Sign in with
Sign up | Sign in
Your question

700w psu sort of exploded is my system ok?

Last response: in Components
Share
March 23, 2012 2:45:36 PM

Hello and thanks for taking time to read this. I posted a thread concerning my 3d mark 11 results. Someone responded indicating that my psu might be inadequate. I have posted this informatio within that thread but feel that as this is a new isue, it warrants a new thread to indicate to people exaclty that.

While playing BF3 sparks began shooting out from where my psu is. It was crackling and popping loudly and I raced to grab some gloves and unplug everything. I was wondering if this has possibly dmaged my mobo or cpu or vid cards? I have not plugged in back in as the cpu had what looked like burnt little components torn off with a lastic coating and small writing on them. Im certain its dead. Im obviously worrried now that when I get my new psu (advise is appreciated concerning that) it will just be a new psu installed on a dead system. So, Im guessing that I basically overloaded the capacity of the psu and it blew as a result. Any information of experience with this i.e. know issues that can result to cpu, memory etc.

My sytem specs are:

2600K ocd to 4.6 (water cooled)
16gb of corsair vengeance @1600
2 560ti sli
4 120mm fans


I have inspected my mobo, vid cards, ram, cpu area (without removing the cooling) and nothing looks or smells defective or burnt. I am wondering what a good psu might be not just for this system but for a possible upgrade to 580/680 type cards as well.

Thanks in advance.
March 23, 2012 3:36:51 PM

There is nothing to protect the system from overvoltage/undervoltage after the the power supply. The systems are totally dependent on the power generated from the psu. Short answer is, yes, you could have fried everything. Please read power supply reviews on say, hardocp, who does very good power supply reviews. Get another PS and see what happens is the only thing to do now. What model PS do you have now that blew up?

I have a 750W XFX Black, you should probably get an 850W or higher though. You should also get a Kill-A-Watt to see your current draw at the plug so you get an idea of what is happening. That being said, a decent PS should have thermal overload prevention, so it is odd you experienced such a drastic failure.
March 23, 2012 4:34:21 PM

The psu came with my pc (which I have basically re-built) from ibuypower. Its a xion 700w. Most of the parts that came with it have been replaced i.e. mobo, ram, vid card. I originally bought a zotac 560ti. It was defective so I rma'd and while waiting I boght an msi twin frozr. When I recently got the rma'd zotac back I put it in for sli as well as crank up my overclock from 4.4 to 4.6. Im not sure if this would have happaned had I not oc'd to 4.6, but I was smelling a burning smell from the pc right after I oc'd to 4.6. Since the overclock sort of coincides with installing the zotac, I assumed it was just the new video card perhaps as I know new componenets can sometimes give that smell when first turning on, then whamo...I was playing BF3 and the crackle and popping and sparks shooting out from what appeard to be the bottom of the pc where the psu is.

Its disconcerting to hear that I could have fried the whole sytems components. I think Im going to go with an 850 or 900w so be safe. I am a little confused on the technicals of how I could have fried the other components. If the psu is being asked to give more power and is attempting to do so beyond its capabilities, wouldnt that mean that the components sort of asking the spu for power simply didnt get any and the psu overloaded in the attempt as none was there to allocate to other parts? Do you (or anyone ) know how this works. I am a project manager and am a stickler for root cause. Thanks in advance.

sorry bout typos...cant find glasses
Related resources
March 23, 2012 4:41:01 PM

http://www.jonnyguru.com/modules.php?name=NDReviews&op=... I just looked this up. Apparently its possible the output on the psu was far less than 700w. I will of course never buy a prefab pc from ibuypower. I ended up replacing the mobo anyway with gigabyte z68 something that ran me about 180 bucks vs the gigabyte one it came with that runs around 89 bucks and it faild on me so I had to rma it only to get a defective one again (the pins in the socket were bent). Lesson learned...build your own is easier. Especially with the help of forums like these for guidance.
a c 1176 ) Power supply
March 23, 2012 4:53:36 PM

It's difficult to say whether anything else was damaged without testing. If you're lucky, only the power supply itself was damaged. If you're unlucky, then that can be an expensive lesson.

For a system using two GeForce GTX 560 Ti graphics cards in 2-way SLI mode NVIDIA specifies a minimum of a 700 Watt or greater power supply that has a combined +12 Volt continuous current rating of 52 Amps or greater and that has at least four 6-pin PCI Express supplementary power connectors.

For a system using a single GeForce GTX 580 graphics card NVIDIA specifies a minimum of a 600 Watt or greater power supply that has a combined +12 Volt continuous current rating of 42 Amps or greater and that has at least one 6-pin and one 8-pin PCI Express supplementary power connectors.

