Announcement

Collapse
No announcement yet.

Asus P5E3 Deluxe and F3-10666CL7Q-8GBXH

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Asus P5E3 Deluxe and F3-10666CL7Q-8GBXH

    Good morning.

    I have the Asus P5E3 Deluxe wifiap@n and recently I bought the G.SKILL F3-10666CL7Q-8GBXH 8GB (4X2GB) kit, but I have some trouble make it run stable.

    I am running Prime95 with default stock CPU settings and I have set the memories manually at 7-7-7-21-2T, with Dram voltage at 1.5v. Every time I was running prime95 for about 1-2 hours, I usually had one core error "FATAL ERROR: Rounding was 0.5, expected less than 0.4".

    I tried auto settings and that was running at 9-9-9-24 (if I remember well), XMP settings that it was running at 7-7-7-21-2T, manual settings at the same as XMP etc, but I was always got this error. Sometimes prime95 was crashing at the first 30 minutes and the program closes. Some other times it was running for 8 hours but one core was giving the error after 90 minutes etc.

    I have run also memtest for several hours without detected any problems.

    Finally I ran MemSet and I found that in SPD board, it was showing that at XMP settings, it should have the tRFC at 107, but as CPU-Z shows, this setting was at 100.
    When I tried to change this at BIOS, it had not 107 value, but instead from 100 the next one was at 110 and that is the maximum I can set, so I have set it now at 110.
    Now I am running prime95 for almost 2 hours, without errors yet and this is the first time that everything seems running well.

    Could this tRFC setting make my system unstable finally?

    What is your suggestion about the NB and DRAM voltages, should I change them or leaving them at auto? What is the best settings for make my system stable 24/7?


    Ps: I forgot to mentioned that the problem with prime95 is happening with blend mode test.
    Last edited by arvacon; 06-06-2016, 11:43 PM.
    Asus P5E3 Deluxe wifiap@n
    Q9550 c1
    G.SKILL F3-10666CL7Q-8GBXH 8GB (4X2GB)
    Windows 10 Ultimate

  • #2
    *Update
    After 3hours and 20 minutes, Prime95 crashed again.. I don't see any error at prime95's log file.
    It seems finally that tRFC helped a little, but it was not the solution to my problem.
    Any suggestions?


    This is the log file error from Windows Event Viewer

    Log Name: Application
    Source: Application Error
    Date: 6/7/2016 10:26:04 AM
    Event ID: 1000
    Task Category: (100)
    Level: Error
    Keywords: Classic
    User: N/A
    Computer: DESKTOP-60S0UO3
    Description:
    Faulting application name: prime95.exe, version: 28.9.1.0, time stamp: 0x56faaa15
    Faulting module name: prime95.exe, version: 28.9.1.0, time stamp: 0x56faaa15
    Exception code: 0xc0000005
    Fault offset: 0x00000000004d04ef
    Faulting process id: 0x1468
    Faulting application start time: 0x01d1c067e756460b
    Faulting application path: F:\My setup files\Diagnostic utils\Prime95 v289.win64\prime95.exe
    Faulting module path: F:\My setup files\Diagnostic utils\Prime95 v289.win64\prime95.exe
    Report Id: 8516dd36-d052-400f-9cd3-2f71bbe7585d
    Faulting package full name:
    Faulting package-relative application ID:
    Event Xml:
    <Event xmlns="http://schemas.microsoft.com/win/2004/08/events/event">
    <System>
    <Provider Name="Application Error" />
    <EventID Qualifiers="0">1000</EventID>
    <Level>2</Level>
    <Task>100</Task>
    <Keywords>0x80000000000000</Keywords>
    <TimeCreated SystemTime="2016-06-07T07:26:04.934424800Z" />
    <EventRecordID>24030</EventRecordID>
    <Channel>Application</Channel>
    <Computer>DESKTOP-60S0UO3</Computer>
    <Security />
    </System>
    <EventData>
    <Data>prime95.exe</Data>
    <Data>28.9.1.0</Data>
    <Data>56faaa15</Data>
    <Data>prime95.exe</Data>
    <Data>28.9.1.0</Data>
    <Data>56faaa15</Data>
    <Data>c0000005</Data>
    <Data>00000000004d04ef</Data>
    <Data>1468</Data>
    <Data>01d1c067e756460b</Data>
    <Data>F:\My setup files\Diagnostic utils\Prime95 v289.win64\prime95.exe</Data>
    <Data>F:\My setup files\Diagnostic utils\Prime95 v289.win64\prime95.exe</Data>
    <Data>8516dd36-d052-400f-9cd3-2f71bbe7585d</Data>
    <Data>
    </Data>
    <Data>
    </Data>
    </EventData>
    </Event>
    Last edited by arvacon; 06-06-2016, 11:49 PM.
    Asus P5E3 Deluxe wifiap@n
    Q9550 c1
    G.SKILL F3-10666CL7Q-8GBXH 8GB (4X2GB)
    Windows 10 Ultimate

    Comment


    • #3
      Originally posted by arvacon View Post
      Could this tRFC setting make my system unstable finally?
      If the tRFC is set lower then specified that might results in memory errors / instability. A value higher then specified does not hurt stability, but will slightly reduce memory performance (but only sub 1% for 107 vs 110).

