Motherboard problems with 1st PCIe slot

dblthunder

Honorable
May 6, 2012
6
0
10,510
Hello, for the last 6 weeks I cannot seem to get a motherboard to boot to the windows desktop with video card in 1st PCIe slot, the one next to CPU. I have 2 Corsair 1200 watt PSU's less than a month old, 1 Coolmaster 750 watt, 2 CPU's i7 950 and i7 960, 12 gigs of G Skill 1600 ram, ran memtest for 6+ hours, no errors, several passes. I have tried various video cards that are new and work in other systems. The boards for the most part will boot to the desktop in slots 2 and 3 but hang on the splash screen if plugged into the 1st slot. My new Big Bang II X79 powered right up with same power supply and video card along with the same new hard drive with fresh install of Win 7 pro OEM yesterday. The boards have all been tested outside the case with 1 stick of ram, one video card and CPU cooler. The 8 pin cpu power supply cable plugged into the MB and 2, 6 pin cables plugged into video card. All my systems run off the same 2 APC 1500 watt battery back ups, which I can see the voltage being provided to the systems. This has happened on 2 brand new and one refurbished MB that I have bought in the last 4 weeks. If anyone can offer any suggestions, I would greatly appreciate the help. As I had stated earlier, I had my new X79 sitting by the way side trying to get my X58 working again, but decided to get it started and I had no troubles with it. I realize that it is a differant cpu and ram, but same psu and video card.


When working, below is my system configuration. At present, I am using one of the motherboards with only 1 video card and 2 monitors. Unfortunately my work requires the use of 3 monitors so this is a real problem for me. The system below is running with one card in the 3rd PCIe slot, if I add another card to the 1st slot so I can use 3 monitors, the computer will not boot to the desktop. This has been the problem with all 3 of the last motherboards I have recieved in the last 4 weeks. When I am testing these MB's it's with a new Western Digital harddrive with a fresh install of Win 7 pro 64 bit and my i7 950 cpu. I am using the lastest drivers for the video cards from nVidia and the board I am running right now has the lastest bios version from EVGA.


CASE: Cool Master HAF 932
MB: EVGA 141-GT-E770 Classified 3
CPU: Intel i7-960
MEM: 12GB G. Skill Ripjaws Series (3 x 4GB) 1600MHz
CPU COOLING: Prolimatech Megahalems Rev.B (Push Pull)
INTERNAL Cooling: Air
STORAGE/OS: HDD 2 x Seagate Barracuda 160GB in Raid 0
STORAGE/DATA: HDD 3 x Seagate Constellation ES 500GB
VIDEO: 2 x EVGA GeForce GTX560 Ti (Fermi) 2GB
DISPLAYS: 3x Acer G245H
DVDE/CD DRIVES: 2 x ASUS DRW-24B1ST
POWER SUPPLY: Corsair Pro. Series AX1200
OPERATING SYSTEM: Win Pro 64-bit


Thanks in advance JG
 

dblthunder

Honorable
May 6, 2012
6
0
10,510
Hoping to find someone that has more experience in these matters than I do, through the research I have done on the matter, I have tried every option that was suggested, bought new components to rule out possible psu, cpu and hardrive along with fresh install of OS. I have not skimped on any of the components in my build to inusre a steaddy and stable computer for my business.
Whether I run the board from my raid array with Win 7 or my new hard drive with fresh install of Win 7, I get the same results, both will load to windows desktop with one card as long as it is not in the first PCIe slot.
If anyone can offer some insight into this problem, or suggestions up and beyond what I have been doing to pin point the problem, I would be very thankful.
Thanks
JG
 

redeye

Distinguished
Apr 29, 2005
225
0
18,710
I had a motherboard that had that problem except it was stuck at 8x... ultimately it did not work.. since my MB was under warrenty, It was replaced.
sorry, you are SOL, and with need to get other MB... file a warrenty claim... i believe that MB is less than 2 years old...
 

dblthunder

Honorable
May 6, 2012
6
0
10,510
Thanks for the reply, MB's are still under warranty, just trying to get to the bottom or root cause of the problem. I live in the US, but not in the 48 so everytime I ship a motherboard it cost me $70.00 one way and I have done this a few times.

Note: When testing these boards, I first set the the ram timings, 9 9 9 24 and voltage 1.65 and ide to ahci prior to testing.

I am just trying to make sure that it is not something on my end, thus the post for information. I have exhausted all my rescources, knowlege wise, and have spent a small fortune shipping motherboards and video cards back to the manufacture then back to me.

While I was waiting to RMA the first MB I dediced to buy a second brand new board, which when it arrived it would not post with video card in the 1st slot. At that point tech support told me it was the video card, a new 560Ti which I then bought 2 more new ones and neither of them would post in the 1st slot and was told I had a bad run of video cards and to return then which I did. They found no problems with the 2 I sent back and the 2 new ones they sent me would not post either while in the 1st slot.

