Is it possible to run a Duron with 133fsb?

[Assimilator1]

Good luck

Btw I found my bro's mbrd ,an Asus A7V133 rev1.05 was also overvolting in auto mode (1.75v!) ,so I set it to 1.65v manually which gave 1.64v.
Hopefully it'll still run at 1266MHz ok.(didnt have time to stability test it)

 

[yodayoda]

Originally posted by: biggiesmallz

Originally posted by: yodayoda
using a $20 duron 600 i picked up in the for sale forum, i hit 6 x 133 = 933 no problem at 1.75V. would also hit 1 GHz via 6.5 x 133 but i did not like to push it there (needed 1.85V). durons are really easy to unlock if you have to lower the multiplier as they are on ceramic--the old pencil trick works on them.

Your numbers don't add up. 6x133=798 and 6.5x133=864.5 according to my calulations.

You must have got them numbers from a Pentium based computer!

agh, 7 x 133 and 7.5 x 133... come on, i haven't run that chip in over a year....

 

[MadTom]

Originally posted by: Assimilator1
Good luck

Well, I think I need it...

I tried to do everything what was necessary to overclock, but made just stupid things...

First, when the Duron didn't work at default voltage on higher speeds (8x133) I tried to update the BIOS on the motherboard (took the right one from MSI's site), but I crapped the mobo. It came to some 60% of the BIOS installation process and it stayed there for more than half an hour... After the restart I got a BIOS CHECKSUM ERROR and that was it for the MS-6330 Lite... (RIP). So, I decided not to be lazy and went to my closest retailer where I bought an QDI Legend Kinetiz 7E-A, since this was the only motherboard nearby with jumper settings for all important things (Vcore, multiplier, FSB...).

Well, I again tried to put the untouched Duron in the mobo. At default settings, everything ran fine (although the board is overvolting a bit...). So, I put Vcore to max (1.85V, actual 1.9V), removed the 100/133 FSB jumper and tried to start the computer - beep, beep, beep, beep... OK, obviously 1,07 GHz is too much for this baby. Took out the processor, connected all the L1 bridges with a defogger kit. Set the multiplier to 7 and the computer booted happily at 700 MHz. OK, tried to remove the 100/133 jumper again - beep, beep, beep... All with max Vcore. OK, I thought, it obviously isn't going to be overclocked. So maybe I should try 6x133 (what corresponds to 800 MHz) and the computer didn't boot again (the beep story). However it didn't boot at 6x100 as well. But, it didn't boot at 9x100 either.

I guess there are the following possibilites:
1. The RAM is crappy (although bought few months ago and said to be PC-133 RAM, I'm not sure if it works at that speed...
2. I haven't unlocked the Duron correctly
3. The mobo isn't so good as I expected
4. I'm stupid

Well, it's probably the last one

If anyone has some true PC-133 lying around and doesn't need it he can send it to me... And, the onboard sound is not working, too (well, will first reinstall Windows to see what then will happen)

 

[Pilsnerpete]

I'm thinking your ram is ok, but definitely check your unlocking job. Did you do it with Scotch tape to isolate one set of contacts at a time? Me neither. It's easier if you do though.

I couldn't boot with 5.5 or lower. I'm gonna try to change the default settings, though I don't know if that'll help cause it boots fine at 6x154. I might just put this in a machine I'm gonna build for a pal and get some money for it.

 

[Assimilator1]

MadTom
Sounds like you haven't unlocked your cpu properly ,especially if the multipliers don't work at 100MHz FSB either.
If they work at 100MHz but not 133MHz (the lower ones that is) then you mbrd has the problem mentioned early in this thread where it boots at the default multiplier for the 1st few seconds before using the one chosen in the bios.

Btw if you've got a steady hand & a mag glass then you don't have to use scotch tape.

