Search
Sign in with
Sign up | Sign in
Solved

M5a78l/usb3 shuts down after few minutes of unreal tournament 3

Last response: in Motherboards
Share
Register or log in to remove
Motherboard Authority

This is a good quality PSU. My first thoughts were that the PSU was not adequate for the intended applications.

You can try monitoring the PSU voltages in "HWInfo" while running Prime95 for about 15 minutes. Check and see whether the voltages are steady and not sagging or jumping all around.

Another option is to try a different PSU for test purposes.
Related ressources
Ask the community

Ubrales said:
This is a good quality PSU. My first thoughts were that the PSU was not adequate for the intended applications.

You can try monitoring the PSU voltages in "HWInfo" while running Prime95 for about 15 minutes. Check and see whether the voltages are steady and not sagging or jumping all around.

Another option is to try a different PSU for test purposes.



I saved a log while running prime95, nothing unusual happened

https://docs.google.com/file/d/0B4iDlAnIpq18a3FnVnQ5ZW1... are you able to get from this link?

Ubrales said:
I couldn't view the log, but if you observed all the voltages to be steady then the PSU is robust.

I don't think that it is your hardware. Look for settings in the game.



I'm taking a look at it right now. Too many data, will take some time.

Is advisable to turn off anti surge in this case? Perhaps... ?

Even disabling anti surge computer crashes (is a computer crash, not a application crash or OS), but only with this game! Upgraded CPU to a FX 8350 Vishera, crash still happens. Any other game is running fine: Dragon Age, Mass effect 1, 2 and 3, Starcraft 2...
  • gusnd has edited this thread

If it's that one game then it could be driver related. if it were more than one game, then I'd be looking at thermal problems and memory stability as well as the PSU. It can be tricky to debug issues like these. It is made easier if you have spare parts, sadly not many users do...

raja@asus said:
If it's that one game then it could be driver related. if it were more than one game, then I'd be looking at thermal problems and memory stability as well as the PSU. It can be tricky to debug issues like these. It is made easier if you have spare parts, sadly not many users do...



Maybe I could try to log when running that game, I didn't think of that.

Now I found a second game which that happens, Anno 2070, but it happened after more than 2 hours of gameplay... I'm starting to bet on driver related issue.

The memory I'm using is corsair




Motherboard Authority

Usually not a CPU problem. CPUs are hard to kill. They will shut down under excessive voltages or excessive temps.

Try easier (no cost) solutions like updating (flashing) the BIOS. Download the latest BIOS from the motherboard manufacturer's website.

Let's get this problem solved.

Thanks Ubrales!

Here what I tested so far:

- Memory:
- Switched memory banks
- tested with memtest, no errors
- tested memories single channeled: when tested this way, the computer crash took a bit longer to happen, and surprsingly I could finish one UT3 mission and then reach another crash, maybe because is single channeled the issue do not happen that fast, not sure though, just a theory

Maybe memory timing issue?

BIOS: latest version already. But when I disabled Memory Hole Remapping happened a lot faster, in seconds!

VGA: is a Gigabyte geforce 650 ti 1gb, found an update for its bios, sounded like a holy grail for my crusade, and did not change anything

What I have connected to the computer:



Onboard network and sound card disabled, using a SBRecon3d PCIe and an old intel nic, since the on board is unable to auto negotiate the connection speed with the router, and manually only at 10mbps

Not sure if it matters, but when it shuts down, the computer sound start to stutter and the mainboard make memory issue beepings
  • gusnd has edited this thread
Motherboard Authority

One of my fellow Tom's Forum members (King) recommended HWInfo; good tool - http://www.hwinfo.com/ http://www.hwinfo.com/download64.html (64-bit)

Try this and let it run while you are gaming. Check data.

Another tool that I like for temps measurements is RealTemp - http://www.techpowerup.com/realtemp/

Try this. Check the temps as well as the max CPU voltage while gaming. If either one of these is over the safe spec the computer will shut off or BSOD. I know that you do not overclock and these shutdowns are usually associated with overclocking the CPU.

