1 5 6 7 8 9 Previous Next 133 Replies Latest reply on May 7, 2012 4:01 AM by Flying_Kiwi Go to original post
      • 105. Re: DG45ID Irregular fan speed


        I already posted a lot about this subject here, and can summarize about my C states experience.

        - I have two DG45ID systems with E8400 and two different PWS (Corsair VX450, Fortron green 400)

        - I got the noise problem on the first one, which disappeared with bios113.

        - I got the fan pump problem with the second one, purchased later when bios 0131 was available, I immediately flashed it with bios0131, and disabling C state solved the fan problem.


        The problems followed the motherboards, whatever the PWS.


        About power comsumption, these systems, with two disk drives, and no graphics card, no other card, and 4GB PC6400 the figures at AC mains plug  ares:

        - something around 60W when just doing "nothing" at windows 7 x64 desktop.

        - something around 10W when either in S3 standby or in shutdown.

        - about 110W at full load using OCCT program.


        C states enabled or not, these figures are the same, at least no difference is visible at the wattmeter. I think that "C states"  interest is mainly for notebooks.

        I kept C1E enabled.


        I also clearly verified that C states disabling removed the fan pumping problem, whatever the power supply.


        Inductor noise and fan pumping were not at exactly the same level of sensitivity with my to boards, and also for the boards of other people discussing here.


        My opinion about the origin of these problem is:

        - Inductor noise:

        Mechanical noise at some frequency of the cpu power converter, the solution is to try to avoid  this frequency by avoiding some (too low?) values in current draw.

        This frequency may be different with different boards, some inductors are maybe better and produce no noise.

        - Fan pumping:

        Electrical noise at the sensor chip level, which cause false board level temperature reading, that cause a  response by the  QST regulation system increasing fan speed to compensate (false) temperature change . Look at my previous posts, there is a reference to a program to observe this.


        Disabling C states, may reduce too deep changes in power converter draw, reducing the general electrical noise at board level affecting the temperature sensor readings, and maybe avoid the converter frequency that causes inductors noise.


        About the bios changes, I suspect bios 113 to have disabling C states without option, then bios 0129 introduced the option (without a note in the releases notes) but was replaced within days by 0131 with the releases notes mention about the option.

        Unfortunately C states were enabled by default with bios 0131.


        Except if some Intel people confirm, there is no chance to know the ultimate reality of the origin of both problems.


        In any case the problems are solved by C states disabling, with no visible consequences on system power consumption.

        • 106. Re: DG45ID Irregular fan speed

          Thanks a lot, parsec and yf38.  Your answers give me some reassurance that it should be OK to run my system with C-states disabled, without too much of a consequence for power consumption.

          • 107. Re: DG45ID Irregular fan speed

            The C-state discussions were appreciated and helpful.  I would expect the inductor problem to normally be associated with high currents such as when the cpu fan is running high, possibly contributing to electromagnetic interference.  Buzz complaints in monitor reviews may be premature since reducing the brightness to normal levels should remove the problem.  On a new monitor there can be a sweet spot to control the buzz.  Two of the caps in my monitor could only be replaced by cutting off the white compound that secured 2 adjacent inductors.  Until I see a problem such as screen blanking, that may have been fixed by reflowing the solders, I believe I don't need to replace the paste.  I can't seem to find a source for this but the last power supply I looked at had a ton of it cradling caps and inductors.  This along with a good fan made it quiet.  It may have had a thermal purpose as well (heat loss by convection rather than conduction) but the outlet air seemed too warm at idle.

            • 108. Re: DG45ID Irregular fan speed

              parsec wrote:


              Curious592 wrote:


              I seem to remember inductor issues with some cpu fans but the SSD is most likely the reason you need to disable C-states.


              Oooh-Kay, and why would a SSD be the reason?  Would you care to explain that?

              I see where you're coming from parsec. I've no SSDs here and I still had this problem. The proper solution for me as far as shutting up the squeal was to have the board replaced under warranty with a newer revision. Similarly, the proper solution for those experiencing this fan speed problem (and the associated shutdown problems) is for Intel to release an updated BIOS which works properly with the CPUs involved. There's a reason why Intel asked me all about my CPU when I first reported the problem to them. Any attempt to measure the power consumption difference with a system afflicted with this problem will be flawed because how can we be sure the fault allows the CPU to actually enter the low power state it should - of course there won't be a power consumption difference if it's not working as it should.


