[SOLVED] shorted gpu from lightning (?) any fix?

pekopekora

Prominent
Nov 14, 2021
7
0
510
a few months ago, i woke up from a really loud thunder while in a thunderstorm, when i tried turning my pc on it just wont boot up anymore.
i checked and both my ups and my psu's fuse were blown. (glass shattered) because of this, i thought that the pc wasnt working because of the
defective power supply so i bought a cheap 700w generic psu online just to check if the other parts were working. when the power supply arrived, i replaced the old
power supply with the new one and the pc still wouldnt turn on , however it now has shown signs of power since the cpu fan spun for half a second along with the gpu's.
i tried uninstalling the gpu to check if just using my apu (ryzen 3 2200g) would work by itself with the power supply i bought but sadly, only the fans spun and there was
no display at all; i also realized that the hard drive was dead and didnt spin anymore. :( at this point, i thought it would be best to ask a technician for help so i asked my
dad who works in the capital of my province to bring it to a technician there. he said that the technician tried a different power supply and other components and told him
that it was the motherboard that was broken. i ordered an asus a320m-k as the replacement motherboard and brought it to a nearby computer shop for them to install.
turns out, the new 700w generic power supply i bought was actually defective and only got to squeeze out a tiny bit of power when the technician tested it. anyways,
i finally got the power supply, motherboard, and my hard drive replaced however the technician said that the gpu wont work anymore as it draws only power and doesnt
work anymore when it was tested.

was the gpu shorted along with the other parts from the lightning strike near me that day and is there any fix for it?
 
Solution
was the gpu shorted along with the other parts from the lightning strike near me that day

Since the rest of your components also got fried (UPS, PSU, MoBo, HDD), it is safe to assume that GPU went along with them as well.

is there any fix for it?

For a "fix", buy a new GPU.

But to fix the fried GPU, well, for that, you need key facts:
  • Do you know exactly which component got fried inside the GPU?
  • Do you have source to buy the replacement component?
  • Do you know how to replace the dead component, which may require soldering?

If answer to any of those questions is NO, you can't fix your GPU.

All-in-all, life lesson: do not let PC run when in thunderstom. Also, check weather forecast daily.
Btw...

Aeacus

Titan
Ambassador
was the gpu shorted along with the other parts from the lightning strike near me that day

Since the rest of your components also got fried (UPS, PSU, MoBo, HDD), it is safe to assume that GPU went along with them as well.

is there any fix for it?

For a "fix", buy a new GPU.

But to fix the fried GPU, well, for that, you need key facts:
  • Do you know exactly which component got fried inside the GPU?
  • Do you have source to buy the replacement component?
  • Do you know how to replace the dead component, which may require soldering?

If answer to any of those questions is NO, you can't fix your GPU.

All-in-all, life lesson: do not let PC run when in thunderstom. Also, check weather forecast daily.
Btw, high-end UPS would've not let the rest of the PC to get fried. So, don't cheap out on UPS either.
 
  • Like
Reactions: Diflexster12
Solution

InvalidError

Titan
Moderator
Sounds like your presumably generic previous PSU freaked out when it got hit by the surge and killed most of your PC's components before killing itself and your UPS' fuse.

Btw, high-end UPS would've not let the rest of the PC to get fried. So, don't cheap out on UPS either.
What would you call "high-end"? I have taken apart a friend's APC SmartUPS Pro (~$300) and that still had far less surge protection than a $40 SurgeArrest Performance. I wouldn't rely on an UPS for surge protection, nothing priced within reason in NA appears to have any to speak of.

The single most important thing for keeping equipment as safe as it can possibly be during lightning storms is to make sure all connected equipment shares a single common ground point for protection. Otherwise, differences in ground potentials will allow surges to jump across connected equipment and still kill stuff.
 

Aeacus

Titan
Ambassador
What would you call "high-end"?

Minimum: Line-interactive, true/pure sine wave.
Preferred: Double conversion topology (online), true/pure sine wave.

I have taken apart a friend's APC SmartUPS Pro (~$300)

You mean APC BR1500G (or similar)? If so, it outputs stepped approximation to a sine wave (aka simulated sine wave) and to me, that isn't high end. Maybe for you it is, based on the cost alone.

within reason

Within reason is extremely vague and subjective, based on personal preferences alone. So, not a good argument to bring forth. Especially since for some, "within reason" is as cheap as possible while for others, "within reason" = great build quality, despite the premium price.

The single most important thing for keeping equipment as safe as it can possibly be during lightning storms is to make sure all connected equipment shares a single common ground point for protection.

No.

THE safest way to keep sensitive electronics (e.g PC) as safe as it can possibly be, during lightning storms, is to unplug it from the electricity grid. And that's it.
 
