Lightning and static damage to computers/electronics

[dakels]

What exactly happens when an electric discharge or some sort affects computers or electronics? For example, a lighting strikes near my house and the power goes off and now my stereo won't work or the computer works but is very glitchy. What are the possible and common effects and problematic symptoms?

What about static from your hand when it dischrges on a RAM chip? What is happening there on the chip or circuitry that causes it to lose function?

 

[Akira13]

The voltage potential in your body over the "zero" voltage of grounded electronics can be in the kilovolt range (causing the spark when you touch a doorknob). Now imagine that your RAM (or whatnot) can have a resistance of only a few ohms. Since I=V/R, you're going to have many, many amperes of current running through the component. That'll kill it.

As for a lightning strike and your stereo... that could be several things. The first is that the lightning caused a surge in your household circuit. If your circuit breakers tripped, there's a chance that it could've overloaded your stereo (assuming no surge protector). The other possibility is that the lightning caused an electro-magnetic pulse that destroyed your stereo. A lightning strike near me caused my computer speakers to pop (but not die). I'm not exactly sure how EMP's kill electronics (yet... first year EE student).

Anybody wanna help me out?

 

[f95toli]

An EMP pulse will induce currents in anything conducting. The mechanism is the same as when a radio-trasmitter sends our a signal that induces currents in the antenna of you radio. You need farily strong pulses to burn modern electronics but I know that for example the US is using EMP-bombs to attack hostile radar.

There are a few devices that can be very sensitve to EMP, high-gain amplifiers (x100000 or so) are sensitive since they will try to amplifie whatever signal is generated by the EMP, effectivily burning themselves.

I know a guy who burned an electronic component by accidently turning on a flash-light in the same room. The EMP pulse destroyed the component, he is on the other hand working on some VERY sensitive electronics.

 

[dakels]

you're losing me guys
I know jack squat about electricity and electrical components.

OK without making you guys go into EE 101, lets look at this in a more practical regard. What parts of an electrical device are getting damaged when it is affected by an overpowering electrical current, like a lightning bolt, or an EMP pulse? I know wires are rated to carry a certain max load before they literally melt down, right? But even if they do melt, aren't they still going to carry that max load thru the wires before they melt? So they are still transferring that current even if it kills them.

Lets put this in regards to a computer. Say a bolt of lightning strikes near my house, the computer is on a surge protector. Now as I understand it the first line of defense would be a circuit breaker in the wall, that should stop an overflow of current? then the surge protector should stop an overflow as well, then the power supply in the computer should have a breaker as well? Only after that it would go to damage the motherboard and components yes? Now if the power kicked off in my house a split second after the bolt struck, I can assume that the fuse/breaker box shut off because of the overload. So it shouldn't have even reached my wall jack and to my power strip, correct? That should stop the electrical current but could the EMP pulse from the lightning bolt also harmfully effect my computer as well? I know many situations of people getting a nearby lighting strike and their computers acted glitchy ever since then. I would assume that was from the EMP pulse. Maybe it damaged the hard drives or possibly the controller chips that contain small amount so f electrically stored data, such as the bios. Am I making any sense?

and btw I imagine your stereo speakers popped because the EMP caused a flash of a current that made your cones react?

 

[Akira13]

lets look at this in a more practical regard. What parts of an electrical device are getting damaged when it is affected by an overpowering electrical current, like a lightning bolt, or an EMP pulse?

Like you said, wires, but consider the fact that circuit boards (PCB's) are simply wires and other electrical components "drawn" onto a stiff surface. So if you melt a wire, there's a chance you'll melt the PCB connected to it as well. (Here's some EE215) Then there are capacitors, which can "blow up" when exposed to too much current. Inductors will melt when exposed to prolonged periods of high current. The power dissipated through a resistor is proportional to the current squared, so they can dissipate lots of power if you have lots of current (causing melting). Diodes and other semiconducting devices can be damaged (I'm not sure exactly how) as well.

if the power kicked off in my house a split second after the bolt struck, I can assume that the fuse/breaker box shut off because of the overload. So it shouldn't have even reached my wall jack and to my power strip, correct?

Not exactly. It takes a split second for the breaker to trip (including that in your surge strip and PS), and so there is a chance that the overload could reach your computer. If you have one of those surge protectors that gives you some $$$ if something goes wrong, then you're in luck.

computers acted glitchy ever since then. I would assume that was from the EMP pulse.

Perhaps. Like f95toli said, the EM wave can cause current to flow in wires (I never considered that... thanks, f95toli). Also, it can magnetize certain things, or change the magnetic orientation in certain things (like hard disks). That could certainly do it.

Funny thing is that I'm studying for my EE215 final right now, so this "playing on the internet" thing has been somewhat helpful.

