[SOLVED] Replaced GPU for losing signal, issue still occuring

[Cramit845]

I purchased my son a pc a couple years back and put in a graphics card for some basic PC gaming. Back in October or so, my son was using my old R9 290 video card and the card was causing the pc to reboot, which seemed related to one of the fans falling off and the GPU over heating.

I purchased a GTX 970 used from a buddy of mine on the cheap and the rebooting issues stopped and everything was fine.

Circle back to a week ago and suddenly the monitor would lose signal but my son could still hear things in the background. This seemed to only happen when he was playing Overwatch or Sea of Thieves, but also started to happen while doing basic stuff like Roblox. We had it on HDMI, tried different cables, same issue, tried VGA, same issue. The PC would boot fine, do Windows basic tasks, web browsing, video viewing, No Problem. However it go to the point that whenever you went into Overwatch or Sea of Thieves, it happened and even started to happen on Roblox.

He has on board video with his Acer TC-780A(KBL) board, and he was able to play the games without issue, however not Overwatch or Sea of Thieves, he could barely load into them because of the basic onboard GPU but still they wouldnt crash out or lose signal.

With the GPUpocalypse going on, I was hesitant to purchase something, but I figured it was the most likely cause, so I purchased him a new AsRock RX570. We received it today, however after doing a clean install, updated drivers, he loaded into Overwatch and same issue again.

Least I can say, things are not great here now, although as a budding PC gamer, a good experience for him to go through with working on PC's.

EDIT Forgot to mention, with RX570 in, when the monitor loses signal, the pc reboots a moment afterwards

Anyway, at this point, I'm kinda at a loss. I want to blame the PCI Express 3.0 on the motherboard, but that means finding a replacement for his current one and moving the processor which I haven't done in a hot minute. Also, after spending a good chunk of change on RX570, I would love to just get either other ideas or verification I'm heading in the right direction.

Any thoughts to the main issue and where I should Head?

PC Specs:
Acer TC-780A (KBL)
GTX 970 / RX 570 (Problem presents with both, currently on onboard video)
8GM RAM
Intel Core i5 7400 @ 3.00 GHZ
Kaby Lake
Socket 1151LGA
Roswell 750W PSU

 

[kiss-method]

After all of that, it sounds like you're dealing with a bad PSU. PC works fine with integrated graphics, which takes a lot less energy, and as soon as you put a GPU that actually draws some power you seem to get a hard reset. Maybe you can try another PSU if you have one?

Not sure buying a new RX570 was a good choice either... but we're not talking about that... something like a 1660super/ti couldve been a better path if you wanted to spend money (although I don't know where you live)

edit: try using another 8pin plug set from your PSU to the GPU. Sometimes they burn

 

[Third-Eye]

I purchased my son a pc a couple years back and put in a graphics card for some basic PC gaming. Back in October or so, my son was using my old R9 290 video card and the card was causing the pc to reboot, which seemed related to one of the fans falling off and the GPU over heating.

I purchased a GTX 970 used from a buddy of mine on the cheap and the rebooting issues stopped and everything was fine.

Circle back to a week ago and suddenly the monitor would lose signal but my son could still hear things in the background. This seemed to only happen when he was playing Overwatch or Sea of Thieves, but also started to happen while doing basic stuff like Roblox. We had it on HDMI, tried different cables, same issue, tried VGA, same issue. The PC would boot fine, do Windows basic tasks, web browsing, video viewing, No Problem. However it go to the point that whenever you went into Overwatch or Sea of Thieves, it happened and even started to happen on Roblox.

He has on board video with his Acer TC-780A(KBL) board, and he was able to play the games without issue, however not Overwatch or Sea of Thieves, he could barely load into them because of the basic onboard GPU but still they wouldnt crash out or lose signal.

With the GPUpocalypse going on, I was hesitant to purchase something, but I figured it was the most likely cause, so I purchased him a new AsRock RX570. We received it today, however after doing a clean install, updated drivers, he loaded into Overwatch and same issue again.

Least I can say, things are not great here now, although as a budding PC gamer, a good experience for him to go through with working on PC's.

