FID Jumpers

[49erinnc]

Hi

I have an ASRock K7S8XE mobo and a Athlon XP 2200+ CPU. My memory is PC2700, so I'd like to overclock my CPU to a 166(333) FSB. However, my computer will just shut down after I get past 145MHz, manually bumping it up in the BIOS. I've noticed that my motherboard has a series of FID Jumpers to adjust the multiplier of CPU but I'm clueless as to what that really means. Does adusting these jumpers have any impact on overclocking? If so, what would I need to do to get my CPU up to 166MHz, because I've heard I should be able to do that without much problem. My CPU fan is rated for 3200+ models so cooling shouldn't be a problem but I'm just curious if messing with the FID jumpers is what's needed to get over the 145MHz hurdle?

Thanks

 

[49erinnc]

BUMP for a little help.

 

[Exodus88]

Try bumping your CPU voltage up a notch and try again. Although your cooler is rated for 3200+ it doesn't really mean that is meant for OCing. stock barton - 1.65vcore Mine 1.95vcore. As long as it is a decent cooler you should be able to up your vcore afely. Post your temps and also a little more info PSU cas of ram etc...

 

[49erinnc]

Well, I figured bumping voltage was the key but I don't think my AMI BIOS allows for it. I found the tab where the voltage is displayed but from what I can tell, you cannot manipulate it.

This was a review about my board posted on Newegg:

"A two stage power regulator limits this motherboard's overclocking abilty. This combined with the board's low overvolt capability (.1 v) prevents me from overclocking my XP2500+ more than 200 MHz above standard"

And this is a quote off a website review of the board:

"There are no multiplier control within BIOS or on board. There are also no voltage options available in the BIOS for tweaking the board."

Those comments may tell you more than what I can about it. Would flashing a new BIOS help with overclocking? My BIOS is ver. 1.0. When I go to ASRock's website, they have several newer versions available. I don't have any problems with my current system so I don't want to screw up anything but if flashing the BIOS to a newer version would help me overclock, then it's something I'd like to consider.

If that's not an option, are there any other ways to increase the voltage without taking a big chance of ruining the board?

 

[49erinnc]

After doing some further research, it turns out that my mainboard is the revision 3.0 which has the following:

In order to provide advanced performance, the Hybrid Booster provides the CPU Multiplier, CPU Vcore voltage adjustment, CPU Frequency Stepless control, and AGP/PCI Frequency control features.

With the CPU Multiplier, power user could change the processor multiplier setting to get the best performance of CPU speed.
The CPU Vcore adjustment helps to power up the CPU core voltage to support the CPU overdrive demand.
The CPU Frequency Stepless control provides a CPU clock frequency upward/downward adjustment interface.
The AGP/PCI Frequency control provides the interface of AGP/PCI frequency value setting.

Based on that, it seems like I should be able to get more out of it but I'm unable to figure out what needs to be done unless it's something that needs to be done with those FID jumpers or something else directly on the board itself and not in the BIOS.

Any ideas? Thanks

 

[tcsenter]

I've noticed that my motherboard has a series of FID Jumpers to adjust the multiplier of CPU but I'm clueless as to what that really means. Does adusting these jumpers have any impact on overclocking? If so, what would I need to do to get my CPU up to 166MHz, because I've heard I should be able to do that without much problem.

It can help if you lower the multiplier.

If you want to try the next highest FSB level (e.g. 133 > 166 or 166 > 200), start with lowering the multiplier so that the total operating frequency is approximately the same. e.g. XP2200+ operates at 1800MHz

Using 10.5x multiplier with 166FSB equals approx. 1750MHz, which the XP2200+ should do. Test the system at that speed, then bump the multiplier up incrementally; 11x, then 11.5x, so on and so forth. Test the system for stability and CPU temps at each multiplier setting. When you reach any instability or glitches, go back to highest stable multiplier. You may also try using the CPU vcore over-volt setting in BIOS to get more stability at higher multipliers.

