Jump to content

FuriousGeorge

Members
  • Content Count

    46
  • Joined

  • Last visited

About FuriousGeorge

  • Rank
    New Member
  1. no im not gonna lie... let me put it this way: it suddenly works using orange dimms (2 + 4) with any ram i throw in there earlier i had been using dimm 1 and 3, ironically 2 and 4 were the two i didnt try till just now. its somwhat disconcerting the way the fan shuts off when my system is idle. i'm assuming this is some default ACPI behavior supported by my linux boot cd.
  2. i was having the same issue, but i found that using DIMM 2 + 4 solved the problem (before i was using 1 + 3). i think you have already tried this, but hopefully it will help someone else. good luck with the new HW, im interested to see how it goes.
  3. UncleDavid, I had tried everything you guys suggested, short of buying new parts. The one thing I had misunderstood was that you guys wanted me to use the ORANGE slot furthest from my CPU. I just assumed you guys were referring to DIMM1 so thats what I did till this time I noticed DIMM1 and 3 were in fact yellow, and i should use 2 and 4.I mean, i guess i would have tried it eventually, but that's not exactly intuitive. Now the thing is posting with pc2100 rosewill or any other excuse for sdram i put in there, value ram included. *Shrug* something about that bother's me. i'm not saying you arent correct, but I thought we had specifications for x86 hardware for a reason. bottom line is it works now, so i'm gonna leave well enough alone, if anyoe asks me i'm saying one of the bays is bad (i've had that happen before on a lanparty nfii ultra). I'm sure you'll disagree, but we dont have to agree on everything, do we?
  4. i forget what the brand my BrandName was. Its intended as a joke. Anyway, its not the rig in question. Praz, i've tried several different types of RAM, and I always memtest memory anyway, so I know its good (at least for other systems) UncleDavid, I dispute your .5% figure, and I have a feeling that an RMA is going to solve both my and this guy'shttp://www.dfi-street.com/forum/showthread.php?t=68318 problem. I appreciate the suggestions, but does anyone have an idea that doesnt involve assuming every stick of RAM and every PSU i've tried will work on other DFI boards, but not this one. Maybe I will mess with the memory timings and see what happens.
  5. hi, thanks for quick reply. i am building it outside the case. ive already tried other memory and no change. the PSU is a 550Watt Rosewill. The beeps go like this. Turn on MB, 1 second beep when 3 LEDs lit. 4 second pause, 1/2 second beep. 4 second pause, 1 second beep, 4 second pause, 1 second beep, repeat lat beep/pause forever. lately its not POSTing at all if I dont regularly reset the BIOS. I think this MB is a dud.
  6. I just got this NF4-D im building for someone, and wouldnt you know it there are issues. Most of the times the board doesnt post at all and beeps intermittently with 3 LEDs, less ofthen it will hang at the splash screen, and rarely it will boot into my OS. When it hangs at the 3 LEDs, it beeps for about 1 second ever 5-7 seconds. If I clear the bios, it will boot right up, and pause at the screen where it informs me that my BIOS settings have been reset to default F1 to continue or F2 to enter BIOS. until I hit either F1 or save my new bios settings it will post reliably, and if I allow it, it will boot into the OS. As soon as I complete the boot however, whatever BIOS settings are being loaded cause the same problem. I have 2 X 512 of Kingston ValueRam but I have tried swapping other ram. I am running everything right now outside the case, with no other components besides GPU,CPU,MB,and MEM. anyone know whats going on here? and if anyone is working at DFI in Cali,USA, either pick up 510-274-8000 or get an answering machine.
  7. Gothic Case: Excellent questions. I don't know the answer, because I haven't had windows for a few years.
  8. Everything works fine, but I have this piece of hardware that likes to have its own IRQ. I put it on Slot 4 b/c I see in the BIOS that it doesn't share an IRQ with the onboard devices. The OS is overriding the settings, and I'm not sure how or why. I have ACPI support built into my kernel, and everything else works perfectly.
  9. i tried loseining my timings to no avail. normally i eun at 2.5 3 3 3 7. tried 2.5 5 5 10, and 2.5 4 4 8.. at those settings i tried raising my fsb to 255 and the computer locked up around 34% of memtest then i tried lowering the fsb to 240 and raising the multi to 10.5 and same result. 265 would be nice but it looks like this combo of cpu, mb, and memory isnt going much farther than 250X10 incidentally i can tweak my t(ras) in either direction and still pass memtest, but i dont understand the effect onmemory bandwidth.
  10. im already at 250 x 10 i just figured im not maximizing my processors potential since my temps are so low, and i wanted to see if perhaps there was some way to bump that speed up. refer to the above posts.
