Before we begin I’ll start with the obvious warning: Working around microwave radiation is potentially hazardous and could result in electric shock, burns or death and may cause interference issues with nearby users of WiFi and other services on 2.4 GHz which could potentially be illegal in some jurisdictions. I’m sure I don’t have to reiterate that you should not attempt the following unless you are aware of the risks, blah blah blah health and safety. Anyway…
Magnetrons are great for outputting large amounts of microwave radiation without resorting to more expensive means hence the reason we cook food with them. Unfortunately their design makes it hard for one to simply attach a coax cable to the end of one and couple it to an antenna. One could mount the magnetron to a length of commercially available waveguide with a waveguide to coax transition but this method is expensive and requires specialized parts. The efficiency however would be near 100% compared to any other means that one could come up with on the fly. It has been said that when powered from a clean source of DC power and not the half wave rectified supply provided in a typical microwave oven, the RF output from a typical household microwave magnetron is surprisingly clean and could be potentially used for communications provided one can build a circuit to modulate the magnetron output.
For experimental purposes I have taken an old 700 watt microwave oven and drilled holes to mount a panel mount N connector on the top of the cooking cavity. This was done to leave the oven and its components (fan etc) untouched. It also avoids having a 2 kV AC supply just sitting in the open unshielded where it could potentially be touched while in operation (trust me it hurts like a bitch).
The center pin of the N connector has a 3 cm strip of thick copper soldered to it. This pickup “probe” needs to be fairly hefty to avoid melting when the oven is turned on. I used the core of some RG6 coax which immediately vaporized the moment the magnetron started up. From this I connected a 3 meter length of RG213 coax to a 1/4 wave ground plane tuned to 2.45 GHz. RG213 is needed in this case as anything thinner will probably melt or catch fire. To be honest I imagine the SWR in my setup is horrendous and contributing to the heating of the coax in this case.
Bringing fluorescent tubes and light bulbs near the antenna would cause them to light up within around 1 inch from the antenna. Touching the glass against the tip resulted in arcing which does an excellent job at blowing pin holes in the sides of lamps along with wood and most other items that come into contact with it. As to the amount of power that makes it into the coax, I’d say it would be less than 25% given how hot the coax becomes after a minute of operation. Picking up the cable during operation isn’t advised either, doing so provides an extremely weird and somewhat unpleasant warming sensation similar to that of placing a hand near a heater but with the sensation of heating right the way through ones hand. Pleasant during the winter, not so much when its 30°C in the middle of a summers day.
Testing to see the effects on WiFi networks, although no signal strength decrease was noted there was a significant drop in throughput while the oven was on and an antenna connected. In the screenshots below you can see that with the oven switched off latency was 19 ms with download speeds of 27.32 Mbps and upload of 24.93 Mbps. With the oven switched on latency increased by around 50% and both download and upload speeds were pretty much zero. Trying to load webpages and apps that require an internet connection failed to work while the oven was on but began to work as soon as the oven was switched off. This was tested in 1 meter increments from the antenna and at 30 meters which was as far as I could get from it and at this distance the connection was unusable but remained connected.
Note that the tests I conducted were against my own WiFi network and I do not advise using this information to disrupt wireless networks that aren’t yours. I can only imagine how disruptive it would have been to analog video senders that use the same 2.4 GHz. Those I advise you to jam as they obliterate nearby wireless networks, Bluetooth and cordless phones. Besides who still uses composite video in 2018 to connect things to their TV when nearly everything has an HDMI port. You never heard that from me. No idea how effective it is against drones being flown over your property although a 12 gauge shotgun is said to be an effective means of eliminating them.
EDIT: 19/02/2018, did some testing regarding jamming drones. I live nearby a local high school who strangely have at least 2 drones for students to use for projects strangely enough. Just happened to be in the shed when two drones were launched from the school and so turned the oven on and put the antenna outside on the roof. All was well until they neared my boundary fence when one auto landed near the fence and the guy flying it looked as though he was struggling to get it to respond while the drone that crossed my boundary looked as though it got stuck in a hover about 30 meters above the shed until the oven was switched off as you could see the guy flying it also looking as though there was issues with the controls. Both drones began to operate normally after the oven was switched off. Although the evidence is inconclusive, it looks promising. If anything this could work as a passive drone deterrent taking advantage of the “return to home” and “hover in place’ functions that most of these consumer drones have when they loose contact with the controller. Basically a way to make them piss off and stay away without destroying them assuming they use 2.4 GHz for control/video.