Microwave direct energy beam and its effect on electronics?

[SMdude]

So, do these failures always happen to the same location/installation, or just in a small area?
Have you thought about installing some dataloggers to record abnormal voltage events on the mains power inputs? Temperature, humidity?
How about some hidden cameras?

I think if you did that, you would at least prove if it is sabotage or not, which I highly doubt it is, but you just never know!!

[chris_leyson]

protection diodes in chips will act as wideband RF detectors and that will have an effect.

Thanks, you mean they will be damaged?

Hi Treez, no I don't think the protection diodes would be damaged but I think I need to think a little bit a more about my blunt statement about protection diodes and RF immunity. It all depends on the silicon, is it bipolar or mos, how are the protection diodes implemented in the silicon but more importantly what is "down stream" after the protection diodes and how does that handle RF pickup. It's not something that is documented, well not in datasheets anyway, but the field strengths would have to quite high to have any effect and that's quite likely at microwave frequencies. Maybe some RF immunity testing is in order just to prove or disprove the possibility.

[radar_macgyver]

The good news is that you can't use multiple magnetrons coupled to the same cavity because they will all work at slightly different frequencies so I think we're all safe for now.

Well whichever magnetron fires first will back-feed energy to the rest of them, which will injection-lock to the first. This effect was exploited in the crossed-field amplifier, which is really just a magnetron with an extra port.

[SeanB]

Multiple magnetron microwave oven do exist, and are common in industrial use as driers for things like laminates, where they have a number of them in a broad antenna box and feed the laminate through, with a stub filter cavity on the inlet and outlet side to reduce radiation.

[stj]

also i said you pulse-drive them in sequence to reduce the duty cycle - you never have more that one emmitting at the same time.

[floobydust]

Inducing a forward-current or reverse-voltage enough to damage a several watt LED, through air... need several kW of microwave power long duration, enough for the thermal effects?

It would be difficult to generate in a mobile rig. Granny would notice a sinister truck driving around her front street and surely call the authorities.

My city uses one dusk sensor to turn on an entire string, several blocks worth. These are crap CdS sensor and thermal-relay units that oscillate (blink on and off) at dusk.

The only way to take out a string of streetlights, as an evil overlord  would be applying an overvoltage at the branch feed.
But really, probably a design/manufacturing/site weakness causing failures. Wonder what the warranty period is.

[eelearn123]

You seem to be in a somewhat tricky situation here. Speaking from experience in other areas and dimensions of life I would say that the truth often is less nefarious that one may think.

Anyway, maybe a solution here would be to build some units with additional logging regarding supply voltages and so on and installing in these troublesome locations. Maybe you could gather additional info this way.

Also, maybe having some external consultant look over your designs for weaknesses.

Good luck!

EDIT: also maybe check out some of the failed units. In what manner have they actually failed?

[stj]

you need to try to eliminate or reduce useage of electrolytic capacitors,
and remember that these things are stuck in the sun all day and exposed to the wind all night.
they need to handle exreme high and low temperatures without (for example) losing control of the current limiting.

i would at the very least, try to test a few by putting them in the freezer for a few days before powering on,
also heating them with a fan heater or hot-air gun for a while till they are too hot to touch before running.

better yet, send them to someplace with a real enviromental testing chamber.

[CopperCone]

ET does not want to be seen, clearly.

[vk6zgo]

inverse-square?
a microwave beam can travel hundreds of miles and still be pretty focussed - that's why they are used as point2point links
reflection is another issue.

Nope!
The signal from a dish antenna as used in point to point links is not really "focused".
It starts out as a point source, is reflected by the dish and ends up a wavefront which appears to come from all parts of the dish.

It is easiest to describe this using the tools of Classical Optics, showing the signal path inside the dish and beyond as individual rays.
The rays are initially radiated in all directions by the antenna feedpoint, then are converted by the dish curvature so they are all effectively travelling in the same direction, parallel to each other.

The wavefront does diverge over distance, so the "rays" are no longer truly parallel at the receiving end, but there are a sufficient number which are "near enough" for practical purposes.
At this point, the dish antenna intercepts some of the "rays" & focuses them back to a point at the antenna feed point.(this is the only time any "focusing" takes place)
 
OK, the "rays" are just a convenient way early Optical researchers invented to allow them to draw those nice diagrams in your High School Physics books, but they make visualisation easier.

Point to point microwave services usually top out at about 30miles or so.
100 miles is really stretching things.