For a system using two GeForce GTX 580 graphics cards in 2-way SLI mode NVIDIA specifies a minimum of a 900 Watt or greater power supply that has a combined +12 Volt continuous current rating of 65 Amps or greater and that has at least two 6-pin and two 8-pin PCI Express supplementary power connectors.

For a system using a single GeForce GTX 680 graphics card NVIDIA specifies a minimum of a 550 Watt or greater power supply that has a combined +12 Volt continuous current rating of 38 Amps or greater and that has at least two 6-pin PCI Express supplementary power connectors.

For a system using two GeForce GTX 680 graphics cards in 2-way SLI mode NVIDIA specifies a minimum of a 750 Watt or greater power supply that has a combined +12 Volt continuous current rating of 53 Amps or greater and that has at least four 6-pin PCI Express supplementary power connectors.

Total Power Supply Wattage is NOT the crucial factor in power supply selection!!! Total Continuous Amperage Available on the +12V Rail(s) is the most crucial factor.

To handle the overclocked i7-2600K I would add at least an additional 5 Amps to the minimum combined +12 Volt continuous current ratings I've specified above.

Stick with reputable name brand power supplies like Antec, Corsair, Enermax, Seasonic, XFX.
March 23, 2012 5:36:50 PM

Thanks for the responses Ko888 and ngoy. Ko888, when you say "lucky" regarding to not only hae this effect the psu, how lucky. Im not trying to by a flip, just trying to understand what the typical results of something like this happening are. Is it more typical for someone to say that they oc'd then their psu blew and had to be replaced or is it more typical to say the psu blew and fried everything...Im hoping the pc fairy was looking out for me. Again, just trying to understand what is more typical...whole system blow firy scenario or psu blows and gets replaced...thanks again.

Also, my question about how?...how can the psu fry anything if it was trying to deliver power and fried itself? Again, doesnt that mean that nothing was deliverd to the components, or does it mean that for an instant, the psu had no sort of protection stopping a fraction of a seconds worth of overload to those componenets as well as frying in the process...im totally guessing here.
March 23, 2012 5:46:44 PM

FMFPAC said:
Its disconcerting to hear that I could have fried the whole sytems components. I think Im going to go with an 850 or 900w so be safe. I am a little confused on the technicals of how I could have fried the other components. If the psu is being asked to give more power and is attempting to do so beyond its capabilities, wouldnt that mean that the components sort of asking the spu for power simply didnt get any and the psu overloaded in the attempt as none was there to allocate to other parts? Do you (or anyone ) know how this works. I am a project manager and am a stickler for root cause. Thanks in advance.


Your Xion 700w power supply has inherent overvoltage and overcurrent protection, so it most likely disables the output voltage if a power rail is overloaded. Your motherboard also has plenty of voltage regulators that will make sure your processor, RAM, etc see the voltages it needs even if there's a little fluctuation coming from your power supply.

The problem of "frying other components" comes when a power supply actually fails. Failure can occur simply because there's a bad batch of parts installed that were destined to fail early. This is usually called infant mortality, contrasting from the more common "end of life" failure. Or if the power supply wasn't designed very well and parts were not derated properly, then you can add extra stresses to your power devices causing them to fail earlier than expected.

Power supply failure can certainly damage components connected to the input or output of the supply. If the AC input rectifer shorted, you most likely blew the circuit breaker in your home. If some DC output rails lost regulation and outputted 30V instead of 3.3V, the motherboard may not be able to handle that overvoltage case and get damaged.

As somebody else pointed out, it's difficult to say what got damaged down the line after your power supply failure. And it's even possible that nothing got damaged because the power supply failed "nicely" and just shutdown all of it's inputs and outputs. You'll have to do a full system test when you start putting back the pieces.

But any respectable power supply will have short circuit protection so you shouldn't be able to damage it if you plug in a DVD-ROM drive that was damaged from your last power supply failure.

I would install pieces one at a time to in order to isolate any suspect parts that could have been damaged.

Good luck!
a c 1176 ) Power supply
March 23, 2012 6:35:41 PM

FMFPAC said:
Thanks for the responses Ko888 and ngoy. Ko888, when you say "lucky" regarding to not only hae this effect the psu, how lucky. Im not trying to by a flip, just trying to understand what the typical results of something like this happening are. Is it more typical for someone to say that they oc'd then their psu blew and had to be replaced or is it more typical to say the psu blew and fried everything...Im hoping the pc fairy was looking out for me. Again, just trying to understand what is more typical...whole system blow firy scenario or psu blows and gets replaced...thanks again.

