I am wondering how a surge affects a computer. Can anyone give me a detailed report of what happens. What would go wrong and why? Lets say to make things more interesting what if the surge was from a short circuit and the power supply really sucked. Yes it happened to me, but I am wondering what happens inside the computer and why.
How do power surges affect a computer?
- Thread starter d3lt4
- Start date
It can wildly depend. Power coming in can overload whatever it touches if it isn't processed properly. For instance, your modem is designed to have a particular level of voltage on the phone line. If the phone line is suddenly carrying a significantly larger voltage, then modem's design is not going to be sufficient to insulate nearby parts. So now you have electricity in unexpected places, in unexpected amounts.
Like in your case, a short circuit basically takes a LOT of the resistance off of a circuit. This creates a big surge of power. A switching power supply is designed to convert 60 hz 100 volt AC into various DC voltages. The transformers that do this, the capacitors to bridge over dips in the line power, all these are rated to handle a certain amount of energy at any given time. You go outside that range, and you overload the parts. Maybe marginal parts fail and die. Maybe they let through more energy than they should. In some cases, maybe things catch on fire. Cheaper PSUs will be using cheaper components - the upshot of this is the components have weaker tolerances, and don't handle going outside their tolerances as gracefully.
The real severe internal damage comes when the input power causes the power supply to push too much electricity into internal parts -- think about it like a mob pushing through a fence, and at some point the fence goes down under the weight. Instead of a few people a minute going throug a gate, you have a crowd surging forward. Some really delicate components -- mainly processors and memory -- are actually damaged by this, and get bricked. Simple parts can be killed too: You take a resistor and dump too much juice into it, it heats up and burns. You dump too much power into an electrolytic capacitor, and it can burst and spew electrolyte fluid all over the place. Think what would happen if your lawnmower suddenly got some super-high-power jet fuel or something. It's just not designed to handle that.
Or, so is how I've always understood it. Electronics experts can probably be MUCH more accurate. Guys with soldering irons, am I out of line here?
Like in your case, a short circuit basically takes a LOT of the resistance off of a circuit. This creates a big surge of power. A switching power supply is designed to convert 60 hz 100 volt AC into various DC voltages. The transformers that do this, the capacitors to bridge over dips in the line power, all these are rated to handle a certain amount of energy at any given time. You go outside that range, and you overload the parts. Maybe marginal parts fail and die. Maybe they let through more energy than they should. In some cases, maybe things catch on fire. Cheaper PSUs will be using cheaper components - the upshot of this is the components have weaker tolerances, and don't handle going outside their tolerances as gracefully.
The real severe internal damage comes when the input power causes the power supply to push too much electricity into internal parts -- think about it like a mob pushing through a fence, and at some point the fence goes down under the weight. Instead of a few people a minute going throug a gate, you have a crowd surging forward. Some really delicate components -- mainly processors and memory -- are actually damaged by this, and get bricked. Simple parts can be killed too: You take a resistor and dump too much juice into it, it heats up and burns. You dump too much power into an electrolytic capacitor, and it can burst and spew electrolyte fluid all over the place. Think what would happen if your lawnmower suddenly got some super-high-power jet fuel or something. It's just not designed to handle that.
Or, so is how I've always understood it. Electronics experts can probably be MUCH more accurate. Guys with soldering irons, am I out of line here?
Knowing how to solder has nothing to do w/ knowing how electricity works. I'm an EE student none of my classes helped me learn how to solder (picked up that skill over summers). That point aside. I think cker's explaination is quite complete. We can get into a more techinical analysis but it's late and my brian doesn't want to think... =)
many ways, here's what ppl run into the most, a voltage spike, where the voltage unexpectedly increases dramatically over a small time.
high voltage makes the insulation useless, effectively speaking, the "capacitor" model of 2 wires next to each other comes into play. at the moment of a voltage spike, there is a discharge across 2 wires. This discharge actually causes a current spike. The current increases the temperature of the location of the spike causing burns and burn marks. It's like a small explosion. a wire can also be a trace on a board. typically we see voltage spikes causing current spike.
