New build problems

S

Spark_Plug

Weaksauce
Joined
Jun 27, 2005
Messages
115
I just built a new computer last night with the following components:

MSI 790GX-G65 Motherboard
AMD Athlon Phenom II x4 945
2 Raedon 4890s in Crossfire
4gb OCZ Reaper DDR3 Ram

And then took all the drives and PSU from my old build, PSU is an HX620.

I installed Windows 7 and have been having a lot of problems. Don't know if it is the hardware or what. But at first it was giving me the blue screen because of BCCode 116, which I read was a video driver problem. Haven't been getting that lately. But I decided to just run OCCT to see if the system was even stable and after 10-15 minutes everything just freezes up. I did notice the temps were pretty high under load, max was 60. I don't know if this problem is due to the heat or what. And then when I try to run 3dMark Vantage everything freezes up to, requiring me to hold the shutdown button in until it shuts down. Any ideas on problems here? Haven't installed any games to try yet though. And the 4890s are an XFX and an HIS.

EDIT: Also tried Prime95 and the system instantly froze up.
 
That mobo has an integrated graphics processor. Have you setup your bios to disable the onboard graphics?
 
Three questions:

1) Are you overclocking?
2) Have you entered the BIOS to ensure your RAM is running at rated speeds & timings with rated voltage?
3) Have you run memtest+86 on each individual DIMM of RAM?

In my experience, a system freezing up completely as you described is most likely caused by an issue with RAM or with the memory controller.
 
lt_shiro, I took out one of the 4890s and it is went though OCCT without freezing up.

Markyip1
I am not overclocking, RAM is not running at the rated speeds. I bought PC3 12800, but according to CPU-Z it is running at 533 mhz, which to my understanding is below what it shoujld be. I have not run memtest on it either.
 
If you're running at stock, then 533mhz is right. 533mhz = 1066mhz double data rate. I would make sure that your RAM is not being undervolted by the BIOS. JDEC spec (which is what you are running at stock) calls for a voltage of 1.5v. However, some RAM may need more than that for stock 1066mhz, especially RAM that is designed to run faster.

You may also want to run memtest86+ to see if you have a defective DIMM. The best way to run this is to take out RAM so that you are running with one DIMM at a time. This way, if memtest86+ fails, you know specifically which DIMM is bad. A few passes should be enough to tell you whether a stick needs to be RMA'd.
 
You need around 40A for both the 4890 cards plus the systems current draw.
You are pushing your PSU to its limit.
 
Alright, so I am running memtest, there has been 1 pass, no errors on the memory.

so its the PSU, which PSU would you guys recommend? I was looking at the Corsair 750HX or the PC Power and Cooling Silencer 750W. I love the 620HX which I why I looked at the Corsair one, and I love the modular part of it especially, but right now the PC Power and Cooling one is at $120. Is the Corsair worth the extra $50?
 
So as I asked before, does problem present itself when you are running one 4890?
 
How many drives do you have? A Ph2 X4 945 system with HD4890's in CF with a couple drives should be pulling around 450W, so you'll need at least 40A on the +12V rail for the whole system. So yes, you're nearing your HX 620W's limit.

The TP New lt shiro linked to above should be more than enough.
 
That’s a good number of drives, i would also recommend you upgrade to that power supply in the link, getting one myself soon. You might want to check to see if one of your pci x slots on the motherboard are bad. .
 
itomwisp said:
That’s a good number of drives, i would also recommend you upgrade to that power supply in the link, getting one myself soon. You might want to check to see if one of your pci x slots on the motherboard are bad. .
Click to expand...

That board has no PCI-X slots. ;)
 
Good grief, I was ok until I saw how many drives you were running...these guys have nailed it totally. You've maxed your power supply. The 750W that was recommended will work beautifully, but if you intend to add anything else into that box, you may want to spring for a 1000W just to have that extra buffer. Just sayin'... ;-)
 
WellI got that new power supply today. Hooked it all up, thought It was working great. 40min into OCCT, freezes up on me again. I had to hold in the power button to get the computer to go off, then when I tried to start it up again, fans would spin-up, stop, then spin-up and would go in a cycle like that, it wouldn't post, no beeps or anything. I decided to take out a piece of memory, and then it booted back up. Swapped that memory for the one I took out and it booted back up, put both pieces of memory in and it wouldn't. So I think it has to do with where the memory is seated. Because of the orientation of the heatsink, I can only put memory into two of the 4 slots. If i try to put memory in the other slot, it does make contact with the heatsink. Would this cause this sort of problem? Or is it a problem with the motherboard?

