8 Replies Latest reply: Jul 15, 2011 1:09 PM by timmy2 RSS

Intel DQ67SW problem with SSD's and PSU’s

timmy2 Community Member
Currently Being Moderated

http://www.highspeedpc.com/http://www.highspeedpc.com/(THIS IS A “GOOD” ONE!  And so far it’s got Intel, OCZ (SSD supplier) and Corsair (PSU supplier) stumped.)


I’ve discovered a repeatable problem with the Intel DQ67SWB3 motherboard.  The same problem occurs whether I use an i3-2100 or an i7-2600 CPU.

The PSU I started with is a Corsair CX430, which I’ve used quite successfully with numerous Intel DQ57 motherboards.

The hard drive is an OCZ Vertex 2.
I successfully installed W7 Home Premium 64-bit, loaded the AHCI driver appropriately, and then installed from a flash drive all the latest Intel drivers and Windows 7 SP1.  It was running great!

Then I connected the system to my gigabit switch to run Windows Update.  The system went dark within a few seconds and rebooted. I left the LAN cable connected and let it continue booting into Windows, but before the desktop completed loading the system rebooted as before.  I subsequently determined:

  • If I booted into Safe Mode with Networking it would remain stable and not reboot.
  • If I connected the system to a basic Linksys 10/100 router it would be stable.
  • If I disconnected the LAN cable from the gigabit router or disabled the internal NIC it would be stable.


I pulled another Corsair CX430 from inventory, tried it, and got the same rebooting behavior.

 

During a long chat session with Intel support it occurred to me to install Intel’s Desktop Utility that measures fan speed and voltages.  IDU immediately began warning me that the +3.3V supply was too low (3.080V). We decided it must be the NIC circuitry (after trying several different NIC driver versions) and assumed that the Corsair CX4320 PSU must be incompatible with the DQ67SW, or at least the +3.3V supply is marginal.


I then tried a new Corsair TX650W and got the same result.  That was a big surprise.

I changed the power supply to an Antec Earthwatts 750W and the system became rock solid!  I reverted back to a system image I’d created before all the rebooting began, performed extensive configuration and software installation work, and because the system had behaved so flawlessly during all that, I delivered it.  It has remained stable on site since then.

A week passed and today I prepared to build the next DQ67SWB3 system. I’d already tried this motherboard as a substitute while troubleshooting the previous problem and had determined that the rebooting problem occurred with both motherboards.

Since I’d made an image of the previous system when it was stable, I wrote this image to a Seagate 160GB SATA HDD drive just to see if the rebooting problems reoccurred.  I didn’t want to use a new SSD and possibly corrupt it if the new system started rebooting, and I figured it probably would because the only PSUs available to me are the three Corsairs that failed with the previous system.

But the new HDD-based system behaved perfectly. Totally stable.  Intel Desktop Manager showed the +3.3V supply to be steady at 3.40V. That's with the Corsair TX650W PSU.  I swapped to a Corsair CX430 and the system remained stable with +3.3v supply at 3.36V.

I then wrote the W7 system image to a new OCZ Vertex 2 and installed it in place of the Seagate HDD. The system booted fine but once the desktop finished loading Intel Desktop Manager started warning me that the +3.3V supply was too low. The +3.3V supply value was moving up and down between +3.20V and +3.080. So we’re back to the same old problem, except now the culprit -- or at least trigger -- is the presence of an SSD.

I've tried the Corsair TX650W PSU but the system barely makes it to the desktop before shutting down and rebooting.
BTW, the "Fan Status and Real-time Monitoring" function in Intel’s BIOS, regardless of PSU, SSD or HDD, shows a reasonable value for the +3.3V supply, between 3.32V and 3.39V. That's before booting into Windows.

Not having another Antec 850W PSU around I just tried an older Antec Truepower Trio 650W PSU and the system won’t even make it into Windows.

If I boot into Safe Mode with Networking the system is stable – to the extent that I can tell, since there’s no voltage measuring software available in Safe Mode.

When I switch back to using the Seagate HDD all is definitely well and the +3.3V supply is steady at 3.40V.

So, in summary, Intel Desktop Utility shows the 3.3V supply to be stead at 3.40V when I use an HDD, but that drops as low as 3.080V (and fluctuates constantly) when using an SSD.  Still using the SSD, I disabled the internal NIC in BIOS, booted, and then the +3.3V supply hovers closer to 3.150V and never triggers IDU to issue a low-voltage warning. Clearly (to me) the NIC circuits and, to a greater extent, the SSD impact the +3.3V supply, whether the real culprit is the SSD, PSU or the DQ67SWB3 is beyond my expertise.