V = IR, that is, the voltage is the current times the current. if you take both sides to change over time, you get change in voltage over time is the change in current over time multiplied by the resistance which is constant over time. the voltage spike is a high change in voltage in time, resulting in a high current change over time, which is a current spike. typically this is what a power surge is, where P = IV, power is the current times the voltage. (all this crap is in magnitude).
the amount of overvoltage depends on the design and materials used in the circuit. the voltage spike is also why circuits get damaged by static electricity. The discharge, when you touch something, is in the order of a few hundred to a few thousand volts over a very small period of time (electrons travel at the speed of light).
why am i talking about current? because for a fixed wire thickness, the more current you run across it, the HOTTER it gets temperature wise. Thicker wires have more mass to dissipate the heat. this is why cables are rated for current (amps) and not voltage. the insulation is rated for the voltage. also, as the temperature of most insulation rises, it's ability to insulate decreases. so, basically this is a lose lose proposition. your current increases, and the temperature of the wire increases. since the insulation touches the wire, the temperature of the insulation increases. The ability of the insulation to insolate (prevent shorts) decreases. So, after a while, the insulation breaks down, and a short happens which causes a power surge through the point of insulation failure, and the insulation catches on fire, as well as anything near the point of insulation failure. this is how most electrical fires happen. the wires get too hot, and the insulation becomes useless. the damage is over time, since insulation doesnt really repair itself. it kinda melts a little if the wire gets overheated repeatedly.
high voltage makes the insulation useless, effectively speaking, the "capacitor" model of 2 wires next to each other comes into play. at the moment of a voltage spike, there is a discharge across 2 wires. This discharge actually causes a current spike. The current increases the temperature of the location of the spike causing burns and burn marks. It's like a small explosion. a wire can also be a trace on a board. typically we see voltage spikes causing current spike.
V = IR, that is, the voltage is the current times the current. if you take both sides to change over time, you get change in voltage over time is the change in current over time multiplied by the resistance which is constant over time. the voltage spike is a high change in voltage in time, resulting in a high current change over time, which is a current spike. typically this is what a power surge is, where P = IV, power is the current times the voltage. (all this crap is in magnitude).
the amount of overvoltage depends on the design and materials used in the circuit. the voltage spike is also why circuits get damaged by static electricity. The discharge, when you touch something, is in the order of a few hundred to a few thousand volts over a very small period of time (electrons travel at the speed of light).
why am i talking about current? because for a fixed wire thickness, the more current you run across it, the HOTTER it gets temperature wise. Thicker wires have more mass to dissipate the heat. this is why cables are rated for current (amps) and not voltage. the insulation is rated for the voltage. also, as the temperature of most insulation rises, it's ability to insulate decreases. so, basically this is a lose lose proposition. your current increases, and the temperature of the wire increases. since the insulation touches the wire, the temperature of the insulation increases. The ability of the insulation to insolate (prevent shorts) decreases. So, after a while, the insulation breaks down, and a short happens which causes a power surge through the point of insulation failure, and the insulation catches on fire, as well as anything near the point of insulation failure. this is how most electrical fires happen. the wires get too hot, and the insulation becomes useless. the damage is over time, since insulation doesnt really repair itself. it kinda melts a little if the wire gets overheated repeatedly.
PowerEngineer
Diamond Member
- Oct 22, 2001
- 3,583
- 756
- 136
Let me add a few clarifying remarks.
What people generally mean by a power surge is really a sudden brief increase in voltage. Short circuits actually drop voltages because of the increased current flow they trigger. Voltage surges can be caused by lightning strikes, conductor failures that drop a higher voltage line onto a lower voltage one, certain line switching operations, etc. Even though surges usually last only milliseconds (and are somewhat filtered by the transformer in the power supply), that high voltage can exceed the withstand voltage of the insulating layers inside the integrated circuit chips. This current that flows then drills a conducting path through the insulating layers, and that conducting path remains even after the voltage returns to normal. "Your chips are fried."
What people generally mean by a power surge is really a sudden brief increase in voltage. Short circuits actually drop voltages because of the increased current flow they trigger. Voltage surges can be caused by lightning strikes, conductor failures that drop a higher voltage line onto a lower voltage one, certain line switching operations, etc. Even though surges usually last only milliseconds (and are somewhat filtered by the transformer in the power supply), that high voltage can exceed the withstand voltage of the insulating layers inside the integrated circuit chips. This current that flows then drills a conducting path through the insulating layers, and that conducting path remains even after the voltage returns to normal. "Your chips are fried."
Power Engineer summed it up best. I'm an industrial electrician. Basically, power surges refer to either transient voltages, or inrushes of current caused by either a failed neutral or a phase-to-phase condition in your supply conductors.
Simplified: parts of your computer's power supply (transformer section coils) will melt. This will cause internal resistance to drop severely, thus increasing current inrush. i do object to one specific element of power engineer's statment:
your chips will actually frikasee, not fry, preferably to a golden brown color. :laugh:
Simplified: parts of your computer's power supply (transformer section coils) will melt. This will cause internal resistance to drop severely, thus increasing current inrush. i do object to one specific element of power engineer's statment:
your chips will actually frikasee, not fry, preferably to a golden brown color. :laugh:
thanx for the input, but on my cp nothing terrible happened. Nothing fried, or (frikasee'd?) It just won't overclock any. What happpened there? did the ground wire in my psu take out all the extra voltage? Why would it not fry anything but leave a few glitches?
