Need an AiMesh (2nd router) recommendation.

[COBOL-Coder]

Hi all. Ok, 6 mos. or so I replaced my older ASUS dual band router with my current RT AC-5300 because I accumulated more devices in need of WiFi. My house, a medium sized 4 bedroom is all lathe & plaster and 2 rooms are far apart seperated by multiple walls. When I purchased the AC-5300 I did not know that the 5 GHz band had such a short range. After trying just about every scenario conceivable to me I've decided to add another ASUS AiMesh router at the far end of the house. I don't want to spend a lot more then I have to but I'll go with the right choice. I would appreciate your suggestions. Thank you…

RT AC-5300 Running Merlins 384.15_0

 

[OzarkEdge]

Hi all. Ok, 6 mos. or so I replaced my older ASUS dual band router with my current RT AC-5300 because I accumulated more devices in need of WiFi. My house, a medium sized 4 bedroom is all lathe & plaster and 2 rooms are far apart seperated by multiple walls. When I purchased the AC-5300 I did not know that the 5 GHz band had such a short range. After trying just about every scenario conceivable to me I've decided to add another ASUS AiMesh router at the far end of the house. I don't want to spend a lot more then I have to but I'll go with the right choice. I would appreciate your suggestions. Thank you…

RT AC-5300 Running Merlins 384.15_0

Metal lathe? I once examined my AiMesh node behind a window with metal screen and behind the brick wall to the side of the window... the backhaul was stronger through the brick wall.

Will AiMesh run on your old Asus router? Model? You do not need Merlin on the node.

AiMesh dual-band candidates are RT-AC86U (router/node; 4x4 5.o GHz) and RT-AC66U_B1 (node; no Smart Connect).

Given your complete upgrade, I wonder how this tri-band AX 2-pack might have worked out:
https://www.asus.com/us/Networking/AiMesh-AX6100-WiFi-System-RT-AX92U-2-Pack/

Have you looked at signal strength around your home using something like WiFi Analyzer to get some idea of current router performance for your conditions? It would also help to confirm that the new 5300 5.o GHz is working properly.

OE

 

[COBOL-Coder]

Metal lathe? I once examined my AiMesh node behind a window with metal screen and behind the brick wall to the side of the window... the backhaul was stronger through the brick wall.

Will AiMesh run on your old Asus router? Model? You do not need Merlin on the node.

AiMesh dual-band candidates are RT-AC86U (router/node; 4x4 5.o GHz) and RT-AC66U_B1 (node; no Smart Connect).

Given your complete upgrade, I wonder how this tri-band AX 2-pack might have worked out:
https://www.asus.com/us/Networking/AiMesh-AX6100-WiFi-System-RT-AX92U-2-Pack/

Have you looked at signal strength around your home using something like WiFi Analyzer to get some idea of current router performance for your conditions? It would also help to confirm that the new 5300 5.o GHz is working properly.

OE

 

[COBOL-Coder]

Hey OC, thanks for the speedy reply. Ok, to your questions. RE the old router, it's a dual band AC66W. Used it flawlessly for several years. I checked and told the processors were different and as a result would not function in an AiMesh environment. With regards to the RT-AX92U 2 Pack, looks great, didn't know about those. I didn't make this clear but at the time I bought the 5300, AiMesh was not a consideration because I thought the 5300 would blow through those walls, not! The RT-AC66U_B1 looks like a possibility, I'll chect it out. With regards to making sure the 5300 is working properly, I don't exactly how to test for that? I did do a factory reset and reinstalled everything, which helped a little. I had hardcoded the channels on 2.4 & 5 using a WiFi analyzer, however, it ran worse. I also set the 2.4 at 20 mhz and the 5 at 80 but found that allowing the router to make those assignments gave me a far better result. Thanks again for your suggestions.

 

[OzarkEdge]

With regards to making sure the 5300 is working properly, I don't exactly how to test for that? I did do a factory reset and reinstalled everything, which helped a little. I had hardcoded the channels on 2.4 & 5 using a WiFi analyzer, however, it ran worse. I also set the 2.4 at 20 mhz and the 5 at 80 but found that allowing the router to make those assignments gave me a far better result.

Using the WiFi Analyzer, you can observe signal strength dBm. If a signal is unusually weak, it should stand out. You may see -30 next to the router and -80 out of range. -68 or less negative is usable, more or less.

You can also see ambient signal congestion on channels you might be using. Generally speaking, I would fix 2.4 GHz on ch1, 6, or 11 at 20 MHz. And 5.0 GHz on ch149 or higher at 80 Mhz. Subject to avoiding any ambient signals as much as possible.

If you intend to try wireless AiMesh, anticipate your node location and check for a decent 5.0 GHz signal there... like -68 dBm more or less. I have found that the router-to-router connection tends to exceed a client-to-router connection for a given distance... due to better radio equipment in the routers, I suspect.

