• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Overclocking and Harddrive Problem

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

TommyDee291

Member
Joined
May 16, 2003
Location
Pennsylvania
I Posted already in the ASUS Motherboard section but got very little help. I am also putting it in here for hopes that I get more responses. . .

Hi, I am new to this forum and to overclocking. However, I have been doing significant research for about a week now regarding overclocking. So I decided to go ahead and give it a try. I have often read that the P4S533 M/B is highly Overclockable. I am running a Pentium 4 1.8a processor on this M/B. I went into the bios, and overclocked gradually. I went to 1.9, 2.0, 2.1, and 2.25. I also increased my voltage to 1.575 when I was running at 2.25 Ghz. I also was running a 3:4 ratio. Although, I will admit, i don't completely understand. I booted into windows successfully at 2.25 Ghz and didn't think to run a stabilty tests. I ran my usual programs when Windows XP popped up a balloon saying that data is corrupted and unreadable in one of my drives. I have 4 Harddrives. I had to backup my data and format the drive and restore the data on that drive. At that particular time, my overclocking did not come to mind to cause this issue. The error message did not appear on any of my other drives and I did not lose any data on any other drives. However, I did notice, when I was at 2.0, that I did not get any error messages. I would like to know why my one drive corrupted itself after an overclock at 2.25 ghz. If it helps, I believe I was using a 125 mhz system bus and an 18 multiplier. Thank you in advance, Tom.
 

fiji

Member
Joined
Jul 14, 2002
Location
berlin
**Edit** ok my guess is that the data just got corrupted, because occasionally does no matte what i dont think it has anything to do with the overclocking ***



whats your system specs?

did you check your mobos site to see if they have data corruption problems? (kinda like the nf7-s sata stuff)

i dont really know about your mobo, but you could get an ide card for a pci slot and lock the agp/pci slots (that is of course if your mobo supports this) and then you could see if it still did it


another thing it mightve jsut done it out of use, or was it fat32 or ntfs because fat32 seems to corrupt more


i dont really know what im talking about, its late and im tried
 

Daemonfly

Member
Joined
Aug 18, 2002
Location
NW Pa
When you raise the FSB to overclock, you're also raising the AGP & PCI bus speed, basically overclocking any cards you have as well as the hard drive controllers/channels. You can prevent this if your motherboard supports PCI/AGP "locking".
 

deathstar13

FSB FRIEK
Joined
Dec 24, 2001
im not familier with intel mobos but it uses deviders like the older kt333 amd mobos.
with those memory can corrupt hard drives also.and by raising the fsb it increases the memory speed.
im not very sure what the memory speed was/is with that ratio.
but increasing the memory voltage or relaxing its timings can really help if this was the issue.
 

bchur83

Member
Joined
Jan 26, 2003
Location
Land of 10,000 Lakes
It could be your PCI frequency. I have heard that if the PCI frequency goes over 38Mhz then you have a good chance of corrupting your drives. Can you lock the PCI/AGP busses, or do you have different ratios to set.
 
OP
TommyDee291

TommyDee291

Member
Joined
May 16, 2003
Location
Pennsylvania
Thanks for replying guys. First off, here are my specs. Intel Pentium 4 1.8a (512k Cache), Intel Stock Fan, Antec Smart Fan, HDD Cooler, Asus P4S533 M/B, 512 MB of Kingston 2100 DDR, ATI All in Wonder Radeon 7500, 4 Harddrives Totaling 160 GB, Creative Sound Blaster Live! Mp3+ 5.1, Allied 350 Watt PSU. The 4 Harddrives are a Western Digital WDJ800 (Special Edition 8 meg cache), another Western Digital which is 20 gigs (This is the one that is corrupting. 1 Maxtor 20 gig and 1 Quantum Fireball 20 gig (IDE). All drives are formated using NTFS. I also use a PCI Ide controller to account for the extra drives. I originally thought this was whats causing the corruption but i moved the drive to my M/B controller and its doing the same thing. This is not a random occurance because at first when I overclocked at 2.25 ghz and it happen, i didn't think that they were related in anyway. But after doing all the work to restore the data it happen again! :mad: So I declocked to my stock speed and its not doing it anymore. My M/B does not have a PCI Lock. Someone in another post reccomended increasing the memory voltage to 2.7 and the CPU voltage to 1.65 and to mess with the memory timings and ratios. Does anyone think that could fix it? I'm sorry if not getting into alot of detail, as I am at school currently.
 

deathstar13

FSB FRIEK
Joined
Dec 24, 2001
well i saw the 125mhz system bus,and the 3:4 ratio.
correct me if im wrong but that would put your ram at 160mhz? or close to it.
anyhow your ram is rated 133mhz and by not increasing the voltage or or at least relaxing the timings,its very easy to corrupt hard drives this way.

as ive stated im not the most intel knowledgable person.but i do know on an amd system that bus speed with stock ram settings and voltage will corrupt a hard drive.

not many know that i dont think.that memory timings and lack of voltage will corrupt a hard drive just as fast as an over clocked pci bus.and with a pci lock it even proves its possable.

increasing the cpu voltage wont affect this.as long as its stable leave it alone.i run my memory 2.9v all day long 24/7 folding so even 2.8 might be enough for you.

could always buy new higher rated ram but thats on you.
 
