Weird Multiplier Experiences

[hunter]

On an Abit KT7A, PC150 ram, super cooling, I ran my 1.33 gig 266 fsb. Cpu , step by step all the way to 10x150=1.5 gig , cas 2 , fast CPU command , CPU drive strength 3, and it is totally stable. Idle temp, 31C. Under load , 46C. Now the weird part is that if I try a multiplier of 10.5 or 11, at a lower FSB and a lower total MHZ, windows will load , and at the desktop the comp will lockup. It is almost like the CPU doesn't like any multiplier but "10" . Is there something wrong w/ the CPU's "supposedly unlocked " multiplier from the factory? It runs good at 1500 mhz and a core voltage of 1.75, I/O 3.5. I tried upping the voltage and it didnt help.

[asmodean]

I've had the same kind of problems... I'm running my tb900 now @8x142. I was able to run 9x126, but it freezes and stuff happens.

Try cleaning the chip (not just the core, all the bridges). Solved one of my problems partially (max. core voltage dropping to 1.725). Now it keeps the voltage at 1.85 (shows 1.91), but bios still reports occasionally a maximum settable vcore of 1.725.

[fallout55]

I had the same problems with the KT7A. I swapped it in for an Iwill KK266 and my Tbird changes multipliers fine now. I think it's the mobo.

[asmodean]

fallout55 (Apr 17, 2001 03:10 a.m.):
I had the same problems with the KT7A. I swapped it in for an Iwill KK266 and my Tbird changes multipliers fine now. I think it's the mobo.

Could be. I was asking about my voltage problem here, and thought it was my crappy 230W PSU. It wasn't, as I got my new Enermax 430W, the problem persisted.

I decided to unlock my tbird permanently with silver paint when I got the Enermax and cleaned the tbirds bridges (and almost everything on the chip) with petroleum (Turpentine, is it?). Solved a part of the problem. But anyway, somebody suggested I have a bum voltage regulator on the mobo. This was before the perma-unlock.

At the moment, I'm not sure which one "sucks", my tbird or my mobo. Or it could be both.

[Hoot]

You have just discovered "The Wall" for your system. If you had a problem with your multiplier unlock job, you would not get as far as you do at 10.5. My 1.0G/266 Tbird does just fine up to 10x150 2-2-2, but refuses to run at 10.5 unless I back off on the FSB quite a bit. That's "The Wall" for my system also. I say system because it is the interelationship of your cpu, mobo, memory, cards, PSU and cooling solution combined. "The Wall" is not insurmountable, but typically it takes a lot more energy to get over it than it took to reach it. Don't be surprised if you have to up the Vcore 4 or more .025V steps to get over. That causes all kinds of ripple effect. The larger increase of Vcore now puts your cpu wattage up in a zone where your cooling soltion may no longer be adequate and maybe your PSU also. You have to decide whether to throw more money at the system or be content to work on the near side of "The Wall".

Hoot

[fog]

ahh, i understand, "the wall" is some kind of pink floyd thing right ? :)

[dem1an]

this is my weird multiplier experience with my 1.33 and kk266-R

it seems to be a problem with multiplier in the bios. if i change it to anything but auto it borks up multiplier. for instance if i change it from auto, it uses 8.5 multiplier. for instance...any setting x 133 = 1133 or any multiplier x 145 = 1233. for example, 10x multiplier yeilds 1233 @ 145fsb and 1133 @ 133 fsb. not even the 10x multiplier works properly. ideas guys???

btw, i'm running stable at max load temp of 38C at 1530mhz (auto x 153fsb) after burning in this cpu...still using the vantec. i've tried backing off significantly on the multiplier...same results. is this how a cpu behaves if it's not unlocked?

[dem1an]

sheet...just found this faq on iwill.

http://support.iwill.net/faq/kk266.htm

[supraway]

HAHAHAHA I love the second faq on the page... Maybe Iwill has a sense of humor?