Now as to the question of which Athlon XP's are multiplier locked, and whether those will permit changing the multiplier without further modification, I don't know. You'll have to ask someone else.

 

[tcsenter]

Those comments may tell you more than what I can about it. Would flashing a new BIOS help with overclocking? My BIOS is ver. 1.0. When I go to ASRock's website, they have several newer versions available.

Hell that's the first BIOS release dated July 2003 and there have been six BIOS updates since then. I would bite the bullet and update it to the latest version.

Asrock states that it doesn't provide the jumper caps for the overclocking jumpers, you may need to get a few.

 

[49erinnc]

If you want to try the next highest FSB level (e.g. 133 > 166 or 166 > 200), start with lowering the multiplier so that the total operating frequency is approximately the same. e.g. XP2200+ operates at 1800MHz

Using 10.5x multiplier with 166FSB equals approx. 1750MHz, which the XP2200+ should do. Test the system at that speed, then bump the multiplier up incrementally; 11x, then 11.5x, so on and so forth. Test the system for stability and CPU temps at each multiplier setting. When you reach any instability or glitches, go back to highest stable multiplier. You may also try using the CPU vcore over-volt setting in BIOS to get more stability at higher multipliers.

Now as to the question of which Athlon XP's are multiplier locked, and whether those will permit changing the multiplier without further modification, I don't know. You'll have to ask someone else.

When you say "lower the multiplier", I'm a bit confused. I don't have my board right in front of me, but from what I can remember, there's just 4 two-pin jumper settings in a vertical line, labled FID 1, FID 2, FID 3, FID 4. Are you talking about putting a jumper onto one of these to lower the multiplier? And if so, which FID setting does what. ASRock's read me files don't tell me anything...already checked. And if I need jumper caps for the overclocking jumpers, are these anything special? Would CompUSA carry them?

Also, from what I can remember, I wasn't able to adjust the volt settings in the BIOS. Not sure if I need a updated BIOS for that or not or if I'm just missing something.

Thanks for your help by the way...trying to take all this in before I go home from work and fry my board/CPU...lol

 

[tcsenter]

When you say "lower the multiplier", I'm a bit confused. I don't have my board right in front of me, but from what I can remember, there's just 4 two-pin jumper settings in a vertical line, labled FID 1, FID 2, FID 3, FID 4. Are you talking about putting a jumper onto one of these to lower the multiplier?

That is correct.

And if so, which FID setting does what. ASRock's read me files don't tell me anything...already checked.

Download the user manual from Asrock. It has a chart for FID jumpers.

And if I need jumper caps for the overclocking jumpers, are these anything special? Would CompUSA carry them?

CompUSA would probably have them. Just plain jumper caps.

Also, from what I can remember, I wasn't able to adjust the volt settings in the BIOS. Not sure if I need a updated BIOS for that or not or if I'm just missing something.

In BIOS under Chipset Configuration, there is a setting called Over VCore Voltage. IIRC, this setting increases the CPU voltage by a fixed percentage (approx. 5%).

 

[49erinnc]

Download the user manual from Asrock. It has a chart for FID jumpers.

When I did that the first time, I clicked on the Version 1.0 and not 3.0 which is why I couldn't locate the info. Found it now and I understand your original post.

In BIOS under Chipset Configuration, there is a setting called Over VCore Voltage. IIRC, this setting increases the CPU voltage by a fixed percentage (approx. 5%).

I'm almost positive that this option is not in my BIOS. I was looking through it last night, specifically for this option and it wasn't there. I can't think of any other explanation other than the 1.0 version of my BIOS I'm using because I think that version was released before the 3.0 version (with Hybrid Booster) of my mobo was even released. I'm thinking I'm going to have to flash my BIOS to the newest version but I've never done this so I'm a bit worried something is going to go wrong. If it doesn't post after doing it, will clearing my CMOS get me back in?

My only other concern (after reading some reviews on Newegg about my particular CPU) is that it's just flat out locked and I'm wasting my time doing the other stuff unless I "unlock" it.