I'll check bios right now. Another thing is, I have unleashed mode enabled, but even when disabled the same issue happens.

I already did the log today when running the game here's what I got at the end of the log before computer crashes

  1. CPU Power [W]	29.654
    

  2. CPU [°C]	64.0
    

  3. Vcore [V]	1.392
    

  4. Vccp2 [V]	1.512
    

  5. +12V [V]	10.343
    

  6. +5V [V]	5.130
    

  7. +3.3V [V]	3.324
    

  8. VIN5 [V]	2.208
    

  9. VIN6 [V]	2.052
    

  10. 3VSB [V]	0.312
    

  11. VBAT [V]	3.312


If I understood it right, the only wrong thing here is the 12v+?

but within the bios is shown as

  • gusnd has edited this thread
  • gusnd has edited this thread

It's clearly a VGA issue (so far)

After I cleaned the VGA, it takes loooooonger (if compared with before) to give tha problem

Is possible to disable the factory overclocking?

  • gusnd has edited this thread
Motherboard Authority

Yes you can adjust the clock frequencies by going into the Nvidia Control Panel. I have EVGA GTX 660Ti and use EVGA Precision-X to adjust the clocks (not that I need to adjust them, but I can if I want to).
Motherboard Authority

gusnd said:
But would this fix the crashing? Or is it a driver issue? Saw some people with similar issues with 650 ti


It could be a driver issue. On my older i7-920 computer with a GTX 275 graphics card I had to roll back to an older driver and then everything was fine. The newer Nvidia driver caused freezes to Photoshop. That's why I rolled back to the older driver.

My i7-3770K has a GTX 660Ti and runs fine with the latest Nvidia driver.
  • Ubrales has edited this thread

I'm trying the latest beta right now, will do some tests, but UT3 really puts the vga to the test, I don't know why, it's an old game, should be fine.

Not sure if I can test an old driver to a newer model, might not have the files or be able to identify the model.

I opened a ticket at Gigabyte :/  hoping for the best :sarcastic: 
Motherboard Authority

Unless you have deleted the older driver, it is still in the Nvidia folder on your computer (probably on the C: drive). If you find it, double-click on it and install it and try the games. (Check your downloads folder too)
  • Ubrales has edited this thread
Motherboard Expert

I may be off-base, since this is not my area of expertise, but +12V [V] at 10.343 V is abominable. While it's possible that poorly made components could draw enough to do that, on a power supply with headspace it seems unlikely.

I did not see anything above about you trying a different power supply, except the suggestion from Ubrales. It's the very first thing that I would try, but I have a spare in my attic.

"VGA: is a Gigabyte geforce 650 ti 1gb, found an update for its bios, sounded like a holy grail for my crusade, and did not change anything." The second thing that I would try is putting the old card back, if you still have it. (I don't have one in my attic; we're cleaning house and I threw it out yesterday.)

Stupid question: does it have extra power connectors? If so, are they connected? The first question shows how little I know about video cards.
  • WyomingKnott has edited this thread

WyomingKnott said:
I may be off-base, since this is not my area of expertise, but +12V [V] at 10.343 V is abominable. While it's possible that poorly made components could draw enough to do that, on a power supply with headspace it seems unlikely.

I did not see anything above about you trying a different power supply, except the suggestion from Ubrales. It's the very first thing that I would try, but I have a spare in my attic.

"VGA: is a Gigabyte geforce 650 ti 1gb, found an update for its bios, sounded like a holy grail for my crusade, and did not change anything." The second thing that I would try is putting the old card back, if you still have it. (I don't have one in my attic; we're cleaning house and I threw it out yesterday.)

Stupid question: does it have extra power connectors? If so, are they connected? The first question shows how little I know about video cards.


No, they don't have extra connectors, is a six pinned connector, but my psu have the connector for a 8 pinned vga.

http://www.gigabyte.us/products/product-page.aspx?pid=4...

@stickg1 asked me to check the values at the BIOS and they are ok, I don't know why when measured by the software in windows they do not match.

