Strange Graphics Problem

puser

Prominent
May 19, 2017
21
0
510
Hello everyone and sorry for my bad english.

After having to deal with a LOT of issues with my PC, I decided to come here, to Tom's Hardware, to try to find help, since this is, if not the best, one of the best support forums for these topics.

My problem started around 4 months ago. I was used to playing Counter Strike GO at around 120 fps. Suddently, the maximum fps rate decreased to 60. As my screen only supports 60 fps, I did not care about it.
3 months later, as my computer was starting to go extremely slow, I decided to save my files to an HDD and hard-reset it. Here begun the big problem.

Since then, my graphic performance has decreased a lot. I cannot play videos on youtube nor using windows media (at some points of the video, like each 4 seconds, the screen freezes and then all the stuff that had to be reproduced during that time comes out extremely fast). I cannot play any kind of game (CS:GO is totally unusable, since I never reach 15fps, and characters seem to teleport). Sometimes, I have difficulties to browse the internet (even Google), since I have the same problem that with the videos (the screen freezes for a while (<1sec) and then everything comes fast). And when I try to execute some stress test, the same problem appears.

This issue is the same in Windows 10 that in Windows 7. I also tried installing Ubuntu, and the problem was different: I could only click the icons from the "status bar"; if I opened any new windows such as a Mozilla session os even the explorer, my PC did not recognise any click inside them.

Just to clarify, I have not made any change in the hardware of my PC since maybe 6 months or so, that I decided to add two new RAM cards.

This is my PC setting:

- AMD FX-4300 Black Eddition
- Asrock 980DE3/U3S3 (latest BIOS update)
- Mars Gaming MPI700 Power Supply
- Samsung 850 PRO SSD
- Sapphire DUAL-X R9 270X OC DirectX 11.2 version (2Gb GDDR5 and 1050MHz)
- 4x4Gbx1333 Kingston Blu Memories

I have been able to test all the 4 RAM modules and the SSD in two other computers, and they seem to work well (I also ran Memtest without trouble). The problem is that I do not own enought components as for testing the Graphic Card, the MoBo or the Power Supply independently.

I attach the results of some video tests and stress tests I made, as they may be useful. I you need any other information, just let me know. I will answer as soon as possible. The may thing I am looking for is to find where the flaw is, since both a new graphic card and the pack processor+MoBo are very (too) expensive. Also, I would like to discard a power supply issue.

Thanks a lot for your help.

(The last GPU-Z screenshot was taken just after finishing a strss test using FurMark)


--------


Video Memory Stress Test Log:

Changing video mode to 640x480x16...OK
[20/05/2017 1:11:32] Test started for "AMD Radeon R9 200 Series"...
Trying 16bpp RGB:565 mode...OK
Trying 16bpp RGB:555 mode...OK
Trying 16bpp BGR:565 mode...NOT SUPPORTED (Code: 80004001)
Trying 32bpp RGB:888 mode...OK
Trying 32bpp BGR:888 mode...NOT SUPPORTED (Code: 80004001)
[20/05/2017 1:52:35] Pass completed (0 errors found).



T8IQM0.png


PaCKUm.png


jXEO82.png


X3gn46.gif


DGDo6a.png
 

puser

Prominent
May 19, 2017
21
0
510


Well. As I already said, that temperature is only reached just after a stress test, where the maximum temperature with the GPU at 99% and fan at 85% is aroud 80ºC. During a common usage, the temperatures are much lower. Normally, they lie between 35ºC and 40ºC. Thanks

 

PistolPaul

Honorable
Feb 15, 2014
83
2
10,665
The sort of issues you describe sound like temps to me too. Having seen my system have a dust on fins problem (graphics card) and again when my water pump on my cooler got stuck some of your systems problems sound eerily similar. You will be surprised how you can quickly go from everything being normal to just being terrible by getting things hot!

While powered down take your graphics card out and give it a good blow out over the vents - put it back in and try it. Verify the fans on your graphics card are spinning up with your eyes! Dont just trust a RPM reading - LOOK.

