R9 390 came back from RMA still broken, goes black after a few minutes of use.

[temporalarcheologist]

My PC has visual distortions and then the screen goes black. Back in December I woke up and my PC seemed to be stuck in sleep mode, the case light was on but nothing was responsive.

I ended up having to cut the power to turn it off and then it didn't turn back on until I replaced the motherboard. I encountered a similar issue to what I have know back in February and solved it with a bit of thermal paste (my first build.), but then in early march the distortions started happening again. Someone on Reddit said that it may be the graphics card overheating. Instead of trying to fix it myself I sent it off to MSI.
A month later it came back without the warranty stickers still not working. The PC will stay on for longer with one RAM stick in and it doesn't matter which one/which port they're in. After removing the GPU everything works normally, the PC stays on and I don't get any driver errors or anything.

Tomorrow I'm going to try using my brother's GPU in my computer to see if it works.

Here are the
specs:
Intel i5-6500k CPU
Gigabyte H110M-A mobo
2x 8 Gb Aegis DDR4 RAM
MSI R9 390 8Gb GPU
EVGA 500B Power Supply
Samsung 250Gb SSD
Seagate 1Tb Hard Drive
Corsair Carbide 200R ATX Case
Acer H6 Series Monitor

 

[maxalge]

My PC has visual distortions and then the screen goes black. Back in December I woke up and my PC seemed to be stuck in sleep mode, the case light was on but nothing was responsive.

I ended up having to cut the power to turn it off and then it didn't turn back on until I replaced the motherboard. I encountered a similar issue to what I have know back in February and solved it with a bit of thermal paste (my first build.), but then in early march the distortions started happening again. Someone on Reddit said that it may be the graphics card overheating. Instead of trying to fix it myself I sent it off to MSI.
A month later it came back without the warranty stickers still not working. The PC will stay on for longer with one RAM stick in and it doesn't matter which one/which port they're in. After removing the GPU everything works normally, the PC stays on and I don't get any driver errors or anything.

Tomorrow I'm going to try using my brother's GPU in my computer to see if it works.

Here are the
specs:
Intel i5-6500k CPU
Gigabyte H110M-A mobo
2x 8 Gb Aegis DDR4 RAM
MSI R9 390 8Gb GPU
EVGA 500B Power Supply
Samsung 250Gb SSD
Seagate 1Tb Hard Drive
Corsair Carbide 200R ATX Case
Acer H6 Series Monitor

Your power supply is awfully weak for your specs, and not very good

test with another psu

need a decent 650w for a 390

 

[temporalarcheologist]

My PC has visual distortions and then the screen goes black. Back in December I woke up and my PC seemed to be stuck in sleep mode, the case light was on but nothing was responsive.

I ended up having to cut the power to turn it off and then it didn't turn back on until I replaced the motherboard. I encountered a similar issue to what I have know back in February and solved it with a bit of thermal paste (my first build.), but then in early march the distortions started happening again. Someone on Reddit said that it may be the graphics card overheating. Instead of trying to fix it myself I sent it off to MSI.
A month later it came back without the warranty stickers still not working. The PC will stay on for longer with one RAM stick in and it doesn't matter which one/which port they're in. After removing the GPU everything works normally, the PC stays on and I don't get any driver errors or anything.

Tomorrow I'm going to try using my brother's GPU in my computer to see if it works.

Here are the
specs:
Intel i5-6500k CPU
Gigabyte H110M-A mobo
2x 8 Gb Aegis DDR4 RAM
MSI R9 390 8Gb GPU
EVGA 500B Power Supply
Samsung 250Gb SSD
Seagate 1Tb Hard Drive
Corsair Carbide 200R ATX Case
Acer H6 Series Monitor

Your power supply is awfully weak for your specs, and not very good

test with another psu

need a decent 650w for a 390

I heard that the quality of the PSU is more important than the wattage (30 amps on the +12V rail needed for r9 390.) What PSU would you recommend? Preferably non-modular (budget is around $60)

 

[mjslakeridge]

You are partially correct about the quality of the PSU being more important than the wattage, but you do need enough wattage to power everything and have a little headroom (maybe 20%). You can enter your build into PCPartpicker.com and it has a wattage calculator.

