My Orbi router's wifi is slower than my satellite

[jorge-07]

Hi, I have a somewhat strange situation. I recently upgraded my internet plan with Xfinity from 200 Mbps up to 400 Mbps and I began to notice a problem. I routinely get wired speeds of 475 Mbps (yes higher than my plan) and I'm also able to get 450 Mbps when in close proximity to my first satellite. It drops a bit near my second satellite but still over 325 Mbps. The problem is actually when I'm connected to the router and standing just a few feet away in direct line of site. The router is maxing out at 250 Mbps.

This makes no sense, how can the router's wifi speed be so much slower than both satellites? I realize 250 is still plenty fast but that is not the point. I would like the router to perform as it should. Any help or advice would be greatly appreciated!

Let me make clear that I have done all basic troubleshooting including but not limited to: firmware update, countless reboots, channel analyzer, disable Circle and Armor, disconnect satellites to test only the router, changed and played with all wifi settings. 

Here is my setup:

Single family one story home (L shape) 5000 square feet on 3/4 acres, very little interference from neighboring wifi.

Running Orbi Router RBR50 and 2 Orbi Satellites RBS50. All units on latest firmware 2.7.1.60

Setup in Daisychain configuration.

My Router is located at the north end of home (lets call it the top of the L) the first satellite is at the south end (the bottom or bend of the L) and the second satellite is at the east end (bottom right of L). Distance from router to satellite 1 is about 45 feet with just 2 walls between. From satellite 1 to satellite 2 is about 35 feet but with cement block wall and other regular walls between which is why speed there is understandably lower.

[CrimpOn]

It is really gratifying to get higher speed than the plan.  The post did not mention that the same device is being used to perform the speed test on each unit.  It is, correct?

Are any "high usage" devices connected to the router while the test is being performed?  In contrast to the ethernet connection, which is full duplex allowing data to flow in both directions simultaneously, WiFi is inherently  "one device talks at a time and the others wait their turn."

I wish there was a tool that could measure how much total bandwidth  is being used on a particular WiFi channel.

[jorge-07]

Thanks for the quick reply. Yes the same devices are being used for the tests, mainly my iPhone 11 and a laptop and both with similar results.

You bring up a good point about connected devices. My router definitely has more devices connected, probably less than 10. These include smart light switches, doorbell, phones, tablets and laptops. But I would not consider any of them to be high bandwidth especially during the time I am doing the test. Plus the satellites probably have around the same number of connections.

[CrimpOn]

The "50" series Orbi's use separate radios, antennas, and 5G channels for the "backhaul" connection between satellites and router.

So, that rules out WiFi backhaul having anything to do with this phenomenon.

The user facing radios are all on the same channel, and thus can interfere with each other.

What about a goofy experiment: power off both satellites and see what the WiFi performance is near the router.

[jorge-07]

Thanks, I actually have tried that already. No change. I’m thinking of reverting to an older firmware to see if it’s related somehow.

[CrimpOn]

I know you have  verified which Orbi the test machine was actually connected to for each experiment.  There are definite issues with certain devices "sticking" to the first Orbi they find (to the consternation of the owner).  Devices like smartphone and laptops usually have network software that constantly scans for a "better connection".  I have no idea what the difference is between "regular" and "fast" roaming.  The Attached Devices display does not update quickly enough to indicate the instant that a device switches.  My tests put fast roaming at under two minutes, but that's not very precise.