• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

*UPDATE - PROBLEM ISOLATED* System shuts down during 3DMark...

brigden

brigden

Diamond Member
This system has been an absolute bitch of a build. One thing after the other has gone wrong. We've finally got it going, but I think the processor heatsink might not be sitting right. I used the stock AMD heatsink with thermal tape. Haven't had any problems with stability until we run 3DMark2001, that's when the system locks up.

Does anyone else thinks there could be a problem with the contact between the CPU and the heatsink, or is there something I could be missing? The CPU was running at 60 degrees celcius without an exhaust case fan.

Here's what we're dealing with:

AMD 2700+
EPoX 8RDA+
512MB PC2700 RAM
ATi Radeon 9800 Pro

UPDATE

OK, I've pulled the system apart completely and put it back together again in an attempt to isolate the issue.

a.) The CPU is on correctly, contact with the HSF is good. The fan is running at full speed.
b.) The CPU is running at about 50 degrees celcius, the system is at 38.
c.) The PSU is an Antec SmartBlue 350. The rails look good.
d.) RAM timmings are 7-3-3-2.5.

The system runs perfectly, until 3DMark2001 SE is run. It almost always craps out during the Max Payne sequence. I'm either kicked back ti Windows and given an error message, or I get the BSOD. I managed a few good runs and scored 15.5k, which is where this system should score.

I loaded up the BF1942 multiplayer demo and after a few minutes the game kicked me to windows.

Guys, I'm at my wits end with this system. Everything seems to run, until I load up a 3D app. What the hell could be the issue?

UPDATE #2

I've isolated the problem. It's the RAM. I swapped in some of my stuff, with lower timmings, and I was able to run 3DMark2001 fifteen times without a problem.

I'm going to instruct the owner to buy some more reliable RAM, probably Crucial.
 
Is the stepped end of the heatsink over the raised, solid-plastic end of the CPU socket (on an 8RDA+, this is the top edge)? If not, you are in heap big trouble, because if it's oriented the other way, the heatsink can't land flat on the CPU core, resulting in very bad thermal transfer from the core to the heatsink. The melt-to-fit nature of AMD's one-shot thermal pad has save you thus far, if this is what's happened.

Besides not being able to get flat contact with the core, it also puts heavy pressure on the bottom edge of the core, which could result in breakage. If this is how the heatsink is attached, then be very ginger about removing it so you don't crack your CPU core in the process :Q

After removing the heatsink, scrape off all the remnants of the used-up thermal patch and use a high-quality thermal grease in its place. Install the heatsink correctly, and you should be seeing temps in the high 40's or so.
 
Since I didn't put the heatsink on myself, and as the machine is away from me, I can't check to see if the owner put the heatsink on incorrectly. But if I'm correct, doesn't the design of the 8RDA+ prevent one from incorrectly installing the heatsink?

The owner told me that when he installed the heatsink, it went on flat and without much difficulty. He also made sure he removed the plastic covering the thermal pad.
 
WHen you go into the bios use by spd. Otherwise try fail safe bios settings or whatever the option is called for that bios. I tried using faster timings for the memory. Computer would run but as soon as any games were started. It would run for a little and
then restart. Go into system properties, advanced, start up and recovery. Under system failure uncheck automatic restart. Then a bsod will appear instead of the computer
just restarting can help to find the problem.
 
BIOS setting for memory frequency is auto, not speed. Auto is default.

Automatic restart is disabled, so we know the system isn't trying to restart. The program is just locking up, kicking us to the desktop, or completely freezing.
 
Originally posted by: brigden
Since I didn't put the heatsink on myself, and as the machine is away from me, I can't check to see if the owner put the heatsink on incorrectly. But if I'm correct, doesn't the design of the 8RDA+ prevent one from incorrectly installing the heatsink?

