System freezing under load

kromboy

Distinguished
Feb 19, 2011
18
0
18,520
Hello folks,
hoping to get some expert opinions about my issue.

I recently took my big rig build and put it into a small Antec Minuet case.

Specs are as follows:It's basically an Asus CM5570, put into a new case, with a new mobo(replaced about 4 years ago, blown caps)
Additional changes: PSU swap, Heatsink swap


EDIT 1:

CPU:

Core 2 Quad Q8200 (95w version)

CPU:
Radeon HD 6570 1gig (slim-ready version, only has a small fan
picture for the gpu
http://www.newegg.com/Product/Product.aspx?Item=N82E16814121591

Memory:
2x2(4g) double stick of ram (Nanya is the brand, DDR2 6200)

Mobo
Gigabyte GA-G41M-ES2L

PSU:
Athena power 300w (it's the highest output TFL psu available nearby)
http://www.microcenter.com/product/382283/300_Watt_mTFX_Power_Supply

Heatsink:
Dynatron K987 (only heatsink with copper that fits the case)
http://www.microcenter.com/product/336076/K987_Universal_CPU_Cooler_For_Intel_LGA_1156_-_1366_-_775


Recently, all of a sudden I have experienced freezing issues under load while playing games.
Specifically, CoD WaW and Ghost Recon Phantoms.

After about 1 minute of playing, the system would simply freeze and sound loop.

I checked the temps and replaced the cpu heatsink to get the temp under load down to 55 range (was 80 before)

GPU temp is fine

I replaced the PSU because it was a 220w unit, my system needs about 200w, 240w at full load.

new PSU is 300w, should be sufficient.

I ran the OCCT, cpu, gpu, and psu, all 3 tests ran fine.

Only culprit left now is the RAM and mobo.

Mobo has no signs of apparent damage, all the capacitors look normal, no bulges.


by swapping out the ram sticks, I found out one of them was faulty and causing the crashes immediately under load.

So I am now sticking to 2gb of ram.

Here is the issue though

CoD WaW: min. req. 512mb ram, runs fine.

Ghost Recon: min req. 3gb ram. Also runs fine, no slow downs. however after about 3 hours of playing, it froze.



I am trying to figure out why it froze after such a long time, it seems to be a different issue.

Can it be possible for system to simply freeze when there's insufficient ram? Or do I just have 2 separate problems, i.e. bad RAM stick + bad mobo?


Last freezing isn't something I can replicate which makes troubleshooting even more difficult.

If anybody has any experience similar to this, I would love a feedback.


Thank you


TL;DR

replaced Heatsink, temps fine
replaced PSU, enough wattage,
took out faulty RAM, runs okay
random freeze after 2-3 hours of running.
Is the issue the other stick of ram? (Same age as the other one, about 7 years old)
Or is it a mobo issue that was just exacerbated by a bad RAM before?



EDIT2: My system specs are now edited.
Also, I would like to run memtest, but because of the legacy usb issue, and not having a P/S 2 keyboard, I can't run it in the advanced mode, memtest passes fine for the default mode.

I used the psu calculator at http://extreme.outervision.com/psucalculatorlite.jsp
It shows 199w for regular use, 240w max for everything under 100% load.
I think my new 300w psu should be okay for that, also the OCCT passes for psu test as well.
 
Solution
There's afair chance the pc is freezing because thew cpu is getting too hot. Can you download and run HWInfo - Sensors and post the results so we can see whether its a high Vcore or low fan rpm or maybe you need to re-seat the cpu cooler.

kromboy

Distinguished
Feb 19, 2011
18
0
18,520


Oh, I just built a low end pc for home (pretty much the system builder marathon q2 2013 $400 build, with updated components)
http://www.tomshardware.com/reviews/mini-itx-do-it-yourself-game-console,3531.html

and decided to change the case for my old pc to have at my business, I don't think I want to invest much more than $50 on the pc to keep its current specs)
 

kromboy

Distinguished
Feb 19, 2011
18
0
18,520


Well damn, I guess the online calculator was way off :(

I think you guys are right, changing the case really only made 2 differences, less cooling, and smaller psu.
I've already changed the cooling situation and system is very very cool.

If that is the case, do you think running with 2gb ram is simply drawing less power and making it a bit more stable with such a minute difference? or maybe I just had a faulty ram AND an inadequate psu?

I think I will try to grab my 500W psu and just use it externally to test it out, if it changes anything I will report back.

Thank you for your help

Also, I plan on grabbing a P/S2 keyboard to use memtest with legacy usb disabled, how good is memtest in detecting faulty ram? and third party software to use instead?

Thank you


EDIT: Also, about the reduced memory situation AFTER installing the game, how do I make the game recognize that the memory has been reduced? is it somewhere in the registries? I think that is also a plausible cause, because right now, CoD WaW runs fine, Ghost Recon doesn't.
 
There's afair chance the pc is freezing because thew cpu is getting too hot. Can you download and run HWInfo - Sensors and post the results so we can see whether its a high Vcore or low fan rpm or maybe you need to re-seat the cpu cooler.
 
Solution

kromboy

Distinguished
Feb 19, 2011
18
0
18,520
Thank you for your continued support, but unless the temp sensors are messed up, my cpu and gpu temps look very good.

I got a better heat sink and it won't go over 50 degrees celsius under load.
Before I put on the aftermarket heatsink, it did go up to about 80 degrees.

So far I've tried new RAM, new (500W) psu, better cooling.

I don't think the symptoms point to my gpu being broken, so I'm leaning towards my motherboard. I just don't think hunting down an LGA 775 mobo is worth the trouble.

My system is very stable though, just not under heavy load, it can even play games like TF2 without freezing, so I've decided to retire it into an HTPC and put down about 200 bucks and use an old case and psu to make an A10-7850k rig.