jonnyGURU Forums
Home Site Search Reviews Articles Contest Links PSU FAQs  


Go Back   jonnyGURU Forums > Computer Hardware > PC Power Supply Discussion

PC Power Supply Discussion Troubleshooting and discussion of computer power supplies

Closed Thread
 
Thread Tools Display Modes
  #1  
Old 09-12-2010
ice445 ice445 is offline
200W User
 
Join Date: Jul 2010
Posts: 22
Thanks: 0
Thanked 0 Times in 0 Posts
Default Just how effective can a surge protector be?

I was thinking of getting this surge protector here: http://www.newegg.com/Product/Produc...82E16812107141 to go along with my new GTX 470.

Now, I'm aware that these things have no hope of stopping a close by lightning strike, but how close does it have to be? within a quarter mile of your house? A mile? What is the likliehood of it not working?

Also, what do these do for noise filtration and general surges (From power company). I think its' a decent investment, but I wanna believe it at least crosses 4 out of my 5 X scenarios.
  #2  
Old 09-13-2010
cypherpunks cypherpunks is offline
1kW User
 
Join Date: Oct 2007
Posts: 440
Thanks: 19
Thanked 25 Times in 16 Posts
Default Using it properly is the most inportant thing.

It is actually fairly straightforward to build equipment that can withstand and keep running right through multiple direct lightning strokes. Every AM radio station has a big tall transmitter tower located on a hill that gets hit a few times every thunderstorm. Yet somehow the transmitter survives.

It's quite easy to understand once you truly purge your mind of all concept of "stopping" the strike. It just came through a quarter mile or more of air; a few inches of insulation (much less a fraction of an inch) is not going to slow it down in the slightest.

Put another way, a typical (and they come much larger!) lightning strike is 500 MJ of energy. An 18 inch battleship shell at 2700 lb and 2500 ft/sec muzzle velocity (figured for the New Jersey) is 355 MJ, only 2/3 of that. Put in nuclear weapon units, 500 MJ is 0.12 tons of TNT, about 1/100,000 of the Hiroshima bomb. That's 436 the M67 hand grenade (with 6.5 oz of Composition B, equivalent to 8.77 oz of TNT).

When dealing with explosives, the one thing you do not want to do is to try to stop it. Wrapping an explosive in steel is how you make a hand grenade; tamping makes the explosion more effective.

On the other hand, how much damage does that shell so to the air it's passing through? Very little, because the air doesn't try to stop it.

You do not want anything you care about between lightning and where it wants to go. In fact, you want as little as possible, because the more you try to stop it, the more you'll get hurt.

The purpose of all lightning arrestors and surge suppressors is to encourage it to go somewhere else.

You want to provide it with a path from whatever wire it came in on to the wet ground outside (remember, lightning starts with raindrops carrying charge from the clouds to the ground) that is about a million times easier (in electrical engineer speak, 1 millionth the impedance) than the path through your equipment.

And the way you do that is to take every wire to your computer, without exception, pull them together into a nice neat bundle and ensure that it's a million times easier for the lightning to jump between the wires than to take the long path through your equipment.

It's just like wanting to live on a street without a lot of traffic: a dead end isn't a short cut to anywhere.

In an AM radio transmitter shack, you generally have a big heavy steel plate on one wall, and the wires are come in through holes drilled in it. Any lightning can jump the gap to the plate and go straight down to ground.

A surge protector is similar, if a lot less heavy-duty: a whole lot of metal oxide varistors (MOVs) connecting all the wires together.

Even if you're catching a surge that is a tiny fraction of a lightning strike (one thousandth is still half a hand grenade), the same principle applies: you want a perimeter around your computer, and maybe your monitor, speakers, USB peripherals, etc. And every single wire that crosses that perimeter is connected to the same surge suppressor.

That way, it provides a short-cut around your equipment that is much easier to take than the path through it.

But while it can prevent a dangerously high voltage from appearing between any two wires that run through it, it cannot do anything for the HDMI cable to your TV, if the TV isn't itself plugged into the same surge protector. Otherwise, a surge can and will pass right past the surge protector, through your computer, your TV, and out some of its other wires. Plugging the TV into a different surge protector does not provide the necessary short-cut.

If you have a bit more electrical engineering background, just remember that a lightning strike is a current source, not a voltage source. Current (I) is fixed at 60,000 A, while voltage is defined by V = IR. The trick to dealing with it is to keep R very, very low!
  #3  
Old 09-13-2010
dlb's Avatar
dlb dlb is offline
1kW User
 
Join Date: Oct 2009
Posts: 271
Thanks: 0
Thanked 0 Times in 0 Posts
Default

So, in short, every component that connects to the PC and requires power from the wall should be plugged into the same surge protector? I'm assuming this applies to DSL/cable modems too, right? I have had people say they have a surge protector plugged into a 2nd surge protector and that is then plugged in to the wall. I have heard that this is "better" or "safer" and I've heard it's worse, and I've heard it makes no difference. Is there any good or bad reasons to plug one surge protector in to another - if, for example, one surge protector's outlets are filled, is it good/bad/indifferent to plug in a second surge protector into the 1st, or should the 2nd simply plug in to a different wall outlet?