Another option you might have is to use an 86U as router and your 5300 as node. Someone here who has worked this equipment out for their AiMesh might comment what they have learned. I have not, but I have my unsubstantiated suspicions.

OE

 

[L&LD]

There are a few posts around these forums where using a WiFi analyzer 'app' didn't give the promised maximum speeds. Why? Because they only 'sniff' out other WiFi radio signals.

When interference by a non-WiFi device is in the environment, those WiFi apps are basically useless and a waste of time for finding the best Control Channel and Channel Width.

That is why I recommend simply trying each channel (they're not infinite) and testing for a reasonable amount of time before writing down some notes and then testing the next one. Refer to the notes when you are finished to determine which is the actual, best channel to use. Regardless of what any 'app' may suggest.

Unless you're using a commercial WiFi program/hardware to determine optimum channels and bandwidth for your environment (i.e. one that can read and display all radio interference, and not just WiFi signal standards), the above method is much faster and much more likely to get you a selection that works as well as possible for your WiFi conditions in the shortest amount of time.

When doing the testing recommended above, I am not chasing maximum throughput. Rather, I am finding the control channel/channel width combinations that give the maximum responsiveness to the network instead. Losing 5% to 15% top end is worth it when the network is twice as responsive on a seeming lessor combination (if just throughput was all that mattered).

Go ahead, use the apps. But only to see how wrong they can regularly be.

 

[OzarkEdge]

Go ahead, use the apps. But only to see how wrong they can regularly be.

Revealing other WiFi congestion in the area is a good start to determining the candidate channel selection for further actual testing.

What puzzles me is my neighbors all use auto channels that bunch up together all over each other in the middle channels 3-8. They do seem to avoid me on 11.

OE

 

[L&LD]

@OzarkEdge, unless the app can actually show channel/bandwidth utilization, it is not important how many AP's are 'seen'. And again, if there is significant non-WiFi introduced interference in the area, the 'clear' channel may still be the worse one if that is where the interference prevails.

I use to use these apps and looked really professional in front of customers. It didn't help when I had to come back and do it all over again for them though.

I quickly learned that interference (constant/random/cyclical) throws out all these measurements that 'free' or $1.99 apps can show out the window. I don't even install them anymore.

Although, WinFi is interesting as a learning tool.

With Control Channel 8 used by your neighbors and depending on if they are 20Mhz/40Mhz bandwidths being used, your network on Control Channel 11 is also affected too (particularly if you're also using 40MHz width, of course).

What I find is that if the best channel/bandwidth for the router (see posts about how Control Channel 1 or Control Channel 6, etc. work better in one router or another) is used and it also happens to give the most responsive network experience too (even if it isn't the fastest), then after a certain period of time, the surrounding AP's on Auto Channels move away and the network is even faster and more stable too. Not all of them have to move, of course. But the ones that do move, seem to be the ones that were the most utilized too.

The WiFi medium is shared, half-duplex, constantly variable and subject to inexplicable curve balls now and then. No sense in using crystal balls, er, I mean 'apps' to tell me what I require from my networks.

And to be clear: that is responsiveness, almost above all else.

Edit: I learned this lesson (network responsiveness) from a few customers that had very low ISP speeds (1.5Mbps down and 256KB up) for which I had tuned for maximum throughput. They reported that the network seemed 'faster' before, even though I had the numbers to show them it wasn't. When I understood that they needed it to perform as close to real-time as possible instead of as fast as possible, they were happy indeed.

When I took that understanding and applied it to my own network with much higher ISP speeds, I too was happier with a slightly slower (peaks) network but a much more responsive one. Now, that is the only way I'll tune a network for.

When some customers tell me that they're paying for 450Mbps speeds and don't want to be getting just 390mbps as its a 'waste', I tell them when they're driving their Masarati are they happiest when they're driving 7 or 8/10ths for hours or 9 or 10/10ths for mere seconds.

 

[COBOL-Coder]

Using the WiFi Analyzer, you can observe signal strength dBm. If a signal is unusually weak, it should stand out. You may see -30 next to the router and -80 out of range. -68 or less negative is usable, more or less.

You can also see ambient signal congestion on channels you might be using. Generally speaking, I would fix 2.4 GHz on ch1, 6, or 11 at 20 MHz. And 5.0 GHz on ch149 or higher at 80 Mhz. Subject to avoiding any ambient signals as much as possible.

If you intend to try wireless AiMesh, anticipate your node location and check for a decent 5.0 GHz signal there... like -68 dBm more or less. I have found that the router-to-router connection tends to exceed a client-to-router connection for a given distance... due to better radio equipment in the routers, I suspect.

Another option you might have is to use an 86U as router and your 5300 as node. Someone here who has worked this equipment out for their AiMesh might comment what they have learned. I have not, but I have my unsubstantiated suspicions.

