Jump to content

stone_cold_Jimi

Members
  • Content Count

    115
  • Joined

  • Last visited

Everything posted by stone_cold_Jimi

  1. In all cases where a CPU has died, the common components have been RAM and PSU. I reckon my Seasonic 600W intermittently blows chunks. I've RAMed the 3700 and am getting another CoolerMaster 550W PSU; my original CM is fine with the other Expert. The new one will be the SLi version, replacing the current one which I'll use on the 3700 rig.
  2. I've had a look round for info - it's the RAM led that hasn't been working. I'm rather bewildered. My other Expert (an AA0) is in my Vapochill LS rig, doesn't miss a beat. The AB0 was what arrived after I RMA'd an AA0 that died in a day. I've been using it since the end of March. I'm no stranger to these Expert mobos now, I should add. And I know that the Asus I have to check with works ok and am pretty certain that both of these CPUs have died. I simply have no idea why. How can that happen? At stock volts too. Edit: when I say 'no go' with the Asus, I mean that the rig sits there like a lemon, fans running but nothing happening. On the Expert, it has one yellow LED by the PCI slots, but hitting the power button does nothing. Got the DMM out on the Asus and there's vcore and vdimm ok; it has to be the CPU.
  3. What does the yellow LED at the edge of the board signify? I've now lost 2 CPUs and just realised that this LED has not been working since before the first chip died. The first CPU, a lovely x2 3800, seemed to degrade over a couple of months until it stopped working. Apart from a couple of boots at 1.42 vcore, it ran at 1.33v or 1.39v for that time. Initially, it was sweet at 2700... then lockups, so I ran it at 2500... then lockups, so 2300... then dead. Tried CPU in Asus. CPU dead. I bought a Sandy 3700+ to replace it, got it last Friday. It ran sweet at stock volts, 230x11. Ran Rosetta with it. Then it too began to play up. Today, memtested at 250x11 with the Ballistix tweaked a bit. 3 hours memtest, lovely. Primed for 55 minutes, prime crashed. Then XP wouldn't boot. Switched off for a while, restarted. Fine. Ran 3DM03, looking good... then suddenly rig switched itself off. Power up again, no red LEDs, one yellow, no go. Tried CPU in an Asus. No go. Tried Expert with no components, 4 red LEDs. Therefore CPU dead. Setup was: DFI Expert Rev AB0 x2 3800 or SD 3700 2GB Ballistix Z503 2x 6600GT SLI Seasonic 600W PSU
  4. Well, I thought mine were dying, but it turned out to be the CPU. When that finally died (x2 3800 at stock), I had the Ballistix at 200MHz 2-2-2-5 1T. Ran 250MHz 3-3-3-7 1T as well with the 250 multi. The CPU mem controller was fine, just something else on the die blew. Also, the one time they gave me real grief, I pulled them and cleaned all the dust out off the board, they were champion after that.
  5. Thanks, that's the easiest too. As it happens, these sticks seem to be loving DS/DDS of 8/1 at the moment. They passed memtest ok like that at 250 MHz 1:1 3-3-3-7 1T and I then used the rig to virus check 4 suspect partitions, then fired up [email protected] which has run faultlessly since. If (when) the wheels come off again, RMA time.
  6. I fear that you are correct, especially as my batch number is CL1115N.2Q and they run at over 60C uncooled. Still, no reason to not extensively test and eliminate other possibilities. I've been playing the drive strength game; so far they like DS at 8 best but came up with a couple of errors on the second pass. They hated 9, 10 and 12, so I've gone from DS/DDS 8/auto to 8/1 and am running memtest again now. At least they're crapping out with real errors now and not locking up. I got these from overclockers on 4th November last year - would I have to ask them for an exchange, or would I just RMA to Crucial direct?
  7. Ah, the plot thickens. Testing the sticks individually is a different experience. With CPU at 2500 and stock vcore, memory at 250MHz 1:1 3-3-3-7, they both pass memtest repeatedly. Can't do 3-3-2-7 without errors, but hey. So suspicion now falls on the mem controller or the mobo. I wonder now if playing with the drive strengths will reveal something.