Maybe I would replace with the older videocard (Radeon 6850) but I sold it, I don't keep these stuff, used the money to not spend that much replacing something, maybe I should start keeping after this incident :heink: , but I could not try the game when I had it, but I did not had any kind of issue besides being a slower card and missing PhysX.

I'm going to test my vga with a friend's computer ASAP and see if that happens too. He also have an OCZ PSU but with 600w but he as EVGA GeForce 650 Ti boost, and nothing happened but it's another computer, comparisons may not be precise.

I'm going to check the voltage values there too and see what I get.
Motherboard Expert

I'm not trying to insult your gear but the 12v, 3.3v, and 5v readings are typically inaccurate via software on the lower end budget motherboards. Heck, a few of the other readings can be a ways off too. With budget boards you have to generally rely on the BIOS for accurate readings.

Finally a "solution", not a exactly solution since the model is defective per se. As mentioned in their site, the vga comes factory overclocked, but their overclock honestly sux.

http://www.gigabyte.us/products/product-page.aspx?pid=4...

They mention 1032 MHz (standard : 928 MHz), I reduced with their own program, overclock guru (awful name btw), and it is pretty stable.

Fact is, there is no sign of temperature rise, so, RMA and wait God knows how long for a fix? Keep it running it that way, underclocked, or put some hardcore cooling (not sure it would fix)?

The right one would be RMA... but I don't have good experiences with that.

Keep the factory clock and put a higher voltage? I'm not used with that, I will need some guidance

I mean, there is any good voltage to start trying or I should keep trying until it burns?
  • gusnd has edited this thread
Motherboard Expert

gusnd said:
Keep the factory clock and put a higher voltage? I'm not used with that, I will need some guidance

I mean, there is any good voltage to start trying or I should keep trying until it burns?


Hmm, it's only a 650ti so chances are the voltage is locked. Nevermind...

stickg1 said:
gusnd said:
Keep the factory clock and put a higher voltage? I'm not used with that, I will need some guidance

I mean, there is any good voltage to start trying or I should keep trying until it burns?


Hmm, it's only a 650ti so chances are the voltage is locked. Nevermind...


Their "guru" allows me to change the voltage as well




Motherboard Expert

I'm not too familiar with your GPU. I do have a GTX 660 that I have been tinkering with. Stock load voltage on that is 1.175v. I would say don't go over that, but chances are you can adjust the slider but you'll see that under load it will go right back to 1.125v, that's how this generation of nvidia GPUs have reacted without modding.

stickg1 said:
I'm not too familiar with your GPU. I do have a GTX 660 that I have been tinkering with. Stock load voltage on that is 1.175v. I would say don't go over that, but chances are you can adjust the slider but you'll see that under load it will go right back to 1.125v, that's how this generation of nvidia GPUs have reacted without modding.


Ok, here's what I did and what happened.

I've set the fan from auto to manual, 100% almost all the time, but I believe is not a good thing, right? Or could I leave that way with no problem?

Added a little more voltage, around +0.0037, kept running the engine it reached 52ºc for a few time and then crashed.

The default voltage as guru shows seems to be 1100

I changed nothing right now, and it seems the gpu gets hot TOO fast and can't stand to something above 50ºc for more than minutes
  • gusnd has edited this thread

Best solution

Motherboard Expert

50C is not even close to hot. If it were 100C then it might get damaged, but it would shut itself off before that happened. You have a bad card. RMA it.
  • gusnd selected this answer as the solution

All I can do now is thank you people for your support on it guys :) , nothing that was just bought should be this problematic, I'll be skipping gigabyte products, any other brand would be safer? (Before the topic is closed)
Motherboard Authority

My choice has been and is EVGA! Get it on sale. I recommend the GTX 660Ti - they have the 2 GB as well as the 3 GB models. And often they have a free game included.

This is the card that I have: http://www.newegg.com/Product/Product.aspx?Item=N82E168... (paid $288 after promos and rebate).

Should you decide to go for this card make sure the PSU has two graphics card power connectors and the case is large enough to accommodate this card. Sometimes, EVGA's site has better pricing than Newegg. Wait for the Memorial Day sale.
Register or log in to remove
Ask the community
Read discussions in other Motherboards categories
!