I got tired of the finger pointing at other compnents in my system so I bought another Corsair 1200, making 2 new ones I own, bought a i7 960 chip and still no go on either MB that I have here. I just bought another brand new MB, number 2 in the last 3 weeks, identical from another store only to find out that it was not new, opened package, dust all over the northbridge and other parts of the board, plus the board had a half inch bow in it, but none the less, tried 2 different psu, working video card that I know works and same thing, no post in 1st slot, but will post in 2 and 3.

So the bottom line question, am I doing something wrong? Had no issues with my MSI Big Bang II a couple of days ago, same PSU and video card, or am I just getting bad motherboards?

Thanks
JG
 
Lets look at 'common elements' - HSF, GPU and UPS. Clearly, a bad new or not PSU can rip apart a 'good' MOBO, CPU, etc in no time flat - but you said (2) things: you could run Memtest and boot though be it from not the first (primary) GPU PCIe slot.

So looking at the GPU - it's easy to test the GPU individually so I am going to assume you did and both are confirmed okay. Now if you have "SLI' enabled either without a 'bridge' or a 'faulty' bridge that can cause a problem. So check that out. Also, a bad driver can also pooch a boot so try booting into 'Safe Mode' in Windows.

Looking at UPS, some UPS 'sine waves' have been known to cause issues, but I also have (2) APC Back-UPS XS 1500 and none of those issues.

Next HSF, the Megahalems Rev.B use a metal backplate; a fundamental problem with metal is that it's conductive which can lead to different forms of shorting including a grounding short. Every build I do I make certain there's absolutely no metal<=>MOBO contact and so to insure this I use thin plastic washers. Example:
Screw -> Metal Backplate -> Plastic Washer -> || MOBO || <- Plastic Washer <- Standoff ... etc

Therefore, IMO try the Stock HSF to rule-out a grounding short.
 

dblthunder

Honorable
May 6, 2012
6
0
10,510
Thanks for taking the time to look at my post, very much appreciated.

The gpu's are both fine and work as long as they are not in slot 1

I guess the thing to do is pull the computer apart and add the nylon washers behind the cooler back plate.

If the back plate was to say short out, it could affect the 1st PCIe slot and not the others? And if the answer is yes, then does that mean I fried the first PCIe slot, or just need to isolate the back plate from the MB and all should be fine? But still confused because until yesterday I did use the pop on Intel cooler that came with the i7 950. When my second new board arrived I did use the Megahalems and just reinstalled everything and stuck it in the case.

I understand your point about metal to metal and just went to the hardware store and bought a couple of dozen nylon washers in various sizes for the next test and install, which I have another board arriving Friday. I will make sure the back plate is isolated from the MB and will use them on the standoffs when I install my X79 board.

Question how do you keep them in place on the standoffs when installing the board in the case? Do you glue them to the back of the motherboard?

Question, if the cpu will fire up and load windows with the card in slot 2 or 3, does this mean the cpu is fine, or could it be semi fried and effect the first PCIe slot?

You are welcome to read the rest of this, but as I was typing, I realized I should at least see if the back plate is shorting to the board.

When I first started testing these boards out of the case, I did use the Intel cooler, but did not like the way that it attached to the MB, had trouble getting all four pins seated and was afraid I was going to break the MB, so the last test I did was with a new Megahalems Rev C. I know the screws need to be turned a certain way on the Intel cooler, but for the life of me was not able to get one secured with all four pins.

The systems / boards will boot to safe mode with card in the first slot, but not to desktop in regular mode. I went through a lot of headaches with the new NVidia drivers but now have working clean installs on both, the raid array and the new hard drive, driver. 296.10 The new hard drive is what I used to get my MSI up and running and the one I have been testing the boards with since my raid array is in the case. So the OS which all that is on the drive will boot to windows as long as the card is not in the first slot

I have also tried adding the 4 pin Molex cable to the MB to get more power to the PCI e slots, but no help. On the last board I tested, I used 2 different PSU’s and when I check system health, both had stable readings of 12.10 and 12.60 with no fluctuation. They volts were holding steady.

I do not use SLI and the option to enable it is not even in my NVidia control panel. I checked just to see if something was set wrong. All the motherboards I have tested in the last 4 weeks have been done with a single video card. Even when my system is running I do not use SLI, I need 3 monitors to do CAD and spreadsheets simultaneously and so on.

I truly appreciate you taking your time to help me with this. I am at my wits end, I do not want to slam the vendor who makes the MB, I just want to make sure it is not something on my end that I am over looking or doing wrong.

Again thanks
JG
 
If it is a grounding short then typically correcting the short also corrects the MOBO problems. However, if the problem is something else than the plastic (nylon) washers is still a good idea.