One other quick question: Should I adjust my multiplier without adusting the voltage (assuming I'm able to flash and get the option)? Or will I likely need to do both? What voltage should I be going to, if you can answer that?

Thanks

 

[tcsenter]

I'm thinking I'm going to have to flash my BIOS to the newest version but I've never done this so I'm a bit worried something is going to go wrong. If it doesn't post after doing it, will clearing my CMOS get me back in?

If you have NERO burning rom, I can create an image file from which you can create a bootable CD that will automatically flash the BIOS for you. This should minimize the risk of something going wrong.

 

[49erinnc]

I have Nero and may need your help. I have been following ASRock's directions for Flashing a new BIOS and it's not working. I have done EXACTLY what it said, but with no luck. I'm on W2KPro and it says to create a bootable system floppy (done). Says to download FLASH.exe to that same floppy. Well, for starters, it requires 4 floppys with W2K so I just assumed it wanted me to save to the first disk, which I did. Then I was supposed to unzip the BIOS file I wanted to that disk as well. Well, after unzipping, there's not enough room to even extract to the disk. At this point, I'm clueless.

Will I be able to flash to BIOS K7S8X_2.10 via the Nero route? If so, you may need to walk me through what I need to do. I'm on here for the next little bit if you're available.

 

[49erinnc]

Double post

 

[tcsenter]

Will I be able to flash to BIOS K7S8X_2.10 via the Nero route? If so, you may need to walk me through what I need to do. I'm on here for the next little bit if you're available.

Check the .PDF manual for Write Protect option in BIOS or jumper on the board and disable (if present) to permit BIOS upgrade. If you use NERO Burning Rom, I created a CD image that you can try:

K7S8XE BIOS Flash Bootable CD image

It contains the latest BIOS (version 2.1), flash executable, and will automatically update the BIOS for you. Open NERO, cancel out of any Wizards, select 'burn image' from the File Menu, and provide the image file. After burning the CD, it will appear blank, because all the files are incorporated into the boot image of the CD. Enter BIOS and set the BOOT sequence to CD, then restart and watch the show.

After the flash completes, exit and restart. Enter BIOS and load Setup Defaults, then customize your BIOS options to taste. Don't forget to set Boot Sequence back to HDD (and remove the CD), else you'll flash the BIOS again.

 

[49erinnc]

Thanks tcsenter...I really do appreciate all your help with this. My best experience at anandtech thus far.

I didn't get your message until I was back at work this morning so I'll give the flash a try when I get home today. Sounds pretty straight forward as far as I can tell so we'll see what happens. Really hoping this works out because flashing it according to ASRock's instructions was turning into a headache.

Anyway, thanks again...much appreciated and I'll just shoot you a PM if I run into any problems. Hopefully that won't be necessary.

 

[tcsenter]

I'll just shoot you a PM if I run into any problems.

No news is good news....?

 

[49erinnc]

I normally have crappy luck with things like this, so I was expecting to run into a problem when flashing via the image. However, I'm happy to say that it was a breeze and it worked without a single problem. MUCH easier than flashing according to ASRock's instructions. And also, the newer version did in fact give me the new option of upping the voltage a hair.

Kind of good news/bad news with what I was hoping to accomplish. Prior to flashing the BIOS, I was only able to get my FSB up to around 145-147MHz before running into problems. After the BIOS flash and voltage override, I was able to get up to 150MHz last night with no problems, which put my 2200+ CPU at an actual speed of 2024GHz and my post read it as a 2400+. Pretty cool. However, when I went up to 153MHz, Windows loaded but I had a total freeze on the desktop so I had to back down to 150MHz. My whole goal with this was to get up to 166MHz, but I don't see it happening without actually unlocking the CPU with a chip-mod (which I don't really want to try).

With that said, I have NOT jumped the mulitpliers yet, which may help out. I did notice that I'm (at factory) on x13.5. Do you think if I jump them to a higher muliplier (maybe x15?) that I could get more out of my CPU? Also, at 150MHz, my CPU temp was 123F. Is that okay or too hot?