OE

 

[COBOL-Coder]

Ok, thank you OE… one question though, you said "-68 or less negative is usable, more or less." Did you mean -68 or more or -68 or less? Is the smaller the number better or worse? Thanks

 

[OzarkEdge]

Ok, thank you OE… one question though, you said "-68 or less negative is usable, more or less." Did you mean -68 or more or -68 or less? Is the smaller the number better or worse? Thanks

I meant less negative... -30 is less negative and a stronger/more powerful/more usable signal than -70 dBm. -70 is 'smaller' than -30.

Of course, this will be obvious when viewed on a WiFi Analyzer graph as you walk away from the signal source.

It's because of the math!

OE

 

[COBOL-Coder]

Thanks for the input OE…

 

[COBOL-Coder]

Hi OE, well, I took your suggestion (everyone agrees that this was the right course) and purchased an ASUS RT-AC86U to be in an AiMesh configuration with my RT-AC5300. I had a standard ADSL modem when I setup the AC5300 but now I have an ARRIS-BGW210 a supposedly fiber modem and it's not the same. I'll let you know how it all goes after I figure out the initial cabling as this modem has what they refer to as a Broadband connection. Thanks again OE

 

[OzarkEdge]

Hi OE, well, I took your suggestion (everyone agrees that this was the right course) and purchased an ASUS RT-AC86U to be in an AiMesh configuration with my RT-AC5300. I had a standard ADSL modem when I setup the AC5300 but now I have an ARRIS-BGW210 a supposedly fiber modem and it's not the same. I'll let you know how it all goes after I figure out the initial cabling as this modem has what they refer to as a Broadband connection. Thanks again OE

Does that Arris gateway have a bridge mode to reduce it to just a modem so that you can use your router as the site router?

OE

 

[COBOL-Coder]

Does that Arris gateway have a bridge mode to reduce it to just a modem so that you can use your router as the site router?

OE

 

[COBOL-Coder]

Hi, OE, exactly, I wouldn't have gone with the service (Earthlink) without that provision. I've gone from an insanely low 6 mbs down (typically 4.7 on the average) 1 mbs (or .7 up) to 75 down (really high 50's to low 60's) & high 7's up. As they used to say in my old neighborhood in NYC, "we be happier then a pig in mud", lol… I'll brief you on the finale unless you opt out in a reply…

Thanks again…

 

[COBOL-Coder]

Hi OE, Ok, I mentioned I would tell you how it all ended up. I was able to install my new RT-AC86U as a node with my RT-AC5300 Router and surprisingly (to me) it was pretty easy to do, great firmware! It wouldn't connect until I did a reset right before I attempted the connection. The instructions are explicit on that point. I placed the AC86U roughly in the middle between the router and the original dead zone. With minor adjustments on the antennae, choice of placement, etc., I'm consistently getting (Speedtest) mid to high 50's down where as I was getting low 20's or worse. There were a few negative aspects to the install, one of which is the AiMesh system takes over the second 5Ghz band for use as a wireless backhaul thereby leaving me with the 2.4 & 5(1) bands for use which is more then adequate considering the increased throughput. As a slight (for me, might effect others more) inconvenience for me, there is no support for utilizing the "Guest" capability. As near as I can tell, the current AiMesh firmware does not support the Guest capability. I also wanted to give Colin a Shout Out for suggesting to another user, the use of Mac Id Filtering to prevent (Accept, or whitelisting, can be used as well) one of my 4K TV's from attaching onto the 2.4 band. It's a great way to use Mac Id filtering that never occured to me before. Thanks everyone for the interest you took in helping me to a successful, but not cheap conclusion,

 

[Grisu]

It is well known that Aimesh nodes wont support guest-wifi, only on master router!
Same as AP mode, here you are able to set a guest, but it will have same (full) rights on master router as using main SSID.
You have to wait, there are rumors that it may come with 386 firmwares.

 

[Klueless]

Wow, my compliments, you sure move fast. I was going to reply last week but I forgot. Just as well, you found something that works so congratulations!

I was going to say for a mere two nodes I don't see the benefit of "mesh". For multiple nodes the benefits are ease of configuration, self-configuring, optimization and self-healing.

But for two nodes I'd have been tempted to pickup up something cheap like a dual-band Netgear 6150. If I could run a wire I would have configured it as a hard wired Access Point. If I couldn't I might have configured the 2.4 GHz radio as a dedicated backhaul to the router and simply advertised 5 GHz services to clients on that side of the house.

Another option might have been one of the dual-band Linksys Range Extenders that supports "Cross Connect".

But, for sure, your solution is a lot more fun!

 

[OzarkEdge]

I was going to say for a mere two nodes I don't see the benefit of "mesh".

... because it's wireless!

OE