800 tbird oc only to 880?

Palpatine

Distinguished
Apr 4, 2001
199
0
18,680
I bought a Tbird 800 and a MSI Turbo-R mobo a few days ago. I cool a chip with a fop38. I tried to overclock it from bios and the best thing i could do is 8 x 110. After 880 the pc wont post. I used the fuzzy logic program that came with the mobo and it oc the cpu at 8 x 118 before it froze giving me 944. What's the problem? Can it go beyond 900? I leave the vcore at auto (1.70). Even though i put the vcore at 1.85 at 8 x 112 the pc wouldnt post again. The temps at 8x110 @1.70 are 43C cpu and 28 case.
Also is there a way to lower my mobo multiplier? It ranges from x8 to x15. And the l1 bridges prevent from raising the multiplier or from changing it (can i put it at x7 without connecting the l1?)

Thanks for your replies
Palpatine
 
G

Guest

Guest
Maybe you have a bum chip? I would think you could get at least 900 out of it. What kind of memory do you have?

To change the multiplier, you have to connect the L1 bridges, no way around that...

<font color=blue>The #1 reason to upgrade your PC - to run faster benchmarks...</font color=blue>
 

Palpatine

Distinguished
Apr 4, 2001
199
0
18,680
No but there is not much to gain even if i do. The minimum multiplier the MSI turbo-r gives is x8 so i prefer raising the fsb so at to improve overall system performance.

When i was trying to place the first fan, a thermaltake super orb, i chipped a little one side of the CPU die. When i took out the fan due to lousy temps (55C idle) i saw that the bottom side of the fan (the one that contacts the cpu) was scratched by the die!!!!!! I thank god that the cpu is still working!!!!!!!! Don't buy Thermatake coolers because they are impossible to install. Now i have a Global win FOP38 which i plan to replace with a silverado because of the noise... :(

Palpatine
 

agpport

Distinguished
Jan 23, 2001
115
0
18,680
Your problem is probably that when raising the FSB speed you also raise the speed at which the PCI-bus is operating.

Now I don´t know about MSI but with Abit the PCI-bus is 1/3 of the FSB when FSB is 100-133 MHz. Above 133 the PCI is 1/4 of FSB.

So, FSB:110=PCI:36,6. FSB:118=PCI:=39,3.

PCI-cards are designed for 33 MHz and often don´t like to be run at faster speeds. 36 is most likely OK, but 39 is a bit high. So it could be that one of your PCI-cards is setting the limit here.

Are you using PC100 memory? Memory is run at FSB-speed and PC100 will most certainly not be able to deal with 118 MHz FSB.

The L1´s prevent you from changing the multiplier. You´ll have to unlock them to change the mult. whether up or down.

You should absolutely try to run the FSB at 100 or 133 MHz to avoid trouble with out-of-spec PCI-buses. Try aiming for 7,5*133 for 997 MHz. Or anything-above-8*100 if you have PC100 SDRAM.

My Abit KT7A+Duron 750+PC133 wont complete 3DMark2001 at 7,5*107=800 but works like a charm at 6,5*133=866.


Hope this helps & good luck!
 

agpport

Distinguished
Jan 23, 2001
115
0
18,680
Sorry I missed that about the minimum multiplier being 8...
But still keep the FSB @100 or 133.
The PCI-bus also control the HDD in some way, don´t know exactly how though.

Yeah the Orb´s suck! chipped my duron too! Is there some possibility that a CPU runs hotter with a chipped core? I mean there´s still SOME less heat being transferred to the heatsink with the smaller area touching the heatsink...

I use my Chrome Orb for a candlestick holder, they do look cool after all...
 

Palpatine

Distinguished
Apr 4, 2001
199
0
18,680
I have one 256 dimm 133 and one 128 dimm 100. I can set the memory clock at 100 or 133 and i currentlu run it at 100. The pci is at 36.3 and the agp at 72,7. I shall try to run it at 8*133 see if this brick will post. Cross your fingers!!
 

peteb

Distinguished
Feb 14, 2001
2,584
0
20,780
I'd bin that crappy 128PC100 as well. It's holding you back - do you need the extra 128Mb right now?

-* This Space For Rent *-
email for application details
 

Palpatine

Distinguished
Apr 4, 2001
199
0
18,680
Well game over for me... :(

I destroyed my mobo yesterday. I was tweaking my bios at 03:00 last night and due to lack of sleep i booted the system with the CMOS jumper to clear data. Now everything is dead.

But before that i tried to boot with 1.85vcore at 8x132 and at 1.85vcore and fsb at 133 (i need to change a jumper to go from 100 to 133 fsb)

Also is the 128/100 a drawback. I can set the dimm clock at 100/133/auto in bios (or at least could before the meltdown). The 256/133 is better than 392 mixed 100and 133? I play flight sims and i think i need the extra 128.

Another thing. After the CMOS thing i tried to boot but everything is dead. Is this the mobo or maybe it is the power supplier? Does it worth replacing the supplier and see (I have one extra lying around)

Palpatine
 

peteb

Distinguished
Feb 14, 2001
2,584
0
20,780
and you have all the small switch and reset connector from the case connected?

