[SOLVED] My PC that was built 2 days ago suddenly wont turn on

Apr 20, 2020
13
0
10
0
My Specs are
Ryzen 5 2600
RX580 8gb XXX GTS by XFX
Asrock B450M PRO4
G.skill 16gb 3600mhz ram
512gb ssd
Had an APEVIA Prestige 600w 80+ Gold now im switching it to a Corsair CV650 +80 Bronze

So i builded my PC 2 days ago and it was working fine apart from a driver issue that i fixed and was working fine and then when i was playing Forza Horizon 4 my pc randomly shutted down suddenly and it wouldnt turn on again. When i disconnect the PSU and connect it back, when i try to power it again the fans on the pc spin for like .1 second and then turn off immediately after. i was checking around the pc and everything seemed normal. No smell of burned cables or anything and did some testing and found out that when i unplugged the GPU from the PSU my PC turns on with all the fans and RGB and everything but as soon i reconnected the GPU to the PSU the PC will do the same fans will move for .1 second and nothing on the pc would turn on but without the GPU connected on the PSU it will turn on fine.

Right now im returning the PSU for a Refund to get a Corsair with 650W idk if its the PSU or the GPU.
 

LinuxDevice

Reputable
May 20, 2017
461
28
4,990
73
i connected the GPU with the PSU PCle cable did that caused the smoke?
Probably. The PCIe slot cannot deliver as much power as the extra connector (which does not go through the motherboard). However, I am thinking the GPU was already bad...when it went bad it probably took out that part of the original PSU. Once you put in a new PSU which didn't go bad something had to give...and it was likely the power along that second connector which made it possible to smoke.
 

LinuxDevice

Reputable
May 20, 2017
461
28
4,990
73
I tried that but in didnt work either :/
If no other power supply to test (and perhaps even if you do have another to test), if you have integrated video, then try without the video card.

You don't actually need RAM or hard drives to get to BIOS POST, so try with all hard drives removed (power as well as data). Then try with only one RAM stick (reminds me of Ramstein, "Du Hast"). If nothing starts, try to see if it gets any further with no RAM at all. If RAM, video card, no hard disk, etc., won't boot, then you pretty much have to try another power supply.

Sometimes thermal creep will take a previously working expansion card (like video) which was previously working to apparently fail until it is reinserted. If things work after removing any component, then try adding the component back in. Any component not mentioned could use reinserting.

Motherboards can be right at the edge of having the mounting screws hit something to ground. Not often, but it can happen. If you have an anti-static surface, then you might try removing the motherboard and running it outside the chassis...again, with no RAM or hard drive.

If it still fails, then remove and examine the CPU pins and socket closely (be careful with static). Remounting the CPU sometimes fixes such issues. Don't forget to use good thermal paste during reassembly.

It is a bad idea to operate without a surge protector, but often the only failed device will be the power supply itself, or some individual peripheral (such as a SATA drive). When devices are connected in circuit that are far less vulnerable than prior to being added to the circuit, but are still vulnerable (especially versus inductive loads, e.g., air conditioner or heating vent fan). I 100% always recommend using surge protectors, and systems do find their way into shops after storms or static electricity hits, but unless this died during a power outage or some unusual circumstance I'm tempted to believe the issue is more likely the power supply than the components downstream to the power supply. FYI, when working in an electronic manufacturing floor, it is a firing offense to walk out on the floor just once without using the correct procedure...entire shelves of parts have been destroyed with a single touch (unmounted parts). I tend to just hold on to the frame of the chassis if I'm being informal, but it does help to have an anti-static surface.

Also, don't forget that before you go and remove or swap or reinstall a component that the power supply should not only be off, but capacitors should be discharged by holding the power button on for about 10 seconds prior to any interaction.
 
Apr 20, 2020
13
0
10
0
If no other power supply to test (and perhaps even if you do have another to test), if you have integrated video, then try without the video card.

You don't actually need RAM or hard drives to get to BIOS POST, so try with all hard drives removed (power as well as data). Then try with only one RAM stick (reminds me of Ramstein, "Du Hast"). If nothing starts, try to see if it gets any further with no RAM at all. If RAM, video card, no hard disk, etc., won't boot, then you pretty much have to try another power supply.

Sometimes thermal creep will take a previously working expansion card (like video) which was previously working to apparently fail until it is reinserted. If things work after removing any component, then try adding the component back in. Any component not mentioned could use reinserting.

Motherboards can be right at the edge of having the mounting screws hit something to ground. Not often, but it can happen. If you have an anti-static surface, then you might try removing the motherboard and running it outside the chassis...again, with no RAM or hard drive.

