Ubiquiti AC Pro and AC Lite Access Points Reviewed

[adrian]

I don't know with anyone else. If cabling is not a problem, Ubiquiti's Unifi AP's is the most cost effective way to fill a location with Wi-Fi. Someone can make it even slightly cheaper by going with Wi-Fi routers with disabled DHCP function or AP mode but cumbersome for management, like changing the Wi-Fi password

 

[DanH]

Looking forward to the AC HD test...

Me too. I am dead set on the HDs regardless of results though. I just need to get 500 friends so they can all use their phones at once. Seriously though, I know they are over kill for a home, but I am getting a couple anyway.

 

[Elisha]

So how is the hand-off on these things? Do each point have to be hard wired or do they have a dedicated wireless backhauls?


Sent from my iPhone using Tapatalk

 

[netwrks]

I have two ac pro's, they hand off fine, with Samsung phones iphones ipads and intel platforms. Better to use ethernet uplinks to the AP's. A mix of .ac and .n. Shouldn't be an issue if you aren't blasting the radios and pick channels.

 

[mdamda]

Me too. I am dead set on the HDs regardless of results though. I just need to get 500 friends so they can all use their phones at once. Seriously though, I know they are over kill for a home, but I am getting a couple anyway.

One anecdotal/report on the ubnt forums talk of a throughput increase over the pros with an ip7+ as a client.

Given most phones are only 2x2, this is actually quite interesting.

 

[DanH]

One anecdotal/report on the ubnt forums talk of a throughput increase over the pros with an ip7+ as a client.

Given most phones are only 2x2, this is actually quite interesting.

Yeah for me I have an ip7+.

 

[Elisha]

One anecdotal/report on the ubnt forums talk of a throughput increase over the pros with an ip7+ as a client.

Given most phones are only 2x2, this is actually quite interesting.

The iPhone 7+ is also only a 2x2 client.


Sent from my iPhone using Tapatalk

 

[mdamda]

The iPhone 7+ is also only a 2x2 client.


Sent from my iPhone using Tapatalk

Which, if the anecdotal report is true, would mean that there IS a tangible difference between the HD and the Pro for clients that can handle less than 4 streams, despite the specsheet/theoreticals saying otherwise.

This reminds me of the article here on SNB that mentions how the 2.4GHz on modern AC routers are noticeably better than the older 2.4GHz N grade routers despite there being no real reason to be any different...

 

[thiggins]

This reminds me of the article here on SNB that mentions how the 2.4GHz on modern AC routers are noticeably better than the older 2.4GHz N grade routers despite there being no real reason to be any different...

AC routers have newer chipsets with improved radios and MAC level improvements that benefit older devices, as the article showed.

More RF chains in an AP can improve receive gain, which can boost uplink performance.

The HD has both 4x4 radio and separate main processor for Ethernet <> WiFi processing. PRO and Lite have main processor shared with 5 GHz radio.

So more processing power can also help improve throughput.

 

[sfx2000]

AC routers have newer chipsets with improved radios and MAC level improvements that benefit older devices, as the article showed.

More RF chains in an AP can improve receive gain, which can boost uplink performance.

Even with single stream clients - there is a clear benefit to a 4*4:4 (or even a 3*3:3) Access Point, as the WiFi chipset in the AP can do nice things like combining the radios using Space Time Block Codes and the like...

AC radios are better N radios - Tim's article pretty much proves that out, and my experience backs it up.

 

[georgedone]