      Originally posted by arvacon View Post
      Any suggestions?
      If all other timings a set appropriately i would try to raise Vdimm a bit. Try +50mV for a start.
      Team HardwareLUXX | Show off your G.SKILL products!

      Comment


      • #4
        In general use, from yesterday that I have change the tRFC from 100 to 110, I feel that my system seems stable now, but the prime95 crash makes me still wonder if something is wrong. Before I change this, I had also some random "Display driver stopped responding and has recovered" errors, so the screen was becoming black for a moment and then it was coming back. This was not happening with my old memories and it has not happen again from yesterday that I changed tRFC to 110.

        I did a test at last night. I found a cd with utilities that they call it "ultimate boot cd", so I burned an iso of that and I ran prime95 via this boot cd. It is including the previous version of prime, 28.5 instead of 28.9, but it did the trick.
        I booted with this and I left prime blend test run via the kind of "dos" environment for the whole night. After 9.30 hours, I didn't see any errors, so I finally stopped the testing and rebooted to windows. If there was a problem with my ram or cpu, shouldn't be normaly discoverable from this test too?

        I have read that maybe the NB voltage must need to increased if you use 4 dimms of ram, because the northbridge controller is pushed more. How much should be set this voltage for everything to be stable? How much dram voltage can I set, to not push hard my memories for 24/7 usage?

        Unfortunately my motherboard doesn't show northbridge and DRam voltage at HWmonitor or similar programs. Instead it shows some values with names as VIN1,VIN5,VIN6. The only one seems close to northbridge is VIN6, but there is a big offset between bios and hwmonitor. When I had set 1.41 volts at bios, the hwmoniror was showing 1.31 if I remember well. The sure is that this VIN6 sensor was following the changes I was doing at NB voltage.

        I have seen some big offset at Vcore voltage even with LLC enabled. I had 1.20volts at settings and I was getting 1.44 at cpu-z or hwmonitor.
        Is it possible the NB voltage to have the so big offset too?

        Here there are some screenshots that might help.
        Attached Files
        Last edited by arvacon; 06-07-2016, 07:40 PM.
        Asus P5E3 Deluxe wifiap@n
        Q9550 c1
        G.SKILL F3-10666CL7Q-8GBXH 8GB (4X2GB)
        Windows 10 Ultimate

        Comment


        • #5
          Oook, I just had an error from display driver... (see screenshot)

          Is this 7-7-7-21-2T too aggresive for my system at all? Should I use looser timings?
          My settings are like at the photos I have attached at the previous message. I have change only what you see, the rest settings are in auto now, even dram voltage and nb too.

          I need your help to set my memories correctly. Can you provide me some settings that they will work stable?
          Attached Files
          Asus P5E3 Deluxe wifiap@n
          Q9550 c1
          G.SKILL F3-10666CL7Q-8GBXH 8GB (4X2GB)
          Windows 10 Ultimate

          Comment


          • #6
            I set the NB voltage from auto to 1.41v but I had that "display driver not responding" error again.
            I went back and set NB to auto again and then I set the DRAM voltage at 1.56v. For the last 2 hours I had not any display driver error.
            Can I increase the DRAM voltage a little more if the system continue be unstable? How much is the maximum voltage I can use for this RAM kit at all?

            Excuse me for the too many posts and words, but its been almost a week since I started trying to make my system stable and I have started feeling hopeless now..
            Sorry for my English, it's not my main language.
            Asus P5E3 Deluxe wifiap@n
            Q9550 c1
            G.SKILL F3-10666CL7Q-8GBXH 8GB (4X2GB)
            Windows 10 Ultimate

            Comment


            • #7
              Do you have the latest BIOS? Which slots are you using? Try one module at a time to see if each performs as it should

              Comment


              • #8
                Hi.

                Long time to check this thread, as this issue had tired me enough, so I didn't search it anymore. But I have get more tired to watching this happen every single day, so if there is any solution, please help me solve this issue.

                As I see, a lot of people's computers suffers from this error, but as I see, it has nothing to do with the graphics card, it just happens when the system is not stable because of overclocking or wrong memory settings.
                I have test the memories with the memtest and everything seems fine, but when I am at windows, in irrelevant times it gives me this black screen issue for 3-4 seconds and then it shows the message "Display driver stopped responding and has recovered".

                I have not overclocked my pc. I want the memories to just run in default settings without having stability issues. I have tried several settings without success. I tried loose the timings to 8-8-8-24 and 9-9-9-24, I increased Vram to 1.6 by steps of 0.2v, but it didn't help.

                I haven't play with the NB voltage very much, so is there any suggestion about how many volts sould I set the NB voltage, or any other settings that I should try?