(BTW- GREAT thread!!! I've always wondered about this stuff but never thought to ask someone that KNOWS what they're talking about!)
__________________
Asus M4N72-E SLI; AMD Phenom II x4 955BE 3.20ghz w/ Spire ThermaxII Eclipse cooler; 6gb DDR2; 750gb+320gb+250gb SATA HDD; 2xDVDRW; Zotac GTX570; Topower PowerBird900; NZXT Beta EVO case; Win7x64
  #4  
Old 09-14-2010
Elledan's Avatar
Elledan Elledan is offline
1kW User
 
Join Date: May 2010
Location: Almere, Netherlands.
Posts: 492
Thanks: 0
Thanked 0 Times in 0 Posts
Default

I use a cheap sacrificial surge protector power strip in which I plug my UPSs. The intention is that if a surge were to come in, it would take out this cheap power strip and leave the UPSs to handle anything that comes through.

Multiple lines of defense, IOW
  #5  
Old 09-14-2010
westom westom is offline
200W User
 
Join Date: Sep 2010
Posts: 26
Thanks: 0
Thanked 4 Times in 3 Posts
Default

Quote:
Originally Posted by dlb View Post
So, in short, every component that connects to the PC and requires power from the wall should be plugged into the same surge protector? I'm assuming this applies to DSL/cable modems too, right?
If a protector is adjacent to an appliance, then what is a shortest path to earth? Often destructively via that appliance or some nearby appliance.

Protection is always about where energy dissipates. Either that surge dissipates harmlessly outside a building. Or goes hunting for earth destructively inside. Either a protector connects short (ie 'less than 10 feet') to earth. Or that protector is inside attempting to stop a shell from the USS New Jersey.

Lightning strikes utility wires down the street. That means lightning has directly struck every appliance inside the house. Which appliance fails? Which appliance also makes a better connection to earth? Both an incoming and outgoing path must exist – otherwise no damage.

Lightning enters on AC mains. DSL modem may be a best path to earth. All phone lines already have an earthed 'whole house' protector. Telephone lines are typically not the incoming path for a surge. Incoming on AC mains. Outgoing on a phone line to earth. If both paths do not exist, then DSL modem damage does not occur.

Damage is when a homeowner permits energy inside his building.

Franklin demonstrated the concept in 1752. Lightning seeks a conductive path to earth. The electrical conductor was a wooden church steeple. But wood is not a best conductor. A large current (20,000 amps) through wood creates a large voltage. 20,000 amps times a large voltage is large energy. Church steeple damaged.

Did Franklin stop that energy? Of course not. His lightning rod connected energy to earth and outside via a conductive wire. 20,000 amps through a conductive wire is a tiny voltage. 20,000 amps times a tiny voltage is tiny energy. No damage.

Protection is always about where that energy dissipates. A protector adjacent to an appliance can only stop, block, or absorb that energy. Protection means all surges - including direct lightning strikes - are harmlessly absorbed in earth. But only if the protector makes a short ('less than 10 foot') connection (with no sharp wire bends, not inside metallic conduit, no splices, etc) to earth.

What does a cheap sacrificial protector do? Exactly what its manufacturer spec numbers claim. Nothing effective. How does that sacrificial protector stop a shell from the USS New Jersey? It doesn't. But if they lie, then many will believe the advertising. A sacrificial protector only protects profit margins. Does virtually nothing to connect energy harmlessly to earth. Will not even discuss earth ground. Even the mutliple lines of defense are mythical.

A protector is only as effective as the thing that absorbs that energy - single point earth ground. Best protector are as close as possible to single point ground and distant (ie up to 50 meters) from protected electronics.
  #6  
Old 09-14-2010
cypherpunks cypherpunks is offline
1kW User
 
Join Date: Oct 2007
Posts: 440
Thanks: 19
Thanked 25 Times in 16 Posts
Default

Quote:
Originally Posted by dlb View Post
So, in short, every component that connects to the PC and requires power from the wall should be plugged into the same surge protector?
No, every wire that connects to your "PC assembly", whether it be power, telephone, network, coax, or whatever, needs to go through a common surge protector. (Actually, I over-simplified: you can use multiples if you know what you're doing, but the rules are more complex, and they don't work as well. Basically, they have to be connected to each other very well, which generally means not via two 6 foot cords.)

Quote:
I'm assuming this applies to DSL/cable modems too, right?
But more importantly, it applies to the cable/phone line, too! You can put the modem outside the protection, and run the network cable through the protector, or put it inside, and run its power & communication like through, but you have to do one or the other.

Quote:
I have had people say they have a surge protector plugged into a 2nd surge protector and that is then plugged in to the wall. I have heard that this is "better" or "safer" and I've heard it's worse, and I've heard it makes no difference. Is there any good or bad reasons to plug one surge protector in to another - if, for example, one surge protector's outlets are filled, is it good/bad/indifferent to plug in a second surge protector into the 1st, or should the 2nd simply plug in to a different wall outlet?
Better to plug one into the other. Although an unprotected power bar would do as well. If you built a second "choke point" which every wire went through, that would actually provide better protection. But the protection comes from the narrow "choke point". With that advantage, 300 men can hold off an army of thousands. If there's an alternate path around the choke point, it's hopeless.