If it still fails, then remove and examine the CPU pins and socket closely (be careful with static). Remounting the CPU sometimes fixes such issues. Don't forget to use good thermal paste during reassembly.

It is a bad idea to operate without a surge protector, but often the only failed device will be the power supply itself, or some individual peripheral (such as a SATA drive). When devices are connected in circuit that are far less vulnerable than prior to being added to the circuit, but are still vulnerable (especially versus inductive loads, e.g., air conditioner or heating vent fan). I 100% always recommend using surge protectors, and systems do find their way into shops after storms or static electricity hits, but unless this died during a power outage or some unusual circumstance I'm tempted to believe the issue is more likely the power supply than the components downstream to the power supply. FYI, when working in an electronic manufacturing floor, it is a firing offense to walk out on the floor just once without using the correct procedure...entire shelves of parts have been destroyed with a single touch (unmounted parts). I tend to just hold on to the frame of the chassis if I'm being informal, but it does help to have an anti-static surface.

Also, don't forget that before you go and remove or swap or reinstall a component that the power supply should not only be off, but capacitors should be discharged by holding the power button on for about 10 seconds prior to any interaction.
The problem is only when the GPU is on the PSU idk if the PSU died on me and its not delivering the Requiered Voltage but ill need to buy a power surge protector as well since i dont have any. Ill see when my new PSU arrives
 

LinuxDevice

Reputable
May 20, 2017
461
28
4,990
73
This indicates your power supply is insufficient or failing. When the GPU is consuming its rated power the rails will not have enough to power the other content on the motherboard. GPUs with the extra power connector should not be run without that connector. So long as the new supply is rated at high enough power this should fix the issue.
 
Apr 20, 2020
13
0
10
0
This indicates your power supply is insufficient or failing. When the GPU is consuming its rated power the rails will not have enough to power the other content on the motherboard. GPUs with the extra power connector should not be run without that connector. So long as the new supply is rated at high enough power this should fix the issue.
Tomorrow arrives my new PSU ill answer here if the problem was solved with my new Corsair cv650w.
 
Apr 20, 2020
13
0
10
0
3600MHz is not a supported memory speed for for Ryzen 5 2600. Leaving BIOS memory settings on [Auto] is also only good for initial start up and not much past that.
I clocked the Ram for 3200mhz so its not at full 3600 and i wasnt having any issues apart from the one i mentioned.
 
Apr 20, 2020
13
0
10
0
This indicates your power supply is insufficient or failing. When the GPU is consuming its rated power the rails will not have enough to power the other content on the motherboard. GPUs with the extra power connector should not be run without that connector. So long as the new supply is rated at high enough power this should fix the issue.
Welp i connected the new PSU and when i started the PC up the GPU started smoking and burned it self guess it was the GPU after all, might need to contact the ebay seller that sold it to me for the warranty it had till next year.
 

LinuxDevice

Reputable
May 20, 2017
461
28
4,990
73
Welp i connected the new PSU and when i started the PC up the GPU started smoking and burned it self guess it was the GPU after all, might need to contact the ebay seller that sold it to me for the warranty it had till next year.
This is why I always buy power supplies with short circuit protection. The last time this happened to me it was a removable hard drive SATA tray which shorted. This took out the power supply. Had the power supply I bought to replace this not been with short circuit protection, then the new supply would have been dead as well and I wouldn't have known about the actual short causing this. On the other hand, since the GPU smoked instead of the power supply, and since the power delivery was not via the PCIe slot, chances are nothing else died.
 
Apr 20, 2020
13
0
10
0
This is why I always buy power supplies with short circuit protection. The last time this happened to me it was a removable hard drive SATA tray which shorted. This took out the power supply. Had the power supply I bought to replace this not been with short circuit protection, then the new supply would have been dead as well and I wouldn't have known about the actual short causing this. On the other hand, since the GPU smoked instead of the power supply, and since the power delivery was not via the PCIe slot, chances are nothing else died.
i connected the GPU with the PSU PCle cable did that caused the smoke?
 

LinuxDevice

Reputable
May 20, 2017
461
28
4,990
73
i connected the GPU with the PSU PCle cable did that caused the smoke?
Probably. The PCIe slot cannot deliver as much power as the extra connector (which does not go through the motherboard). However, I am thinking the GPU was already bad...when it went bad it probably took out that part of the original PSU. Once you put in a new PSU which didn't go bad something had to give...and it was likely the power along that second connector which made it possible to smoke.
 

ASK THE COMMUNITY

TRENDING THREADS