If that doesnt fix things pay attention to your CPU cooler. Is it water - air ? Clean up your rads / heatsink fans, whatever and again make sure fans are actually running when the system powers up. Dont trust RPM readings, visually check!

Good luck!
 

puser

Prominent
May 19, 2017
21
0
510



Hello PistolPaul and thank you for your interest. My graphic's card cooler (a double fan) seems to be ok. It is totally clean (it's the third time I clean all fans since the problems began), and the noise it makes during the stress tests seems to be normal. My CPU fan is a Cooler Master Hyper 212X, more than enought for an AMD 4300 without OC.

By the way, I attach here two screenshots of GPU-Z and HardwareMonitor during a casual game in CS:GO (screenshot of in-game settings included in spanish, everything at "low" and resolution at 1920x1080, since lowering the resolution does not fix the issue; both gpu monitors were started once the game had already begun).

Thank you.

UPDATE: Playing/watching videos/using my PC with the case open does not solve the problem (I did it to check temperatures). My box is a Nox Hummer MC

muWKb7.png


1rvfLm.png


PLfsLw.png


 

PistolPaul

Honorable
Feb 15, 2014
83
2
10,665
OK, so fans are spinning, heatsinks are clean and blown out. You've eliminated airflow by running with the case open so it is not that and temps seem to be ok. I know AUXTIN looks high but apparently is not reliable on all motherboards anyway. Marece la pena - check the PSU/Fuente Alimentacion is able to breathe too!

I know we are limited to what we can test / swap as you say. One thing you may be able to test is your Disco Duro - there are various tools, see here.

https://www.howtogeek.com/134735/how-to-see-if-your-hard-drive-is-dying/

**EDIT** Of note - la marca MARS GAMING has a poor reputation. Is there any way you can test with a new PSU?


Suerte!
 

scout_03

Titan
Ambassador
try this on speed fan main page put fans on auto then press configurethen check box and pres to read wanted temp for the lower one to be 35 c with at max of 50 c then press log boxand show in notification zone for all of them on fan check all box and log you could do the same for voltages , then on speed check all box and put 1000% value for all of them for max one leaving min at 0% check the automatic variation and log press ok close speed fan and recheck temp on main page . use hwinfo on sensore page to check all fan speed voltages and others readings .
 

puser

Prominent
May 19, 2017
21
0
510



Thank you for your ideas and nice Spanish!

I have proved my SSD in 2 other computers (one of them was a laptop) and graphics work very fluently, without problem. I also tried to install 2 other hard drives (not SSD but HDD) in my PC, one of them running WIndows 10 and the other one running Ubuntu, and the problem is there.

I attach a screenshot of CrystalDiskInfo. Thank you.

pfIWsl.png

 

puser

Prominent
May 19, 2017
21
0
510


Well. If I understood well what you told me to do, the problem continues there. I attach a new screenshot with data of both SpeedFan and HM during a normal web browsing use, in which I suffered the lagging/freezing problem. They are pretty much the same that before.

Thanks.

YlqM6Z.png
 

PistolPaul

Honorable
Feb 15, 2014
83
2
10,665
Ok, so PSU? is there any way to borrow a fuente from an amigo? I have no experience of what a PSU can cause to happen if it is broken but from your parts list the Mars Gaming PSU is the weakest link in the chain.
 

puser

Prominent
May 19, 2017
21
0
510


I will try to test it independently. On the other hand, voltages obtained using HardwareMonitor seem to be OK.

I will post here the results as soon as possible.
 

PistolPaul

Honorable
Feb 15, 2014
83
2
10,665
If your problem is not there then I do not know where it is. Your thermals are good and you have eliminated software side issues trying fresh installs on cross checked drives and eliminated memory from the equation too. Now you are left with physical parts, CPU / GPU and PSU and motherboard. The more I look at that PSU you have the less I like it and the marca in general!

http://www.realhardtechx.com/index_archivos/Tacens.html
http://todopsu.com/2016/08/18/que-es-lo-que-pasa-con-tacens/

I wish you the best of luck. I know you can measure voltages but it won't tell you how clean that voltage is. If you can really do try and physically swap it for another to test. Beyond that I am afraid you will be left eliminating the remaining parts and figuring out which one is responsible for your problem. Do let us know if you ever figure this out.