              Although disabling CPU C-states is an effective workaround, it should be remembered by all viewing this thread that it is just that, a workaround and not a solution. The only way we'll get a solution is if enough people raise this issue (expecting more than just a workaround)  with Intel and those responsible at Intel, listen and act.

              • 109. Re: DG45ID Irregular fan speed

                Thanks Kiwi, and well put sir!  It certainly is a workaround and not a   fix, but something you mentioned makes me wonder.  You said you had   your squealing board replaced with a newer revision.  Good for you, but   that makes me think that it's faulty components on the noisy boards  that  are the culprits, not the BIOS.  I'd say you agree.  OTOH, please  confirm for us that the fan surging or sleep/shutdown problems remained  with your new DG45ID.  That is not clear to me in your post.


                The philosophy or stance corporations assume when  dealing with  defective products can be strange IMO.  Those may include;  repair or  replace products only when the customer complains.  Post a  voluntary  recall notice on the products info page.  Ignore the  situation, it's no  big deal, and/or blame (rationalize) it on "stupid  user" syndrome.   Provide a customer/non-factory fix for the product, if  possible.  The  first and third options are the cheapest for the  corporation in the  short term, the second and fourth more expensive in  the short term.  Of  course there is a possible overriding motivation  above all else, to  never admit responsibility or guilt for a products  problem, to avoid  legal liability at all costs.


                But why don't these people get it, they can simply ask   themselves, will you purchase another similar product from a company   that either ignores a fault or is not proactive in repairing it?  Will I   purchase another Intel mother board after this experience?  Extremely   unlikely at this point.  Have I purchased mother boards since this  one?   Yes.  Are any Intel?  No.  So apparently a few bucks in hand now  is  worth far more than the potential for more in the future.


                I don't enjoy criticizing Intel, my family and I are  Intel shareholders, I want them to do well.  In my experience, complaining in forums serves little purpose besides venting, employees  are not  scanning them looking for issues and offering assistance, beyond  the  advice of the moderators.  Is there even one moderator response in  this  thread?  Official ones (an Intel employee) are few and far  between.   Another company's forum I visit attempts to recruit moderators with no remuneration beyond the prestige of being one.


                So you were able to get a replacement 'board based  upon which issues,  inductor noise, fan surging, shutdown/sleep bugs, or  what if I may  ask?  I think I have a three year warranty, I just may use  it.  Do you  have anything specific in mind when you state "... if  enough people  raise this issue..." in order to have Intel listen and  act?

                • 110. Re: DG45ID Irregular fan speed

                  parsec wrote:


                  Thanks Kiwi, and well put sir!  It certainly is a workaround and not a   fix, but something you mentioned makes me wonder.  You said you had   your squealing board replaced with a newer revision.  Good for you, but   that makes me think that it's faulty components on the noisy boards  that  are the culprits, not the BIOS.  I'd say you agree.  OTOH, please  confirm for us that the fan surging or sleep/shutdown problems remained  with your new DG45ID.  That is not clear to me in your post.......





                  So you were able to get a replacement 'board based  upon which issues,  inductor noise, fan surging, shutdown/sleep bugs, or  what if I may  ask?  I think I have a three year warranty, I just may use  it.  Do you  have anything specific in mind when you state "... if  enough people  raise this issue..." in order to have Intel listen and  act?

                  I obtained the replacement board because the so-called BIOS fix didn't fully cure the squealing problem for me (but I'm satisfied that a combination of the newer board revision and latest BIOS is quiet enough as far as scaring away any dogs in the neighborhood is concerned).


                  I don't know for a fact what the difference between the newer and older revisions is but it's quite feasible that the newer boards chokes have better varnish impregnation to help reduce inductive audio noise. I suspect that the BIOS update was only a workaround to try and avoid the situation where the chokes are being powered in such a way that the audio noise is generated e.g. juggling voltages and frequencies to try and eliminate any harmonic resonance (if that was actually what was going on). I'm always weary of BIOS 'fixes' to physical problems and there's no denying audio noise from components is a physical problem.


                  What I mean by that last sentence is exactly that. My most recent rev 310 board does still have the fans locking into high speed problem and associated problems shutting down the PC when this occurs. I've tried raising this with Intel but as my replacement board is no longer under guarantee (a long saga I won't go into here), the Intel techs won't accept my issue as a case for escallation (not even as a 'heads up' to fix this for other users similarly affected). Any users with boards which are still under warranty and which preferably have retail CPUs with original cooling fans however should report this to Intel tech support (preferably lots of people will do this) as the squeaky wheel (and noisey fans) get the BIOS developers time.


