Connection issues with 2.4 ghz going through several walls

Linksys WRT1900ACS with OpenWrt 19.07.6.

Lately I noticed that on the 2.4 GHz wifi, once i go a couple of walls in my house (made of concrete), the connection completely dies even though my phone the wifi signal indicator is good. Yet in the same room with 5 GHz, there's no problem. I don't understand. I thought 5 GHz is suppose to have issues going through walls.

Also the bitrate for 2.4 Ghz is stuck at 72.2 Mbit/s and i've seen it drop as low as 20 Mbit/s

EDIT: After hours of troubleshooting. It turns out that the Hard Drive dock that I have plugged in the USB 3.0 interface is causing interference with 2.4 GHz. Switching back to Linksys firmware with the HDD dock plugged in also suffers the same issue. When I turned it off, the problem went away.

Then researching on this brought me to this


Yes, interference is a real radio problem in all frequencies!
I guess thats why USB3 cables so often are so thick because they have a screen inside.

And the 5GHz actually interferes with the weather radars...

You have thin concrete with small amount of steel in it I must say. Where I live if you go two concrete walls from the transmitter (all transmitters) all signals in all communications systems have dampened so much that nothing works.

1 Like

...and depending on the spacing of the steel, the smaller 5 GHz waves may fit through, while the longer (in wavelength) 2.4 GHz ones cannot. :wink:

It's called a Faraday cage; but I'm not so convinced that's what's happening though... :thinking:

Inside a apartment the radio waves should move thorough the door openings and move through the apartment quite easy.

It is the outside walls that should stop the radio waves with concrete walls.

Will buying a shielded USB 3.0 cable mitigate this? After going through this troubleshooting, I decided to just plug it in to the USB 2.0 port and 2.4 GHz is usable again. Sucks that I had to compromise my speed to drive by 30%

very possibly.

One day my entire zigbee network just went offline after working fine for months... After a full day of debugging, I discovered it was because I had to debug something on a Raspberry Pi and I had plugged a cheap keyboard into the USB3 port because the USB2 ports were already in use... leaving it sitting there absolutely destroyed the zigbee communications. immediately upon unplugging it it went back to normal.

Needless to say not only was the keyboard a thin cheap cable, but also the keyboard itself is just an unshielded antenna plane.

Using quality USB3 specific cables should help a lot.

Only some parts of my house can receive 5 GHz. Presumably ones near the router and only going through one wall. But as soon as it passes to two walls, 5 GHz has a hard time so falling back to 2.4 GHz is the only option.

After investigating, it turns out that my HDD dock with USB 3.0 was causing interference with 2.4 GHz. Plugging it to USB 2.0 makes the problem go away but at the cost of reduced performance.

1 Like

Alright. I'll see if i can contact the manufacturer that made this HDD dock and see what my options are. Otherwise, I'll have to look around for good quality shielded cables

something like this:
I'd use that kind of thing wherever you have USB3 just routinely. don't use older usb cables with 3

1 Like

I find it unlikely that a keyboard will have radio interference with zigbee with enough transmitting power that makes them stop.

A keyboard is USB1 standard device regardless of physical connection. So the keyboard will never get to the high speed frequency in combination with the high current that would interfere with either 2,4 or 5GHz. And the access point transmitter is a relative high power transmitter in regards to a bad design keyboard so a jamming transmitter on a keyboard is simply outpowered.

Did you run the zigbee transmitter on the RaPi also?
It sounds more likely to be a driver issue.

Don’t forget that Linux based O/S has a really poor hardware management, in practical terms actually non existing hardware management. So whatever things can stop working when connecting hardware to a Linux unit.

Windows worked like this before XP but is today the only O/S in the world that has all drivers for all hardware that exist and you can connect what ever you want in any hole you find and it works with everything else.

The keyboard was about 100 to 1000 times closer to the zigbee controller than the zigbee devices trying to communicate to it so with inverse square working against them the devices couldn't be heard by the controller. I promise you within less than a second after I unplugged it all the zigbee stuff began working. Although the keyboard may not operate at high frequency, a hard drive dock was in fact operating at usb3 speeds. Perhaps it's just acting like a big antenna and picking up the hard drive connection

Are the zegbee and keyboard connected to the same device?

If they are connected to the same device then move the keyboard to a completely different device. If your hypotises are correct then the zigbee device should be scrambled again.

100-1000 times closer.

What distance do we talk about in the first place?

20cm will be 20-200m away.
That is on the working edge of the zigbee system.

yeah they were probably 10 to 15 cm away from each other. The zigbee devices are 10m away so more like 100x than 1000x of course the path to the zigbee devices is likely not straight, more like reflecting off walls and going through doorways etc than going through the thick walls... so it might well be attenuation as if 1000x away line of sight.

All I can really tell you is that by plugging a keyboard into a USB3 port on a Pi4 which is acting as a file server using an external USB3 dock, and having it near my zigbee controller which is a USB connected dongle on a USB2 port, the zigbee system crashed out to the point where only about 1 in 5 to 1 in 10 commands was working. Unplugging the keyboard immediately fixed everything (after a full day of debugging). I don't really want to test further because too inconvenient.

Also I have a cluster of two RPI4s and I don't actually remember which of them the keyboard was plugged into. It might have been the same one as the zigbee or it might have been the other one right next to it.

Try to focus on radio problems and frequency allocation, for example building a mesh network. I have a mesh network with 5Ghz upling and 2.4Ghz APs downling to stations. In my other house, I have 2.4Ghz and 5Ghz APs with roaming and wired links providing uplink. This is how you can fix your problems.