Printable View
Wow, this ds3 really seems to be on last legs. I orthos from last night at 2am until just now at 8x450=3600 no problem. I take setfsb and try to bump to 460. Locks up. Reboot, try to drop ram timings to 4-4-4-12. Windows locked up at logon screen. Fails to post, and needs to sit for 10 minutes off. Drop back to 5-5-5-15. Still running into issues with it being stable. Any sort of load in windows sends it into a crash. Voltages don't dip below working values. Wonder how long gigabyte's rma is
Haha, and you are going to rma the board because of that :clap:Quote:
Originally Posted by ziddey
Nono, well yes, but there's more. Even at stock speeds, I can use it for hours on end and be flawless, but then all of a sudden, it'd go berserk on me and lock up and fail to post for a while. Nothing has been changed except for motherboard (ds3 rev2 c2 changed into a ds3 rev1 c1). It was perfectly stable with the first board. I'm thinking if it's not consistently stable at stock, then something's up eh? And it also does the silly turnoff-on on startup and reboot even at 266fsb. Kind of weird
edit: I might need to retest the psu, even though I already did a month ago and it checked out fine. Really strange since the board works miraculously at 8x450 albeit 5-5-5 for hours on end but then eventually goes into the :banana::banana::banana::banana:ter and won't come back to life for a while. For testing purposes, I've focused a lot of airflow on the board itself, but it didn't help. So I'm not sure what the time is doing to help it, but it doesn't seem to be temperature
Update: aaagh. I've been letting it sit for a while and it came back at stock speeds. It froze at logon screen at stock. Then it went in and hung at prime. Third time it's priming now at stock. Really weird.
Well I guess a lot of the C1 boards don't clock very well. No matter what I can't get anything stable over 425mhz fsb. Although my memory limitations should limit me to below 500mhz fsb, I've reached 455mhz FSB 910mhz ram (running 4 sticks too) on my P5W DH. The board is VERY picky about ram voltages and timings. Would be nice if I could RMA it for a C2 board, but oh well, it's not worth it. Better off waiting til Q3 when bearlake comes out and I have the money!
Hutch, what have you tried running for timings? I've found with my c1 board that a lot of times it refuses to post until I remove a stick (though I'm only using 2 sticks). Also, even though I know my ram is good for 500 4-4-4, I'm only able to run 5-5-5 at say 450 even. So I dropped down voltage since there's no sense feeding the 2.3v for higher speed. My board is being very tempermental with higher fsb, but after manipulating some more, I've got it temporarily stable at 450fsb 1:1 5-5-5 again. Don't want to bother pushing anymore since I don't think it could possibly end well. Also, I'm back to two sticks. Hope this helps.
Edit: One more thing that might help. Not sure if you have access to a dmm (looks like you would since vmch mod in progress), but you could try measuring vdimm and see what it actually is. On my c1 board, it's actually undervolted by about 0.1v. Or, based on the two boards I've used, vcore2 in speedfan is pretty accurate vdimm (might be 0.02v too low though compared to actual).
Yea it won't boot into windows if I use more than 2.0v on the ram, Have tried 4 sticks, 3 sticks, 2 sticks and even 1 stick. The ram isn't high clocking ram (Elipda chips, 4-4-4-12 @ 667 is spd settings.) I've gotten it to run 4-4-3-4 880mhz @ 2.25v on the P5W DH but over 900mhz required 5-5-4-15 @ 2.25v. I'll check out vr-zone as they probably have the vdimm measuring point..
Ahh, elpida. I hear some of that doesn't like higher volts? Then again 2.2v isn't really considered high though. I think it was more referring to 2.5+.
As for measuring point, it's really simple on the rev1 boards. If you look to the left of the dimms (assuming you tilt your head 90' ccw such that you're properly reading the silkscreen on the motherboard), you'll see an inductor (round and upright with wireloop around). Stick negative to chassis or whatever ground you prefer, and red to the end of the inductor (I used the side facing the dimm). I got readings about 0.02v higher than what vcore2 reports, and 0.1v less than what I set.
When I had my rev2, I didn't look to figure out how to do vdimm checkpoint (really different components and no conventional looking inductor). But the vcore2 value was consistently about 0.1v higher than what I set, so I just assumed it to be more or less true. Much better board than my rev1, but I recall the higher vdimms being instable for whatever reason. Not sure if I hadn't done enough fiddling at the time or if it's the case, but in the end, I've found that I never needed to go higher than 2.3v (+0.4v for that board) to achieve results I was looking for.
sorry u r having all those problems , i do have the same board u have C1 and god i love this DS3 is running at 500fsb 24/7 no problem VERY stable, had a 680I b4 and sold it to go back to my DS3, i was doing a bit better on the memory department with the 580i but only 50 mhz faster than my ds3, im runnng 1000mhz now, any way , im getting tomorrow a bad axe 2 cause i always wanted to have one of those , lets see if i can beat my ds3 , so far ive done 533fsb maybe i can do more with the bad axe 2Quote:
Originally Posted by Hutch
f10 bios was terrible for me. after about 8 different trys with various settings it showed no higher memory speeds and no better timings. the option 1 or 2 for the memory dll thing did nothing. neither option booted at 1000 speed no matter how loose the timings were. anyone else with micron d9 having any luck? wait till official release i guess, don't waste your time AGAIN with this bios!!!!!!!!!! this was on my rev. 1, C2, B2 rig that's been doing 450x8 rock solid.