                  Crucially, if the workaround of disabling any options in the BIOS is suggested, the users should ask for and expect from Intel, a proper long term solution to the problem in the form of a BIOS which works correctly with their hardware giving full functionality and full power savings when 'Load optimal defaults' is used to configure the BIOS. I'd be interested to hear back what Intel have to say about this if you do decide to report this matter to Intel tech support parsec. I really don't like having a go at Intel either (and I don't even have any Intel stocks ) but they need to take responsibility and properly fix this matter. Given the response I received, I suspect they will not be interested unless you can prove full entitlement to warranty support so that means having your retail CPU and CPU cooling fan details etc available - of course the retail CPU fan will be fully installed at the time too, wont it (although Intel Desktop Utilities won't register its speed when it's locked into high speed mode)

                  • 111. Re: DG45ID Irregular fan speed

                    Flying Kiwi,

                    The cold boot problems I had turned into sleep/hibernation problems due to progressive monitor failure.  Other issues such as USB key/memory retention may not exist in that context.  The signal simply doesn't return within the timeout period.  This problem could clearly affect numerous users where the monitor could be failing at any time.  I won't be holding my breath waiting for a sampling algorithm that circumvents this problem.  The monitor manufactures may not do this either since LED monitors are likely to last longer.  Compatibility with Apple and IBM systems is another roadblock.

                    If you still have not installed your OS followed by INF, audio, graphics (and network?), and then physically inserted the TV tuner card I can only suggest that you visit the Microsoft Answers site or the tuner site (assuming that fixes the problem).  If you can provide registry conflicts that prevent proper installation of your card in other ways I am sure they would appreciate it.  The issue with my VGA to DVI adapter installation sounds like a similar driver problem, even w/o a tuner.  I could probably hook up other monitors via VGA to VGA-DVI connector but there is no motivation.

                    I use the same validation testing logic employed in most complex cases, if less extensive and refined.  My conclusions do differ regarding installation but my system works best for this DIY.  Most of the posters here are worse than extensive Intel and Microsoft support pages.  Many struggle balancing viability and adoption issues.  Few seem to remember that a concise Help system exists in Windows.

                    • 112. Re: DG45ID Irregular fan speed

                      Curious592 wrote:


                      Most of the posters here are worse than extensive Intel and Microsoft support pages.

                      Thanks for the earlier suggestions re removing TV cards etc but as I mentioned at the time earlier, that was something I tried as a long shot (with absolutely no difference to the problem, I might add). The fact that my motherboard was subsequently empty of other hardware means it's fair and square an Intel problem though, no if's or buts.


                      Re your comment above, as long as those seeking help here realise that things like the number of posts made by the person, their screen name or even how long they've been posting here is not necessarily an indication that what's posted is correct (or relevant to any problem at hand) they can't go wrong.


                      Those reading this long thread will have plenty of ideas for what they should do to get this issue properly sorted, also what the can do in the interim to work around it.

                      • 113. Re: DG45ID Irregular fan speed

                        It may still be to early to say but I've just updated my BIOS to rev 135 (late Feb release date) and so far, so good. Even with Load Optimum Default options and both C States and C1 states enabled, things seem to working well (with no sign of the locking into high fan speed mode or faulty shutdowns etc). The readme for the BIOS lists CPU compatibility was an issue that they've worked on so this is a VERY good sign.


                        How are all of you other folks with this issue finding the latest BIOS revision?

                        • 114. Re: DG45ID Irregular fan speed

                          Really Kiwi, C-States enabled and no fan-freakout?  Great!  I installed the new BIOS and noticed IDU ran better, but did not change the C-State settings.  Then of course I was sidetracked by other things, so I haven't had that PC on for days.  I forgot what CPU you use with your DG45ID, I use a Q8300.  At one point I thought for EIST to function, you needed to have C1E enabled, but if C-State is disabled, then no EIST?  I know EIST works for me, I see the CPU frequency and multiplier drop at idle and change back and forth.  I still don't see the need for deep C-States in a desktop PC, as there always seems to be some minor activity going on, so I believe it will never enter those states.  Plus I use sleep mode, not hibernation.


                          I'll keep you posted.  Happy to see a BIOS update on this "old" board.

                          • 115. Re: DG45ID Irregular fan speed

                            parsec wrote:


                            Really Kiwi, C-States enabled and no fan-freakout?

                            Alas, I spoke too soon parsec and it was to good to be true. Although it was fine left running overnight, I left it on all today while I was away at work and when I came back the fans were at 'higher' (although not full) speed constantly and the problem where it wouldn't power off (O/S Shuts down but fans keep running) had returned.