Thanks for all you help and hopefully I can get to 166MHz via mulitplier adjustment but I don't feel too optimisitic about it without connecting a bridge on the CPU itself. I'd just assume buy a new CPU than try that one.

 

[tcsenter]

I normally have crappy luck with things like this, so I was expecting to run into a problem when flashing via the image. However, I'm happy to say that it was a breeze and it worked without a single problem.

WOOT! Good to hear.

With that said, I have NOT jumped the mulitpliers yet, which may help out. I did notice that I'm (at factory) on x13.5. Do you think if I jump them to a higher muliplier (maybe x15?) that I could get more out of my CPU?

No, I think if you set a lower multiplier, it will help you achieve 166MHz, provided the multiplier can be changed via jumper.

If you meant keeping the FSB @ 133MHz but increasing the multiplier, that is a viable option.

The weak link here may not be your CPU, but the PCI BUS clock going too high as you increase the FSB through BIOS. I don't know how ASRock designed the K7S8XE, whether it features a PCI lock that keeps the frequency of the PCI BUS from going too high out-of-spec. Or it may be your memory, if it is configured by SPD or manually to run @ 166MHz, then you may also be increasing the memory clock beyond 166MHz as you increase the FSB. e.g. At 150MHz FSB your memory may be operating @ 183MHz.

It may be more productive to lower the multiplier (if able) and jumper the board for 166MHz so the PCI divisor is set to 1/5 from the get-go.

Also, at 150MHz, my CPU temp was 123F. Is that okay or too hot?

Its not too high. If it starts to approach 150' F (65' C), you should think about a better cooling solution or quitting while you're ahead.

 

[49erinnc]

No, I think if you set a lower multiplier, it will help you achieve 166MHz, provided the multiplier can be changed via jumper.

Okay...I was under the impression that the higher the multiplier, the better chance of overclocking. And yes, I can set the multiplier jumpers on the mobo (picked up some jumpers the other day). My board allows for 5x up to 24x. I haven't bothered changing the FSB jumper on the board...only adjusted incremently through BIOS. I can try just disabling the manual adjustment in the BIOS, going back to default settings and then just move the jumper on the board over to 333MHz. I guess I can do that and lower the multiplier some more. If that doesn't work, then I'll just cough up the $70 to get the 2500+ so I can run it naturally at 333MHz. Only thing that concerns me is if my CPU is just flat out meant on only run at a specific multiplier setting and if it will fry should I lower the multiplier to a setting the CPU isn't too thrilled about.

Its not too high. If it starts to approach 150' F (65' C), you should think about a better cooling solution or quitting while you're ahead.

I ran it for a while at 150MHz and the CPU temp wants to just hover at 132F and the case tended to stick at 95F. Both never really went beyond those temps, but I didn't test under gaming stress...just internet and idle. I think one problem I have is that my tower is set inside a tight cabinet opening of my computer desk, so there's not much air circulating. I do have the cover case off though. I have a 9800Pro on the way and I'm planning on getting an Arctic Cooling VGA Silencer which should help move some of that hot air out of my case.

I guess I should just be content knowing I got a 2200+ to run stable at almost 20MHz over normal speed but where's the fun in that?

 

[tcsenter]

Okay...I was under the impression that the higher the multiplier, the better chance of overclocking. And yes, I can set the multiplier jumpers on the mobo (picked up some jumpers the other day). My board allows for 5x up to 24x.

You can approach overclocking from either side of the equation; base frequency (FSB), multiplier, or a combination of both. All things being equal, there is more performance in higher FSB. e.g. 1GHz (997.5MHz) CPU-X @ 133MHz BUS performs better than 1GHz (1000MHz) CPU-X @ 100MHz BUS, though effective operating frequency is the same (within an insignificant 2.5MHz).