Some of the really low end PSUs may advertise 500 Watts, but only put out 250 Watts until they go up in a cloud of smoke. Attached is a link to the Tom's Hardware PSU Tier List. If you are overclocking I wouldn't go below Tier 3, Tier 1 or 2 is recommended if your budget allows.

http://www.tomshardware.com/forum/id-2547993/psu-tier-list.html

 

[temporalarcheologist]

You are partially correct about the quality of the PSU being more important than the wattage, but you do need enough wattage to power everything and have a little headroom (maybe 20%). You can enter your build into PCPartpicker.com and it has a wattage calculator.

Some of the really low end PSUs may advertise 500 Watts, but only put out 250 Watts until they go up in a cloud of smoke. Attached is a link to the Tom's Hardware PSU Tier List. If you are overclocking I wouldn't go below Tier 3, Tier 1 or 2 is recommended if your budget allows.

http://www.tomshardware.com/forum/id-2547993/psu-tier-list.html

Based on the guide I'm split between these two
https://www.newegg.com/Product/Product.aspx?Item=N82E16817152070 (Tier 4)

https://www.newegg.com/Product/Product.aspx?Item=9SIABP94N12287 (Tier 3)

 

[rcald2000]

No sir. Tier 1 is the highest quality. Tier 3 and 4 are poorer quality.

I'm not sure if your issue is related to your PSU, but this is a top quality choice.

EVGA P2 650
$99.99 ($89.99 after $10.00 rebate card)
https://www.newegg.com/Product/Product.aspx?Item=N82E16817438055

 

[mjslakeridge]

Stay away from the Tier 4 PSU. Also it has poor reviews on Newegg.

Can your budget support this one?

https://www.amazon.com/SeaSonic-650-Watt-CrossFire-Certified-SSR-650RM/dp/B00918N4A0

Tier 2 Semi-modular.

 

[temporalarcheologist]

No sir. Tier 1 is the highest quality. Tier 3 and 4 are poorer quality.

I'm not sure if your issue is related to your PSU, but this is a top quality choice.

EVGA P2 650
$99.99 ($89.99 after $10.00 rebate card)
https://www.newegg.com/Product/Product.aspx?Item=N82E16817438055

I don't think I could afford cables with it

 

[temporalarcheologist]

Stay away from the Tier 4 PSU. Also it has poor reviews on Newegg.

Can your budget support this one?

https://www.amazon.com/SeaSonic-650-Watt-CrossFire-Certified-SSR-650RM/dp/B00918N4A0

Tier 2 Semi-modular.

That one fits within my budget but I don't have cables

 

[mjslakeridge]

It comes with the cables, just doesn't show in the pictures. I bought one a few months back and like it. I only used the cables I needed for better cable management inside the case.

 

[Karadjgne]

Seasonic psus are the standard every other psu is judged against, they are and have been basically top rated. Not many are better, most are worse. Seasonic is also the OEM for many brands such as some Corsair, Antec, XFX, etc.

That Evga 500b is in no way capable of handling the 390 at any kind of load more than web surfing. Good choice on new psu and good idea to get the old one replaced.

 

[temporalarcheologist]

It comes with the cables, just doesn't show in the pictures. I bought one a few months back and like it. I only used the cables I needed for better cable management inside the case.

Aight it's coming in the mail. Thanks!

 

[mjslakeridge]

Great, let us know how it goes when you get it!

 

[Rexer]

I have an R9 390x and I found it can gobble up a lot of power depending on the application, program, game, whatever. Plus, overclocking to get as much performance as I can, really taxes a psu. I started with an XFX XTR 750w experiencing crashes time to time and replaced it with a EVGA 850 G2. Lol. Believe me, when 390 makers recommend a 750w psu, they're thinking about you.
Everything was fine for a couple of months till total blackouts started. After a few weeks of random blackouts, my UPS unit suddenly died. ? It was an 9 old APC VA 800 unit. So old, it was only suited as a work station unit. I replaced it with an APC NS 1080. First time I had a UPS unit fail without it being the battery.
Hope this helps.

 

[Karadjgne]

Minuteman 700 Pro, changed up the 1200ma/h battery for a 1500ma/h few years ago. Still kicking. It's currently 12 yrs old. But it's only pushing a Evga 550 G2, so alls good.