Again, remember: a surge protector DOES NOT STOP THE SURGE. Purge the heretical concept from your mind. It provides an alternate path for the surge, and the fact that your equipment is an electrical cul-de-sac with no other exits is what prevents the surge from going through it. The only way for a surge to get from one outside wire to another is to come into the surge protector and leave the same way. If there's a path that only goes through the surge protector once, you're vulnerable.
  #7  
Old 09-14-2010
Elledan's Avatar
Elledan Elledan is offline
1kW User
 
Join Date: May 2010
Location: Almere, Netherlands.
Posts: 492
Thanks: 0
Thanked 0 Times in 0 Posts
Default

Quote:
Originally Posted by cypherpunks View Post
.

Again, remember: a surge protector DOES NOT STOP THE SURGE. Purge the heretical concept from your mind. It provides an alternate path for the surge, and the fact that your equipment is an electrical cul-de-sac with no other exits is what prevents the surge from going through it. The only way for a surge to get from one outside wire to another is to come into the surge protector and leave the same way. If there's a path that only goes through the surge protector once, you're vulnerable.
Hence MOVs: with a large enough surge, they start conducting. In a surge protector they create a path from the hot wire to ground, thus allowing the surge the easy path directly to the ground it's seeking, instead of a cumbersome, high-resistance route through a series of applications
  #8  
Old 09-14-2010
cypherpunks cypherpunks is offline
1kW User
 
Join Date: Oct 2007
Posts: 440
Thanks: 19
Thanked 25 Times in 16 Posts
Default

Quote:
Originally Posted by westom View Post
If a protector is adjacent to an appliance, then what is a shortest path to earth? Often destructively via that appliance or some nearby appliance.

Protection is always about where energy dissipates. Either that surge dissipates harmlessly outside a building. Or goes hunting for earth destructively inside. Either a protector connects short (ie 'less than 10 feet') to earth. Or that protector is inside attempting to stop a shell from the USS New Jersey.

Franklin demonstrated the concept in 1752. Lightning seeks a conductive path to earth. The electrical conductor was a wooden church steeple. But wood is not a best conductor. A large current (20,000 amps) through wood creates a large voltage. 20,000 amps times a large voltage is large energy. Church steeple damaged.

Did Franklin stop that energy? Of course not. His lightning rod connected energy to earth and outside via a conductive wire. 20,000 amps through a conductive wire is a tiny voltage. 20,000 amps times a tiny voltage is tiny energy. No damage.
This is the key point. We tend to use voltage-driven circuitry, where power is supplied at a fixed voltage from a low-impedance source. Low impedance means that dV/dI is a small number: the voltage doesn't change (dV is almost zero), even if the current changes a lot. A high resistance across the power supply means a low current means little energy, while a low resistance means a high current and lots of energy.

Lightning and related surges, however, are current sources, with a high impedance: dV/dI is a very large number: the current doesn't change (dI is approximately zero) no matter what the voltage. A lightning stroke will deliver, say, 20,000A. Ohm's law, V=IR still applies, but a this time it's the current that's fixed. Low resistance means a low voltage means little energy is dissipated in the resistor. (When dealing with a full lightning strike, you need seriously thick wire, like AWG 0 or 00.) A high resistance means high voltage and lots of power.

Quote:
Protection is always about where that energy dissipates. A protector adjacent to an appliance can only stop, block, or absorb that energy. Protection means all surges - including direct lightning strikes - are harmlessly absorbed in earth. But only if the protector makes a short ('less than 10 foot') connection (with no sharp wire bends, not inside metallic conduit, no splices, etc) to earth.
Here is where I disagree. The whole point is that you cannot stop or block the energy. What you are trying to do is divert it. Aikido, not karate. You want to give it a path from whatever wire it's coming in on to ground that does not pass through your equipment. And you want that path to be thousands of times easier to take, so less than a thousandth of the incoming current flows through the protected equipment.

It's not about "stopping" the surge, it's about providing an attractive short-cut. That's why it's so important that all the wires go through the same bottleneck: you want the short cut from wire to wire to be as short and easy as possible. If it's only 100 times easier, than 1% of the energy is going to pass through your equipment, and that might not be good for it.

You can do a few things to make the path through your equipment more difficult, but what matters is the ratio between that and the short-cut you provide with MOVs or equivalent. So you bring all the wires within inches of each other and provide the shortest short-cut that you can.

There are reasons to provide a good, direct ground path, but for the protected equipment it actually doesn't matter how good the ground is after leaving the protector; all that matter is the ratio of impedances through the protector.

And MOVs really do work quite well for such cheap little things. here's a datasheet showing how they're rated for hundred or thousands of amps in the short term.

If you want to get fancier, try a gas discharge tube. It can take 40,000 A for a few microseconds.
Closed Thread

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -4. The time now is 05:04 AM.


Powered by vBulletin® Version 3.8.1
Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.