 

[f95toli]

I don't think you have to worry too much about inductors and other passive components. If a chip is destroyed due to an EMP pulse it is probaby because one or several transitors were burned.
A transitor is made up of several layers, at least one layer is a dielectric (an insulator, for example silicon-oxide). This insulating layer can only withstand a certain electric field (which is proportional to the voltage divided by the thickness of the layer) before it breaks. If the voltage over this layer is too high the currents will burn (litterally, you can see the burn-marks in a microscope) through the insulator, this will either short the transitor or cause what what is known as leakage currents (depends on the technology and transitor type).

 

[Mark R]

A transitor is made up of several layers, at least one layer is a dielectric (an insulator, for example silicon-oxide).

Not all transistors are made like this - those without the insulating layer are not easily damaged by static. However, most high-speed modern electronics are built on a CMOS process, where field-effect transistors are used. These are highly static sensitive.

I've got a couple of heavy-duty MOSFETS lying around, they claim to be able to switch several hundred amps at 100 volts. Despite they're apparent strength, they are actually highly static sensitive - I destroyed one, by applying 30 volts to the 'gate' pin (the insulated one).

if the power kicked off in my house a split second after the bolt struck, I can assume that the fuse/breaker box shut off because of the overload. So it shouldn't have even reached my wall jack and to my power strip, correct?

The breaker did trip because of overload - however, these breakers have a built in time-delay before they trigger - things like lightbulbs and motors, take a huge surge of current as they warm-up/spin-up: a 500 W halogen bulb can take 70 A at switch on. Without this delay, they would be to prone to false triggering.

Any mechanical breaker is not going to be quick enough to protect against a lighting strike. You need to use a surge supressor. The most common type of surge suppressor is called a metal oxide varistor (MOV). It is essentially a resistor, where the resistance changes with voltage. At low voltages (less than e.g. 500 volts, the resistance is very high - virtually infinite), but at high voltages (e.g. 1000 volts) the resistance becomes very low, so it draws a huge amount of current and therefore absorbs the energy from the surge. As you might expect, they dissipate this energy as heat, and during a major surge may well explode or be burned.

A power surge is simply an excessively high voltage on the line, for a short time. These cause damage by burning the components that are connected to the supply. In a stereo system, this might be a small transformer.

In a PC PSU, there are delicate transistors connected directly to the mains - these will withstand up to about 600 volts, but will fry instantly with much more. Therefore, all PC PSUs have surge supressors built in to protect the circuits, although none would be adequate to absorb a nearby lighting strike. In a PC PSU, if the main switching transistors are damaged, the most likely effect is for the PSU simply to die, although it is conceivable that some designs may carry on working, just very badly. This may explain a malfunctioning PC.

In fact, in installations where power quality is critical you often need several levels of surge suppressor:

1) - Main suppressor (Very high surge rating, ultra-fast acting) at point where power cable enters the building.
2) - Circuit suppressor (High surge rating, fast acting) at the room supplied by each individual circuit.
3) - Power-strip type suppressor (Low surge rating, slow acting) for each individual PC, appliance, etc.

The main suppressor absorbs the energy from the main lighting strike. However, as it absorbs it, the resulting EMP causes 'ringing' in the nearby circuits, causing surges in those circuits. Therefore you need a seperate suppressor on each circuit, as well. Similarly, this suppressor causes ringing in the wires to each individual socket outlet. Not that the most powerful suppressor has to be the fastest - If it was too slow, the other surge suppressors would try and dump it, and would probably fry in the process.

 

[everman]

I think this is somewhat related:
I know you can short out electronics as discussed above, but I"ve also heard damage can be done by supplying too little voltage such as to a motherboard? That is to say if the power for some reason didn't entirely cut out, but not enough voltage is getting to it for it to run correctly. Why would something be damaged because there isn't enough electricity?

 

[DeathByDuke]

Anyone got any pics of it happening or has happened?

 

[dakels]

This is all good stuff. I'm learning alot but still have alot of questions.

Anyways here is is what I have witnessed:
Worst case was a person who's house got a direct lighting strike. They were a neighbor of my friend who's house I was over while this happened. There was electrical damage ALL over the place. Wires inside the wall were fused and melted. You could see burn marks all along the walls. The wiring in the house was new (5 years or so) with safety wall jacks. Still most of their plugged in electrical devices such as stereo, computer, TV, appliances, were ruined. Most of them wouldn't turn on at all. The fire marshall stated it was the worst case of lightning damage he had ever seen.

More typical stuff I have seen was computers that got some sort of electrical damage.
I was standing next to someone when they picked up a RAM DIMM chip. When she touched it I heard a small crackle and the chip didnt work after that. I would assume it was the static discharge that was mentioned above.

Another was a friend who brought me a computer. They said they had a lighting strike nearby and the power went off. Ever since then he has had some major problems with his computer (Dell machine) and his stereo. The computer crashes on startup, video problems, such as it won't sync correctly (the monitor was fine on other machines though), all sorts of odd problems. I had him just send it back to Dell. The stereo's problem sounded to me (I didn't directly see it) that the amp had burned out or somehting.

