On the specs for the TP-Link EAP660 HD | Omada WiFi 6 AX3600 and the TP-Link EAP670 | Omada WiFi 6 AX5400 ....

[Bridges1]

What does the AX3600 vs the AX 5400 mean ? Is that coverage the AP is capable of ?

 

[tgl]

What does the AX3600 vs the AX 5400 mean ?

Those numbers are typically the sum of the max theoretical Tx rates for all the bands the router has. They are largely marketing BS. For instance the EAP670 spec sheet says "Simultaneous 574 Mbps on 2.4 GHz and 4804 Mbps on 5 GHz totals 5378 Mbps WiFi speeds", which they then round up to 5400 because they can. You will not get anywhere near those speeds in real life. 4800Mbps on the 5GHz band is the theoretical best case Tx rate if you have a 4x4 client (which you don't) and are using a 160MHz channel width (which you can't, unless you live somewhere far from any airport or weather radar that will hog the "DFS" channels). On top of which, real-world data throughput is typically no better than about two-thirds of whatever Tx rate you do manage to reach. So that "4804" is probably more like "800" for most of us, or at best "1600" if you live in the sticks where you can use DFS channels. Derate the 2.4GHz number similarly, although few makers provide 4x4 radios on 2.4GHz so there's one factor-of-two you might not be losing on that band. And you're probably not saturating both bands simultaneously in day-to-day use, either.

It's best to drill down to the actual spec sheet details. Look at the number of antennas (typically 2 or 4 per band, often written "2x2" or "4x4"); more is better, but it helps more for serving more clients than for getting better speed to any one client. Check whether 160MHz channel width is supported (not that you care if you're in an urban environment). The EAP660 has 4x4 radios on both bands but no 160MHz, so it ends up with the lower marketing number but might nonetheless outperform the EAP670 (2x2 on 2.4GHz, 4x4 on 5GHz, with 160MHz) if you don't care about 160MHz but do care about supporting a lot of 2.4GHz clients.

Also pay attention to whether the device has backend bandwidth sufficient to support whatever performance you're hoping for. These two TPLinks both have 2.5Gbps ethernet ports, but there's a lot of gear out there with just 1Gbps ports, which tells you something about what the manufacturer's engineers think you'll actually get out of it.

I didn't look at the Omada-nameplate devices to see if they are different, but they might be.

 

[Bridges1]

Those numbers are typically the sum of the max theoretical Tx rates for all the bands the router has. They are largely marketing BS. For instance the EAP670 spec sheet says "Simultaneous 574 Mbps on 2.4 GHz and 4804 Mbps on 5 GHz totals 5378 Mbps WiFi speeds", which they then round up to 5400 because they can. You will not get anywhere near those speeds in real life. 4800Mbps on the 5GHz band is the theoretical best case Tx rate if you have a 4x4 client (which you don't) and are using a 160MHz channel width (which you can't, unless you live somewhere far from any airport or weather radar that will hog the "DFS" channels). On top of which, real-world data throughput is typically no better than about two-thirds of whatever Tx rate you do manage to reach. So that "4804" is probably more like "800" for most of us, or at best "1600" if you live in the sticks where you can use DFS channels. Derate the 2.4GHz number similarly, although few makers provide 4x4 radios on 2.4GHz so there's one factor-of-two you might not be losing on that band. And you're probably not saturating both bands simultaneously in day-to-day use, either.

It's best to drill down to the actual spec sheet details. Look at the number of antennas (typically 2 or 4 per band, often written "2x2" or "4x4"); more is better, but it helps more for serving more clients than for getting better speed to any one client. Check whether 160MHz channel width is supported (not that you care if you're in an urban environment). The EAP660 has 4x4 radios on both bands but no 160MHz, so it ends up with the lower marketing number but might nonetheless outperform the EAP670 (2x2 on 2.4GHz, 4x4 on 5GHz, with 160MHz) if you don't care about 160MHz but do care about supporting a lot of 2.4GHz clients.

Also pay attention to whether the device has backend bandwidth sufficient to support whatever performance you're hoping for. These two TPLinks both have 2.5Gbps ethernet ports, but there's a lot of gear out there with just 1Gbps ports, which tells you something about what the manufacturer's engineers think you'll actually get out of it.

