Initial posting probs only..pls help!!

[ziggatron]

Hi, I have recently built a system using the following components..

Q6600 cpu (used)
Asrock G41M-S3 Mobo (new)
PKR 8gig 1333mhz ram (2 x 4gig) (new)
EZ cool case and EZcool 450W PSU (new)
Geforce 8800 GTS grx (used)


The system was fine for a week and then all of a sudden I have a problem where in the morning, when I first turn the system on, the POST fails, starts to post, then cuts out and recycles to try again.

This will go on for about 40times!!!!!! .. maybe 5 mins...

Then it will post, boot into win 8 and then be chipper for the rest of the day.. no problems, posts, boots, no stability issues..?????

Tricky one , because once it has jumped that first fence in the morning, I can no longer test it, because it works fine..

I have, from cold today(while the problem persists) disconnected, one by one, the HDD's, the grx card and the DVD rom...

After each device was unplugged, the system was tested and the problem was still there - would not complete the post..

(Yesterday I ran mem test on the ram for a few hours (once the system had made it to up and running status..) and there were no errors reported..)

having unplugged everything, except RAM and CPU, I swapped the PKR ram for 8 gig of corsair xms3 1600mhz ram from another build and the system posts correctly and boots straight away with no noticeable stability issues anywhere...

Here's the thing...

Having swapped to the 1600mhz RAM for a singe post and boot, system running for maybe 3-5mins.. I close down and replace the original PKR 1333mhz ram an the system posts first time, boots and runs like clockwork...

The jumpers are 2-3, 3-4, 4-5.

I am a bit stuck on this...and I am wondering ..

Is it a temperature thing? (something is failing until it reaches a better temp...)
Is it the jumper settings.? (if the system runs all day, including reboots etc, the jumpers must be correct????)
Is the cpu not seated completely correctly.. (temp using Realtempis approx. 34 deg across four cores.)


This system is built for sale, I can't sell it on the basis you have to wait for 5 mins every morning for it to jump the initial post fence before it performs normally..

I need to get this fixed asap and any advice or ideas would be much appreciated..

Thank you.

 

[cmichael138]

"Is it a temperature thing? (something is failing until it reaches a better temp...)
Is it the jumper settings.? (if the system runs all day, including reboots etc, the jumpers must be correct????)
Is the cpu not seated completely correctly.. (temp using Realtempis approx. 34 deg across four cores.) :

Doesn't sound like it's any of those. Sounds more like a voltage or timing issue with the RAM.

 

[ziggatron]

"Is it a temperature thing? (something is failing until it reaches a better temp...)
Is it the jumper settings.? (if the system runs all day, including reboots etc, the jumpers must be correct????)
Is the cpu not seated completely correctly.. (temp using Realtempis approx. 34 deg across four cores.) :

Doesn't sound like it's any of those. Sounds more like a voltage or timing issue with the RAM.

Hi, thanks for the reply..

When you say voltage issue, do you mean overloading the PSU..

I did think of a voltage issue.. thought maybe the psu was being overworked due to the extra power for the grx card.. I thought by removing the componnts one by one, hence reducing the load on the PSU , that would fix any overload issue, but even ultimately with everything disconnected, still kept failing...

Memory timings are on auto.. unfortunately on the PKR memory there is no indicated voltage or timing info, otherwise I would have set that because I know the system defaults can sometimes create system instability..

Bear in mind that once it does initially post (40 attempts later...) it will post and boot all day long with no system issues what so ever...

That's why it's such a bitch to fix..

 

[Richard3rd]

It could be temperature related. If you test the memory modules individually and it only happens with one of the modules, then maybe there is some micro fracture in this module.

 

[Rockdpm]

I'm just making a supporting statement here. But I do believe good so it's your power supply. Though I wouldn't think your full system would need anymore than 500 watts

 

[ziggatron]

@Richard3rd

thx for the reply. I must admit I didn't try the modules individually when it was happening this morning.. but I did run mem test on the pair for a few hours yesterday with no errors..

With a microfracture would the warmer temp help conductivity allowing the RAM to perform normally.. is that the idea..?

I had thought along those lines, but I was wondering if maybe one of the contact pads on the CPU would be the culprit.. poor connection by a few microns.. until it gets a bit warmer..

then again, equally true for the mobo I suppose...

Real problem is I can't do a thing because the system is smugly 😀 working perfectly and will continue to do so until I leave it over night...

edit: maybe it's scared of the dark, has unsettling nightmares or some thing.. 😱 😱

I will try the RAM individually tomorrow morning based on what you said, see if it throws some light..

@Rockdpm

thx for the reply. I have checked the recommended minimum Watt PSU for this grx card and it is 400W. I am running 1 x 250g, 1 x 80 gig hdd and a cd rom..

I made sure I used a direct feed from the PSU to the two molex connectors needed for my 6 pin grx power adapter...

Also the mobo houses on board graphics and I basically disconnected everything, one by one, doing a retest each time and even with everything disconnected i.e no HDD's, no CD rom, no Grx card, just the mobo, cpu and 8gig RAM...

still could not get the horse to jump the fence..

With all this in mind do you still think it is a PSU problem? What type of PSU problem do you think it is?

its worth a try anyway.. I will swap it out tomorrow morning and see..

thx you both for your assistance..

 

[ziggatron]

Hi, following on from yesterday, as expected the system will not post first thing...

I think I have solved the problem though..

@Richard3rd

looks like you were exactly correct..

Been using the laptop so the desktop wouldn't be used, stay cold and still show the fault...

Initially wouldn't post with both PKR sticks in situ or with one (In retrospect I didn't try both sticks individually a this point..)

Removed PKR sticks and replaced with one corsair XMS3 1600 MHz and the system boots first time.

Then removed the XMS3 and replaced with 1 stick of PKR and it posted straight away..

I put two sticks of PKR in and it wouldn't post again..

Then I made sure I tried both PKR sticks individually and hey presto...

Basically one of the sticks of PKR is faulty and when in situ it will consistently prevent successful post when the machine is cold.

One of the PKR sticks is good and if used on it's own the system posts straight away, consistently from cold.

If you try the faulty PKR stick on its own, even when the system has just posted successfully with the "good" PKR stick, the system, consistently fails to post.

If you use the "good" stick combined with the "bad" the system will not post.

Thank you for your advice. PSU is brand new and passed every test.

As I said in the original post, I had previously run Memtest for a few hours with no errors.. but this was after the PC had jumped the first POST hurdle and was well into operating temperature...

I will need to return this Memory, do you have any advice about returning??

I bought it from ebay, the modules have sold really well 800+ sold, his feed back looks good, so I assumed there were no complaints...

Do you think he should replace the memory with no hassle based on my tests?

Hope so...

 

[cmichael138]

Most reputable sellers on ebay will replace an item or give a refund. A printout of this page should be more than sufficient.

Glad you finally have a solution to your problem

 

[ziggatron]

Hi cmichael138, that is a really good idea mate..

I have sent him another message with the link info...

Nice one.

thx.

 

[Rockdpm]

if you still have issues let us know so we can help. So one is on 24/7 to help