Question about Overclocking

Trance-Lord-SnyderTrance-Lord-Snyder Eastern PA Member
edited October 2006 in Hardware
I've used the ultimate guide to overclocking for AMD Athlons, and it was very helpful, but I have a question that wasn't covered in the guide.

First my rig:

CPU: AMD Athlon 64 3500+ Venice
Mobo: K8N Neo4
Mem: 4x 512MB PC 3200 Corsair XMS
Video: eVGA 7900 GT 256MB
Sound: Soundblaster Audigy ZS2
HDDs: 250 GB SATA Seagate, 74 GB Raptor
PSU: Corsair HX520 Modular Power Supply

I used to have my cpu overclocked to 2.5 Ghz, but ever since I installed a western digital raptor, my computer will not boot at 2.5 Ghz. The highest stable cpu speed I can get is about 2.43 Ghz. I thought maybe my old power supply couldn't handle the load, and I knew it was a cheap free one I got with my case, so I did research on power supplies and thought I'd give the HX520 a shot. It did actually give me a slightly higher overclock when the Raptor was attached, but it was only about 20 Mhz. If I disconnect the raptor and boot of the seagate drive I can overclock higher again. If I disconnect the seagate and boot off of the raptor only, I still can't overclock any higher than 2.43 Ghz. If I try to boot at 2.5 Ghz, I hear a click noise come from the raptor hard drive. So my question is...

Is it possible that increasing the HTT Mulitplier is causing instability on my raptor hard drive and that's why I can't overclock any higher? Would a new motherboard fix this problem? Does anyone have any suggestions?

Also, I don't think heat is a problem, my case stays pretty cool, I've got plenty of fans to keep a nice airflow going, and the hard drives each have a fan bringing in cool air from the front of the case and blowing the air over them.

Thanks guys.

Comments

  • OrianeOriane Turn around.
    edited October 2006
    Trance Lord Snyder wrote:
    If I disconnect the seagate and boot off of the raptor only, I still can't overclock any higher than 2.43 Ghz. If I try to boot at 2.5 Ghz, I hear a click noise come from the raptor hard drive...

    While I have no suggestions for you about overclocking this, I do have a suspicion I think you should think about.

    Have you run diagnostics on your HD?
  • muddocktormuddocktor
    edited October 2006
    Man, that sounds kind of like your motherboard isn't locking th pci bus speeds when you are overclocking the HTT and the Raptor can't handle the out of spec pci bus speed as well as the Seagate. Is there a setting in bios to allow you to hard lock the pci bus at 33 MHz? Other than that, I can't think of a thing right offhand that would account for this difference in overclocking speed between the 2 drives.
  • ThraxThrax 🐌 Austin, TX Icrontian
    edited October 2006
    muddocktor wrote:
    Man, that sounds kind of like your motherboard isn't locking th pci bus speeds when you are overclocking the HTT and the Raptor can't handle the out of spec pci bus speed as well as the Seagate. Is there a setting in bios to allow you to hard lock the pci bus at 33 MHz? Other than that, I can't think of a thing right offhand that would account for this difference in overclocking speed between the 2 drives.

    Exactly my thought too, Mud.
  • Trance-Lord-SnyderTrance-Lord-Snyder Eastern PA Member
    edited October 2006
    There is no option to lock the bus speed, but Clockgen says that my PCI bus speed is running at 33.34 Mhz.

    And diagnostics... what would you suggest I run?
  • muddocktormuddocktor
    edited October 2006
    You can try running the diagnostic suite that Western Digital has on their website for checking out their hard drives. I'm not positive what they call it, but it might be called "Data Lifeguard". Boot with your Raptor at it's max overclock setting and run the suite and see if it returns any errors.

    And Clockgen is showing that your PCI bus is running at spec too, so I guess testing the drive itself out would be a good next step. Let us know how it turns out and we'll go from there. I'm not familiar enough with your mobo to know the settings in bios, but usually there will be some kind of setting that makes a board using an Nvidia chipset lock the PCI bus for sure. My 2 socket 939 boards use an NF3 Ultra chipset on one and an ULi M1697 on the other. I don't own a NF4 mobo, so I'm not sure what locks down the PCI bus but on my Epox board setting the AGP bus to 67 locks everything and if I remember right, the Asrock board has a setting to lock the PCI bus at spec.