That’s my story.  Any suggestions?
------
The day after I posed to this forum I had an inspiration that maybe the fluctuating +3.3v supply with the SSD is a grounding problem.  I do all my initial system building on an HSPC Tech Station. This has never caused a problem before but just to see what would happen I connected the Seagate mechanical HDD to the DQ67SW and the OCZ Vertex 2 SSD via SATA data and power cables. (Before this I had been connecting only one of the drives, not both at the same time.)   With the Seagate HDD present there is no voltage fluctuation, regardless of whether I boot from the SSD or the HDD. So something about the presence of the HDD, whether it's the connection to the PSU or to the DQ67SW I don't know, solves the problem of +3.3V fluctuating wildly between 3.08V and 3.2 that occurred when only the SSD was connected. In fact, with the HDD present IDU shows the +3.3V remaining steady at 3.36 or 3.40 depending on which Corsair PSU I use (and they all work fine in this configuration).
.
My next step is to install the DQ67SW into a case, which is metal, and I'll try the OCZ SDD alone and see if the problem arises or the metal case supplies a better ground than is available on the Tech Station.
-----
After installing the DQ67SW, PSU and OCZ SSD in an all-metal Antec case I booted the system and it started crashing immediately. I'm sure the voltage problem was present but I didn't have a chance to run IDU. I then added the Seagate HDD to the system, rebooted, and it stabilized. IDU shows 3.40 V for the +3.3V line.  So I'd say the problem could be that the DQ67SW has trouble with a lone SSD attached, or the OCZ SDD has some floating circuit that creates the instability, or the Corsair PSUs have problems when connected to a system having only an SSD.
.
Is this a problem with the Intel DQ67SW?  I'm not qualified to judge because I don't know the extent to which the DQ67SW affects power to these devices.  It's notable that the system was stable when booted into Safe Mode (though IDU was unavailable for measuring the +3.3V line, so maybe it was lower than normal but not so low as to cause a crash).  Remember that after plugging in a gigabit NIC connection the system would crash, whereas disabling the NIC in BIOS stabilized it (but this, too, was before I began using IDU so, again, maybe the +3.3V line was low but not so low as to cause a crash).  Then there's the fact that when booting into BIOS and checking the power and fan status the +3.3V line looked fine.  It seems like the system become unstable once drivers are all loaded and all devices are operating.  Ultimately, adding an HDD totally stabilizes the +3.3V line regardless of where the fault lies.
.
I'll leave this to Intel, OCZ and Corsair to sort out.
  • 1. Re: Intel DQ67SW problem with SSD's and PSU’s
    AnybodyM Community Member
    Currently Being Moderated

    Wow, thank you for posting such an extensive descriptions of the problems you've encountered!

     

    I feel I'm suffering from the same problems!

     

    My system

     

    - DQ67SW

    - i7-2600k CPU

    - SSD - tried both an Intel 160GB Postville and the OCZ Vertex 3

    - Seasonic S12II-430Bronze PSU (which has always worked perfectly in the old Core i5-661 System)

    - some Ram (which has always worked perfectly in the old Core i5-661 System)

     

    intermittently and completely unexpectedly reboots. But it's not a bluescreen reboot (I switched that off the very first time it happened to be sure thats not the case) and it "feels" more like powering down for a second and then going back on when I catch the system doing this.

     

    It's definitely not cpu-load or heat related, if anything it seems to happen more when the PC is idling then when it's actually in use.

     

    Sometimes it happens twice in an hour, sometimes it takes 6 hours until it happens. But it's definitely not working at all like it's supposed to be and is very very frustrating...

     

    I've switched the SSDs between the Intel and the OCZ, which didn't change anything. I'm also using GBIT Ethernet, have not yet tried if going back down to 100MBit makes a difference. It may very well be that the problems only occur on my system with no HDD connected because during the initial installation I did not have problems - and I also had my Backup HDD connected at the time...

     

    I'm really wondering that to do now, throw the board in the trash and get a DH67BL? Buy and try other PSUs until I find one that accidently works?
    RMA the board ? (I'll probably get another one that's just the same)...

     

    Yours sincerely,

    (a slightly frustrated) anybody (who is nevertheless incredibly grateful for your post here)

     

    PS: My voltages in the IDU utility look fine though. Perhaps they only drop just before the reboot? And those seem to be way more sporadic than in your case!