The owner told me that when he installed the heatsink, it went on flat and without much difficulty. He also made sure he removed the plastic covering the thermal pad.
Click to expand...
I don't know of any SocketA board that can prevent the reversed installation of the heatsink, sorry to say. The clip has one end with a screwdriver-tip receptacle, and that corresponds to the stepped end of the heatsink's base, so the screwdriver end of the clip should be at the top edge, facing the power supply. Can he confirm that's the case?
 
That's good then. Try with the case open, so it doesn't get so hot, and see if it handles 3DMark2001 when it's cooler.
 
Yeah, the molex is connected to the video card. However, the HDD shares the same line. Could the video card and HDD on the same line be too much for the PSU?
 
No thats probably ok. I think Mechbgon is probably on the right track. If that thing is running 60's then it might be unstable do to overheating, especially if the 60 is idle.
 
I'll ask him one more time, but he told me the screwdriver edge is orientated towards the top of the case. If that is true, then the heatsink should be on correctly.
 
Forget the auto memory timings. If you want to eliminate memory as the culprit, it is simple. set the timings to 3-3-3-8 (CAS 3, and at stock speed, 166 and stock multiplier) and try to run it. If it works, then gradually set the timings more agressive until it won't run, and then you know what the problem is. If I am wrong, and the memory timing doesn't solve your problem, then try another approach.
 
My 1.8Ghz Pali never hits above 47C under full load, 40C idle. 60C is way too high in my book. Even tho the threshold is 90C, you'll get stability problems way before you get anywhere near 90C (get out the pot n boil some water 😛)

I would say set memory to 2.5-4-4-8-1T for safety, you can change it to more aggresive timings if it's stable then. Also get a better HSF, perferably an all copper one they work wonders. No need for a 80mm swiftec if ur not gunna do any heavy overclocking.

The 9800Pro and HD shouldn't be a problem, i have 2 Raptors + 9800Pro sharing the same line from my 480w Antec Blue, so it should be ok. on the 350w too w/ less strain

Hope u figure it out, peace. And what ^ said, CL 3 is ridiculas unless your doing some heavy OC'in



 
Originally posted by: OverVolt
My 1.8Ghz Pali never hits above 47C under full load, 40C idle. 60C is way too high in my book. Even tho the threshold is 90C, you'll get stability problems way before you get anywhere near 90C (get out the pot n boil some water 😛)

I would say set memory to 2.5-4-4-8-1T for safety, you can change it to more aggresive timings if it's stable then. Also get a better HSF, perferably an all copper one they work wonders. No need for a 80mm swiftec if ur not gunna do any heavy overclocking.

The 9800Pro and HD shouldn't be a problem, i have 2 Raptors + 9800Pro sharing the same line from my 480w Antec Blue, so it should be ok. on the 350w too w/ less strain

Hope u figure it out, peace. And what ^ said, CL 3 is ridiculas unless your doing some heavy OC'in
Click to expand...

Yes, CL3 is rediculous, but I suggested it as a start point. I have actually seen many times that (crap) memory won't run well, unless set to 3, and I suggested to set it at that, and if it worked, then at least you can be sure of the problem, and start from there with timings, and more more agressive until you max out. Plain old PC2700 can be bad (especially Fry's memory specials)

 
Do you guys really think the memory is causing the system to crap out?

Everything seems to run smoothly until we attempt to run 3DMark, which leads me to believe the system is running too hot and crashes.

Once I get the system in my hands, I'll play with the memory settings like you have suggested. The owner isn't too bright and getting him to run these tests over the phone is a royal bitch.

I'm going to pull the entire thing apart, make sure he put the heatsink on properly, and that all his fans are hooked up and running.

If anyone else has any ideas, please feel free to contribute. I'll update this thread once I get my hands on the PC.

Thanks a million.
 

Has anyone come to the conclussion that 3DMark might be a piece of sh1t?

I hear (and have experienced) more grief about 3DMark on an otherwise perfectly stable and fast system. It's crap and it has little to no bearing on real world performance.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top