  • Like
Reactions: Tac 25
Sounds like your presumably generic previous PSU freaked out when it got hit by the surge and killed most of your PC's components before killing itself and your UPS' fuse.


What would you call "high-end"? I have taken apart a friend's APC SmartUPS Pro (~$300) and that still had far less surge protection than a $40 SurgeArrest Performance. I wouldn't rely on an UPS for surge protection, nothing priced within reason in NA appears to have any to speak of.

The single most important thing for keeping equipment as safe as it can possibly be during lightning storms is to make sure all connected equipment shares a single common ground point for protection. Otherwise, differences in ground potentials will allow surges to jump across connected equipment and still kill stuff.
I am less knowledgeable on UPS's, but can you plug in a UPS to a quality surge protector and be fine, or does the surge protector mess with the UPS's ability to work properly? For instance a Tripp Lite isobar surge protector.
 

InvalidError

Titan
Moderator
THE safest way to keep sensitive electronics (e.g PC) as safe as it can possibly be, during lightning storms, is to unplug it from the electricity grid. And that's it.
That is unrealistic since thunderstorms will occur when you aren't there to unplug stuff.

An UPS being line-interactive and "pure sine wave" has zero bearing on how much surge protection it has and under normal operating conditions, line power passes straight through to outputs. The surges from a lightning strike are on a microseconds scale while the relays in an UPS operate on a milliseconds scale, far too slow for the AVR boost/buck transformer or inverter to do anything with.

Under medium to light loads, stepped approximation UPSes are far more efficient than either "pure sine wave" UPS I have owned (about half as much run-time for the same battery size) and none of my equipment has any issue with stepped approximation. As far as I am concerned, sine wave is a waste of money and efficiency. BTW, the APC BR1xxxG aren't a sine wave models or at least not in their Canadian incarnation - I'm using a BR1300G with my PC right now.

I am less knowledgeable on UPS's, but can you plug in a UPS to a quality surge protector and be fine, or does the surge protector mess with the UPS's ability to work properly? For instance a Tripp Lite isobar surge protector.
Wall -> surge protector -> UPS -> PC, monitor and whatever else that is connected to your PC and requires backup power like external HDDs
The LAN, telephone, coax and whatever else connects to other stuff in your house should also pass through a surge protector connected to the same ground as the UPS/surge-protector since the goal here is to bring all of the equipment up/down at the same time so an incoming surge cannot jump between equipment.

Just another sacrificial component in the line, shouldn't do anything to the incoming power.
Depends on the sort of surge and what sort of filtering is being put in the line. The isobar for example has some pretty beefy chokes in it that will slow down how quickly a voltage surge can raise current and cause most of the surge energy to go elsewhere.
 
  • Like
Reactions: helper800
Wall -> surge protector -> UPS -> PC, monitor and whatever else that is connected to your PC and requires backup power like external HDDs
The LAN, telephone, coax and whatever else connects to other stuff in your house should also pass through a surge protector connected to the same ground as the UPS/surge-protector since the goal here is to bring all of the equipment up/down at the same time so an incoming surge cannot jump between equipment.


Depends on the sort of surge and what sort of filtering is being put in the line. The isobar for example has some pretty beefy chokes in it that will slow down how quickly a voltage surge can raise current and cause most of the surge energy to go elsewhere.
So to be clear, a surge protector does not interfere with a UPS in a way that matters? Thanks for the information, been putting off researching for a UPS. The time is nigh. The reason I brought up the Tripp Lite isobar is because that was what I personally bought a couple years ago when I wanted something decent for its intended purpose. My modem, WAP, monitors, OLED tv, and PC are all connected to 2 Isobars from the same 2 plug outlet. 1 monitor, PC, modem + WAP on one isobar and 1 monitor and the OLED + other miscellaneous chargers, etc on the other. I didn't wanna hit the ~1550w (possible with my PC + monitors and OLED tv) limit of a single plug on the outlet so I got 2 Isobar's and split my components load across the 2 plugs on the outlet.
 

InvalidError

Titan
Moderator
So to be clear, a surge protector does not interfere with a UPS in a way that matters?
Between the wall and UPS? There is no reason for it to cause any problems there.

Where things can get dodgy is if you connect something like an isobar which has power line filtering (full π filter) to the output of an inverter, then you can have weird issues from interactions between the surge protector's filter, the UPS' own inverter output filter and the inverter controller itself.
 
Between the wall and UPS? There is no reason for it to cause any problems there.

Where things can get dodgy is if you connect something like an isobar which has power line filtering (full π filter) to the output of an inverter, then you can have weird issues from interactions between the surge protector's filter, the UPS' own inverter output filter and the inverter controller itself.
I see, thanks for the information.
 

