F2-6400PHU2 failure

[coeddlemon]

Motherboard: Asus P5W DH
Memory: F2-6400PHU2-2GBHz
Processor: Core2 Quad Q6600

I built this machine up in October 2007 and it's worked quite well under XP Pro with everything stock and no overclocking. I did have to set the RAM voltage at 2.1V though before I could get several hours error free runs on memtest86+.

A couple of days ago I turned the computer on one afternoon and it wouldn't POST. So after trying to get the system to start up several times I pulled the AC plug off the back and removed the case panel to look things over. I reseated the the video card, checked all the cables, and tried to power up again but still no POST. The MB power light was on so I checked a few of the voltage lines on the power supply (Corsair 620W) and the 12V and 5V lines all appeared normal. I powered down again, reseated each of the two RAM chips and after powering up this time the system completed POST and XP booted normally. So after letting it run for an hour or so I shut it back down.

The next morning there was a repeat performance of the "no POST" situation, ARRRRrrrrg! Well, I knew where to start this time. I powered down and reseated the two RAM chips just as I had before and powered up ... still no POST. So after a couple more tries I pulled the stick at socket DIMM_B1 to try booting on the single stick at DIMM_A1. Again, no POST. I then replaced the stick at A1 with the one I had pulled from B1 and the machine booted.

After seeing the box running on that particular single stick I tried replacing it with the one that wouldn't POST singly the first time and verified that stick still wouldn't let the machine do the POST. So now for a couple of days I've had this machine limping along on the single 1GB stick w/o any other changes.

Any suggestions for anything else I might need to do to verify the problem here? I purchased the RAM NewEgg.

Regards, Charles

[Tradesman]

Sorry to say, it sounds like the one stick has gone south (died). What you could try, would be to clean the contacts, you can buy any of a multitude of electronic cleaning solutions, but plain old rubbing alcohol works fine. Many will suggest using a Q-Tip, but I lean towards a lint free cloth or a foam makeup applicator, Q-Tips can leave unseen cotton fibers that mess things up. If it doesn't work, you'll probably want to take a look at the RMA board or just request a RMA from rma@gskill.com or possibly faster ustech@gskill.com.

[GSKILL TECH]

We can definitely have that issue resolved. Please visit our website at http://www.gskill.com/rma.php for complete RMA procedures. You will need to download and complete the RMA Form then email it to me for an RMA number to return the defective memory sticks. If you have any further questions or issues, please feel free to let me know.

Thank you
GSKILL SUPPORT

[coeddlemon]

Thanks Tradesman! You said," ... What you could try, would be to clean the contacts, ..."

I got back to the problem and thought about your suggestion. I had 91% alcohol and some Q-Tips so I cleaned the contacts on both chips. The suspected "bad" stick booted up by itself! So I stuck in the other stick and ran Memtest86+ for a complete pass with no errors.

Now although this looks great so far I want to wait and see if it boots up tomorrow. I saw this problem go away before while troubleshooting. I'm really hoping there's NOT an intermittent somewhere that's developed.

Thinking back about my having to up the voltage on the RAM to 2.1V: I now remember that was to solve an earlier problem with stalls in running XP setup/repair. I came upon a discussion where others were having the same sort of problem. The VIP's were suggesting this was most likely caused by faulty RAM. After checking the BIOS and RAM specs I realized that I had been running those sticks at only 1.8 volts! I had to raise the voltage to 2.1V before I could get 2 or 3-hour error-free runs completed in memtest86.

Everything is stock here so I think I'll be checking temps in and around the box and to try to see if I need more air flow.

COE

[coeddlemon]

Thanks GSkill Tech:

I'll take you up on the RMA suggestions if this situation presents itself to be a definite RAM problem.

COE

[Tradesman]

Hope it works out OK tomorrow....alot of times when people upgrade their RAM, something (very often a finger or two comes into contact with those connectors and can smudge it, doesn't have to be food or anything, just the natural oils in the skin can do it.

Another possibility would have been to used compressed air in the slots, but with the one working in either/or, that wouldn't have been it....but it's something you can file away..just in case for the future

Have a gud 1!