Also, my question about how?...how can the psu fry anything if it was trying to deliver power and fried itself? Again, doesnt that mean that nothing was deliverd to the components, or does it mean that for an instant, the psu had no sort of protection stopping a fraction of a seconds worth of overload to those componenets as well as frying in the process...im totally guessing here.

A well designed power supply would protect the components that are attached to it if it should fail.

A poorly designed power supply with no or inadequate protection circuitry will allow attached components to be damaged when a power surge happens when it fails.

Your XION 700W power supply is based on an old ATX v1.3 design.

When JonnyGURU said that the pair of S30D60C Schottky rectifiers for the +12V rails is only 10 Amps away from disaster, I believe that you experienced that disaster.

The values shown on its label are max peak values and not continuous values.

OCP (over current protection) is all that was implemented in that PSU's design.
March 23, 2012 6:37:39 PM

nthreem said:
Your Xion 700w power supply has inherent overvoltage and overcurrent protection, so it most likely disables the output voltage if a power rail is overloaded. Your motherboard also has plenty of voltage regulators that will make sure your processor, RAM, etc see the voltages it needs even if there's a little fluctuation coming from your power supply.

The problem of "frying other components" comes when a power supply actually fails. Failure can occur simply because there's a bad batch of parts installed that were destined to fail early. This is usually called infant mortality, contrasting from the more common "end of life" failure. Or if the power supply wasn't designed very well and parts were not derated properly, then you can add extra stresses to your power devices causing them to fail earlier than expected.

Power supply failure can certainly damage components connected to the input or output of the supply. If the AC input rectifer shorted, you most likely blew the circuit breaker in your home. If some DC output rails lost regulation and outputted 30V instead of 3.3V, the motherboard may not be able to handle that overvoltage case and get damaged.

As somebody else pointed out, it's difficult to say what got damaged down the line after your power supply failure. And it's even possible that nothing got damaged because the power supply failed "nicely" and just shutdown all of it's inputs and outputs. You'll have to do a full system test when you start putting back the pieces.

But any respectable power supply will have short circuit protection so you shouldn't be able to damage it if you plug in a DVD-ROM drive that was damaged from your last power supply failure.

I would install pieces one at a time to in order to isolate any suspect parts that could have been damaged.

Good luck!


Thanks for the reply. I didnt short my homes circuit breaker so your scenario of 30 v frying my motherboard is making me wonder. You confused me with your last statement "But any respectable power supply will have short circuit protection so you shouldn't be able to damage it if you plug in a DVD-ROM drive that was damaged from your last power supply failure." Im simply not sure what this means. My psu is already damaged wo what am I not damaging? and I dont know if my dvdrom was damaged yet but if it was damaged it shouldnt damage...I dont understand.
March 23, 2012 6:45:42 PM

ko888 said:
A well designed power supply would protect the components that are attached to it if it should fail.

A poorly designed power supply with no or inadequate protection circuitry will allow attached components to be damaged when a power surge happens when it fails.

Your XION 700W power supply is based on an old ATX v1.3 design.

When JonnyGURU said that the pair of S30D60C Schottky rectifiers for the +12V rails is only 10 Amps away from disaster, I believe that you experienced that disaster.

The values shown on its label are max peak values and not continuous values.

OCP (over current protection) is all that was implemented in that PSU's design.


Thanks for getting back to me however Im not sure what your saying here:
"OCP (over current protection) is all that was implemented in that PSU's design.[/quotemsg]" are you saying that the overcurrent protection was all that was implemented here so Im screwed or even though its all that it has for safety, at least it is supposed to in theory protect parts from overcurrent and if it did , Im ok?

Any way, driving to microcenter probably tomorrow and getting new psu...will advise...and thanks. Byh the way, I like going to the microcenter newar me now as I found they will take back anything for a whole year (if you pay like 20 bucks for a plan) and can use what you bought for value towrds anything else.
a c 1176 ) Power supply
March 23, 2012 6:54:24 PM

FMFPAC said:
Thanks for getting back to me however Im not sure what your saying here:
"OCP (over current protection) is all that was implemented in that PSU's design.
" are you saying that the overcurrent protection was all that was implemented here so Im screwed or even though its all that it has for safety, at least it is supposed to in theory protect parts from overcurrent and if it did , Im ok?

Any way, driving to microcenter probably tomorrow and getting new psu...will advise...and thanks. Byh the way, I like going to the microcenter newar me now as I found they will take back anything for a whole year (if you pay like 20 bucks for a plan) and can use what you bought for value towrds anything else.[/quotemsg]
OCP only shuts down the power supply if you were to draw too much current from a particular rail. I see it being used as a power supply self-preservation feature rather than for protecting attached devices.