Aeacus

Titan
Ambassador
Under medium to light loads, stepped approximation UPSes are far more efficient than either "pure sine wave" UPS I have owned (about half as much run-time for the same battery size) and none of my equipment has any issue with stepped approximation. As far as I am concerned, sine wave is a waste of money and efficiency.

Simulated sine wave UPS produces a zero output state during the phase change cycle resulting in a power “gap”. This gap may cause power interruption for active PFC PSUs when switching from AC power output to simulated sine wave output (battery mode). Meaning, that one of 3 things can happen:
  1. UPS displays error resulting PC to shut down immediately.
  2. UPS shuts down resulting PC to shut down immediately.
  3. UPS switches to battery power resulting PC to power off from UPS (PC stays on).
Not a good chances, if you ask me. :rolleyes: That's why i don't suggest nor use simulated sine wave UPS.

I'm using CyberPower CP1300EPFCLCD, which is true/pure sine wave,
specs: https://www.cyberpower.com/hk/en/product/sku/CP1300EPFCLCD

One for Skylake build, another for Haswell build and i have no issues with the efficiency. On normal use (web browsing), i can get ~30mins of runtime out of my UPSes, while on full PC load, i can get ~10mins of runtime out of it. Which is more than enough to safely stop what we're doing, save our items and shut down the PCs. And to extend the UPS runtime, both PCs also have 80+ Titanium efficiency PSUs (Seasonic SSR-650TD for Skylake and Seasonic SSR-650TR for Haswell).

I guess different companies (e.g APC) haven't fine-tuned their true/pure sine wave UPSes to be more efficient. Instead, are focusing on simulated sine wave UPSes.

As far as I am concerned, sine wave is a waste of money and efficiency.

I guess you find 80+ Titanium PSUs also waste of money, right? Since why pay premium price for 80+ Titanium unit when you can be fine with 80+ Gold or even 80+ Bronze unit. And that brings us back to the subjective, based on personal preferences criteria, which is individual for each person.

That is unrealistic since thunderstorms will occur when you aren't there to unplug stuff.

No, it's not. It's negligence since what stops you checking weather forecast beforehand? I'd say laziness. And if you go out for longer than 4 hours, unplug your hardware. And what stops that? Again, laziness.

Being lazy is not an excuse.
 
Last edited:

Tac 25

Estimable
Jul 25, 2021
1,391
421
3,890
All-in-all, life lesson: do not let PC run when in thunderstom.

I learned that lesson the hard way. Kept playing game during a thunderstorm, and boom. Fortunately, only the GPU and the AVR got fried. The rest of the pc still work to this day. https://forums.tomshardware.com/threads/lightning-strike-fried-my-gpu-and-avr.3720913/

lesson learned, a single rumble of thunder (even a faint one) is now enough for me to turn off pc. Does not matter who I'm chatting with, does not matter what game I'm playing. Will turn off pc right away. LOL
 
  • Like
Reactions: Aeacus

InvalidError

Titan
Moderator
Simulated sine wave UPS produces a zero output state during the phase change cycle resulting in a power “gap”. This gap may cause power interruption for active PFC PSUs when switching from AC power output to simulated sine wave output (battery mode). Meaning, that one of 3 things can happen:
  1. UPS displays error resulting PC to shut down immediately.
  2. UPS shuts down resulting PC to shut down immediately.
  3. UPS switches to battery power resulting PC to power off from UPS (PC stays on).
Not a good chances, if you ask me. :rolleyes: That's why i don't suggest nor use simulated sine wave UPS.
Yeah, right.

I have been using stepped approximation UPSes with my PCs for over 20 years and have yet to run into a combination of UPS and PSU that has given me any problems and that includes 80+ Bronze-Titanium PSUs. Properly designed APFC circuit will detect non-sinusoidal input and simply boost whatever the input is to whatever the main caps' target voltage is. Even the crappiest PSUs still have 10+ms of hold-up time and at 60Hz, you have power for about 4ms of the 8.3ms off-time between alternating pulses, so the capacitors only need to hold up for about 5ms.

BTW, my own experience with Cyberpower UPSes is that they are garbage that fails within 3-4 years of purchase. One of them I was previously using for my PC thoroughly cooked its battery because its battery voltage reading got stuck at 24V even when the battery is removed while the battery itself reads sub-20V, the other one (a PFCLCD model) only gives a few seconds of backup likely because it has cooked its battery too. I've never had to replace batteries in my APC UPSes within the first six years. (And I doubt replacing batteries in my CP1500 would do any good since it looks like its BMS is dead.)
 

TRENDING THREADS