Ajfina, that's very impressive. Is that with "normal" vmch or what? how about vfsb?
is your corsair micron d9? i can't even sniff those fsb speeds cause of my mem.Quote:
Originally Posted by Ajfina
all voltage to max here ,except for cpu vcore and memory is +4Quote:
Originally Posted by ziddey
mines are 64ooc4 , i think they are micron or is it only the 6400c4d the ones d9?Quote:
Originally Posted by burn what?
Measuring vdimm when set to +.2 = 2.02v actual voltage... gmch set to +.1 actual voltage reading is 1.337v (WOW 1337 MAN!) Lol
Is there a problem with running an SATAII drive on the external sata ports (purple ones) and having SATA ports running in IDE mode in bios? It's weird that the overclocks dont fail until it starts to load windows and it gives me a c:/windows/system/ cannot find a certain dll message...
Wow Hutch, I got that cannot find error yesterday too with the rev1 c1 (never saw it with rev2). Scared the :banana::banana::banana::banana: out of me since I'm used to high pci corrupting hdd's in ide day. Never narrowed that down to something, although I was thinking possibly vfsb or vpcie.
As for running off the purple jmicron connectors, I don't really see anything wrong with that. However, I'm not sure if it'd be slower when accessing optical media, since I'm not sure how the jmicron operates with respect to the ide controller and those two raid ports.
Yea the first few times it happened to me it scared the **** outta me... But it just about always happens when I try to boot over 430mhz FSB and more than +.2 on the vdimm...
Something interesting was just posted in the biostar thread that might possibly hold true for us too. One guy was setting cpu temp warning of 65'c and even though readings didn't show him coming close to it, it was causing bad behavior when overclocking. Once he removed it, it was good to go. That's kind of interesting because I never enabled that option (albeit 70'c) on the ds3 until I got my rev1 board. That's the one and only thing I did different. Perhaps it's worth a shot. Not sure if it'd apply to you though.Quote:
Originally Posted by Hutch
Things here are weird, it's freaky man! :eek:
So i did that orthos run, 400x8=3.2GHz with 1.375V in bios..this was the result (attahced image), F8 bios used for that.
So i then flashed to F9, this time i tried flashing using a USB flash drive...and it worked! :banana:
The flashing process was even faster, y'all know how slow it is to read from a Floppy drive, using USB flash was super fast!...and the flash was sucessful, this is a really good addition to the bios and QFlash.
My USB problems i think are related to my RangeMax USB dongle, i think it's fooked...it's like touch and go with it, sometimes it works other time it won't get detected by Windows or lockup or won't make an internet connection.
F9 is dated January 11th 2007, but the Award bios date still reads "2006" when you enter bios. It seems like my hardware is going kaput...i've now asked OCZ for a second RMA for the 850W Gamextreme...the rails are very bad.
Under 100 Orthos load i get 12.61V on the 12V rails and 4.89V on the 5V, on idle it's 12.65V and 4.88V...thats surely is out of ATX spec :confused:
It's a bloody noisy PSU, sensitive ears be advised, strange frequencies, inductor/coil/electrical noise from within the PSU...none of which was present on the 700W model :mad:
I'm still in diagnostic mode right now.... :fact:
Coretemp showing 3600 eh? haha
CoreTemp also shows my 425x8 as being 478.13x8...Orthos is also wrong and assumes a 9x multi, thus 3600MHz.
Another screenie for y'all. My bog standard setting (for now) 425x8 with 1.4Vcore set in bios. As you can see Vdroop is no better with F9, it still reads 1.344V whereas revion 1.0 PCB was 1.360V.
(G)MCH and FSB set to normal, DDR2 voltage set to 2.0V, PCI-E and others set to normal. My Orthos run is still going...but i'm upto this stable. I am gonna try for 487x7 once i fell things are stable, then try some multi options and memory at 1000MHz.
Can i just say that the new Arctic Silver method of AS5 application works perfect, done the routine four times now and each time the IHS is 100% covered, a thin line down the center of the IHS.
Got my new case/artic cooler freezer pro 7 today, running stable 1.86ghz @3.2ghz stock voltage on everything, only thing i had to change was FSB and memory at 1:1 :D
Temps are only ..50c under load! vs 60c at 2.8ghz before in old case, stock HSF. :D
Orthos locked up after 5 hours, so it appears i need more Vcore. My USB problems continue...it's also effecting my Logitech USB keyboard/mouse...could it be a USB problem related to the Intel INF Software?
You know how there is a box to turn USB devices off to save power etc.
I'm starting to think it's either hardware related to the mobo or some driver issue.
http://www.ijib.com/~ziddey/rig.html :) finished 24hr.
Next time I'll try to see how far I can tune down voltage. I'm currently running at what my other e6400 needs. Maybe this ES can do it with less
What's your mouse/ usb issue? I used to have to goto device manager and disable power save for the mouse. Otherwise, every minute of inactivity will cause it to lag for a sec while it starts up again. Haven't seen that happen lately though, and never with the gigabyte board.
I *think* it's bios related, i don't remember any issues with F7..basically my attached USB devices stop working, no data being moved and many times my Neagear Rangemax USB 2.0 dongle has a yellow warning symbol in Device Manager saying "error code 10 - this device cannot start", the only way to get it working is to unplug the device from the USB post and it starts working again.
It's effecting my wirless USB dongle more then my Logitech MX3100 USB keyboard/mouse.
They are all USB 2.0, has anyone else experienced USB issues with these recent bios releases?