It sounds like your psu or mobo has definitely kicked the bucket....

-* This Space For Rent *-
email for application details
 

Palpatine

Distinguished
Apr 4, 2001
199
0
18,680
The switches are connected. I pretty sure that the mobo is dead because the pc was working (trying to post) with the fsb experiments. After the CMOS everything was dead. My supply is 350W. And i prefer the psu to be dead than the mobo because its cheaper.
 

jg38141

Distinguished
Dec 31, 2007
832
0
18,980
Yeah I have to dissagree with you on the no gain for unlocking the L1 bridges. Makes no sense. I have 3 800's all running at 1ghz or 1050 and it does make a difference. You can still up over all with the fsb even after upping the mutiplier, but my nic card shuts down when the pci bus hits over 34.

My Jesus is whiter than your Jesus.
 

Palpatine

Distinguished
Apr 4, 2001
199
0
18,680
My dead motherboard's minimum multiplier is x8. The factory settings of 800 tbird is 8x100. As i read in the posts here the pci is 1/3 of the fsb from 100 to 132, then resets at 133 (is again at 33) and starts to climb again. So 132 is PCI @44 and 133 is PCI @33. The best thing is to cross the l1 and lower the multiplier to x7 or x7.5 and set the fsb to 133. Mine is stuck to x8 lower with l1 crossed or not. So i can overclock as far as my pci can handle. At fsb 110 and PCI @ 36,3 the pc posts at 112 does not. Even in 110 sometimes (especially when i reboot) the pc makes 10-15 secs to start (fans are working). So i think the 110 is the critical number. I bought a shitty fan yesterday, a spire (www.spire.nl) because i can't stand the noise of the fop38 and i am going to run with it even if temps are 55C until i get a silverado (two months if i am lucky).

The bottom line is dont experiment too much if you dont have the money to support your tests.

Thank you all for your replies
Palpatine
 

Makaveli

Splendid
HEy I have the same board. U need the overclocking Bios from MSi's site to get the lower multipliers. And also
there is a jumper on the board below the first memory slot
u must remove it to get the 133 FSB! I'd try unplugging everything let it sit for a couple hours then reconnect and try to reboot if not the board is dead but remember my settings above!
 

Makaveli

Splendid
Another thing if u get it up and running with the O/C bios
U'll have to leave the 100 fsb jumper on lower ur multiplier to 7 save and exit. Shutdown the PC remove the FSB jumper. Then when u reboot ur chip should be @ 133x7
and your 800 T-bird should be running 1000mhz now!
 

Palpatine

Distinguished
Apr 4, 2001
199
0
18,680
I see two different bios, the 2.7 and the 10b16 for the turbo-r. Does the 10b16 and the 2.7 have the same fixes? Or is the 10b16 in a completely different path? I see useful fixes for both bios and i dont know which one to pick.

Thanks
Palpatine
 

Makaveli

Splendid
the 2.6 bios is a newer bios and not beta so it would probably be the best choice for stability the overclocker bios is for overclocking so if u plan on overclocking u will get best results with that one. But it is BETA so don't forget that.
 
G

Guest

Guest
Hi there . I have MSI K7Tpro 2 a and the only bios ver which let me to lower the multiplier was the last beta 10b16. The other versions 2.4,2.5,2.6 & 2.7 couldn't lower the multiplier. Warning ! If u are plannig to update your bios with 10b16 read this (this was for my motherboard)
-----------
Special Note:
During BIOS flashing procedure on K7T Pro2 or K7T Pro2A. Screen will show "The program file's BIOS lock string does not match with your system"
answer : Use /nbl parameter
For example: awdfl789 w6330vms.240 /nbl
----------------
I don't know if this is the same thing on your motherboard ?!
----------------

Also standard bios can be updated online with msi live bios from msi site.
Good luck !
 
G

Guest

Guest
Dead motherboard on MSI K7T Raid, possible, but might not be dead after all. Happened to me and a few friends quite a few times.

Hope might still be around. Try the following thing.
Unplug all you hard drive and stuff and leave just the 256MB RAM, video card and power supply, keyboard and mice (optional) only.

First of all, make sure that you set the jumper next to the agp slot is set to 100, not 133.
Now for the most obvious thing, make sure that your cmos jumber is set correctly as this is the biggest cause for no power (dead). Looking at the mb from the front side (where slots/keyboard/etc) is away from you, the jumper should have 3 pin, make sure that is connected to the 2 pin on the right side, not the 2 on the left side (system will not come up when you have it on that). Another thing you can try which I do all the time because this mb tends to go out because of a bad setting is pull the jumper off and plug it on, you should instantly see the computer take off. Sometimes it's good to wait 5 to 10 minutes after a bad setting.

That's what I have discover working with MSI mb. Abit mb don't have this problem as it will come back after 30sec or so.

Consult your mb manual if you are not sure what jumpers I am talking about, but it's very important that you put it in the right place.

There's always hope at the end of the light!!!!!!!!!!