If you ran your XP2200+ at the same effective operating frequency of [approximately] 1800MHz, by using 166MHz FSB with a multiplier of 10.5 (yielding approx. 1750MHz), you would net a performance gain. In this case, overclocking is somewhat of a misnomer, as you would be over-bussing. Overclocking has become a catch-all term for operating a CPU out-of-spec to enhance performance no matter the method used.

FSB and multiplier are inextricably linked. If you increase the FSB to the point of instability, you have reached the maximum FSB attainable at the given multiplier. Lowering the multiplier may (and often will) permit you to further increase FSB. If you increase the multiplier to the point of instability, you have reached the maximum multiplier attainable at the given FSB. Lowering the FSB may (and often will) permit you to increase the multiplier.

Core voltage is also tightly correlated. e.g. if you increase the FSB to the point of instability, raising the core voltage may (and often will) restore stability at that FSB. So on and so forth....

 

[49erinnc]

You can approach overclocking from either side of the equation; base frequency (FSB), multiplier, or a combination of both. All things being equal, there is more performance in higher FSB. e.g. 1GHz (997.5MHz) CPU-X @ 133MHz BUS performs better than 1GHz (1000MHz) CPU-X @ 100MHz BUS, though effective operating frequency is the same (within an insignificant 2.5MHz).

If you ran your XP2200+ at the same effective operating frequency of [approximately] 1800MHz, by using 166MHz FSB with a multiplier of 10.5 (yielding approx. 1750MHz), you would net a performance gain. In this case, overclocking is somewhat of a misnomer, as you would be over-bussing. Overclocking has become a catch-all term for operating a CPU out-of-spec to enhance performance no matter the method used.

Okay...I officially understand how it all works after your latest post. Makes perfect sense to me now. So, based on what you said, I should be better off (performance-wise) by lowering my multiplier and raising my FSB instead of just jacking up my FSB while sitting on 13.5x. I think where I was screwed up was thinking that higher my processor speed (with stability), the better off I'd be. But now I'm thinking that I can actually get a better performance boost without having to run my 1.8GHz CPU up to 2.0GHz speeds, as long as I can raise my bus speeds higher, while maintaining stability.

I think that tomorrow, I'm going to set the bus speed to default in the BIOS, put the mobo FSB jumper on 333 (166) and set my multiplier at 11x. That will have my CPU speed at roughly it's natural state of 1.8GHz but I'll have the increased FSB speed I'm looking for. Hope it works...it's my last attempt.

Thanks again for the education...love to learn about this stuff.

 

[49erinnc]

Update:

First attempt: Set FSB jumper on mobo to 333MHz and jumped multiplier at 11x. The result was that I had to boot about 4 times before I could even get it to post. And once I did get it to post, I got a message saying that the system couldn't boot and to either enter the BIOS or system would boot up with defaults. I tried it both ways and the only way it would boot, was through the defaults. Once booted, I ran my mobo monitor app and noticed it had my CPU running at 1350MHz and the bus speed was at 100MHz.

2nd attempt: Left the FSB jumper on 333MHz and jumped the multilplier at 12x. Got the same problem as my first attempt in that I couldn't get it to post and once it finally did, it would only boot with system defaults.

3rd attempt: Set the FSB jumper back down to 266MHz, jump multiplier to 10.5x and hope to raise the FSB manually in BIOS. Same problem...several attempts before posting and same message about only loading with defaults. Went into BIOS, set the FSB to 166MHz and memory to 166MHz. Wouldn't boot. Had to load defaults which at least showed my 2200+ running at 1.8GHz but the FSB is back to default (133MHz).

One thing I did notice however, is that no matter what I set the multiplier to, when I did actually get the system to post, it's posting the multiplier at the factory 13.5x setting. I'm wondering if the manual jumps I made on the FID pins actually even took effect because never once did I get a posting of anything other than 13.5x. That may be why I was having booting problems but if that's the case, I have no idea why the multipliers won't post correctly unless there is some lock on them of some sort. I'm completely baffled on that one. I jumped them correctly (according to the chart I have from ASRock) but 13.5x is all it will post at. And my mobo monitor app shows it at 13.5x every time as well.