 

[MadTom]

Well, as mentioned 7x100 worked... But 6x, 9x or others didn't. However, I would be happy with getting 7x133 (933 MHz) what should be possible. And the RAM is OK - it is running async with 133MHz and that's fine (CL2 or CL3, doesn't matter)...

Will try to change BIOS today, since I get problems with my RADEON 7500 in this box.

 

[Assimilator1]

Yeah your cpu isn't properly unlocked ,933 may be achievable though

 

[MadTom]

Originally posted by: Assimilator1
Yeah your cpu isn't properly unlocked, 933 may be achievable though

I hope, too. However I maxed the voltage and was curios how I could achieve 7x but not any other (lower) multi. So I checked the QDI Kinetiz 7E-A manual and on page 16 (30 of 60) the corresponding multiplier table. As it seems J9 has the same function as the L6 bridges (more info here) and J19 probably takes over some functions of the L3 and L4 bridges. However after checking the multiplier table I found out that, in theory, the only 2 multipliers I could achieve (from original 8x) with just closing one or two more jumpers are 7x and 11x. So I maxed the voltage and tried 11x. Strangely it posted at 1100MHz, but it couldn't load Windows. OK I said, there obviously is a chance to overclock, then.

I would be happy with getting the 7x 133MHz to work and then raise the FSB to around 140 MHz. I think that would be close enough to 1GHz... Still, it doesn't want to boot or even POST at 7x133... I have read on a page that for this mobo, to really unlock the processor, it is necessary to paint (connect) all L1, L2, L3 and L4 bridges. So, I probably will paint the L3/L4 bridges, but for the beginning won't touch the L2, as I haven't found any real info about it's function.

Will keep you informed.

 

[Assimilator1]

Don't touch your L2 bridges you'll reduce your L2 cache size!

And you don't just randomly connect all your L3 & L4 bridges ,god knows what you'd end up with:Q
The L3 & L4 bridges set your CPUs default multiplier ,for you to change it from x8 to x7 would involve cutting the 4th L3 bridge (& closing a few others) ,I wouldn't recommend cutting bridges if you can help it ,not easy & rather risky!.

Look here to see a table of multipliers settings

Your best off getting the L1 bridges to work if you can

Btw will it boot at a FSB between 100-133? ,say 115MHz.

[update]
I've just had a look at the manual & the page you refer to,J9 & J19 do not take over L6,3&4 functions ,these jumpers select a multiplier IF the cpu is unlocked ,in other words these jumpers 'control' the L1 bridges.
Also your mbrd has a 100/133MHz jumper ,AFAIK all mbrds that have this will boot at the default multiplier for the first few seconds before the bios over rides it with your chosen setting.
So what is happening is that your cpu is trying to boot at 8x133=1064 intially ,I think you said you had tried with max voltage too so it seems your Duron won't run at that speed even briefly (don't get how it posted at 1100MHz though! )

To get a full description of this problem read this article by Anandtech

[update2]
2nd thoughts ,scrap the L3-4 bridge changing just alter the L6 bridges.
In fact all you have to do is to join the 2nd L6 bridge & that changes the default multiplier to x7

 

[MadTom]

Originally posted by: Assimilator1
Don't touch your L2 bridges you'll reduce your L2 cache size!

And you don't just randomly connect all your L3 & L4 bridges ,god knows what you'd end up with:Q
The L3 & L4 bridges set your CPUs default multiplier ,for you to change it from x8 to x7 would involve cutting the 4th L3 bridge (& closing a few others) ,I wouldn't recommend cutting bridges if you can help it ,not easy & rather risky!.

Look here to see a table of multipliers settings

Your best off getting the L1 bridges to work if you can

Btw will it boot at a FSB between 100-133? ,say 115MHz.

