Windows 7 64bit is not impressing me right now

G

greyeyezz

Weaksauce
Joined
Jan 26, 2008
Messages
119
This is a brand new build,
Coolermaster CM690
Corsair 650TX 650W
Asus M4A87TD EVO
AMD, Athlon™ II X3 450 Triple-Core 3.2GHz
Corsair 4G XMS3
Saphire Radeon HD5770
Western Digital 500GB Caviar Black
Creative Soundblaster X-fi Titanium

I have programs crashing constantly, opera, explorer, outlook, commodo, sidebar, creative audio, dhcp client, etc. My bios is up to date along with all my drivers. Reliability monitor looks like the stock market.

I have all the same software installed on my old machine running W7 32 bit and its rock solid. I'm about to send this W7 64 disk back to the builder and request W7 32. I will sacrifice a gig of ram just to have a stable machine. This is ridiculous.
 
Last edited:
any hardware issues? Failing hard drive? ran memtest? 64-bit has been rock solid for me.
 
It's not Windows 7. It is your hardware. Something is unstable with your new build. Run memtest first to check memory. Set all bios options to default, don't overclock until you know the base build is stable.
 
Nothing OC. My heatsink is extremely close to one ram module, almost touching it, if that means anything. I'll run memtest.
 
Start with one memory module at a time (swapping them if it still crashes) and the on-board graphics (if present). Set all options to default and work your way out. Many people run W7x64 with no issues, so it has to be a hardware fault somewhere.
 
So then either memory is bad or the motherboard itself is broken and needs to be RMA'ed. I have my CPU fan sitting on top of my ram as well but no issues cause of it.

Trying doing what was mentioned above, put only 1 stick of RAM in and test it, if it freezes, move to another slot and test again with the same RAM. This way you can figure out if it is the motherboard itself or the RAM.

Can you swap the RAM with a known good set of RAM?
 
greyeyezz said:
Memtest ran for 2:42 and locked up.:mad: I set bios to defaults.
Click to expand...

This tells me that you very likely have one or more bad sticks of RAM. Please re-run Memtest with only one stick of RAM, then shut down and run Memtest again with the other stick of RAM. Memtest should be run with only one stick of RAM at a time to isolate the bad stick(s).

That above scenario is likely because I have an i3-2100 system do the exact same thing with two 2GB Patriot DDR3-1333 modules. It turned out that one of the modules ended up locking up Memtest after just a few seconds. Replaced them with two 2GB standard Crucial DDR3-1333 modules.
 
Why when something goes wrong with a built, or bad hardware happens, ppl inmediately blames WIndows??
 
Not sure where you got the RAM but now would be a good time to take it back and get it replaced, as it's most likely the culprit here.
 
AVA direct built this rig so I'll probably just drop it off there. Thats why I had them build it so I wouldn't have to pull what hair I have left out trying to troubleshoot.
 
greyeyezz said:
AVA direct built this rig so I'll probably just drop it off there. Thats why I had them build it so I wouldn't have to pull what hair I have left out trying to troubleshoot.
Click to expand...

Then why make a thread about it? :confused:

Just to blame windows. :rolleyes:
 
Can we officially request that a mod change the title of this tread to "AVA direct build is not impressing me right now"

:D
 
Alright with the windows banter. I reset the bios to defaults and nothing has crashed lately. I did initially try overclocking using the OC auto utility but blue screened so I thought I set everything back, maybe not. I will run it at default settings and see whats up.

Reason I was quick to blame windows as I stated before the same software is installed on my W7 32bit machine and its flawless. This is all brand new hardware and was supposedly tested after assembly.

Appreciate the help kidding aside.
 
Windows 7 x64 seems to be better at discovering defective hardware, as in its less tolerant to defective hardware.

A related example.....years ago intel recalled both the 1.2Ghz PIII and the 3.8Ghz P4 when they could not compile the kernel using 64 bit linux systems.


Mackintire's rule #1

Don't even bother installing windows if it doesn't pass memtest + tests: 3, 4, 5 and 7 with 10 passes (min) each.
 
greyeyezz said:
Reason I was quick to blame windows as I stated before the same software is installed on my W7 32bit machine and its flawless. This is all brand new hardware and was supposedly tested after assembly.
Click to expand...