Que tengas suerte amigo..
 

scout_03

Titan
Ambassador
now on speed fan main page put all 6 box to 100% the other thing on the second hwmonitor you post the cpu fan run at 665 rpm ans on the first one it was at 1711 rpm this is more a correct speed if it run to low wont cool cpu and could cause boothing error also did you check all this with hwinfo for temp and others .
 

puser

Prominent
May 19, 2017
21
0
510



Well! I have some news!

I am now using an almost-new EVGA G3 550W PSU, and the problem continues. Moreover, I found a cheap GeForce 340 of 1Gb to try, and, yes. The problem is the same with my original card than with this one.

Then, in conclusion, the failing component may be either the MoBo or the processor. Is there any way to find the flaw without using other components, only by some software tests?

BTW, I am planning to get a new Intel Mobo+processor, but I would like to fully repair this PC for me to sell it working properly.

Thanks!
 

puser

Prominent
May 19, 2017
21
0
510


I did not unplug my processor from its socket since the first time I used it, so it can't have any bent pin. Isn't there any software solution?
 

puser

Prominent
May 19, 2017
21
0
510


Well. The PSU is new, and its a reputable brand, so there's no point for it to fail.

I have just stress-tested my PC using Aida64. When I run either CPU, FPU, Memory, Disk or Cache tests, everything works great. However, when I try to run the GPU stress test, my PC gets crazy: keyboard stops working and my mouse works with very high lag (>5 seconds?), so that it is difficult even to click the sop button.

I am going to get crazy. Is it possible for the failures to appear only when stressing the GPU but be a flaw of the MoBo or processor? Because I have tried 2 different good video cards, 1 AMD and 1 NVidia (with no drivers' problems) and the problem has not vanished
 

PistolPaul

Honorable
Feb 15, 2014
83
2
10,665
We knew that at some point it would come down to eliminating components one by one. It is VERY unfortunate that it seems it will come down to the very last thing you can test before you realise that it was that. Out of what is left, Mobo or CPU? I would say Mobo. A mobo is more complex (in terms of having things that can break) - VRMs - Sockets and slots that can bend and warp with age - Capacitors - Resistors - Diodes - on and on and on. I doubt your CPU has a bent pin if you have not taken it out. Note that while you may think your new PSU was not a good buy because it did not fix your problem be aware your old PSU maybe the cause of whatever may have gone wrong with your machine. Again I wish you luck, you have been very unlucky not to have found your problem so far.
 

puser

Prominent
May 19, 2017
21
0
510


Well, this is my new PC:

-AMD Ryzen 1700
-MSI SLI Plus
-G.Skill 8Gb 2133 DDR4 RAM module
-GeForce 340 (Tested in another PC)
- EVGA G3 550W PSU

As you can see, I have change EVERY SINGLE COMPONENT (including the SDD/HDD, I have tried a buch of them) and, yes. The problem continues.

:(
 

PistolPaul

Honorable
Feb 15, 2014
83
2
10,665
You are saying that with a completely new system - every component is different - with a completely new operating system installation - when you go to run games you get exactly the same problems that you had with your old system? Have you been cursed by a witch? It's kind of getting to the point I'd switch even the mouse, keyboard and monitor just to be sure!!

Are you using the same PC case as the old system? Does it use a vertical riser slot or something - with a PCI-E extension cable from the motherboard? This problem is just crazy!
 

puser

Prominent
May 19, 2017
21
0
510


Well, I think that we have 3 options:

1. There is a stupid error that I am repeating over and over when installing some specific component
2. 2 components of the same type have the same issue (i.e.: 2 graphic cards, 2 mobos...)
3. Witches exist