  8. Awwww.. I bought my Z503 kit back in October, but I've only used them on and off until about 6 weeks ago. Since then, I've been running them 24/7 on [email protected] or was until a couple of weeks ago. The rig is: DFI SLI-DR Expert AB0 AMD x2 3800 (x2 4600 stock HSF) 2GB Ballistix Z503 (120mm Vantec fan) 2x Leadtek 6600GT SLI Seasonic 600W PSU For about 3 and 1/2 weeks, the rig ran 2700 MHz / 1.39v with the sticks on a 180 divider, about 245MHz 3-3-3-7 1T. Originally they used to run 1:1 up to 285 3-3-3-8 and at 250 I could run 3-3-2-7, but since I set this rig up they would fail memtest at that, hence the divider. Then I started getting lockups, so I raised vcore to 1.42v. That worked for a couple of days, then lockups again. Dropped the CPU to 2500 @ 1.32v and set the sticks to 1:1; that worked for another day or so and then lockups. Now I'm running memtest with 2500 / 1.32v with sticks at 250MHz 3-4-4-8 2.8v (i.e. stock, all but base timings on auto). They pass all tests until #8, then it locks up with the wall time at 400 or 700 hours or screen corruption. If I loop #8 it passes for about 30 or 40 minutes and then lockup. No errors though, just lockups. Is this my mem controller going bad or the Ballistix? Uncorrected CPU temps on load about 38C, btw.
  9. HZ ftw. Good stable memory, best price. Mine do 270 3-4-4-8 and not a MHz more, fair enough (edit: @2.6v too). I'm thinking of getting HV if my Ballistix finally bork, they look nice.
  10. No, you're right - it doesn't match my findings either. The odd numbers 1-3-5-7 are definitely grades of weak, as I recall, just can't figure out the rest yet although I know they're weird.
  11. So does this go (weakest to strongest): 2 sticks: 1-3-5-7-2-4-6-8 4 sticks: 9-11-13-15-10-12-14 ??
  12. Ummm.... A64 is DDR(1), DDR2 is strictly Intel until AMD socket AM2 comes out. Get the G.Skill HZ, so boringly reliable you just forget it's there. I keep reading that more recent batches of OCZ EB, G.Skill HZ, Mushkin Redline all overclock less than earlier batches. So you might as well get the cheapest.
  13. Note that I clocked my 4600 at 415 x 8 just for the thrill. Posted this before, but here it is again: http://www.xtremesystems.org/forums/showpo...58&postcount=63
  14. x2 3800 Manchester. Had one at 3100 phase, another at 2800 stock air. 06xx chips are high chance of very nice.. and nice price.
  15. 99% chance that it's a RAM and mem controller issue I bet. Save your last stable settings, raise HTT and memtest, SuperPi. This is where you have to work with these, I've found (especially with N substrates). Sometimes I've set everything to auto and read back the settings off A64Tweaker to see what the system likes.
  16. The only 0605 in that list: http://www.xtremesystems.org/forums/showpo...42&postcount=66 doing very nicely.
  17. It's a Venice, not a Manchester. Don't apply the dual-core patch.
  18. Yeah, I start hitting the wall between 3100 and 3200. I always think of phase as stock air +300 MHz for a given voltage, so I'd expect 2700 to 2800 around the 1.5v mark on stock HSF with mine, which isn't far from the truth. Beyond that, voltage has to rise exponentially and it just ain't worth stressing them that much (for long, anyway). Some people with 0601s reckon on 3g with air... very few though.
  19. Vapo LS. Have a look, two 0607s, a T and an N: http://www.xtremesystems.org/forums/showpo...2&postcount=147 http://www.xtremesystems.org/forums/showpo...8&postcount=122
  20. Hi to the LDBFE owners club. So far, although the sample is smaller, while LDBHE 0601 have been carrying the swing, LDBFE 0607 have had consistently good results, whereas not all 0601s are equal. I'm very impressed with these. Neither of those I've tried has struggled to get 3000 / 1.5v on phase and one has bested my x2 4600 at that speed by being stable with 1.47v against 1.5v. Nice! :cool:
  21. Known. They're used in the Team Xtreem PC4000 Cronus kits (unavailable in UK afaik). As good if not better than 5B-D; I've seen 3-2-3-6 at 300 MHz I believe (Onepagebook and his cascade OC was using them).
  22. In theory, yes. Mine and my Expert(s) really didn't get on well at all though. Other people may have different experiences. It certainly handled 2x GTs in the Asus A8N32 ok.
  23. It crapped out at full power in one of the respected reviews (can't remember where). They had the good grace to test one of the SLI versions when they came out and said the problem was fixed (CoolerMaster had acknowledged the issue) - actually, the SLI version looks like an excellent PSU and I may yet get one for this rig.
  24. 12/07 for both. A BIOS shot will take longer lol! 405x8 had sensible things like chipset voltage one under the max, LDT 2x, LDT volts a couple of notches up. The G.Skill was CPU/12, 3-3-3-8 1T, trfc 3120, DS/DDS 3/auto and everything else auto (lazy, but it works equally well as tweaking it). I think vcore was 1.525 + 104%. Some things just didn't work, like the 7.5x (forced CMOS clear) and 7x (didn't boot) multis and the DDR240 divider just left the RAM at DDR266. This is the highest I've clocked the chip anyway... hmmm, 270x12 looks like it could be sweet.
×
×
  • Create New...