  11. the default cas for the ram is 3, i dropped it to 2.5 as i heard the nfii ultra b doesnt like 3. i havent tried losening my timings, ill give that a shot now. this is all at default memory and chipset voltages. if i raise the fsb to, say, 253, even raising my dram voltage from 2.7 to 3.1, and my chipset voltage from 1.6 to 1.9 wont make it prime/memtest stable, and even if it did im not confortable with raising my volages that much. if i dont raise my voltages, it has trouble posting at all. what happens is prime95 fails after a fem minutes at most, and memtest starts to fail around 40%. whats wierd to me, or at least what i didnt expect, is that raising my multiplier to affect memtest so much. i always assumed the ram didnt care about the ratio, but i guess it does.
  12. just got a new motherboard, decided to wait a bit longer before going 64 bit. i got pny ddr 600 memory anyway for when i decide to upgrade, which i run at ddr 500 (was that a bad idea?), to go with my xp-m 2500+ anyway when all is said and done, my timing are 2.5, 3, 3, 7. my "cpu interface" is on "aggressive" under advanced chipset settings. i ended up at 250 X 10. at vcore 1.825, and default voltages for chipset and ram. i cant go any higher on the fsb or multiplier, not because of the chip will run hot, but because my memory wont pass memtest. my cpu's temp never gets to 50C. somehow i feel like im underachieving. i feel like i should be able to raise the speed until my temp gets into the 60s at least, but I guess the nforce chipset on my mb doesnt want to go faster than 250 mhz on the fsb, regardless of my multipler. is that my memory's fault? its not that im not happy with 250 X 10 for what i payed for this hardware, but i am curious about what you guys think.
  13. ok guys here's the update: i wasnt convinced it was the bios, and i didnt want get a new bios and have it not work so i cross shipped the board. as to whether it was any given component causing it not to post before, i tired swapping PSU, CPU, GPU, DIMMs, and doing all of the above outside of the case on an antistatic surface. i reaseated the cpu a few times, tried pc2100, tried every 4led fix trick i could find online, no luck. so i got the new mb, and my headache continues i never /really/ overclocked/. i have a mobile barton, so i guess the stock setting is 133 X 14. i ran at 133 X 16 for 2.133ghz at stock voltage. on this new mb, i cant even do that. it crashes (by crashes i mean shuts off) after a few minutes of running prime95. i got new memory which should run fine at 200fsb, but 200x10 is out fo the question. the memory is ddr550. i can run prime95 a few minutes longer at 250x8, but i cant go to x8.5. it seems im limited by my fsb. i cant raise it past 1.6 or the mb shuts off, despite the temp not even being at 51C. 200x10 is the default speed for a barton 3000+, and i think i should be able to run at that speed. i'd consider that stock, i guess. what do you guys think the stick speed for my chips should be? what would you do about this board.
  14. my ceiling seems to be my vcore. even though my temp never reaches 60C, if i raise my vcore past 1.55 the righ will hard shut off when running prime95, i have a mobile barton 2600, which is not really overclocked at all when run at 200X10, but all i can get out of it is 5 minutes of prime95 before it shuts off ive tried swapping back in all my old components. the only variable that has changed is my mb, and im afraid it just isnt cutting it. i guess ill call DFI on monday and ask for another RMA. i know i cant expect it to overclock, but i can expect it to run stable at stock speeds.
  15. perhaps i spoke to soon, though it remains to be seen after my last post i cleaned the arctic silver 5 up on my chip and hsf, and reapplied it (and obviously reseated my cpu). Then i passed test1 of prime95 for the first time at 15 X 133 @1.65 vcore / 2.6 dcore so i tried to do the same at 10 X 200 @ same voltage and instant hard shutdown again. on a whim i decided to turn the memory speed up and run memtest. with some tweaking i got it to pass one test @ 250mhz (lowering cas latency from 3 to 2.5 made it pass, my ram voltage and chipset voltage are still at 2.9 and 1.7, though i dont know how necessary those are). fully expecting my system to grind to a hard shutdown within 5 seconds again, but instead it took a few minutes, which is progress, nonetheless. (i got to third itineration test). in my experience when a system you are oc'ing just shuts off like that during prime95 (rather than fail a test) its usually an inadequate PSU. my PSU is brand new and better than the last one I had, which i was able to crank the vcore to 1.825 on before it would start to shut off. not that i run unstable. my system is a part time server and it was up for 100 days at one point (not running prime95 the whole time obviously) @ 2.4ghz. even then it only went off due to a power outage.
×
×
  • Create New...