EDIT Forgot to mention, with RX570 in, when the monitor loses signal, the pc reboots a moment afterwards

Anyway, at this point, I'm kinda at a loss. I want to blame the PCI Express 3.0 on the motherboard, but that means finding a replacement for his current one and moving the processor which I haven't done in a hot minute. Also, after spending a good chunk of change on RX570, I would love to just get either other ideas or verification I'm heading in the right direction.

Any thoughts to the main issue and where I should Head?

PC Specs:
Acer TC-780A (KBL)
GTX 970 / RX 570 (Problem presents with both, currently on onboard video)
8GM RAM
Intel Core i5 7400 @ 3.00 GHZ
Kaby Lake
Socket 1151LGA
Roswell 750W PSU

As was stated already, it may be a PSU issue that can be fixed with a new PSU. However, you should run a ram test with memtest86 and see what happens. If you get errors on the first test pass, you could possibly have a bad module causing issues. You might have sifted CPu pin issues. Normally the system would lock up because of a shifted pin related to ram and other components or if the ram is faulty, programs will crash or BSOD and reboot Windows.

A possibly more likely issue is one or more CPU socket pins that are related to PCIE are shifted out of place. If CPU pins are the issue, it could be caused from a heat-sink that is too tightly mounted or there were already pins shifted out of position that are affected by heating and cooling cycles from CPU usages.

Don't attempt the following if you don't trust yourself to not damage anything and take it to a computer shop instead if you are fine with paying for the service.

You could start by loosening each of the CPU heat-sink screws by a quarter to a half turn and see what happens. If it actually improves things, you may need to remove the heat-sink and check the CPU socket pins for damage from different angles and with good lighting. Then reseat the CPU in the socket if you don't see any damage or pins out of position and repaste it after cleaning the heat-sink and CPU. Don't do this until you have thermal paste, something like Arctic MX-4 would work fine.

 

[kiss-method]

As was stated already, it may be a PSU issue that can be fixed with a new PSU. However, you should run a ram test with memtest86 and see what happens. If you get errors on the first test pass, you could possibly have a bad module causing issues. You might have sifted CPu pin issues. Normally the system would lock up because of a shifted pin related to ram and other components or if the ram is faulty, programs will crash or BSOD and reboot Windows.

A possibly more likely issue is one or more CPU socket pins that are related to PCIE are shifted out of place. If CPU pins are the issue, it could be caused from a heat-sink that is too tightly mounted or there were already pins shifted out of position that are affected by heating and cooling cycles from CPU usages.

Don't attempt the following if you don't trust yourself to not damage anything and take it to a computer shop instead if you are fine with paying for the service.

You could start by loosening each of the CPU heat-sink screws by a quarter to a half turn and see what happens. If it actually improves things, you may need to remove the heat-sink and check the CPU socket pins for damage from different angles and with good lighting. Then reseat the CPU in the socket if you don't see any damage or pins out of position and repaste it after cleaning the heat-sink and CPU. Don't do this until you have thermal paste, something like Arctic MX-4 would work fine.

would assume ram would be okay since he is not getting blue screens or errors as they would usually be accompanied by, as well as OP said that he tried integrated graphics and it worked fine, which would load up system ram to work. Test would be good to do, but doubt he'll find anything there.

again, why would you say it's it's more likely that it's the CPU? using integrated graphics runs through his CPU, and it works fine. Could it be.. possibly? sure there's always a chance, but in the scenario of putting load on PC with a hard reset/ GPU power cut, I seems like the wrong place to start. CPU moving scenario is a more of a once in a blue moon solution, once clamped down into their guide slots, it's not an easy task to move it, it's not like the heat sink is bolted to the CPU itself either.

After having several of my own PSU failing on me, this fits the bill.

 

[Archaic59]

I purchased my son a pc a couple years back and put in a graphics card for some basic PC gaming. Back in October or so, my son was using my old R9 290 video card and the card was causing the pc to reboot, which seemed related to one of the fans falling off and the GPU over heating.

I purchased a GTX 970 used from a buddy of mine on the cheap and the rebooting issues stopped and everything was fine.

