Can a damaged graphics card or HDD damage other components?

[SonicZap]

Hi,

Today a family PC suddenly shut down while it was in use. It wouldn't power on afterwards, so I suspected that something had died. I swapped the PSU to another unit and upon pressing the power button, there were sparks and smoke coming from near the CPU socket (from the VRMs I think). I pulled out the plug as fast as I could.

So, from that it seems clear that at least the motherboard and CPU are dead, and likely the PSU as well. However, the HDD in the PC contains some rather valuable data, and it'd be nice if the GPU could be saved as well. But I have no idea if they also got damaged.

We have another PC in the house, but we can't afford losing it. So, the question is, is it safe to try the GPU and HDD in the other PC to see if they work? Or could a fried GPU or HDD also damage other components in the system (like the motherboard and PSU, since the components are connected to those)?

Thanks in advance.

 

[Final_Fighter]

short answer, yes. if the other computer in the house is your only pc and is absolutely needed then don't chance it. Is it likely that those two components are bad? no. if i were in your position i would look the hard drive over for any burnt electronics and if you don't find any then its properly good. the graphics card i would test in an old system or find a friend that has an old system they dont mind testing sketchy hardware in.

hope things work out. also please list the hardware specs of the system that burnt.

 

[SonicZap]

hope things work out. also please list the hardware specs of the system that burnt.

CPU: AMD Phenom II X4 955
MB: MSI 870A-G54
RAM: 4 GB (2 x 2 GB) 1333 MHz, can't remember the brand
GPU: Asus Radeon HD 5830 (EAH5830 DirectCU)
PSU: Enermax 500W (I always suspected that it was low quality..)
HDD: 500GB, can't remember other details right now

The PC was mostly used for office work and light gaming, which the HD 5830 handled fine. While it's not very valuable, the current state of the GPU market is rather bad so a replacement would cost significantly. I'm planning on building a cheap Ryzen 3 based PC to replace it, so a GPU is necessary. (The light gaming isn't light enough for an Intel IGP to handle it)

But the most important thing is the HDD, it contains some 3 weeks' worth of work that hasn't been backed up.

 

[Steevo]

Put the HDD in a caddy to isolate it during power up, then if it sounds normal and doesn't die in a shower of sparks, plug the external caddy into the USB port of whatever computer to see if the disk is still useable. Depending on the GPU you could try it with another computer if you are OK with the possibility that it might damage the PCIe slot or motherboard. Or you could try a riser card and see if it powers up without any worry about damaging anything.

https://www.newegg.com/Product/Prod...=Pci-e_riser_cable-_-9SIABMK5BS2329-_-Product

https://www.newegg.com/Product/Prod..._re=External_enclosure-_-17-182-155-_-Product

 

[Komshija]

As Final_Fighter said, check HDD for any burns and inspect HDD connectors. If your other PC works fine, connect your HDD as a secondary drive and copy important data to USB or to an external HDD. Prior to that arrange boot priorities in BIOS, so that PC boots from the original HDD and not from secondary which might be damaged. I recommend copying porn to an external HDD if you have one.

 

[de.das.dude]

if it smells its likely dead. RIP

you can probably salvage the CPU and RAM. Since these are indirectly powered by the motherboard, they might be alive. HDD is probably dead since the VRMs were sparking, i.e. some over voltage issue on the 12V rail.. HDDs use rthe 12V rail so..

you can try getting a hdd of the same exact model and swap its PCB with yours to try and access data. I dont think the platter inside was damage too badly.

 

[EarthDog]

What? Cpu/ram is directly powered by the vrms...fed from 8 pin and 24 pin.

Hdd is fed power through psu directly and doesnt touch the board/has nothing to do with vrm.

What are you trying to say?

 

[de.das.dude]

What? Cpu/ram is directly powered by the vrms...fed from 8 pin and 24 pin.

Hdd is fed power through psu directly and doesnt touch the board.

What are you trying to say?

the more semiconductors between the part and the PSU, the better, np /pn junctions like to blow out and become non conductive.

it might be that the mosfets in the vrm sacrificed themselves.

 

[EarthDog]

Youx typically see the psu shutsown bedore it starts cooking vrms, no. Psu has ovp/ocp...

 

[Vya Domus]

CPUs typically survive from blown-up VRMs. Along side everything else.

Word of advice : for most AM3+ boards put a fan on the VRM heatsink it will dramatically reduce the risk of something like this happening.

 

[silentbogo]

Dead/burning VRM is a frequent problem on cheap AM3 motherboards.
Your HDD is most likely alive and well, so as your GPU. Your PSU is probably fine too.

If there is no visible damage or burnt PCB traces - it's safe to put your GPU+HDD into another system.
If you are going to repair your PC, you should get a decent motherboard, preferably with 8VRM phases w/ some kind of passive cooling.
The cheapest adequate option is probably MSI 970A Gaming Pro Carbon. For a few more bucks you can get an "enthusiast-grade" 990x or 990FX board, which usually comes with 8 or more phases.
Even though your old MoBo specification claims to be compatible with most 125W CPUs, I doubt I can personally risk putting anything over 95W into it (overclocking is also a big no-no).

- Check your PSU : unplug from your PC and try the forced start by shorting green and black contacts on 24-pin connector. If it starts - check all voltages.
- Check your GPU, RAM, HDD and other components in another system. Start with visual inspection, and if it looks and smells OK - connect it.
- If everything works, get a new motherboard, put your system together and enjoy your overpriced space heater for another few years.

Youx typically see the psu shutsown bedore it starts cooking vrms, no. Psu has ovp/ocp...

OCP and OPP only kicks-in when you reach PSU's limit, so if one of your VRMs pulls 5 times it's normal current draw and glows red - PSU may not notice it and may not even break a sweat.
Your only hope is that motherboard detects an anomaly and shuts down in time.

P.S. Are you typing from your phone again, @EarthDog ?

 

[EarthDog]

90% is from my phone... sorry for the typos.

 

[de.das.dude]

Dead/burning VRM is a frequent problem on cheap AM3 motherboards.
Your HDD is most likely alive and well, so as your GPU. Your PSU is probably fine too.

*cough* MSI *cough* ASUS *cough*

i had issues so i got an 990fx asrock extreme 4, it has a fan on it

 

[silentbogo]

*cough* MSI *cough* ASUS *cough*

In my experience not so much on ASUS, but mostly AsRock, MSI, Gigabyte.
My neighbor's Gigabyte AM2+ board was the latest and the weirdest. A mosfet on the first phase died, but not completely, so the PC would start, boot, work for 10-15 minutes and only then shutdown, while leaving a slight wiff of burning.
By the time it got to me the entire PCB trace came off the board, because all adhesive between layers got burned.
Some AsRock 970 boards I've seen exploded with a big crater through multiple PCB layers (also 4+1 VRM in combination with 125W CPU and/or overclocking). One of those instances happened on my desk, and that was a 970 Extreme3 with heatspreader!

 

[basco]

the hdd you can try whithout probs on your other pc.
-never heard of pc damaged trough hdd.