Updating FSB?

damnyanks

Hey,

I've a AMD 2500 Barton and a Asus A7n8X Deluxe mother board. The current listed clockspeed of hte CPU is 1.1 ghz and FSB rate in the bios is set at 100.

Anytime I try to increase this the computer either stops booting up fully or crashes constantly. Does anyone know if its a heating or power issue or what?

The machine also has a Radeon 9600 and 1gig of 3200 ddr ram.

Thanks for any help!

unkel

That's desperate

What PSU do you have?

KdjaCL

Tried clearing CMOS and using optimal defaults?

Also make sure RAM is set to SPD for now and not AUTO or other.

kdjac

Col_Loki

Looks alot like power supply.

Try it again but use memory divider say 80% or 5:4 (FSB:RAM) , since its Pc3200 i doubt its the ram unless its faulty (memtest might be an option either).

If you can get to windows run asusprobe and monitor temps and voltages.......... run prime95 and you should either see the temps getting very high or the voltage rails dropping off.

damnyanks

Ok I'm not so good with all this so some questions

Just checked case spec. Its a 360w power supply (Chieftec?). That is too low ?

Tried clearing CMOS and using optimal defaults

Where would I do that ?

Try it again but use memory divider say 80% or 5:4 (FSB:RAM) , since its Pc3200 i doubt its the ram unless its faulty (memtest might be an option either).

My head hurts
How exactly would I do that?

Col_Loki

How exactly would I do that?

Done in the BIOS.

Where would I do that ?

Its a matter of changing a jumper on your motherboard , the manual should tell you where. I dont think this will help at all TBH as your system is prob stable before the FSB change.

That PSU should be fine, memory might be a problem. Try the divider and see if she will boot........then run prime95 (mabye memtest also) as above!!

Mutant_Fruit

first check the voltage going into your CPU. If it's less than about 1.6, then increase it to 1.6. Thats what its supposed to be at.

Then try increasing the FSB again. Also, try increasing it directly from 100 to 133. And if that works, straight to 166.

If niether works, see if you can test another processor in the mobo, or your processor in another mobo. Or just RMA the lot and let your supplier test it all.

jozi

sorry for changing the subject

Is the voltage for the barton 2500+ 1.6 volts? mines at 1.65 in the bios, lowest option.

The jumper for clearing the cmos is near the bat on my board, asus a7n8x-e.

jozi

Col_Loki

Is the voltage for the barton 2500+ 1.6 volts? mines at 1.65 in the bios, lowest option.

The stock voltage for the 2500's is 1.65v and not 1.6v so dont worry your fine.

If niether works, see if you can test another processor in the mobo, or your processor in another mobo. Or just RMA the lot and let your supplier test it all.

Thats a serious pain as he will have to do without his PC and your RMAing parts that are perfect. Atleast if he can isolate the problem mabye a spare CPU or Stick of memory will keep him going.

Try the couple of tests and see how you get on.

Mutant_Fruit

i couldn't remember if the stock voltage was 1.6v or 1.65v, so i said about 1.6v

Anyway, i had this exact problem before with an Epox motherboard, and it turned out the board was dodgy. I RMA'ed the board and processor, and the processor was grand.

damnyanks

The parts were bought before xmass so I think its too late to RMA ? Got it from overclockers.co.uk

Col_Loki

You should be fine for an RMA , just say youve been running at a lower speed than normally an Xp2500+ would and thats why you didnt notice the problem.

Did you run any tests? Check the temps in the BIOS after a reboot?

Mutant_Fruit

you should have 1 years warrenty, so RMA'ing won't be a problem.

damnyanks

I have ASUS prove working.

CPU Temp = 38C
Motherboard = 23C

This is jsut after playing Warcraft 3 (Which it constantly crashes out of!!!)

Col_Loki

Download memtest and Prime95. You can run these in windows without any problem and can watch the asus probe at the same time ........... links are here........


http://www.boards.ie/vbulletin/showthread.php?s=&threadid=57342

(memtest86 and prime95) - there very small programs btw.

What are your voltages like? ie 12v , 5v , 3.3v?

damnyanks

Also make sure RAM is set to SPD for now and not AUTO or other.

When I did that I got an error saying system overclocking or something along those lines and it would keep restarting. (From that voice of the chcik with the motherboard)


Downloaded both those programs. I don't have a floppy drive, I use USB Keys. Any idea how I'd get memtest to work properly then?

12V = 12.352
5V = 4.99
3V = 3.66
VCore = 1.68

TacT

Uninstall asus probe and read this page and follow the instructions to get MBM (motherboard monitor) installed to get some proper readings on your cpu diode temperature, instead of your cpu socket temperature, which is what you have been reading in asus probe/bios.

http://www.nforcershq.com/forum/viewtopic.php?t=16364

Your cpu is most likely overheating and/or your voltage rails are unstable, I can tell because WCIII is crashing.
You are reading your socket temperature readings in the bios or in asus probe, which is not the important bit. The diode temperature on your cpu is the important bit.
If your socket is running at 38-45 then your diode is running between 50-65+ which will crash you out, mine crashes out if it hits 61 on the diode.

What sort of HSF have you got on the cpu and any other case cooling, does the gfx card heat up, does the nothbridge chipset get warm?

My current temperatures at idle/load respectively are:-
diode 38/40-50/53
socket 27/29-35/38

case 21/23-25-28.

Previous to that the cpu had a hsf on it with crappy thermal paste hence, shooting up to 61 diode temps (in conjunction with poor case air flow) it now has arctic silver-5 and a nicer hsf with an 80mm fan gently taking hot air out alongside my psu which has managed to keep things going quietly and cool enough, I don't need it arctic cold for overclocking

Edit -- I don't like what asus did on the northbridge chipset with the heatsink and thermal pad underneath it as well as motherboard clips, they tend to hold heat those thermal pads do and I shall be removing mine

damnyanks

I got the retail 2500 barton so it uses whatever heatsink comes with that. I've only got the fans that came with the chieftec case (1 Side panel one). I'll go pick up some fans today see how that works out.

Never chceked the graphics card at all. I havent messed with any of the settings on that.

TacT

Put your finger smack bang in the middle of the gfx card gpu core and see if it's scorching hot or just warm after playing games. You might also put your finger to the top left of your HSF on the cpu and just under the PSU and take note of any trapped hot air there, if there is any and is the HSF warm or hot? I'm guessing the latter but you won't know till you run MBM or have a feel around in there yourself!

damnyanks

I bought 2 new fans today both are working and the airflow is in the right way.

Case = 32c CPU = 41c cpu Socker = 25c

Thats on the fsb 100 and sfd ram on auto.

Sound about right ?

Mutant_Fruit

FSB should be at 166 for a 2500+ barton.... try bumping it up now.

TacT

yeah, that's a bit better but still quite warm -- any luck hitting 166FSB?

damnyanks

133 and it buggers up.

What else could I do ? I bought 2 fans today so there are 3 in there now. I havent run the memory test on the ram yet (Need to get a blank cd, no floppy drive). Is there anyway to test if hte motherboard is dodgy without sending it back to the suppliers?

I dont have access to another AMD processer or motherboard so I cant put another one in to see how it works.

TacT

yikes, could be faulty, ram/mobo or psu. More likely one of the first two