I agree, education would be warranted. All protection systems seem to provide very robust power-line filtering while apparently ignoring the grounds. Failure analysis seems to involve a condemnation of the entire assembly which on the face of it seems to be appropriate as the assumption is nothing could survive so significant an overvoltage. My experience suggests the majority of the damage to circuits comes from overvoltage on the ground line which frequently destroys only some diodes or capacitors and can be repaired. Lightening striking the ground will dissipate vertically or horizontally depending on topography. Horizontal dissipation will provide a fairly significant voltage difference per unit of distance and even in salt water the gradations can be significant over as much as a quarter of a mile. Even factories that use motor-generators to totally isolate incoming power have been damaged by lightening power spikes coming in through the bonded ground line. I've never designed a lightening protection system but I have been the victim of some sorely mis-designed ones.
I design electronics for aircraft that get qualified for flight through a battery of tests called out in RTCA/DO-160. Induced Lightning is just one of the tests our circuitry has to pass. We typically shunt both lightning and RF currents to chassis ground. This is done as close to the connector as possible. Lightning isn't as fast as RF, so it's not quite as important, but we intentially make the current path to the chassis very low inductance. For that reason, you must use a ground plane, and you can't run it through a wire from the PCB to the chassis. We use standoffs, and the electrical connection must be intentional, not accidental. Types of metal and mechanical attachment all come into play because the connection must be good through temperature extremes and vibration. We don't use MOV's because they have a wearout mechanism and aren't fast enough for ESD. This would all get pretty tricky for something in a plastic box connected to household 120AC. We have had problems with our wireless router being taken out in a thunderstorm. It might be better to filter it where power comes into the building because the infrastructure just isn't there to do it at the unit.
It is worth note that lightening doesn't strike very often, and most of us aren't running 24 hour data centers.
Most industrial designers are using 3rd party power supplies that typically have designed in protection of some sort. It may not completely protect against lightening strikes, but it usually becomes the sacrificial lamb for the expensive parts downstream.
Should we design in flood protection too? I have seen more equipment damaged by high water than lightening. I say plan for the 99.99% and get insurance for the 0.01%
This is one area where the White Goods manufacturers have always been rather diligent. Controls were always specified to withstand very high level discharges and fast transient bursts. The specifications for oven controls often exceeded the capability of the oven's cabinet wiring.
Of course when it comes to huge amounts of energy there's no way to completely protect a system. I used to work on safety systems for cranes. As you can imagine, a crane extending over 100 feet in the air is going to get hit by lightening quite regularly. I used to have a collection of lightening keepsakes, my favorite was a tubular capacitor that had a small hole blown clean through it and then into the aluminum chassis that had a hole melted about 2" in diameter. The capacitor still measured in tolerance, though.
Perhaps a pragmatic reason to ignore this is a lack of understanding of where to put this current if I shunt it away from my precious circuits. Don't really want to shunt this into AGND or DGND do I ? And where exactly is I.E. or P.E. on a Beaglebone Black? Might be a good topic to stuff into a webinar?
"When you bring up lightning protection, it bothers them. They say, 'I didn't plan on spending extra money for that.' "
What a typical and even dangerous sentiment from people more worried about business than creating solid, secure products. This is a really interesting post, Chuck, on a topic it seems engineers and electronics designers often want to sweep under the carpet.I hadn't thought of it myself but it is incredibly relevant. Very creative of you to bring it to light (no pun intended!).
A new service lets engineers and orthopedic surgeons design and 3D print highly accurate, patient-specific, orthopedic medical implants made of metal -- without owning a 3D printer. Using free, downloadable software, users can import ASCII and binary .STL files, design the implant, and send an encrypted design file to a third-party manufacturer.
For industrial control applications, or even a simple assembly line, that machine can go almost 24/7 without a break. But what happens when the task is a little more complex? That’s where the “smart” machine would come in. The smart machine is one that has some simple (or complex in some cases) processing capability to be able to adapt to changing conditions. Such machines are suited for a host of applications, including automotive, aerospace, defense, medical, computers and electronics, telecommunications, consumer goods, and so on. This discussion will examine what’s possible with smart machines, and what tradeoffs need to be made to implement such a solution.