                            Although it makes less fan noise in this state than earlier BIOSes, it still enters the state so it looks like it's back to disabling C States with my E8500 CPU powered DG45ID


                            Your mileage may differ with other CPUs as it's clear they've been doing something to try and address this issue - they just haven't quite got there with the E8500!

                            • 116. Re: DG45ID Irregular fan speed

                              Kiwi, Oh well, that's a shame, I'm glad you still feel up to trying whatever you can.  Ya know, your E8500 is one click away from the top Core 2 Duo CPU, does that make it exotic or temperamental?


                              I have an E8400 in a Gigabyte P45 mother board, which has been such a low maintenance PC for me, even with Vista on it.  I would just love to understand what causes that nonsense, it's just so crazy.  I guess I'll try my DG45ID/Q8300 combo with the new BIOS with C-State enabled just for "fun", although I wonder if this BIOS update had anything to do with this issue.

                              • 117. Re: DG45ID Irregular fan speed

                                parsec wrote:


                                Ya know, your E8500 is one click away from the top Core 2 Duo CPU, does that make it exotic or temperamental?

                                No but it was considerably cheaper than the E8600 - about £70 difference if my memory serves me correctly . I wasn't too concerned about it being temperamental because of its speed. I did have a cutting edge Pentium Pro 200 (loaded up with a whopping 64 MB of SIMM memory) when they first came out and that was very reliable so I assumed this should be fine. The thing that has dissappointed me is that this was my first Intel branded motherboard and despite the fact it has exactly the features/specs I wanted, there have been quite a few bugs/issues with it - more so than other brands I've owned in the past! This issue and the IDT Audio Driver Dolby Digital stability problems are 2 outstanding niggles - both of which I've had to work around (in the case of the latter by using the Windows 7 SP1 audio driver which completely clears that up but doesn't offer the same bells and whistles).


                                As you mentioned, this BIOS release is for an 'elderly' board so I'll just get used to the idea of having C-States disabled rather than Intel fixing it


                                Good luck with your testing.

                                • 118. Re: DG45ID Irregular fan speed


                                  I upgraded my two DG45iD from 131 bios to 135 bios and the fan speed issue remains the same.

                                  My 310 rev. card which had fans speed problem with 131 bios when C states enabled (and no issue when disabled) has exactly the same behaviour.

                                  My 308 rev. card had no issue and still have no issue with 135 bios C states enabled or not.

                                  For your information I upgraded using F7 key at startup with ID0135.BIO file on a USB key. It tooks about 7 minutes, the 3 first ones with no message on the display, the LED key indicating that file was (slowly) reading, then I saw the usual messages showing the process of flashing, followed by a restart.

                                  My processors are both E8400 (complete story available at page 1 of the thread).

                                  As usual I was capable to record the trace of the issue using SIW freeware (hardware sensors menu) I already explained in this thread.

                                  DG45iD will continue their life with C States disabled.

                                  I now also own a DH67GD with a core i5 2500 and it is a nice follower for those with a  DG45iD.

                                  • 119. Re: DG45ID Irregular fan speed

                                    You're right about the price difference between the E8600 and E8700, Intel lists the "1k Unit Price" as $183 and $266 respectively.  I really don't see this as a CPU compatibility problem, I see it with my Q8300, and check this thread for all the CPUs used with this board that experience this.  IMO, it's just some weird side affect thing with C-State.  I can say that with the new BIOS, my CPU fan does not run wide open during POST anymore, without C-State enabled of course, but I haven't tried that yet.


                                    I just added 4GB of memory to my DG45ID, a different brand but same basic speed, and it worked fine without the slightest burp, at latencies better than spec'd for the new DIMMs. (For all the love of Corsair memory, they only had two SPD profiles populated, without an EPP profile, my Patriot memory has four.)


                                    Kiwi, forgive me but why are you so in love with C-State?    I put my PC's in sleep mode when not in use, but I guess that can be considered minor abuse of components going in and out of that state, like shutting down and restarting a few times each day.  Sure if the PC is running at idle for long periods, you're saving CPU power, and everything else is idling but still using energy.  Just different philosophies, to each their own.


                                    I just saw that all the E8000 CPUs are being discontinued, and that the P43 and P45 MCHs are being discontinued.  But then a Pentium CPU on the 775 was introduced in late 2010, so who knows when this board will be end of life?

                                    1 5 6 7 8 9 Previous Next