Tman

[coeddlemon]

Sorry to say were are back to square one! That box wouldn't boot this morning. So I've got to rethink this thing. There could be a lot of reasons so I've just got to try to run through another elimination process. First thing off though I think I'll see if it will boot on a single chip ... the only thing that seemed to work before.

At least I have this little HP Presario/Vista box available with a wireless connection back to the router. Even if XP boots up on a single chip it's too heavily loaded to run on 1GB.

Peace, COE

[Tradesman]

Might see if you could borrow some from a friend/co-worker and start the RMA. From the sounds and look of what you've said though it appears to be a bad stick, especially if it won't boot by itself in either slot, but the other does in both.

Only other thing I can think of off hand would be to run some third party testing i.e. CPU-Z and/or others that will approximate the voltage going to the RAM to ensure the 2.1 that shows in the BIOS is what's getting there, if it's iffy, try the RAM at 2.2, that would be within tolerances.

[coeddlemon]

Thanks for the suggestions.

I don't know of anyone that would have compatible memory to load for testing but I'm thinking of purchasing something to have on hand after all this!

The 'puter has booted up several times today! The only thing I intentionally changed was on the timing: from 4-4-4-4-12-5 to 4-4-4-4-12-5 and the RAM frequency from 400MHz to 800MHz. I made these changes last night after looking at the GSkill RAM package and the ASUS P5W manual.

In looking through the P5W DH manual several places show BIOS settings as [Auto] along with other selectable parameters but with little specific info about the selection. The only specs I have on these sticks are the ones on the box they came in so I don't even know about the an upper limit for RAM voltage above 2.1V. Thus far I'm mostly running into OC'ing info that's posted at various websites where someone is using the P5W MB with these sticks.

Any chance GSkill has a searchable setup archive for these older MB's and sticks somewhere? I've got CPUID CPU-ID installed so I can capture post those settings or others as necessary (digital pix of the BIOS settings) if that or other programs are needed.

I've not tried any performance benchmarks to see if the performance for a stock setup is where it should be and I'm a little skeptable that the POST problem has cleared up.


COE

[coeddlemon]

Oops! "The only thing I intentionally changed was on the timing: from 4-4-4-4-12-5 to 4-4-4-4-12-5 and the RAM frequency from 400MHz to 800MHz."

Should have been: The only thing I intentionally changed was on the timing: from 4-4-4-12-5 to 4-4-4-12-4 and the RAM frequency from 400MHz to 800MHz.

COE

[Tradesman]

Hmm, interesting, will it boot up w/ only the 'bad' stick?

[coeddlemon]

The faulty stick seems to test defective again after days of both sticks working together in this box. I was about to suggest we ask for this thread to be decommissioned until something showed up again. And today the problem chip seems locked in with a problem ... where have I heard that before!

I ran with the sticks posting and booting every day several times with no evidence of a problem. The problem though seems to have popped back in now with the stick that was marked defective always failing to post and the other stick always posting and booting ... I'm working with that single stick now as I reply.

I'll go ahead and try to get the RMA request form now unless you think otherwise? Do you know if GSkill has any loaner sticks (for a fee maybe)? Else I'll have to purchase something to get by unless they have a fast turnaround.

COE

[Tradesman]

Believe they turn over pretty quick, you can request the RMA number through their site - just e-mail to

rma@gskill.com or for a faster turnover send it to ustech@gskill.com - they probably won't get to it this weekend (they go Mon-Fri), but you'll get the paperwork - believe the ustech address goes to the guys that work these boards, so you might reference your post here.

Next, might check with friends/co workers to see if anyone could let you borrow a stick. Where abouts are you, i.e. citywise, might know someone in the vicinity..if you don't want to post it, can PM me

[coeddlemon]

This box was so loaded down running XP from a single 1GB stick I couldn't even get to the RMA download site!

I was able to borrow a pair of 1GB sticks from the grandkids HP Presario/Vista machine. Had forgotten it was running on an ASUS MB and had Kingston DDR2 CL5 RAM. This XP box booting right away on the borrowed sticks seems to verify the test machine works well on a good set of RAM sticks. I was able to email the RMA request form off a few minutes ago.

Thanks for your help and suggestions.

Regards, COE

[Tradesman]

Take a look at your PMs, dropped you a note
Tman