If can understand that logic if you were dual booting x86 and x64, but it sounds like these are two completely different pieces of hardware.
 
Ran each RAM stick separately Memtest

Mod 1, 20:56 , 1 pass no errors, locked up.
Mod 2, 2:03, locked up

I did have issues with the front USB ports, plugging anything in would freeze windows, had to reset. It was intermittent. I plugged the top ports cord into a different header and that seemed to fix that issue.

AVA tech said to run without my AV software:rolleyes:
 
AVA might be back to their old shit again. Couple years ago there was a huge rant thread about them in another part of this forum. AVA posted and sort of cleared it up. I still trust myself over anyone else for building my own systems though.
 
Running without AV isnt a fix for bad hardware, its a good laugh that they suggested it
 
I'm not that knowledgeable about the hardware aspects, but this doesn't look right, that heatsink is 1/64 away from the stick. That can't cause some sort of short?

P1030375.jpg
 
No, that is fine, and in fact normal. There seems to be quite a bit of incentive to have the RAM slots as close to the CPU socket as possible. As long as it's not actually touching it will be okay.

RAM is probably the most vulnerable part of any build, especially if the RAM has just been shipped, either by itself or as a system. Bad RAM happens quite a lot. Testing RAM is the first thing that should be done, whether it is a new system or an older system that has begun crashing.
 
I had similar issues with an AMD build and it turned out to be the video card (HIS HD5570). It was BSOD after BSOD. Removed the card and used the onboard video and problem solved. Card is currently in RMA.

If both mem sticks fail memtest, it seems to be the motherboard. Do you have another rig you can test the ram in?
 
No my old rig is socket 754.
I never seen windows services crash like this, Network location awareness, Services and controller app. plug n play. ever
 
greyeyezz said:
No my old rig is socket 754.
I never seen windows services crash like this, Network location awareness, Services and controller app. plug n play. ever
Click to expand...

Anything will act strange with hardware errors. You need AVA to just take it back and fix it. If they won't, do a credit card chargeback.

Can some mod rename this? Really?
 
It's either bad ram or bad mobo or incorrect settings for the ram.
What about the RAM voltage?
 
greyeyezz said:
I renamed the title. Thanks for the help.
Click to expand...

You can't, a mod has to.

FYMbyIce said:
It's either bad ram or bad mobo or incorrect settings for the ram.
What about the RAM voltage?
Click to expand...

This also. Check that the voltage is high enough, and the timings are right. I know my older crucial RAM tried to run at too low of a voltage and wrong timings...memtest failed within seconds until I manually set the voltage and timings.
 
send it back thats what you paid for, a running system you don't have a running system. end of story. if you tinker with it they may blame you for the problem.
 
greyeyezz said:
I don't know how to set RAM timings
this is all i got

8-15-20118-28-44PM.png


666

i punt
Click to expand...

That doesn't say what the motherboard is supplying voltage wise...just what the memory manufacturer burned into the chip...which may not even be the spec (it wasn't on my crucial RAM.)

Like Jack says, just send it back in. You paid to have a system without these hassles, so get that.
 
Most of the time when you have 2 bad sticks its a memory setting. Don't worry about timings, just use the auto detect. It is more likely voltage more than likely as others have said.

For some reason when you auto detect memory settings in the BIOS it always ends up at the wrong voltage. Look at the chips or manufacturer/model number and find it specs. Usually there is a sticker on the ram saying which voltage. Alot is 1.5 these days but in the past it could be different. There is sometimes a voltage range. I've seen 1.6 - 1.8 listed on the sticker before. Usually the rule of thumb is if you are not stable, you should increase the voltage. I remember with one of my sticks that said 1.6 to 1.8, I had to set it at 1.75 in order to make it stable, as anything less would fail. That thing I'd look at if you have the time. If it doesn't end up stable with the listed rating, or within the range listed, its faulty.
 
You must log in or register to reply here.
Back
Top