I also had another friend that had some sort of electrical problem during a lighting storm. His power cut off, presumably due to a lighting strike, now his computer was acting up along with other things in his house, even appliances. He had alot of things replaced by his homeowners insurance.

Most recently my friend's Mac G4. The power in my house cut off due to a felled tree on a power line in the neighborhood. My machines are on a battery backup. His machine downstairs was not but it was on an APC surge strip. He turned on the machine later and all went well for a minute then he saw smoke coming out. He then powered it down and it would not restart again. I looked inside and the firewire controller chip had literally burned. The chip itself was scorched thru and buckled in spots. The PSU was not respondng either.

So here are some examples of rel life instances. I just want to understand what is happening and how to prevent it. Do you think my battery backup (APC smart-ups) was better protection then the APC surge strip? They both have fuses or some sort I noticed. I think the strip did but maybe not. I know for certain my UPS has one. Also in a related situation, the electric heater reset my fuse box last night again. What's happening there? No other items were on that wall jack but that "fuse switch" is for 2 small rooms which have outlets of course. There was other stuff (computer) on the other outlet that is part of that breaker. I guess that means I am adding load on the breaker that the jack the electric heater was on, which I assume is a bad idea.

Also when you mention the damage to the capacitors in a chip, could it be just some damage that would screw it up, BUT still be able to partly function, or is it like all or nothing? Or could the BIOS chips be getting the data "scrambled" in a sense from the EMP?

 

[everman]

What if you had 2 surge protectors (real ones, not glorified extension cords), one plugged into the wall, one plugged into that one, and then put a ups plugged into that? Should be safe But you'd have to make sure nothing else connects the pc to an unprotected power source, which can be phone lines, speakers, network cable. You could use a wireless network I guess.

 

[dakels]

like I said, I know nothing about electricial junk but I would think that adding multiple surge strips woudn't really help that much if I understand what was said before. O unless they are suppressors like Mark R was talking about. I guess it would depend on the amount of excess voltage is coming thru the line and how much each suppressor can "dampen". I guess if you had a series of 10 power strips that all could absorb 1000 volts a piece then by the last strip you could have absorbed a 10,000 volt charge? (not including impedence and stuff?) and as for the wireless, i guess that would be a good way to protect the system but those systems are also super sensitive to EM interference. I doubt it would damage it though.

sorry but I had to try and utilize what I just learned. Hope I am not totally off the "mark".

 

[Mark R]

Why would something be damaged because there isn't enough electricity?

I wouldn't have thought a PC motherboard would be damaged by insufficient power - I would simply expect it to operate incorrectly until adequate power was provided.

However, some devices can be damaged by low power - examples include fridges and AC units. It is quite common for these devices to burn out during a 'brown out'. The reason is because, when the motors in the compressors switch on, the work to turn the compressor is very high, and the motor needs to overcome this (once the compressor is actually running, the load on the motor falls). During a brown out the maximum torque the motor is capable of is reduced, and the motor may simply stall while attempting to turn the compressor. The amount of electricity a motor uses depends on load - if the motor is stalled the power draw is massive. The resulting massive power draw, and the lack of movement to cool the motor leads to rapid burn out.

I guess if you had a series of 10 power strips that all could absorb 1000 volts a piece then by the last strip you could have absorbed a 10,000 volt charge?

It doesn't quite work like that. Surge suppressors are rated in Joules (the amount of energy that they can safely dump) - they will, however, try to dump any surge that appears on the line - if they are not big enough then they burn out in the process. Once the device burns out, then it can no longer offer protection - so a large surge will destroy a surge-strip before most of its energy has been dumped. You do gain extra protection by having additional suppressors - you don't even have to connect them in series (or even to the same socket outlet). However, maximum benefit is gained by having a suppressor as close to the appliance as possible.

A typical power-strip type surge suppressor will typically have a rating of 2,000 joules (although the advertising will often quote more impressive figures like 35,000 amps - which is how much current it can safely take while absorbing a surge).
A massive surge protection unit (suitable for a whole building like a data centre) may have a rating of 5,000,000 joules or more. This would survive something only slightly short of a direct hit by lighting. You could have as many power-strips and UPSs as you want, but there is no way that they would survive a similar surge. The real question is how much you want to be protected against.

 

[Akira13]

Originally posted by: everman
I think this is somewhat related:
I know you can short out electronics as discussed above, but I"ve also heard damage can be done by supplying too little voltage such as to a motherboard? That is to say if the power for some reason didn't entirely cut out, but not enough voltage is getting to it for it to run correctly. Why would something be damaged because there isn't enough electricity?

I asked a friend in my EE class about this today. He came up with an operational amplifier circuit that would be prone to damage by low voltages. The non-inverting input would be connected to a really high (kilovolts) voltage source, and the output would be some equipment. The feedback loop to the inverting input would be a low resistance (~0), and the inverting input itself would be the small voltage in question. For an ideal op-amp, if the voltage supplied to the non-inverting input was too small, it would cause a huge current to run through the output and the feedback loop. Of course, this is all idealized.