There is an OCP set point (i.e. threshold/trigger value) for each rail that determines when the circuit gets tripped.

20 bucks for that plan sounds like an excellent deal to me.
March 23, 2012 7:05:54 PM

Yeah it was worth it. I got the plan when I bought the 560ti twin frzr. I can simply take back the vid card and use the 260 bucks towards anything I like. Again not sure what your saying when say "only shuts sown power supply" is there anything else? It sounds like your saying if the power supply shuts down...current can still be derived somehow? Anyway...thanks.
March 23, 2012 7:06:13 PM

FMFPAC said:
Thanks for the reply. I didnt short my homes circuit breaker so your scenario of 30 v frying my motherboard is making me wonder. You confused me with your last statement "But any respectable power supply will have short circuit protection so you shouldn't be able to damage it if you plug in a DVD-ROM drive that was damaged from your last power supply failure." Im simply not sure what this means. My psu is already damaged wo what am I not damaging? and I dont know if my dvdrom was damaged yet but if it was damaged it shouldnt damage...I dont understand.


Ah yes, that was a little confusing. If your new power supply has short circuit protection on its output, then you shouldn't worry about testing a DVD-ROM drive on your new power supply.

If the DVD-ROM drive was damaged from your previous incident, most likely it will either not power up or would "short out" whatever you were using to power the DVD-ROM. You would find this out when you to go test out all of your components.

Since good power supplies have short circuit protection, you don't need to worry about damaging your brand new power supply when you're in this testing phase.
a c 1176 ) Power supply
March 23, 2012 7:16:28 PM

FMFPAC said:
Yeah it was worth it. I got the plan when I bought the 560ti twin frzr. I can simply take back the vid card and use the 260 bucks towards anything I like. Again not sure what your saying when say "only shuts sown power supply" is there anything else? It sounds like your saying if the power supply shuts down...current can still be derived somehow? Anyway...thanks.

What I'm saying is that the OCP circuit is not there to protect the attached devices. Its purpose is to protect itself from burning out.

When your power supply failed the I would go as far as to say that the OCP circuit wasn't even tripped. If it was tripped and working properly the power supply wouldn't have failed (i.e. self-destructed), it would just shut down and you should be able to resume use of it after removing the cause of the over current.
March 23, 2012 7:50:30 PM

I was really hoping someone would say " oh I have seen that happen a million time and rarely is any other components effected...mostly likely your fine but do check to be sure". Anyone want to validate that statement??? :D 

Thanks again for the replies.
March 23, 2012 8:19:03 PM

You have to be VERY careful on how you select which cables get plugged into whatever card when you SLI.

The power supply you listed has the following specs from Newegg:

+3.3V@24A,+5V@30A,+12V1@18A,+12V2@18A,
+12V3@18A,+ 12V4@20A,-12V@0.5A,+5VSB@2.2A

IIRC, most power is drawn from the 12V rails on newer systems, very little from 3V and 5V.

An overclocked 2600 can easily use 200W or more when stress testing, so 200W/12V=16A.

This Zotac board uses 175W EXTRA when under load, from a base of 145 to 320. 6 Pin PCI is allowed 75W per connector, 8 Pin is 150W, although a 6+2 connector can deliver 150W, as it is a subset of an 8. So say 200W again @ 12V is another 16A. So one of your 12V rails is shared with the 12V for the MB/CPU, which means at full bore, you are using a minimum of 32A on two rails. If we assume 12V4 is the one also linked to the MB/peripherals, so rail 3+4 =38A max 12V, -32A used=6A for everything else that uses 12V, like fans etc, could easily push one rail to max or near 100% capacity, which is NOT good. You only want to be at 70-80%, and under "normal" operating conditions, should ideally be less.

Your stuff isn't under warranty? Even though you were overclocking, a power supply should never explode. That is a lawsuit waiting to happen.
March 23, 2012 8:56:11 PM

ngoy said:
You have to be VERY careful on how you select which cables get plugged into whatever card when you SLI.

The power supply you listed has the following specs from Newegg:

+3.3V@24A,+5V@30A,+12V1@18A,+12V2@18A,
+12V3@18A,+ 12V4@20A,-12V@0.5A,+5VSB@2.2A

IIRC, most power is drawn from the 12V rails on newer systems, very little from 3V and 5V.

An overclocked 2600 can easily use 200W or more when stress testing, so 200W/12V=16A.