A friend of mine has a whole house surge protection system. There are 5 or 6 LED readouts that show instant average voltage (that's what it says!) it's always 119-120 volts. There is another that shows spike/surge voltage, and that one is the one to watch, the numbers on that readout are pretty scary. A kind of neat feature is where you can "play back" the last 20 or so surges, and see the voltages come up one after another. The numbers on that display are pretty big.
There is an ammeter readout that tells you the current being drawn, updated once a second, along with average current.
There is a readout that has a power LED, a "fault", "ground fault", EFI filter status, and surge status.
When there is a storm and the lights flicker, it really makes some odd noises. It has a nice low 60 Hz hum that gets very loud when surges hit.
There is an ammeter readout that tells you the current being drawn, updated once a second, along with average current.
There is a readout that has a power LED, a "fault", "ground fault", EFI filter status, and surge status.
When there is a storm and the lights flicker, it really makes some odd noises. It has a nice low 60 Hz hum that gets very loud when surges hit.
something I've had happen to my local boxes.
terrible lightning and thunder storm durng the night.
I forget to unplug cat5 from doxport cable modem. (in comes a surge thru the rg6 cable! oh yes!)
later in the morning I discover:
1 damaged network switch (LEDs still worked, but it wasn't functioning properly at all).
2 killed ethernet network PCI cards (both in a different mobo, both mobos still working fine today).
1 killed internal mobo ethernet port. Motherboard still works to this very day.
the cable modem itself survived, and still worked fine. Kept it for around a year, but ditched it, and got a newer one.
it probably could've been far worse, but it wasn't.
other bad stuff happened that same night, heh, the furnaces PCB got fried. Black marks on it. The surge for that came thru lightning striking and obliterating the top of a metal light post outside (quite loud, one of the outside dogs flipped out). In came the surge, which actually blew off the plastic panel in the kitchen, and destroyed the light switch. Some smoke and smell, but no fire I saw. Insurance is your half-friend. In that regard, I'm happy I don't live there anymore.
terrible lightning and thunder storm durng the night.
I forget to unplug cat5 from doxport cable modem. (in comes a surge thru the rg6 cable! oh yes!)
later in the morning I discover:
1 damaged network switch (LEDs still worked, but it wasn't functioning properly at all).
2 killed ethernet network PCI cards (both in a different mobo, both mobos still working fine today).
1 killed internal mobo ethernet port. Motherboard still works to this very day.
the cable modem itself survived, and still worked fine. Kept it for around a year, but ditched it, and got a newer one.
it probably could've been far worse, but it wasn't.
other bad stuff happened that same night, heh, the furnaces PCB got fried. Black marks on it. The surge for that came thru lightning striking and obliterating the top of a metal light post outside (quite loud, one of the outside dogs flipped out). In came the surge, which actually blew off the plastic panel in the kitchen, and destroyed the light switch. Some smoke and smell, but no fire I saw. Insurance is your half-friend. In that regard, I'm happy I don't live there anymore.
skyking
Lifer
- Nov 21, 2001
- 22,622
- 5,730
- 146
high voltage can wreak havoc for another reason: The typical circuit breakers upstream in a power system rely on excessive amperage to trip them, and only the most sophisticated systems will trip on over voltage.
Example:
A friend was taking three folks who had never flown before for the first plane ride, in a C182.
Shortly after departure, the voltage regulator failed in a unique way, and the 14 volt alternator started producing 90 volts or so. The cabin quickly filled with smoke as all the radios and electronics started to burn up. He tripped the breaker for the master and alternator, only to find the surge had welded the remote contacts. Meanwhile, the circuit breakers in the panel were passing this current just fine until each device dead shorted and caused the amperage to rise high enough. He was flying the plane, clearing the smoke and trying to stop the electrical flow, but many of those breakers never failed.
He is a professional pilot, and made a safe landing. I don't think those people ever flew again
Example:
A friend was taking three folks who had never flown before for the first plane ride, in a C182.
Shortly after departure, the voltage regulator failed in a unique way, and the 14 volt alternator started producing 90 volts or so. The cabin quickly filled with smoke as all the radios and electronics started to burn up. He tripped the breaker for the master and alternator, only to find the surge had welded the remote contacts. Meanwhile, the circuit breakers in the panel were passing this current just fine until each device dead shorted and caused the amperage to rise high enough. He was flying the plane, clearing the smoke and trying to stop the electrical flow, but many of those breakers never failed.
He is a professional pilot, and made a safe landing. I don't think those people ever flew again
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 24K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 19K
-
-
-
Facebook
Twitter