Double Post when overclocked

First post in a Forum.... Been fixin and buildin my own for years. Here goes...
When ever I change any settings from stock to OC and reboot or cold start my computer starts for about 3-4 seconds then reboots and most of the time posts on up and runs normal. Every now and then it will reboot a 3rd time but then goes back to stock settings and continues posting. I never get a bsod (bios F2), all power saving options disabled in bios, PCI-E at 100, memory timings loose. Once it posts it runs perfectly. I tried all kinds of different memory timings, goosed the cpu vid to 1.3, raised mch by .1 and fsb volts by .1 in an effort to find out why this posting phenomenon occurs. No matter what I have tried, when I OC this posting regimen occurs. I play Supreme Commander Forged alliance alot and no issues the game just rocks with all the goodies at max and 60fps with fraps. Any smoking grey matter out there with a clue?

Have you tested the OC by the usual means?

Run memtest, Prime95, Intel Burn Test etc.

I usually don't run the "stress tests" due to time involved. I have run memtest to completion with no errors. However the problem I am experiencing is only occuring when I do a cold start. I have everything on power surge protection and turn off everything each night and then switch off the surge protectors isolating my system from power completely. In the morning or first start I power on the surge strip wait a few seconds and then hit the power button. It starts up normal, meaning the fans and lights and hardrives start up but just before bios posts (about 3-4 seconds) it reboots like I had hit the reset button manually, and then powers on up at whatever FSB I had selected. I have gone up to 3.2ghz where I am now. I have never BSOD'd in Vista 64 or XP. This seems to be isolated to the mb post up routine. Every now and then it will reboot a 3rd time and if that happens my bios resets itself to stock. If I hit delete and then go back and change it all back to my oc parameters and reboot its fine????? Something in the bios doesn't like what it sees on cold startups. I have my setup go to sleep after 30 minutes, and when I come back and hit the power button it comes to life without any rebooting. I did rma one stick of memory a month ago but the problem still remained with the new stick. Does removing the psu from power have anything to do with this? It never happens unless I have oc'd the bios. Perplexing?

Only thing I can think of is what you said, and it might be something to do with removing power totally from the computer when you turn the surge strip off.

If you turn your PC off for the night, but leave the surge strip on, do you have the same problem?

It sounds kind of similar to what happens to mine occasionaly. If I turn the PSU off while doing some inside work etc and then start the comptuer cold, my system powers up for a few seconds, at pre-POST it 'turns off' again and then restarts and continues to boot normally. It only ever happens after the MB has been without power for any length of time. I too am overclocked and can't remember if this occured at stock speeds.

I also have it connected through a surge protector, but I don't switch that off.

As to what causes it, I don't know. I just put it down to being a feature, not a bug

Do you have a Gigabyte motherboard by any chance?

Yep, its a GA-X48-DS4 with F2 bios. Is this a Gigabyte anomaly? I have had no other problems but the occasional first post reboot and an extremely rare 3rd boot that resets bios to stock. I have absolutley no problems once I boot into Vista64, or XPpro or Windows 7.:huh: I play Supreme Commander, and Sins of Solar Empire maxed out with no problems.

Mines an asus and it does the same-ish thing, so I doubt it's just a gigabyte issue.

It is Gigabyte thing yes , mine did it when i first got my cpu to 3.4 gig , it jsut went away on its own?? lots of gigabyte owner/overclockers seem to suffer from it , it doesnt seem to do any harm

smarteyeball said:

Mines an ASUS and it does the same-ish thing, so I doubt it's just a gigabyte issue.

Click to expand...

...it's not just a Gigabyte thing

Thanks everyone for the info. Just like some of you, my board suddenly stopped doing the double post on every start and I haven't changed the way I power off as I stated in my earlier posts. I just flashed to the new f3c bios on my GA-X48-DS4 (rev 1.3) due to the new intel ahci rom bios update from the stock 1.07 to the new 1.20 and now my post is much quicker on the ahci rom end of things as the intel rom used to take an additional 5-7 seconds to post.:D

Good to hear you sorted it.

I think I might have found a link to the double post phenomena. I just bought a MS sidewinder mouse and then a couple of weeks later a MS 4000 keyboard (yesterday). I haven't got a double post since... maybe the ps2 ports had something to do with it as now I am usb entirely. I took my old MS natural multimedia and moved it to my other computer with a p35 gigabyte and e6600 on it. It had usb keyboard (mouse on ps2 w/adapter) prior to this and never doubleposted on oc to 2.7ghz, until I booted up with the ps2 keyboard port in use. I got a double post immediately and then reset bios to stock. It wouldn't take any effort to oc after that and just double posts and resets. It must be something on the isa bus that powers the ps2 keyboard port. The keyboard functions fine with intelletype6.3. That system runs xp64 btw. I would like to know if you other guys with this phenomena are running ps2 or usb keyboards?

USB keyboard here. But it also occurred with only a ps/2 plugged in or only a USB bluetooth dongle. :huh:

I stumbled across another possible explanation:

when the system is switched back on what remains of that power in the capacitors try's to power the system on, but can't because of the "break" in the flow of power, hence why the system shuts off, and moments later powers up again.

Click to expand...

It makes as much sense as anything else. However it's a bit worrying that even though you think you've discharged power before removing/installing components, that there might actually still be a charge in the mobo