Opinions to expand my coverage

[Tech9]

Sorry if that somehow offended you.

Well... you ask a question, get an answer and then consider something cheaper. Few times already. If you ask me, I wouldn't play with any of those home routers and install 3-4 centrally managed SMB class PoE access points where needed. Do it once properly and focus on more enjoyable things in life. This AiMesh you're looking at is a marketing name of wired access points or wireless repeaters with very limited control. Asus afterthought solution to join the "mesh" market trend. The roaming is slow or not working, no "node" channel control, no "node" power control, the devices are with ugly external antennas and need individual power supplies. Yes, it's cheap. It works in a cheap way too. There are better "mesh" systems around. Nothing under $200 though.

 

[bcm00re]

@Tech9
I had an Engenius device like you mentioned, but they stopped supplying firmware updates for it. So from my experience those aren't future proof either. I do appreciate the suggestion; if this Asus experiment doesn't work out I will look into that.

 

[Tech9]

So from my experience those aren't future proof either.

Access Points are wired-to-wireless bridges. They don't do routing and are not affected by most vulnerabilities discovered for Internet facing devices. Hence they don't need very often firmware updates. Access Point firmware updates are mostly focused towards stability and added features.

 

[bcm00re]

Access Points are wired-to-wireless bridges. They don't do routing and are not affected by most vulnerabilities discovered for Internet facing devices. Hence they don't need very often firmware updates. Access Point firmware updates are mostly focused towards stability and added features.

That is a good point -- thanks for pointing that out. I unearthed my old one, and it turns out it's just a b/g unit so it probably won't be very useful at this point. In your scenario I would still need something to do the routing, but I guess it's easy to find a router that performs well when you don't have to worry about the wireless side.

 

[bcm00re]

If you don't do (or even listen/agree) with the poster you think you're having a rational conversation with, you'll be thrown under the bus too by him. It's an ego thing (they need to be right, or they'll melt).

The RT-AX68U should offer greatly improved throughput and coverage of your home's stated size. Testing is what will determine if it works for you and your environment or not.

RT-AX82U or RT-AX86S

Guys, I own such AiMesh system: RT-AX86U (main) + RT-AX82U (node) + RT-AC86U (node). All backhauls are wireless. I'm looking for the AC86U replacement to have the whole AiMesh system that is fully WiFi 6 compatible. What you advise to get - another AX82U or AX86S? Is it worth upgrading my old...

www.snbforums.com

I picked up a RT-AX68U today for $140, and I even found a coupon that got me a free $20 Visa card too -- so it was effectively only $120. I plan to try it and the RT-AX86S during the holidays to determine which one performs best in my environment.

 

[bcm00re]

I am trying the AX68U even though it has a questionable past for a few reason. First it has been reported that the firmware released a month ago has fixed the 5GHz radio issues that some have seen. I am also intrigued by reports that it has really good range. It is generally accepted the AC86U has better range than the AX86U -- and thus AX68S too. It has also been reported that a single AX68U outperformed two AC86U units in one scenario. I understand that might not translate to my setup/environment, but it still looks mighty impressive in paper! Lastly, the AX68U seems to have all the features I need as well as the same processor and memory as the AX86S -- and its 33% cheaper.

 

[Tech9]

It has also been reported that a single AX68U outperformed two AC86U units in one scenario.

From a single user actively advertising AX68U and backing own opinions with "customers" stories. I know. He's in my ignore list for quite some time.

Just think about it - most clients are 2-stream AC class, up to 866Mbps link speed in ideal conditions. Do you have any AX clients and what type? AX can do 1200Mbps link speed in ideal conditions to 2-stream AX client. AX68U doesn't have 160MHz wide channel support (AX86S does). How it is possible for a single 3x3 AX mid-range router to outperform 2x best performing 4x4 AC routers in both range and throughput? Do you believe in fairy tales?

and its 33% cheaper

Nice. You've got cheap. The thing you wanted.

 

[bcm00re]

Nice. You've got cheap. The thing you wanted.

I don't think you are following along. I bought both the AX86S and the AX68U and plan to compare them in my environment. Are there any disadvantages to trying two different models?

I'm not sure I have any AX clients; maybe one of our phones is, but not sure it would care/need all that speed.

 

[jsz]

I don't think you are following along. I bought both the AX86S and the AX68U and plan to compare them in my environment. Are there any disadvantages to trying two different models?

I'm not sure I have any AX clients; maybe one of our phones is, but not sure it would care/need all that speed.