Circle back to a week ago and suddenly the monitor would lose signal but my son could still hear things in the background. This seemed to only happen when he was playing Overwatch or Sea of Thieves, but also started to happen while doing basic stuff like Roblox. We had it on HDMI, tried different cables, same issue, tried VGA, same issue. The PC would boot fine, do Windows basic tasks, web browsing, video viewing, No Problem. However it go to the point that whenever you went into Overwatch or Sea of Thieves, it happened and even started to happen on Roblox.

He has on board video with his Acer TC-780A(KBL) board, and he was able to play the games without issue, however not Overwatch or Sea of Thieves, he could barely load into them because of the basic onboard GPU but still they wouldnt crash out or lose signal.

With the GPUpocalypse going on, I was hesitant to purchase something, but I figured it was the most likely cause, so I purchased him a new AsRock RX570. We received it today, however after doing a clean install, updated drivers, he loaded into Overwatch and same issue again.

Least I can say, things are not great here now, although as a budding PC gamer, a good experience for him to go through with working on PC's.

EDIT Forgot to mention, with RX570 in, when the monitor loses signal, the pc reboots a moment afterwards

Anyway, at this point, I'm kinda at a loss. I want to blame the PCI Express 3.0 on the motherboard, but that means finding a replacement for his current one and moving the processor which I haven't done in a hot minute. Also, after spending a good chunk of change on RX570, I would love to just get either other ideas or verification I'm heading in the right direction.

Any thoughts to the main issue and where I should Head?

PC Specs:
Acer TC-780A (KBL)
GTX 970 / RX 570 (Problem presents with both, currently on onboard video)
8GM RAM
Intel Core i5 7400 @ 3.00 GHZ
Kaby Lake
Socket 1151LGA
Roswell 750W PSU

• Before replacing any parts or taking anything apart, try a complete wipe of the display drivers using DDU and installing the latest direct from AMD.
• DDU is a great freeware program that's helped countless people fix driver issues. Only takes a few minutes and costs nothing. Download the latest AMD drivers before running DDU so you have them handy.
• If your issue persists you can try more aggressive measures. It could also be a windows issue. Try software fixes first.

 

[Third-Eye]

would assume ram would be okay since he is not getting blue screens or errors as they would usually be accompanied by, as well as OP said that he tried integrated graphics and it worked fine, which would load up system ram to work. Test would be good to do, but doubt he'll find anything there.

again, why would you say it's it's more likely that it's the CPU? using integrated graphics runs through his CPU, and it works fine. Could it be.. possibly? sure there's always a chance, but in the scenario of putting load on PC with a hard reset/ GPU power cut, I seems like the wrong place to start. CPU moving scenario is a more of a once in a blue moon solution, once clamped down into their guide slots, it's not an easy task to move it, it's not like the heat sink is bolted to the CPU itself either.

After having several of my own PSU failing on me, this fits the bill.

That is why I said PSU first, but I read of GPU failures or bugginess like this before, because of Intel LGA CPU socket pins being shifted from previous work on a system or from transport. The suggested test was to loosen the heat-sink to see if it changes the behavior during testing and then do a full socket check and CPU reseat.

I personally have only seen system boot looping and ram channels becoming unsable on Intel LGA sockets as well as recently discovering two melted pins on one of my Z77 motherboards that was causing system lock ups when it did run or not booting up and having to do a full power button shutdown and pray it would boot up.

 

[Cramit845]

Thanks for all the replies guys. Today my son did some testing with the other 6-pin cords with my supervision, and the PC lost signal then a moment later rebooted again. I also noticed that the internal PSU fan doesnt seem to be spinning at all. We checked and the fan is unobstructed and will spin easily when pushed. I'm more thinking the PSU at this point all though the CPU reseat seems like a last ditch effort. This PC doesnt move really, other than a little giggling at times from my son.

The PSU we currently have is a Rosewill RX750-S-B

I am gonna start pricing a replacement although if anyone has any suggestions, definitely open to it.

I also agree purchasing another GPU was a bit fool hardy so soon but I can live with that at this point if we can get to an actual fix.

If anyone has any other suggestions, please let me know, but moving forward from now I am going to look at replacement PSU's and see if that does the trick.

