[jrockway]

> Since WiFi and a microwave operate at the same frequency (2.4GHz), we ran our leaky microwave at full power to block out any persistent network signals in the air.

Incidentally, WiFi tries to intentionally avoid this interference. Microwaves output no power during the zero crossing of the AC line that's driving it, and in this interval, there is no signal in the air to jam things. WiFi listens before sending (so as to avoid stepping on other stations), and the microwave's signal is enough to trigger this. (I forget if microwave ovens are "half wave" and you get 1/120th of a second 60 times a second, or if there is just a threshold near the zero crossing where there isn't enough power to interfere.)

I would say it's likely that the microwave oven didn't really do much here.

[subarctic]

Whatever the theoretical analysis tells you, I've been able to reproduce a microwave interfering with a wifi signal with at least one microwave and router. I've had other times where it didn't have a noticeable effect though.

[jrockway]

The key to happiness in WiFi is that all the stations have to "hear" each other, or the listen-before-talk algorithm can't work and you end up stepping on valid transmissions that you can't hear. This ruins it for the stations that are the targets of that transmission but can also hear the first station. This is why WiFi tends to degrade at longer ranges; with a topology like <computer A> <---> <access point> <----> <computer B>, both computers can hear the access point, but can't hear each other. This means that they step on each other when talking to the access point; when this happens, the access point sees the sum of the two signals which is has to discard as garbage. As a result, whenever you see enterprise WiFi that actually works, you'll probably be able to see a ton of access points covering a large room. This is so that they can transmit at low power, causing devices in the above topology to roam to a different access point before they enter the failure mode of not hearing other stations connected to that access point.

Now that I think about it, in OP's case, it's quite possible that the iPad can "hear" the microwave, but the access point can't, so the access point will send out its SSID broadcast while the microwave is interfering. This is great because you WANT that packet to get corrupted. So maybe the microwave does help!

[shagie]

> As a result, whenever you see enterprise WiFi that actually works, you'll probably be able to see a ton of access points covering a large room.

IETF attendees reengineer their hotel’s Wi-Fi network - https://www.computerworld.com/article/1448494/ietf-attendees... ( https://news.ycombinator.com/item?id=3771876 31 comments)

> “There was no WiFi signal when on the desk in front of the window in my room, but after some experiments, I discovered that the signal was quite good… on the ceiling of the bathroom,” emailed Marc Petit-Huguenin.

> “I have a Nexus S phone, so I taped it on the ceiling of the bathroom, and used tethering over Bluetooth to bridge the gap to the desk,” he explained. This is a slow connection, but good enough to send emails over SMTP or use vi [the popular Unix text editor] over SSH.”

> ... Working behind the scenes, a team of IETF attendees negotiated with the hotel and were granted access to the wireless network by Sunday night. ...

> The changes made by the IETF makeover team included:

> – Decreasing the AP receiver sensitivity ([changing] HP/Colubris configuration “distance” from “large” to “small”);

> – Increasing the minimum data and multicast rate from 1Mbps to 2Mbps;

> – Decreasing the transmit power from 20dBm to 10dBm;

> – And, turning off the radios on numerous APs to reduce the [RF] noise.

> ...

> Each floor now has approximately two access points on each of these four channels, with the channels staggered on adjacent floor. That design maximizes the distance between access points on the same channel. “I hope this will significantly improve the coverage in some rooms that had marginal or no signal while also improving the signal to noise ratio for all,” he said

----

Note that the changes were being made to decrease the power being used.

[withinboredom]

You'll also see this in wireshark as retransmissions of packets. It's really great for discovering that a service you've written isn't filling packets and can only get 50-ish bps throughput due to head-of-line blocking and sending lots of tiny packets.

[xanderlewis]

I’ve had a microwave oven interfere quite clearly (correlated with turning it on and off) with my AirPods before.

[anonomousename]

Fun story: I’ve had a speaker which would connect to a Bluetooth device as soon as it was turned on, and only enter pairing mode if a paired device wasn’t in range. I couldn’t figure out what device it was connecting to, so I put it in the microwave (while it was off!) briefly to trigger the pairing timeout.

[rollulus]

> Microwaves output no power during the zero crossing of the AC line Why is this? Do microwaves by design modulate their 2.4gHz on top of 50/60Hz?

[RF_Savage]

The magnetron needs about -4.4kV to work and food does not care about the purity of the signal, only net energy delivered.

So the transformer based microwave oven power supplies have a 2.2kV transformer and then double that to 4.4kV for the magnetron.

There is no filtering or smoothing, as those parts would cost money and present a danger to the service technician.

So the voltage feeding the magnetron is not even sinusoidal.

[jrockway]

Yup, exactly. Compare the price of a PC power supply, which outputs a very smooth constant voltage, to a microwave oven. The microwave oven is cheaper, and uses all 1800W of your circuit. A DC power supply that does that is much more expensive.

It is truly amazing how cost-optimized microwave ovens are.

[RobotToaster]

So two leaky microwaves on different mains phases would block (2.4ghz) wifi?

[Etheryte]

In theory, yes, but in practice microwaves messing with wifi is such a well known phenomenon that there's an XKCD about this.