  • 2. Re: Intel DQ67SW problem with SSD's and PSU’s
    timmy2 Community Member
    Currently Being Moderated

    Your symptoms sound familiar. Connect a traditional HDD to the system and see if the rebooting problem goes away.

  • 3. Re: Intel DQ67SW problem with SSD's and PSU’s
    Currently Being Moderated

    Disable the smart power setting in the bios.

     

    Let me know if you still have a problem after doing that.

  • 4. Re: Intel DQ67SW problem with SSD's and PSU’s
    timmy2 Community Member
    Currently Being Moderated

    Where is Smart Power setting in BIOS?

  • 6. Re: Intel DQ67SW problem with SSD's and PSU’s
    AnybodyM Community Member
    Currently Being Moderated

    @n23mc:

    "this thread" doesn't in any way reveal what option this is supposed to be. I've searched the entire BIOS and was unable to find anything that even sounded similar.

     

    @timmy2:

    "Connect a traditional HDD to the system and see if the rebooting problem goes away."

     

    I have since bought a new power supply to test if swapping the power supply to something completely different (different brand, 2 years newer than the one I was using before) makes any difference: It did not. After not-switching-off for 2 days (on the old Seasonic S12II power supply) the computer did it again this morning. I immediately switched to the new PSU (be quiet! Straight Power E8 500W), and only a short time later it happened again.

     

    PSU did not make a difference in my case.

     

    I then connected a trusty WD Raptor 36GB SATA "traditional HDD" to my system. This HDD was active the entire time, it did not go into sleep mode, and it was in use. And the "switch off => switch on" happened again within quite a short time.

     

    So, basically I did everything I could to eliminate the "problem" and it's still here. I've now loaded the Standard BIOS settings. While I had not changed anything of importance, perhaps some stupid thing like me disabling firewire or serial ports causes the problem? I'm now running the BIOS with 100% default settings except "Numlock OFF". I'll report back if the problem persists (which I would expect).

     

    ---

     

    UPDATE: It has powercycled again. BIOS defaults didn't help. I will try to exchange the Q67 Board on monday. If I won't get a direct replacement for it I'll buy a DH67BL and RMA the Q67.

     

    Since I can't buy a replacement board for the weekend I'll also try if limiting Ethernet to 100MBit helps, but i'm not too optimistic about it.

     

    ---

     

    UPDATE2: 100MBit/s didn't make a difference.

     

    ---

     

    UPDATE3: Several months later it is still unclear what caused my problems. I had then changed motherboard to the DH67BL and just kept the DQ67SW around. Never ever had those reboot/powercycle problems any more. Then, 1-2 months later I had to build a PC for someone... I used the DQ67SW and tested it for some time before giving him the PC - it worked fine. And this PC is running the new location almost 24/7 with the DQ67SW in it and I've never heard any complaints at all. So basically my problems went away after exchanging the motherboard for something else, yet the motherboard in another system works fine too. Yet, with all my troubleshooting and component-exchanging, I was never able to get it to work in MY system (and i really changed EVERYTHING). Nothing really makes any sense here - but the problem is gone without really being solved.

  • 7. Re: Intel DQ67SW problem with SSD's and PSU’s
    Currently Being Moderated

    Might be a long shot, but try double checking that the heatsink/fan assembly is really snapped into the motherboard.  I was seeing random reboots with a i7-2600K that resisted all HW swap attempts, until I realized that not all 4 plastic HSF tabs were completely locked in place behind the motherboard.  Fortunately it seems I avoided any long term damage to the CPU.

     

    -Brian

  • 8. Re: Intel DQ67SW problem with SSD's and PSU’s
    timmy2 Community Member
    Currently Being Moderated

    Here is Intel's reply:

     

    I received an answer from our engineering team. They conclude this issue has to do with a compatibility problem with those particular Corsair power supplies. The OCZ SDD drive might be also conflicting with those power supplies as well but that is totally out of our scope.
    A few of these motherboards with the same revisions and BIOS versions were tested  with Cooler Master, Antec, and Ultra power supplies and Intel Solid State Hard Disk Drives and no problems were found. The Intel® Desktop Utilities behaves fine and the BIOS shows stable power numbers. The NIC is always enabled and running at 1000mbps and no post failures at all.
    The best recommendation we can provide is using the same Antec power supply model you used in one of the machines that worked. 

More Like This

  • Retrieving data ...