[update]
I've just had a look at the manual & the page you refer to,J9 & J19 do not take over L6,3&4 functions ,these jumpers select a multiplier IF the cpu is unlocked ,in other words these jumpers 'control' the L1 bridges.
Also your mbrd has a 100/133MHz jumper ,AFAIK all mbrds that have this will boot at the default multiplier for the first few seconds before the bios over rides it with your chosen setting.
So what is happening is that your cpu is trying to boot at 8x133=1064 intially ,I think you said you had tried with max voltage too so it seems your Duron won't run at that speed even briefly (don't get how it posted at 1100MHz though! )

To get a full description of this problem read this article by Anandtech

[update2]
2nd thoughts ,scrap the L3-4 bridge changing just alter the L6 bridges.
In fact all you have to do is to join the 2nd L6 bridge & that changes the default multiplier to x7

Thanks for all the info, but I've read through all the info at anandtech, Tomshardware and so on, even before I started my overclocking journey...

For sure I won't touch the L2 bridges. Regarding the L3 and L4 bridges I probably will give it a try (got this Duron for free, so if it goes up in smoke I will survive it...) since I found few pages which mention this issue (one in Czech, one in Serbian - since I'm Croatian I can read and understand it...) where they managed to get all multipliers by connecting all L1, L3, L4 bridges...

It boots without problems (and without raising voltage) at 8x112 MHz (just kept it there since I didn't want to cross the 35 MHz on the PCI.

Don't know when I'll have time to play around with the Duron, since it's my wifes computer and I have to find a time when she is out and I'm home... BTW, will keep nail polish handy...

Thanks again, Assimilator1

 

[Tectron]

I have my 1100 running 166 X 7 @ 1.7v on a msi 6561 (745 Ultra). I dont even have a symptom of instability, Even after looping 3dmark for hours. I get 42C under load using a sk-7 with a quiet 31cfm fan.

 

[Assimilator1]

Nice

MadTom
I take it you missed my last line
Don't bother with the L3 & L4 bridges ,do the L6 bridge instead ,no cutting then

 

[Pilsnerpete]

Gotta get my lower multis working so I can beat that! 166fsb with a duron? Crazy.

 

[MadTom]

Originally posted by: Assimilator1
Nice

MadTom
I take it you missed my last line
Don't bother with the L3 & L4 bridges ,do the L6 bridge instead ,no cutting then

I have understood everything you were talking about. However, I was never talking about cutting and playing around with the L3, L4 bridges, just about making them all connected (by painting them). It seems that there is a problem with the KinetiZ boards (overclocking features not implemented correctly), and what I'm reffering to is:

1. See KinetiZ 7T down there
2. See text: "zavøít piny L1,L3,L4" - means: close pins L1, L3, L4]http://svethardware.cz/sh/sh1.nsf/0/191CBCD5BEB911BDC1256960003C872C[/L][/L]
3. Same thing here
4. And here

Although those are Czech and Serbian sites I believe it can be done. However, I probably will have enough time to try it out this weekend, so we will know.

 

[Assimilator1]

I can't read Serbian

But from what I've read at Anandtechs & Toms HW do not join all L3 & L4 bridges ,there is no cpu that exists that has all the L3&4 bridges closed.
Do the L6 bridge & then you can join the L1 bridges

 

[calvink]

Ran a Duron 1000 10x133 = 1333 in a crappy ECS K7S5A mobo, stable with no voltage increase.

 

[MadTom]

Originally posted by: Assimilator1
I can't read Serbian

But from what I've read at Anandtechs & Toms HW do not join all L3 & L4 bridges ,there is no cpu that exists that has all the L3&4 bridges closed.
Do the L6 bridge & then you can join the L1 bridges

Tried everything. Nothing works for me. Will it leave as is. I guess it's more of a mobo problem.

 

[Assimilator1]

Wierd ,L6 bridge trick should work as that is how the default clock is set (along with L3&4)

 

[cow123]

im running a 1.3ghz morgan duron with a 133 fsb still locked.... well i had to change the multiplier on the processor to 10

probly one of the riskiest things ive done with comp stuff, getting a power supply/wires/needles and breaking some bridges on the processor + pencilling others