Oh and the heatsink is an Thermalright Ultra 120 Extreme.
 
Last edited:
Spark_Plug said:
WellI got that new power supply today. Hooked it all up, thought It was working great. 40min into OCCT, freezes up on me again. I had to hold in the power button to get the computer to go off, then when I tried to start it up again, fans would spin-up, stop, then spin-up and would go in a cycle like that, it wouldn't post, no beeps or anything. I decided to take out a piece of memory, and then it booted back up. Swapped that memory for the one I took out and it booted back up, put both pieces of memory in and it wouldn't. So I think it has to do with where the memory is seated. Because of the orientation of the heatsink, I can only put memory into two of the 4 slots. If i try to put memory in the other slot, it does make contact with the heatsink. Would this cause this sort of problem? Or is it a problem with the motherboard?
Click to expand...

with some of my old boards, there were 'right' and 'wrong' ways of placing your multiple sticks of RAM...

*edit - just d/l'd the manual

If you are using two sticks, they need to be in matching colors (right next to each other)
either DIMM 1&2, OR DIMM 3&4

i.e. both blues or both pinks...
 
Last edited:
Yes, it could cause a problem if the RAM comes in contact with the heatsink, especially if your RAM has heatspreaders, so its not advisable.

It sounds like your RAM needs more voltage. Did you raise the vDIMM?
 
my RAM does have heatspreaders, I got the OCZ reaper ones. There isn't a way to get to my BIOS but now that I can actually start my computer I will bump up the voltage, I have not changed a thing in the bios. And I was trying in both pink ones since I was not able to get to the blue ones.
 
Spark_Plug said:
WellI got that new power supply today. Hooked it all up, thought It was working great. 40min into OCCT, freezes up on me again. I had to hold in the power button to get the computer to go off, then when I tried to start it up again, fans would spin-up, stop, then spin-up and would go in a cycle like that, it wouldn't post, no beeps or anything. I decided to take out a piece of memory, and then it booted back up. Swapped that memory for the one I took out and it booted back up, put both pieces of memory in and it wouldn't. So I think it has to do with where the memory is seated. Because of the orientation of the heatsink, I can only put memory into two of the 4 slots. If i try to put memory in the other slot, it does make contact with the heatsink. Would this cause this sort of problem? Or is it a problem with the motherboard?

Oh and the heatsink is an Thermalright Ultra 120 Extreme.
Click to expand...

To my knowledge, dual channel memory must be seated in different colored slots on the motherboard. If you're only seating them in the far two slots that are the same color then that might be causing the problem.

*EDIT* According to indokyne, what I said is wrong. But he downloaded the manual so I would go with his explanation. Plus I'm too slow anyway. ;)
 
what drivers you using?

did you test each of your video cards to see if the issue was absent ? or just remove one card and it worked? i would test each alone as well, since you are trying to narrow to the exact cause, should be easy enough to swap to the other card,..retest
 
Im using Catalyst 9.7, and I am pretty sure it is a memory problem, since I can boot and run programs at 100% load fine when I am at 2gb, but when I stick the other stick in it freezes up or just doesn't boot at all.
 
+1 on ram voltage...

i was looking on Newegg at RAM and started reading peoples "con's" or complaints.

whenever something bad was posted, the manufacturer would respond and point out that sometimes the default voltage that the BIOS assigns for RAM is inadequate and that you should make sure that the RAM voltage is up to spec w/ your sticks.
 
Well I did the voltage thing, didn't help at all. still wouldn't POSt when I put the second stick in. I switched out the heatsink for the one that came with the processor, everything botted up fine but temps were extremly high. They were like 45 idle and as soon as the processor got up to load, temps were at like 80 so I shut it down right away. I have no idea why temps are getting so high.

I am thinking it is a problem with the motherboard now. I can only get it to post if the memory is in specific spots, which is the last blue and the last pink, any other combination does not work. Also when I had the stock cooler on, I got a little farther in the OCCT test before temps were extremly high, but then it just froze up on me again.
 
Last edited:
You must log in or register to reply here.
Back
Top