Just compare and play around. Theres no real point in arguing schematics unless you're planning on using 1 router for your entire house and even then, diminishing returns is quite high with most 4x4 AC/AX hardware.

Like I said in another post, my GT-AC2900 almost matches the GT-AX6000 with it's newer gen/wave 2 AX 4x4 hardware at range/walls for me. This isn't a concrete fact or truth, but just my subjective experience with my AX clients and router location.

Throughput was definitely higher on the newer radio but I was also sitting at lower QAM modulation relatively. Weird trade off.

1. If you don't care about updates a $100-110 BF sale on a GT-AC2900/AC86U would do very well and should pair up with current router. Lowest price was $91 on Amazon. Two of those wired backhaul will dumpster any high end router.. AX/AC.
2. AX68U will perform worse than the cheaper 4x4 AC router (GT-AC2900) for AC clients @ distance, but likely match 86S for AX clients due to default bonding being lower (80mhz) and subjective testing i've seen from someone I "trust". 5G 3x3 has always been an outlier that can perform better or worse than 2x2/4x4 alternatives depending on config.
3. AX86S is the most consistent/updated option, but it also cost more. It compromises on the other two routers shortcomings, but you pay more.

All 3 share the same 4906 dual core A53 platform . Just different radio config and EE design.

 

[L&LD]

I am trying the AX68U even though it has a questionable past for a few reason. First it has been reported that the firmware released a month ago has fixed the 5GHz radio issues that some have seen. I am also intrigued by reports that it has really good range. It is generally accepted the AC86U has better range than the AX86U -- and thus AX68S too. It has also been reported that a single AX68U outperformed two AC86U units in one scenario. I understand that might not translate to my setup/environment, but it still looks mighty impressive in paper! Lastly, the AX68U seems to have all the features I need as well as the same processor and memory as the AX86S -- and its 33% cheaper.

And I'll restate that I have not seen those issues for any RT-AX68Us I've installed for customers (lousy batch for those reporting problems?).

The AX class routers consistently outperform AC class routers where it matters; in throughput and lower latency. I have not seen a situation where an RT-AC86U had a better range than the RT-AX86U that replaced it.

And thank you for using good judgment and testing both in your environment. I haven't advertised anything here on SNB. People will make up lies to make themselves look good though.

 

[coxhaus]

If it was me a would buy a 2 pk or 3 pk Cisco 240ac wireless APs and set them up. You will have a lot higher throughput than any mesh system. Cisco is great about updating their firmware. I see updates about once a year.

 

[jsz]

And I'll restate that I have not seen those issues for any RT-AX68Us I've installed for customers (lousy batch for those reporting problems?).

The AX class routers consistently outperform AC class routers where it matters; in throughput and lower latency. I have not seen a situation where an RT-AC86U had a better range than the RT-AX86U that replaced it.

And thank you for using good judgment and testing both in your environment. I haven't advertised anything here on SNB. People will make up lies to make themselves look good though.

I generally agree.

My GT-AC2900 was a little worse than the GT-AX6000, but it wasn't worth the $300 upgrade subjectively. QAM modulation was lower on GT-AX6000, but throughput was higher. Where the GT-AC2900 had lower throughput, but hovered at a higher QAM tier. Cant win

I think he can't go wrong with any of them you factor price/performance, but he needs to TEST them in his environment. my2c.

 

[bcm00re]

The AX class routers consistently outperform AC class routers where it matters; in throughput and lower latency. I have not seen a situation where an RT-AC86U had a better range than the RT-AX86U that replaced it.

Was that with AC clients or AX ones? Maybe that's the difference -- or perhaps I just misinterpreted a couple statements I read here.

The only AX clients I will have (at least initially) are a 2nd gen iPhone SE and an iPhone 12. I cannot imagine why either of those would need/use more than 800 Mbps of throughput -- or even a fraction of that really. Or am I missing something perhaps? I think my wireless clients demanding the most bandwidth will be my Rokus. One is still just a b/g/n unit, but the others are b/g/n/ac and dual-band (or wired). Again my internet service only 300Mbps Charter, but we haven't felt constrained by it so far. I do have a HTPC running an Emby server that has a couple HDHomeRun units (networked OTA tuners) attached.