 

[Rexer]

Minuteman 700 Pro, changed up the 1200ma/h battery for a 1500ma/h few years ago. Still kicking. It's currently 12 yrs old. But it's only pushing a Evga 550 G2, so alls good.

12 yrs. old! I just might look Minuteman up. Thanks.

 

[temporalarcheologist]

Great, let us know how it goes when you get it!

So I got the new power supply as well as a mobo speaker and plugged them in. Dark Souls 3 ran smoothly with the GPU for about 15 minutes before having the same issue as I had in the past (freeze then black screen. Sound stays the same but keyboard inputs stop working). I took out the RAM and put my brother's in and was unable to POST with 5 long beeps. I googled this and saw that this means there's a RAM issue (Gigabyte is shitty about telling their customers about things), so I took one of them out and I was able to boot up with 1 stick in the 2nd port. The PC is still running now but I haven't ran anything 3D yet. 1 of the RAM ports is probably messed up (2nd Gigabyte motherboard in 6 months, never buying from them again) and I may have some bad RAM (my brother says his computer is a bit slower with it but he hasn't gotten a crash yet) Here's the temp logs before the crash and the logs from event viewer:

Reset to device, \Device\RaidPort0, was issued.
The IO operation at logical block address 0xaca7bc0 for Disk 0 (PDO name: \Device\00000034) was retried.
The start type of the Background Intelligent Transfer Service service was changed from auto start to demand start.
The IO operation at logical block address 0x26679130 for Disk 0 (PDO name: \Device\00000034) was retried.
DATE TIME MHz CPU_0 CPU_1 CPU_2 CPU_3 LOAD%
05/14/17 15:14:10 3591.00 54 58 54 51 10.1
05/14/17 15:14:20 3591.00 59 64 54 56 17.8
05/14/17 15:14:30 3591.00 60 65 55 56 16.4
05/14/17 15:14:40 3591.00 62 69 57 58 15.7
05/14/17 15:14:50 3591.00 62 67 56 59 14.6
05/14/17 15:15:00 3591.00 64 72 61 61 18.6
05/14/17 15:15:10 3591.00 68 72 60 63 15.8
05/14/17 15:15:20 3591.00 67 73 60 63 15.3
05/14/17 15:15:30 3591.00 69 75 65 65 17.2
05/14/17 15:15:40 3591.00 75 82 69 69 21.2
05/14/17 15:15:50 3591.00 72 77 64 67 15.1
05/14/17 15:16:00 3591.00 68 80 63 66 8.0

I'll attach the excel document I got from OpenHardwareMonitor(it doesn't include times though??? it's in 30 second increments.) and I'll start running something hardware intensive after I post this.


(Excel document, I had no idea how to format this, sorry.) https://pastebin.com/Surh0qvZ

 

[temporalarcheologist]

Great, let us know how it goes when you get it!

Okay so with the working stick of RAM I was able to run Metal Gear Rising for about 10 minutes before getting the same issue. I removed the GPU and now it boots fine (even with the RAM in the broken slot). The side of the GPU that is closest to the CPU felt really hot when I took it out but the CPU felt fine. Someone recommended replacing the stock cooler but I'm on a tight budget and I'm not sure that that would fix it.

http://imgur.com/a/J1c0N
Album with cable management pictures

 

[Karadjgne]

You have an area broadcast type of cpu cooler. When you run that thing hot, and you will with such a small, ineffective cooler, a considerable amount of the cpu exhaust will be going straight down and heating up the backside of your gpu. It's also going out towards the ram. It's looking like your entire issue is with that cooler. Quick and easy test for that. Get a chunk of cardboard and cut it to the width of the gpu, let it run 1/2 way towards the front then upwards between cooler and ram, like a capital L but backwards, isolating the cpu cooler from ram/gpu. This'll still let airflow reach the upper area from the sides, but will stop any exhaust from the cpu cooler reaching the gpu or ram. See if that helps with stability / ram / gpu usage. You also have bare minimum airflow, you need not only to clean up the wiring, but get air moving in that case. The lower front by the hdds and the optical bays are stifling airflow, and a single rear exhaust on a area cpu cooler are doing next to nothing for flow. With what you've got to work with, you'd be better off turning that rear exhaust into an intake and blowing air directly across that cooler, there's a ton of venting above for the heat to natural rise. Either that or stick 2x fans on top and forget about the rear exhaust placement, there's not that much exhausted that way with that cooler.