I didn't look at the Omada-nameplate devices to see if they are different, but they might be.

I have the TP-Link TL-SG2016P Switch with 10/100/1000 Mbps RJ45 ports .
I have a lot of clients for 2.4GHz.
So should I downgrade on the APs.
I have a 500mg connection and can upgrade to 1gig later on.

 

[Piotrek]

They are completely different devices with completely different purposes. Asus routers are toys for the home, The TP-Link AP are APs for serving multiple clients in a business environment.

I have a lot of clients

How much specifically? 100, 200, 500? What type? Computers, phones, IoT?

 

[Bridges1]

Ok the 3 Computers are wired, 8 cams wired and a music server. On the wireless side I have 4 ip cams,6 phones, 8 Ipads, 6 tvs on wireless, 7 Alexas, and a few other IOT devices.

 

[Piotrek]

more:

High-Density Environment | TP-Link

Equipped with enterprise chipsets, dedicated antennas, advanced RF functions, and auto channel selection and power adjustment, Omada Wi-Fi 6 and Wi-Fi

www.tp-link.com

 

[Piotrek]

I have a 500mg connection and can upgrade to 1gig later on.

AX1800 is enough up to 1GbE.

 

[Tech9]

EAP620 HD is much smaller unit compared to EAP 660 HD and better for home use.

AX1800 is good enough especially if multiple access points are used around the house.

 

[Tech9]

Another excellent option for home use - EAP615-Wall:

AX1800 class with 3x LAN ports for wired devices. Most homes don't have ceiling wire drops.

 

[Bridges1]

Covering 3 floors and outside at 100 to 150 yards.

 

[Bridges1]

Average 50 to 60 clients

 

[Tech9]

You asked for AP specs differences. Got the answer plus some suggestions. Don't expect complete network planning. This is something you have to do. If you go with Omada solutions - they have different APs for indoor/outdoor use including wireless mesh options. Only you know what do you need.

How to Build a Mesh Network via Omada EAP in Controller Mode | TP-Link

How to Build a Mesh Network via Omada EAP in Controller Mode

www.tp-link.com

 

[Bridges1]

I have the Omada controller OC200, TP-Link ER7206 router, 1 TP-Link TL-SG2016P Switch, and 1 TP-Link TL-SG2008P switch.
Just trying to figure out the AP part.

 

[Tech9]

We can't figure out this for you. You have to select the ones best for your application. There is nothing wrong in EAP660 HD. They are very capable APs with good range, but in larger physical size package. They are also more expensive. If you have the budget - one of the best you can get.

 

[GHammer]

I'll just chime in here.
I installed the Omada controller on a Linux machine, installed 2 EAP670, TL-SG2008P for PoE
First, the selected switches did not have 2.5 Gbe to serve the APs. I guess I could have used 2.5 capable injectors or the included power bricks, but power was unavailable. The cost for any Omada compatible switch with 2.5 Gbe ports was astounding.
Ok, I have some devices that would like the higher speeds, but luckily they are wired. Plus, my Internet is 900 down, 35 up. So, let's install.
First the 670s are HUGE! Think about a stack of dinner plates nailed to your ceiling. My wife took one look and asked where I thought they were going. Yeah, yeah, they'll look better once installed, trust me. They didn't.
Ok, got things installed and let them run a few days to do the auto power/channel adjustment Omada can do.
Great, I have a number of IoT devices and some were on the ragged edge of connectivity, like -75 dBm. The other thing I noticed, though I had Omada configured to force a roam if the device didn't, no device in my house/yard ever changed APs. I only have a 1900 sq ft house, 2 floors and a basement. Perhaps the signal was too great, but given the IoT devices I'd say not.
In short, it all went back to the retailer for a refund and I returned to my AX86 and AX58 running as APs. The problematic IoT device still only connect at -70 dBm, but that's a way better/stable connection than the Omada APs gave.
Since WiFi is a black art, perhaps you'll have much better results than I had. For me, it came down to why spend more to have worse results.