I don't plan to do my testing until second to last week in December, but I will come back and report my results. I plan to mainly use the Wifi Analyzer app on my Android phone for testing, but I might move my laptop around and see what numbers I get when I go to speedtest.net. Of course, I'll see how well my existing wifi clients do in their usual places; I do have one Roku device that hasn't worked so well lately. I am certainly open to suggestions as well.

 

[L&LD]

With both AX and AC clients, the improvements are there. Of course, the AX clients will see a higher number.

What was immediately noticeable to me with the RT-AX86U when I first used it was how fast the network seemed vs. the RT-AX88U (the former had a newer RF design/hardware, and a newer SDK at the time).

RT-AX86U vs. RT-AX88U

That got me to see how valuable higher ISP speeds were. The raw speed itself didn't really matter. The lower latency though was very noticeable (not just to me, but to most of my customers too).

For testing, I don't suggest using any handheld device to do so (except as an additional data point). Nor do I recommend using an app either (waste of time for little real information that helps directly).

Tune your network for range and speed.

I look forward to your results. Be sure you keep good notes!

 

[bcm00re]

@L&LD
I plan to compare to the 86S not the 86U. I suspect that benefit is from the quad core proc in the 86U model. The 86S has the same dual core as the 68U. If the 86U drops to below $200 during holiday sales, I will probably buy it and just return both the routers I have.

Thanks for the testing tips!

 

[Piotrek]

If you're not using CPU-intensive features (e.g. VPN, QoS) then CPU cores won't make a noticeable difference to you.
If you are looking for a cheaper option, you can test cheap models like RT-AX56U (ver.1) or RT-AX58U (ver.1)/RT-AX3000 (ver.1).
They are satisfactory for most regular users.

 

[jsz]

@L&LD
I plan to compare to the 86S not the 86U. I suspect that benefit is from the quad core proc in the 86U model. The 86S has the same dual core as the 68U. If the 86U drops to below $200 during holiday sales, I will probably buy it and just return both the routers I have.

Thanks for the testing tips!

Radio performance should be 100% reliant on the radio itself. I personally don't think the main CPU (4096/4098 in this case) has much impact on WIFI, but I could be wrong... Maybe if you're running layers of software in addition to the base radio.. IE: VPN, AI protection, traffic analyzer etc.

From what I've gathered, the older AC W2 BCM4366E Radio runs at 800mhz, while the Gen 1 BCM43684 (main 5G in 86S/U) bumps the internal clock to 1.5ghz, though no one can confirm or deny that.

Broadcom seems to want to keep that confidential although some OEMs/brands sort of leaked or indicated that 1.5ghz spec on radio itself. And no, I'm not talking about the 6755/6750 SoCs which are indeed 1.5ghz.

I subjectively couldn't tell much of a difference vs the newest gen 2 hardware (technically better than whats in 86U), but to be fair, my QAM modulation did hover at a lower tier relative to my AC router..

 

[bcm00re]

@Piotrek
Thanks for the suggestions, but I believe both are only 2x2 and therefore would yield lower performance for my N and AC clients which is mostly what I have (compared to my current 3x3 router). On top of that neither appears to be as cheap as the $120 I paid for the RT-AX68U which is also 3x3.

 

[Piotrek]

I don't think you'll get lower performance for yours N and AC clients.
As far as I remember, the AX56U was always the cheapest Wi-Fi 6 model with Merlin support - I bought it for less than $100 equivalent (regular price, I see the AX58U at a price of around $110 equivalent). However, prices in your country may be different.

 

[Tech9]

but I believe both are only 2x2 and therefore would yield lower performance for my N and AC clients

Your believe is wrong. Your clients are 2-stream at best. AP with 4x4 radio may have better sensitivity and offer more MU-MIMO streams, but certain conditions have to be met and the clients have to support Beamforming and MU-MIMO. None of N clients and only some of AC clients do. The number of eventual MU-MIMO clients support is limited by the number of available streams (5GHz band only). In most cases in low client density home environment it just doesn't work or doesn't make any real life difference. Just stop reading "reviews" from people with no good understanding of how Wi-Fi works. Two 2x2 APs working on different channels will outperform any single 3x3 AP in aggregate throughput, number of supported clients and coverage area.

I'm trying to help you, but you always return to cheaper options and read wrong sources of information. A cluster of 3-4 EAP225v3 APs will crush all the home routers solutions discussed above to your existing N and AC clients, offering up to 600Mbps speeds to common single 2-stream client, plus coverage of your entire area on both 2.4GHz and 5GHz bands, plus much faster and better roaming for your mobile devices. What else do you need?