As a 'test' I would also disconnect all but the Power button from the case, i.e. temporarily disconnect all Front Panel, USB, etc from the case. Heck I've seen a bad mouse cause a boot/post failure -- so when I have 'oddball' stuff going on I examine ALL common elements. The chance of a bad MOBO is 2%~3%, so having multiple MOBO issues it has got to be something else.


If the stock HSF also produces a failure then it's probably not a grounding short. To kill a MOBO requires in most cases a bad PSU, but otherwise folks always want to 'blame' the MOBO. Again, keep in mind MOBO failure rate is 2%~3%.

Many HSFs can be installed on the LGA 2011 using only (4) standoffs with no e.g. backplates.

On really high-end rigs I'll use a washer between the standoff and the MOBO ans when I do I use a dab of silicone between the standoff (silicone) plastic washer to secure it. In most instances a 'real standoff' isn't a problem IF properly aligned -- also there's no need to screw it in any more the snug + 1/4 to 1/2 turn -- the MOBO isn't going anywhere and the screws just need to be tightened enough so they aren't loose and won't fall out.

It's rarely the CPU especially IF another GPU PCIe slots 'works', and the handful of times times I've run across this problem:
1. Bad MOBO
2. Older MOBO's with an oddball PCIe slot card that required
3. Bad PSU PCIe connectors -- try a different pair of 6-pins
4. MOBO PCIe dip switch disabling a 'PCIe Slot' ; diagnostic PCIe switch, apparently your MOBO does have 'PCI-E Disable Jumpers' - see http://www.evga.com/products/pdf/141-GT-E770.pdf Therefore this would be my next place to look!
5. BIOS INIT the wrong PCIe slot, so Load Defults.

The stock HSF is fairly straight forward and very difficult to screw-up.

Strongly suggests a Driver Problem -- boot back into Safe Mode and Uninstall ALL GPU drivers, and if you ever installed an AMD/ATI GPU then that can be especially troublesome. The boot into Windows Normally, if the 1st GPU is fixed then download the driver prior 296.10 and use it until the next nVidia driver is available which should be soon; I've had issues with 296.10 'driver crashes' with Chrome + Flash.

The extra power shouldn't hurt, but it's generally only needed with 3 or 4 GPUs.

The option 'should' only appear if the Bridge is connected, and if the Bridge is connected then you can toggle on or off the 'SLI' option.


My next two moves would be to (1) check the 'PCI-E Disable Jumpers' and make damn sure ALL PCIe slots are on, (2) Boot in Safe Mode and uninstall the nVidia drivers. Then please update and good luck!
 

dblthunder

Honorable
May 6, 2012
6
0
10,510


Well not yet, I am supposed to have a new board here today, but looks like it will not make it till tomorrow. If I get it tomorrow, I will try to get it up and running by Sunday, will deffinately let you know.

Will deffinately be checking everything you wrote to insure this to work....

Thanks for checking....

JG
 

dblthunder

Honorable
May 6, 2012
6
0
10,510
Well almost 7 weeks later, I finally have a working motherboard / computer with 3 monitors. The new one I ordered which was number 3 in the last month that I got today will finally boot to windows with a video card in the first PCIe slot. Thank you for your help!!! The only two things different I did was, 1st- I did install the nylon washer on the back side of the board before I put the back plate on, deffinately really isolates the back plate from the board. Good tip! The 2nd thing I did when I tested the board was use my new Clasdified GTX 580 video card instead of the 560 ti, the 580 had never been taken out of the box. Plugged it all in and bam right to windows desktop. Even though it had the correct drivers insalled, I deleted them all, used driver sweep in Safe Mode then loaded them back up. No problems what so ever.

So then I pulled the card and used the 560 Ti, would boot to windows, but would not recoginize the drivers, so I had to unistall them again and reload them and all was well.

Final test, two 560 Ti's (Not in SLI) board would not recoginize 2nd card video drivers, had uninstall / reinstall them again and was able to get 2 cards and 2 monitors working.

So after I saw the 2 cards working, I pulled the old board out of the case, which is only a month old, and installed this board. Once everything was installed with the OS now running on the Raid array, I once again had to unistall / reinstall the video drivers, after a couple of boots now have everything back to normal. Even the placement of all my desktop icons went back to where they were located on the 3 differant monitors which I was surprised after doing a clean install on the nvidia drivers.

The other 2 new boards and the one RMA refurbished board were all tested outside the case and with the original Intel cooler until I got the one last week which I did use the Megahalems cooler to test. Not really sure what was causing the problem, everything I did today was the same as every other board I have recieved, diffenant psu's, cables, and so on accept the washers and starting with a differant video card.

When I have sometime I am going to put the nylon washers on the back of my X79 board, let them dry a day or two before I install that board in the case, that will have to wait, big mess in office after the last month and a half that I need to get cleaned up... lol

Anyways, I am fully aware of time restraints and full plates, truly Thanks for taking the time to look at my post and offer all the help...

Many thanks JG