    EDIT: What do you think of that Corsair psu? I've been thinking of buying myself that model you have for my Epox rig because the Antec TC 550 I presently have in it has problems dealing with the startup draw of the monster 120mm Delta GFB I have on my heatsink.
  • DonutDonut Maine New
    edited October 2006
    The NF4 should have the PCI locks. 33.34 is what Clockgen reports on my dual Opty rig (K8N-DL).
  • lemonlimelemonlime Canada Member
    edited October 2006
    Glad you found the overclocking guide helpful :)

    The behavior you describe does indeed sound like some kind of bus locking issue, but SATA on the NF4 is not prone to that. The NF4 chipset has 'chipset' level SATA control, and the PCI-E and PCI bus speed does not impact it. A high LDT or HTT multiplier will cause that type of issue, however. What is your HTT/LDT multiplier currently set to? Try decreasing it by a notch. If you are using the 11x CPU multiplier, you should definitely be at the 4X HTT/LDT multiplier (800/1600MHz) to maintain stability. You can get all sorts of 'board' level issues with an out of spec HTT bus.

    Let us know how that goes..
  • Trance-Lord-SnyderTrance-Lord-Snyder Eastern PA Member
    edited October 2006
    My HTT speed is currently set to 221 Mhz, and the multiplier is set at 4, and the board supports 1000 Mhz, so I'm within the range it should be in...

    Something I noticed with CPU-Z is that my CPU voltage fluctuates a lot. According to the program, it's saying my core voltage is anywhere between 1.466 volts, and 1.52 volts. I don't think that kind of fluctuation is ok, but maybe CPU-Z isn't reliable for that measurement? I have no voltmeter to actually test the CPU myself, but maybe I need to invest in one. Maybe the board is causing the instability. My previous power supply showed the same kind of fluctuations, and now I have what I think is a pretty good quality power supply. Hmm...

    On another note, I did play with the voltage and speeds a little bit, and I tried loosening up the timings on my ram from 2.5 3 3 5 to 3 3 3 8. When I increased the HTT speed, I wasn't able to get any higher of a CPU clock speed that was stable with the loosened timings.
  • lemonlimelemonlime Canada Member
    edited October 2006
    Trance Lord Snyder wrote:
    My HTT speed is currently set to 221 Mhz, and the multiplier is set at 4, and the board supports 1000 Mhz, so I'm within the range it should be in...

    Something I noticed with CPU-Z is that my CPU voltage fluctuates a lot. According to the program, it's saying my core voltage is anywhere between 1.466 volts, and 1.52 volts. I don't think that kind of fluctuation is ok, but maybe CPU-Z isn't reliable for that measurement? I have no voltmeter to actually test the CPU myself, but maybe I need to invest in one. Maybe the board is causing the instability. My previous power supply showed the same kind of fluctuations, and now I have what I think is a pretty good quality power supply. Hmm...

    On another note, I did play with the voltage and speeds a little bit, and I tried loosening up the timings on my ram from 2.5 3 3 5 to 3 3 3 8. When I increased the HTT speed, I wasn't able to get any higher of a CPU clock speed that was stable with the loosened timings.

    Looks like you have ruled out the ram. You could try a divider, but I doubt that would be the issue. What is your CPU vcore currently set to in the BIOS? I'd be curious to see what happens when you increase it. If you boost it up to 1.5V or even a little higher (even if it is just temporary for testing), I'd be curious to see if the instability remains.

    TBH, it still sounds like a board issue to me. The CPU vcore instability would likely give you some nasty bluescreen issues long before you saw problems with your hard drives.

    Does your board have more than four SATA connectors? it is possible that only some of them are on the NF4 chipset. Many boards have a second external SiI SATA controller. If so, try your raptor on another SATA port. Give that a try regardless of the number of ports.
  • muddocktormuddocktor
    edited October 2006
    I just looked on the MSI site and the only board they show is the K8N Neo4-F, which has only 4 SATA ports. So those are all controlled by the NF4 chipset itself, which shouldn't have any problems locking the pci bus speeds.

    Trance Lord Snyder, have you tried running the Data Lifeguard tools from WD on your Raptor yet? That shouldn't take more than an hour or so to run. :)
  • Trance-Lord-SnyderTrance-Lord-Snyder Eastern PA Member
    edited October 2006
    I haven't tried that utility but I will as soon as I get a chance, I'm not at my home PC at the moment. I'm not sure if I'm hoping that something is wrong with the hard drive or the mainboard. I was already frustrated with Western Digital that I didn't get my 30 dollar mail in rebate on the hard drive...

    I'll also try switching serial ports and see if that solves anything.

    I'm sure that the Ram isn't the problem because I've lowered memory divider to make sure that it was running stable. I couldn't get any higher of a HTT speed with the divider.

    This seems to be a motherboard problem to me, but I don't know if I want to upgrade quite yet, or stick with socket 939. I want to upgrade to dual core, but I don't know if it's worth buying a socket AM2 board or sticking with a 939... too bad it's not easier to sell old parts.
Sign In or Register to comment.