This Zotac board uses 175W EXTRA when under load, from a base of 145 to 320. 6 Pin PCI is allowed 75W per connector, 8 Pin is 150W, although a 6+2 connector can deliver 150W, as it is a subset of an 8. So say 200W again @ 12V is another 16A. So one of your 12V rails is shared with the 12V for the MB/CPU, which means at full bore, you are using a minimum of 32A on two rails. If we assume 12V4 is the one also linked to the MB/peripherals, so rail 3+4 =38A max 12V, -32A used=6A for everything else that uses 12V, like fans etc, could easily push one rail to max or near 100% capacity, which is NOT good. You only want to be at 70-80%, and under "normal" operating conditions, should ideally be less.

Your stuff isn't under warranty? Even though you were overclocking, a power supply should never explode. That is a lawsuit waiting to happen.



I appreciate the response however, I dont know what to make of this:+3.3V@24A,+5V@30A,+12V1@18A,+12V2@18A,
+12V3@18A,+ 12V4@20A,-12V@0.5A,+5VSB@2.2A
and how I would know what line does what output. Im certain I can understand it if told however, I would just be guessing if I tried to define it. Also, my 560ti is not a 448 core...I think the 448 uses a 6 and 8 pin vs the 2 6 pins for regular 560ti. I guess Ill be googling psu voltages and definitions to understand how the voltages and amps work. Your statement was extremely nebulas for me. Thanks again though for responding.
March 23, 2012 9:19:52 PM

2491066,16,417408 said:
You have to be VERY careful on how you select which cables get plugged into whatever card when you SLI.

The power supply you listed has the following specs from Newegg:

+3.3V@24A,+5V@30A,+12V1@18A,+12V2@18A,
+12V3@18A,+ 12V4@20A,-12V@0.5A,+5VSB@2.2A

IIRC, most power is drawn from the 12V rails on newer systems, very little from 3V and 5V.

An overclocked 2600 can easily use 200W or more when stress testing, so 200W/12V=16A.

This Zotac board uses 175W EXTRA when under load, from a base of 145 to 320. 6 Pin PCI is allowed 75W per connector, 8 Pin is 150W, although a 6+2 connector can deliver 150W, as it is a subset of an 8. So say 200W again @ 12V is another 16A. So one of your 12V rails is shared with the 12V for the MB/CPU, which means at full bore, you are using a minimum of 32A on two rails. If we assume 12V4 is the one also linked to the MB/peripherals, so rail 3+4 =38A max 12V, -32A used=6A for everything else that uses 12V, like fans etc, could easily push one rail to max or near 100% capacity, which is NOT good. You only want to be at 70-80%, and under "normal" operating conditions, should ideally be less.




Can you reccomend a PSU based on my specs with a little room for upgrades like bigger vid cards etc.

March 25, 2012 4:57:24 AM

Status update:

Bought an OCZ 1000w modular psu. So far, everything works perfectly so, you can tell your friends that you can blow up your 700w psu and still very much have a complete, non-blown our or fried system. Even the sli works better in BF3...not a stutter or fram rate drop on all max except aa shich is set to 2x. nice. thanks to those who responded.
a b ) Power supply
March 25, 2012 7:21:16 PM

You are a lucky man..
March 25, 2012 7:46:05 PM

AlderonnX said:
You are a lucky man..


I suppose so. I do have a question about my new psu and if anyone knows what direction it should be? I have a bottom mount and the little window where I guess the fan sucks are into from the fan is nowhere near covering the width of the larg 140mm fan. It allows for about 1/4 of the size of the fan to pull air through, not too mention the tiny rubber stands my pc is on barely lifts it of the deck. Anyone know if it would be better to flip it upside down so the fan is not so confined and if so, does this just suck the hot pc air into the already hot psu? Thanks again.
March 27, 2012 1:25:55 PM

FMFPAC said:
I suppose so. I do have a question about my new psu and if anyone knows what direction it should be? I have a bottom mount and the little window where I guess the fan sucks are into from the fan is nowhere near covering the width of the larg 140mm fan. It allows for about 1/4 of the size of the fan to pull air through, not too mention the tiny rubber stands my pc is on barely lifts it of the deck. Anyone know if it would be better to flip it upside down so the fan is not so confined and if so, does this just suck the hot pc air into the already hot psu? Thanks again.



Fans should never be blocked, you will overheat the psu. It is better to pull in lots of somewhat warm air and exhaust it, rather than mostly block the airflow into the psu and make the fan work inefficiently. So flip it over, or raise up your pc a bit and cut out the bottom so it can get a full intake.
!