 

[Third-Eye]

Thanks for all the replies guys. Today my son did some testing with the other 6-pin cords with my supervision, and the PC lost signal then a moment later rebooted again. I also noticed that the internal PSU fan doesnt seem to be spinning at all. We checked and the fan is unobstructed and will spin easily when pushed. I'm more thinking the PSU at this point all though the CPU reseat seems like a last ditch effort. This PC doesnt move really, other than a little giggling at times from my son.

The PSU we currently have is a Rosewill RX750-S-B

I am gonna start pricing a replacement although if anyone has any suggestions, definitely open to it.

I also agree purchasing another GPU was a bit fool hardy so soon but I can live with that at this point if we can get to an actual fix.

If anyone has any other suggestions, please let me know, but moving forward from now I am going to look at replacement PSU's and see if that does the trick.

Get at least a 550watt 80+ Bronze to Gold rated with 2x8pin PCIE for your current GPUs or 650-750watt for future GPUs. Corsair or Seasonic are good brands, I have no current experience with others like EVGA or newegg's Rosewill brand.

 

[Cramit845]

Get at least a 550watt 80+ Bronze to Gold rated with 2x8pin PCIE for your current GPUs or 650-750watt for future GPUs. Corsair or Seasonic are good brands, I have no current experience with others like EVGA or newegg's Rosewill brand.

He currently has a 750watt in there, so I am trying to stick with that. It seems it should be enough. I'm currently looking at this:

https://www.newegg.com/seasonic-focus-750-gold-ssr-750fm-750w/p/N82E16817151201?Item=N82E16817151201

In my reddit thread someone linked a linustech tips site with a listing from november 2020, this one seemed to be good according to that but still researching.

PSU List: https://linustechtips.com/topic/1116640-psucultists-psu-tier-list/

 

[kiss-method]

He currently has a 750watt in there, so I am trying to stick with that. It seems it should be enough. I'm currently looking at this:

https://www.newegg.com/seasonic-focus-750-gold-ssr-750fm-750w/p/N82E16817151201?Item=N82E16817151201

In my reddit thread someone linked a linustech tips site with a listing from november 2020, this one seemed to be good according to that but still researching.

PSU List: https://linustechtips.com/topic/1116640-psucultists-psu-tier-list/

you'll only be throwing money away by replacing with 750watt PSU. Your system really does not need it, and there is very little benefit to having 750w unit (unless you can get a sale price for cheaper than counterpart 650s or what not). As Third Eye has mentioned you want to get 550-650w unit 80Bronze and higher, a decent 650w will be able to handle a strong GPU and I don't see a reason to go above that for your rig.

Here's a PSU tier list that basically lists units from best > worst. Use it as a guide and find the best price match/unit. For the price range you're looking at, I would suggest that you buy a FULL MODULAR unit. You're already paying that price tag, it's just less cabling to deal with.

>>> PSU TIER LIST LINK <<<

Here is another useful link, PC Part Picker. Compares vendor prices for selected items, everyone around here uses this website to compare and complete builds for others.

>>> Part Picker Helper <<<

 

[Cramit845]

you'll only be throwing money away by replacing with 750watt PSU. Your system really does not need it, and there is very little benefit to having 750w unit (unless you can get a sale price for cheaper than counterpart 650s or what not). As Third Eye has mentioned you want to get 550-650w unit 80Bronze and higher, a decent 650w will be able to handle a strong GPU and I don't see a reason to go above that for your rig.

Here's a PSU tier list that basically lists units from best > worst. Use it as a guide and find the best price match/unit. For the price range you're looking at, I would suggest that you buy a FULL MODULAR unit. You're already paying that price tag, it's just less cabling to deal with.

>>> PSU TIER LIST LINK <<<

Here is another useful link, PC Part Picker. Compares vendor prices for selected items, everyone around here uses this website to compare and complete builds for others.

>>> Part Picker Helper <<<

Ok, well I will take a look. I see the 650w version of the same one I linked for 107 on newegg which doesnt seem all that bad and is listed tier 1 on that list. I guess I will look around some more and see what else is possible to save some money.