OP
TommyDee291

TommyDee291

Member
Joined
May 16, 2003
Location
Pennsylvania
Thanks for your response, I'm gonna try the memory voltage increase, i only think it goes up to 2.7 volts but i could be wrong. And like I said earlier, my m/b doesn't support PCI/AGP locking. So Hopefully my problem is fixed! :cool:
 

Sonny

Senior TIFOSI
Joined
Aug 3, 2001
Location
VARIANTE ASCARI
To add to what deathstar13 has already posted,

125MHz / 3 (4) = 166MHz

Well your clocked to 166MHz from 133MHz (DDR266|PC2100) so this could be one of the issues you need to look into. Does it pass Prime95 at this speed? What are your timings? Were you at default voltage, 2.55V, on your memory?


125MHz / 3 = 41.66MHz PCI

Another thing to look into is your PCI BUS Speed, 41MHz is pushing it for most HDDs.
 
OP
TommyDee291

TommyDee291

Member
Joined
May 16, 2003
Location
Pennsylvania
When I was overclocked, I was running the memory at the default voltage because i totally forgot about it. :rolleyes: And when I was overclocking, I had my timings at 2.5 5... and i can't remember the rest, but under memory settings in my bios, i had it set at "by SPD" whatever that is, but it had the timings set fairly high, like i said, 2.5t 5... i can't remember exactly. Anyways, the only thing im not clear at for the moment is what should I set my Memory and CPU Voltages at to reach 2.4 ghz, (This is my goal). keep in mind, I have to and am using Stock Cooling. So far people have suggested 1.65 for CPU and 2.7 for memory. In order to hit 2.4 ghz, I have to run a 133 mhz system bus and that puts my memory perfectly clocked at a 1:1 ratio and my PCI and AGP at 33.25 and 66.5. Is that safe?
 
OP
TommyDee291

TommyDee291

Member
Joined
May 16, 2003
Location
Pennsylvania
*Update*Ok guys, I attempted another overclock.. Mixed feelings... Um... Ok, I followed your guys advice and Upped my CPU Voltage to 1.65 and strangely my RAM was already set to 2.9 volts.. weird, because the default is 2.5 and I don't remember changing it. Anyway, I also relaxed the memory timings to 2.5 4 4 7 set the system bus to 133 and saved the bios and rebooted, it wouldn't post, turned it off, turned it back on, the bios recovery came on saying it didn't properly post, blah blah blah. I can't understand why it won't post, I had the ratio set to 1:1 too which should work by all means. I beileve we solved the Harddrive corruption issue because at 2.25 ghz, I was over the 38 mhz barrier for the PCI bus which is known to cause corruption. So I am thinking that my Kingston 2100 RAM simply cannot handle the overclock? I am overclocked currently at 2 ghz, 111 FSB, 1:1 ratio, 2.9 v for ram, and 1.575 for CPU. Im gonna run stress tests and such. My PCI bus is running at 37 mhz, so thats right below where corruption could occur. So far so good! I look forward to hearing from everyone inputs again!
 
OP
TommyDee291

TommyDee291

Member
Joined
May 16, 2003
Location
Pennsylvania
I was thinking a bit.. and I think its either my Video Card or RAM now. My Video card is an ATI All in Wonder Radeon 7500. and the RAM is Kingston 2100. I just ran some benchmarks, my CPU score in PCMark was of course better, but my memory was lower. This of course is because I have very loose memory timings. And Since I cannot go to 2.4 ghz for the moment. How tight can I get my memory timings for 2.0 ghz without risking data loss?
 
OP
TommyDee291

TommyDee291

Member
Joined
May 16, 2003
Location
Pennsylvania
Ok, my Memory score was lower because in my BIOS, I had my ratio to 1:1 and 111 FSB which would de clock my RAM. So its now at 3:4 with tighter memory timings. I researched my RAM and video card and found that they are not the issue with my overclocking problem. I believe its now because of my power supply. Its 350 watts and I have alot of things in my system. I.E: 4 harddrives, 2 optical drives, Video Card, RAM, Soundblaster Live Mp3 + 5.1, Another IDE controller on my PCI bus. So I am thinking that my computer won't post at 2.4 ghz because I don't have enough power when I upped my voltage to 1.65. Any thoughts?
 
OP
TommyDee291

TommyDee291

Member
Joined
May 16, 2003
Location
Pennsylvania
My RAM voltage was set at 2.9 volts by default for some wierd reason. Im now at 3:4 though. But, Like I said earlier, I think my power supply is not enough power for my over clock of 133 mhz and 1.65 voltage.
 
OP
TommyDee291

TommyDee291

Member
Joined
May 16, 2003
Location
Pennsylvania
I read today an article on AnandTech's website that having your memory at 2.9v is not a good idea, that your ram will not live too long. Is this true?