EP45-UD3P powers up twice

**codeslinger** · 02-02-2009

Just built a new system [Q6600, 8G] and am having a weird problem. When I power on, the mobo phase lights all light up, some of the fans turn on and then everything turns off for a couple seconds. Then everything starts again successfully. All devices are running, memory and stress tests are fine. This also happens with everything unplugged except the mobo and the CPU so not an overload but could potentially be a bad PSU though runs fine in the end. I read in one forum where this was normal for another brand mobo w/ a P35 chipset. This mobo has a P45 and I can't imagine this is normal. RAM is at proper 1.9V, CPU is overclocked at 8*400=3.2 rather than stock 9*266=2.4 and 1.36Volts. Ideas? Thanks!

P.S. I am concerned about the PSU some as I somehow fried my UPS by shutting down and unplugging the PSU but forgetting to turn the power supply switch off as when I plugged the cord back in, the UPS went into overload a couple times till now that is ALL it will do and it has gone bad so am currently not having safe computing at the moment till I can decide if I need to order a new PSU along with a new UPS.

**SpeedEuphoria** · 02-02-2009

When you change settings in bios, the board powers down then restarts which is normal for these boards. This is not what your describing, just for reference.

Now if you change settings and most likely when OC'ing. Power on, then it starts, turns off and starts again. This is normally when some setting is not correct or stable.

If you go back into bios on the restart under the MIT section there is normally a Red box that pops up and says that your overclocked settings didnt allow it to boot correctly or similar.

If you dont go into bios it usually sets your cpu speed back to default and keeps booting. So can you verify in windows that your OC'd settings are still applied??? If they are infact going back to default then you need to tweak your setings b/c they are not stable

**codeslinger** · 02-03-2009

SpeedEuphoria,

You were right and saved me some $ for thinking my PSU was going bad. Thanks!!!

I did not find any warning in MIT and I was being overclocked just fine when checking with various hardware monitors. So then I reloading the optimized settings and reset all my desired options except the overclocking and the machine powered up just fine and not twice. I checked this several times doing a complete powerdown each time. Then I used ET6 to set overclocking to level 1 for my Q6600 which set FSB=400, Mult=8 and Vcore=1.4625 and it still powered on normal. Weird part is that in the BIOS, it shows the FSB and Mult as I said but Vcore is set to 1.3125/auto and not 1.4625. Also weird is that earlier I had set just the FSB and Mult as above but forgot to bump the Vcore and could not boot until I had bumped the Vcore to 1.3625 and when I checked with hardware monitors they still showed the Vcore set to 1.3125 and not what I had set. Seems between the BIOS, ET6, the hardware monitors and my understanding there is some inconsistency or my memory is failing so I am going to play with this some more. In any event, thanks, as this in some shape or form is the problem! Dave

**SpeedEuphoria** · 02-03-2009

Voltage set in bios is always higher than voltage seen in windows with software, its known as Vdrop and Vdroop.

Basically voltage in windows will be lower(vdrop), then if you stress the cpu the voltage goes even lower(Vdroop), like running OCCT or Prime95.

I highly recommend that you do NOT us the gigabyte built in overclocking settings, they are know to over volt things much more than needed or safe.

You can take note of what changes when you try that OC, but then change it manually in bios to lower some voltages that are not needed. It will take some testing to find the sweet spot settings but your cpu /board will thank you for the lower voltages in the long run.

**codeslinger** · 02-03-2009

Thanks for that info. I have found 2 things weird. 1) After ET6 set level1 OC the bios still showed an auto setting for Vcore though showed 4.625 in the HW monitors. And a normal power on worked on the F6 bios. 2) Not being able to leave well enough alone, I upgraded the bios to F7 and now the problem is back and I can't seem to get rid of it this time. I even loaded fail safe values after flashing after optimized values did not work, set most of my desired bios options to make sure that power on worked OK, which it did, then set Mult=8, FSB=400 and Vcore=4.625 manually and now have the double power on problem again. I have gotten the bios popup screen a couple times that complained about potentially bad OC values but don't always get it. I continued with the boot and used ET6 to set the OC to normal, shutdown, no problem powering on, use ET6 to OC to level1 (3.2GHz) and shutdown and get the power on problem. I have flipped back and forth a few times, always shutting down to a power off state and always getting the double power on with the ET6 level1 settings but not with the normal settings. Maybe I need a new ET6 version to match the new F7 bios?

**codeslinger** · 02-03-2009

OK after another couple rounds with ET6, and a final setting to level1 OC [8/400/1.4625]and shutting down and letting sit for a while, the machine powered on normally and has done so 3 times in a row after a complete power off shutdown. ET6 Tuner shows Vcore 1.3125 power on and 1.4625 target and 1.408 actual. Bios shows 1.3125/auto and HWmonitor show 1.41 as well. I have to admit I'm totally confused as to why ET6 would have a setting of 1.4625, the bios not show it at all and then get an actual Vcore of 1.4 in the end which I guess is maybe the Vdrop? But why does the Bios show 1.3125 and auto for Vcore when ET6 set it to 1.4625? Sorry, am a newb to this OC'ing stuff but just want a nice stable 3.2GHz out of my Q6600. Thanks for all your help!