                Ps: I have the latest bios and I use all the 4 slots.
                Attached Files
                Last edited by arvacon; 09-21-2016, 07:23 AM. Reason: Ps addition
                Asus P5E3 Deluxe wifiap@n
                Q9550 c1
                G.SKILL F3-10666CL7Q-8GBXH 8GB (4X2GB)
                Windows 10 Ultimate

                Comment


                • #9
                  Can you send pictures of BIOS? Memory timings page as well.

                  For that memory kit you shouldn't need to play with NB Voltage. A test you can try is using one stick at a time and see if they perform the same.

                  Comment


                  • #10
                    Hi, thanks for your reply.

                    I did something different at last Friday. I increased the voltage at the graphics card from 1volt to 1.1volt and the system was stable for 4 days without any issue. But at the 5th day I had the display driver crash once again. From that time, I have set the NB voltage from auto to 1.31volts and I am waiting now to see if this will happen again. The graphics card is the Sapphire hd5750 vapor-x and I did these changes at the bios of the card, via RBE utility.

                    As I have read, this motherboards had issues when you was loading the 4 memory slots, probably a voltage drop at the controller I guess, so as I saw, some people were suggesting to increase the NB voltage and they have mentioned that their system was becoming more stable.

                    But what is confusing me now is, if the increasing of the graphics card voltage helps finally to prevent this error of happening so often, then is this problem related with the pci-e voltage or with the northbridge finally? Or is it just wrong timing memory settings at all?

                    Weird situation..

                    What's the best settings that you suggest me about the memories, from what you see at my bios pictures?
                    Attached Files
                    Asus P5E3 Deluxe wifiap@n
                    Q9550 c1
                    G.SKILL F3-10666CL7Q-8GBXH 8GB (4X2GB)
                    Windows 10 Ultimate

                    Comment


                    • #11
                      Set REF Cycle Time to maximum value

                      What is the production date of the modules?

                      Comment


                      • #12
                        REF Cycle Time is already set to maximum, that it is 110.
                        The production date of memories, is into 2015 if I remember well.

                        It's been 5 days now without problems. Windows 10 also did that huge anniversary update and it seems like they use the half memory now than before and the system runs smoother.
                        In any case, if I get this error again, I will report it.

                        If I don't have any issues for, let's say 10 days in the row, I will try lowering the graphics card voltage again and I will test it for a few days to see if this will happen again. If not, then I will put also the NB voltage to auto and afterwards the memory voltage to auto too.

                        If everything works fine after all, then it means it was windows 10 bug and after the anniversary 1607 update, they solved it. If not, then I have to reverse the changes one by one..

                        Let's see what will happen...
                        Last edited by arvacon; 09-27-2016, 02:43 PM.
                        Asus P5E3 Deluxe wifiap@n
                        Q9550 c1
                        G.SKILL F3-10666CL7Q-8GBXH 8GB (4X2GB)
                        Windows 10 Ultimate

                        Comment


                        • #13
                          It seems finally that windows 10 anniversary update, just changed the way that the system is informing you about the display driver error, so nothing solved finally.. They just hide the error, so now you don't get that annoying notification at the right corner of your screen, but you can see of course the screen to becoming black and coming back again after this happening.. Now you can find the error only at the event viewer, so nothing changed at all.

                          I forgot to mentioned also, that today after I turned on the pc from sleep mode (as I usually put my pc to sleep mode every time I close it), I had a BSOD that it was informing me, that I have a problem about "video TRD failure" (see screenshot).

                          I haven't do any changes at my last settings yet, everything is as I have described it at my earlier posts, so what's next to try now?
                          Attached Files
                          Last edited by arvacon; 09-30-2016, 04:41 PM.
                          Asus P5E3 Deluxe wifiap@n
                          Q9550 c1
                          G.SKILL F3-10666CL7Q-8GBXH 8GB (4X2GB)
                          Windows 10 Ultimate

                          Comment


                          • #14
                            Good evening.

                            I have moved to windows 7, 64 bit system, because I thought I would get rid about the display driver issue. The system overall seems to work much better than with win10, it consumes the half memory now and cpu is almost 10 degrees Celsius less than before.

                            The problem with the display driver error still exists, so finally this has nothing to do with the operating system. But something else that I have mentioned is, that if I open a program that it runs in Java, then this crash will usually happen at random times, even I am at win10 or win7, so that's weird.

                            What I have mentioned during all these tests, is that the only setting that seems to reduce the error from happening, is the increasing of VRAM. When I have this at 1.5v, then this error is happening very often. When I go to 1.6v, then this happens rarely, but I hadn't set it more than 1.6v until some minutes ago that I set it at 1.62v.

                            For the moment I don't have any crash yet, but I would like to know, how far can I go with the ram voltage, to be in a safe zone and not fry them?
                            Last edited by arvacon; 11-27-2016, 08:00 AM.
                            Asus P5E3 Deluxe wifiap@n
                            Q9550 c1
                            G.SKILL F3-10666CL7Q-8GBXH 8GB (4X2GB)
                            Windows 10 Ultimate

                            Comment

                            Working...
                            X