I am a user of UAP-AC access point. Before I used 2 Buffalo WZR-HP-450GH (450n, 2.4GHz) with DD-WRT, DHCP disabled (just as access points).
The UAP-AC is mounted on the wall, not on the ceiling (prefer not to drill into ceilings). The location of the Ubiquiti gear is slightly different than the Buffalo, which actually is still on for a smooth transition (WIFI is still enabled on Buffalo but no device is connected to it from our 25+ devices anymore, only some wired devices are still using them). I have to say that I am very happy with UAP-AC, I guess testing range and throughput does not say the whole story, the problems with many routers start when 10+ devices are using them concurrently, a dedicated AP deals much better with it than even a router with DD-WRT in AP mode. I value very much the tests done here and very few places remain which are truly independent (like here) but there are certain scenarios which are relevant and yet impossible to test (such the 25+ clients in a modern house). I bought 3 UAP-AC and installed only one so far in the center of the house (3 levels, concrete floors, about 60 square meter per level), I do not know what to do, to sell or to install the other 2 APs since the coverage from the installed one is sufficient for us. Even on 5GHz I have excellent coverage on the same floor and 2 lines (out of 5) on the other floors. The 2.4Ghz is better in coverage but I'm living in a area with 2.4Ghz overcrowded.

 

[d-m-z]

still doesnt explain why the orbi didnt or wasnt effected

This has turned into quite an interesting discussion.

To the particular concern (quoted above), if I'm interpreting things correctly the issue is that the Orbi was affected. And that the Orbi is benefiting from RF signal leaking past the artificial attenuation that's introduced to lower the signal level in the test. The results that the Orbi shows (in the 2.4 GHz test) simply don't jive with theory.

I don't think anything sinister is going on.

I look forward to a more in-depth explanation. More information is always better.

 

[thiggins]

Update. I found the source of the leakage. The high gain chamber antennas in the Rev 9 process move the received signal level up to the point where I saw DUT overload in 2.4 GHz. So I stacked two pair of attenuator channels, setting one to a fixed 10dB and varying the other for the throughput vs. attenuation test. The diagram in the Revision 9 "How We Test" article shows this configuration.

If the device can stay connected until the signal level is low enough, then attenuator-to-attenuator coupling keeps the signal level from going lower. This was the case with Orbi and some other products used for comparison.

You don't see long tails on 5 GHz plots because devices disconnect before attenuator coupling takes effect.

I reported this first to Brandon and he agreed with the assessment.

 

[DanH]

I just read that last post out loud. And now there are like 100 women banging on my door trying to jump my bones. Seriously though, glad you got it figured out.

 

[thiggins]

I just read that last post out loud. And now there are like 100 women banging on my door trying to jump my bones. Seriously though, glad you got it figured out.

Huh? I'm not seeing what you are seeing.

 

[DanH]

Huh? I'm not seeing what you are seeing.

You have to read it out loud trust me they will come.

 

[ilkevinli]

LOL . !!!

You have to read it out loud trust me they will come.

 

[pete y testing]

If the device can stay connected until the signal level is low enough, then attenuator-to-attenuator coupling keeps the signal level from going lower. This was the case with Orbi and some other products used for comparison.

so all in all this was a bit of a storm in a tea cup and the actual results for the unbifi's is the same , eg as in the orbi is still better in the graphs overall , it was just a minor tail end issue that TBH wouldnt effect most of us

here is me thinking it was some huge issues he was pointing out that would effect the whole testing procedure !!!!

 

[DanH]

so all in all this was a bit of a storm in a tea cup and the actual results for the unbifi's is the same , eg as in the orbi is still better in the graphs overall , it was just a minor tail end issue that TBH wouldnt effect most of us

here is me thinking it was some huge issues he was pointing out that would effect the whole testing procedure !!!!

I don't think there was any storm, just some overzealous forum people who basically get all huffed over nothing, but you know... internets...people ready to rumble like the outsiders anytime their favorite product or their favorite website isn't mentioned in 100% glowing terms.

I think I will wait until all the results are in, and then read it. Still not going to change my mind on getting what I want anyway. I just really enjoy reading tims reviews, he legit is the only person (like some other person already mentioned) who at least puts numbers in his reviews, not just "feels". Kudos to him and the UBNT fellas to making it better.

 

[sfx2000]

Just keep in mind that Orbi (and other consumer oriented multiple point wireless) are focused towards a specific market segment, and UBNT has a product in that category.

The AC-Pro/AC-Lite are good devices, and for small business, they're an excellent value, esp. when used within the UBNT verticals (Edge Router, their managed switches, etc).