 

[mjslakeridge]

To test whether heat buildup is causing your issues, leave the side panel of the case off, and point a room fan directly into the case and run the game that was causing the crashes. If no crashes, then what Karadjgne suggested is probably the issue with airflow.

 

[Rexer]

I forgot to mention this. Make sure after each crash, that you reset the game to default settings. It's sometimes referred to 'optimal settings', 'reset game' or 'start in safe mode' Sometimes games remember a crash as part of their program and they need to reset to a default state to erase it. Call of Duty was notorious for remembering black screens. Crashing all the time corrupts the game, creates a lot of gpu heat and the routine uninstalling/reinstalling's really super aggravating.

 

[temporalarcheologist]

To test whether heat buildup is causing your issues, leave the side panel of the case off, and point a room fan directly into the case and run the game that was causing the crashes. If no crashes, then what Karadjgne suggested is probably the issue with airflow.

I don't have a personal fan but I breadboarded it with similar results
https://www.youtube.com/watch?v=TtN5xtHDt7o (Freeze at 3:04)

 

[temporalarcheologist]

You have an area broadcast type of cpu cooler. When you run that thing hot, and you will with such a small, ineffective cooler, a considerable amount of the cpu exhaust will be going straight down and heating up the backside of your gpu. It's also going out towards the ram. It's looking like your entire issue is with that cooler. Quick and easy test for that. Get a chunk of cardboard and cut it to the width of the gpu, let it run 1/2 way towards the front then upwards between cooler and ram, like a capital L but backwards, isolating the cpu cooler from ram/gpu. This'll still let airflow reach the upper area from the sides, but will stop any exhaust from the cpu cooler reaching the gpu or ram. See if that helps with stability / ram / gpu usage. You also have bare minimum airflow, you need not only to clean up the wiring, but get air moving in that case. The lower front by the hdds and the optical bays are stifling airflow, and a single rear exhaust on a area cpu cooler are doing next to nothing for flow. With what you've got to work with, you'd be better off turning that rear exhaust into an intake and blowing air directly across that cooler, there's a ton of venting above for the heat to natural rise. Either that or stick 2x fans on top and forget about the rear exhaust placement, there's not that much exhausted that way with that cooler.

So if I were to replace the CPU cooler would that fix the problem? Are these good?
https://www.amazon.com/Premium-Quality-Technology-NF-A14-PWM/dp/B00CP6QLY6/ref=lp_11036281_1_6?s=pc&ie=UTF8&qid=1495172453&sr=1-6
https://www.amazon.com/Cooler-Master-Hyper-RR-212E-20PK-R2-120mm/dp/B005O65JXI/ref=lp_11036281_1_1?s=pc&ie=UTF8&qid=1495172453&sr=1-1
https://www.amazon.com/DEEPCOOL-GAMMAXX-400-Heatpipes-Compatible/dp/B007JEMXYQ?tag=bom_tomshardware-20&ascsubtag=site:thus|tid:89471495172459215

 

[Karadjgne]

The gammax 400 is slightly better than the hyper212 evo, both quite good coolers. The bonus there is the tower design will push a good majority of the heat right at your rear exhaust fan, and then out. It'll go a long way to solving case heat temps. The addition of the other fan above the rear exhaust will provide greater draw on the remaining case heat. Because of where your storage drives are, clearing out as much space for intake airflow from the front fans is important for the gpu temps especially.

 

[temporalarcheologist]

The gammax 400 is slightly better than the hyper212 evo, both quite good coolers. The bonus there is the tower design will push a good majority of the heat right at your rear exhaust fan, and then out. It'll go a long way to solving case heat temps. The addition of the other fan above the rear exhaust will provide greater draw on the remaining case heat. Because of where your storage drives are, clearing out as much space for intake airflow from the front fans is important for the gpu temps especially.

got new CPU cooler, facing same issue
https://youtu.be/lQQZkczh7qc