the ubiquiti web site and ask for some firewall setup help. Take a look here:

http://ubnt.com/forum/showthread.php?t=26164&highlight=firewall

It would be nice if the device shipped with some basic firewall instructions or a basic pre-stored firewall setup that could be tailored a bit to secure your local network from unwanted outsiders.

My unit remains bulletproof (only one power cycle in 5 months of use) and it has great range and steady throughput performance for all of my client devices; it is much superior to my Netgear and Cisco predecessor routers.

Let me know if you come up with some basic or intermediate firewall setup instructions.

 

[dyskinesia]

I too now have the PowerAP N and I will say it is very easy to setup, (much easier than the E1200 I struggled through all the goofy menus. Plus the guest AP was not functional in bridge mode so half the reason i purchased it was lost.

The range on this thing is awesome. I ended up putting it upstairs and the coverage from the 2nd floor fully covers the entire 2nd floor plus the patio/backyard that I was lacking in my other setups.

The cisco had an issue where it was 5 bars in the room with the router but dropped off to 1-2 as soon as I went onto the patio (the adjacent area). and the drop-off was exponential after that. The ubiquiti device maintained 3+ bars in the same location but still didn't stretch to the pool like I wanted. (full disclosure: large stone arches and fireplace between the patio/pool that hindered the signal) moving the device to the 2nd floor better covered that portion of the house and now bathes the backyard/pool as well. I didn't even need to worry with antennae aiming or transmit power. this thing is a beat out of the box.

my old wrt54gs2 that was upstairs now is downstairs in the trouble spot from before and fills that small gap caused by all the stone blocking in the main part of the house.

I say if you need the coverage this device did what 2 others could not. now I'm flush with wifi across 1,6,11 with no holes I can detect.

glad I found this site and the recommendation for this device.

Shawn