VOGONS


First post, by retro games 100

User metadata
Rank l33t
Rank
l33t

I'd like to see if my Asus P2B i440BX chipset based mobo can handle 133 FSB. In it, there's a Voodoo3 AGP video card, a couple of sticks of CL2 PC-133 SDRAM, and a Powerleap adapter with a Celeron 1.4 Ghz processor. If it set the mobo's jumpers to 133 (from 100), what components could get damaged?

Say if the mobo POSTs OK, and I can get to the DOS prompt, then run a game for a few minutes, then power off the PC. What harm could this do? Is the only problem extra heat? Or are there other components that could deteriorate, given the OC'd FSB? Could the capacitors be put under extra strain? When I say capacitors, I mean the ones on the mobo, and on the Powerleap adapter, and also on the video card! Would the PSU be put under any greater pressure? (I'm using a good one, and so there shouldn't be a problem providing the extra "juice".) Thanks for any info.

Edit: Also, the mobo's IDE ports have a compact flash disk for the HDD, and a DVD-ROM drive.

Last edited by retro games 100 on 2010-09-12, 13:32. Edited 1 time in total.

Reply 1 of 14, by Mau1wurf1977

User metadata
Rank l33t++
Rank
l33t++

It will virtually do zero harm, but it might not be fully stable...

My website with reviews, demos, drivers, tutorials and more...
My YouTube channel

Reply 2 of 14, by Old Thrashbarg

User metadata
Rank Oldbie
Rank
Oldbie

It's very unlikely that it will be anywhere close to stable... 1.8ghz is a bit much for most Tualatins, without a significant voltage boost and some good cooling.

But just generally speaking, at 133fsb, everything else should be OK. Voodoo3 cards are usually fine with an 89mhz AGP speed, and the PCI will still be running at the standard 33mhz, so no problems there. As far as extra strain on the VRMs and PSU, it won't be anything to worry about so long as the components are in good condition... it's not that big of a difference in power draw, and the P2B was one of those 'overkill' designed boards anyway.

If you wanna run a higher bus speed, I'd suggest looking at a 1.2ghz Celly. 1.6ghz would be much more attainable, still probably need a minor voltage bump, but a quick search tells me a lot of people had luck with 1.65 or 1.7v.

Reply 3 of 14, by ux-3

User metadata
Rank Oldbie
Rank
Oldbie

Chances are, nohing happens. However, this stuff isn't pristine any more, is it? So who knows? Caps may blow, the AGP gets a heavy OC on a BX board. The PSU won't care what you fry with it. If something gives way, it may blow a fuse.

On the other hand, why bother with a valuable antiquity?

Retro PC warning: The things you own end up owning you.

Reply 4 of 14, by retro games 100

User metadata
Rank l33t
Rank
l33t

Thanks very much people.

Old Thrashbarg wrote:

If you wanna run a higher bus speed, I'd suggest looking at a 1.2ghz Celly. 1.6ghz would be much more attainable, still probably need a minor voltage bump, but a quick search tells me a lot of people had luck with 1.65 or 1.7v.

I've just spotted a 1.2ghz (100 FSB) Celly on ebay, with a CPU code of SL68P. I notice that cpu-world.com says it's just 1.5v:

http://www.cpu-world.com/sspec/SL/SL68P.html

In terms of successfuly OC'ing, is this a good or a bad thing? I appreciate your comment about 1.65v and 1.7v OC'ing, but I wondered if a 1.5v CPU would be even better? This is a complete guess on my part. If I'm wrong, I'll simply hunt about on ebay for either a 1.65v or a 1.7v 1.2Ghz Celeron. Thanks a lot.

Reply 5 of 14, by retro games 100

User metadata
Rank l33t
Rank
l33t

I put a socket 370 Pentium 3 1000 Ghz CPU (SL5QV, 100 FSB) in to the Powerleap adapter. Its voltage is 1.75v, according to cpu-world com. I set the mobo's jumpers to 133 (I also tried 124), but I couldn't get it to POST. There seemed to be some kind of "power issue". When I switched on the mains power at the wall socket, the Powerleap adapter's green light came on, and the PSU began working. Normally, I have to press the power button connected to the mobo's power pins to get the mobo to work.

I did try bumping up the Powerleap's voltage by 1 notch, but that didn't help. Incidentally, with this CPU requiring 1.75v, there's only 1 more notch you can increase the Powerleap adapter's voltage jumpers to, and that's 1.8v. There's also a couple of "power jumpers" on the mobo labelled VCore and VIO. I guess the VCore jumpers can be ignored, because that's for a CPU without the Powerleap adapter. I tried setting the VIO jumper to its alternate setting (presumably "more juice"), but that didn't help. Perhaps this 1 Ghz Coppermine cannot tolerate being OC'd to either 124 FSB or 133 FSB, and the mobo simply won't POST.

Reply 6 of 14, by Tetrium

User metadata
Rank l33t++
Rank
l33t++
retro games 100 wrote:
Thanks very much people. […]
Show full quote

Thanks very much people.

Old Thrashbarg wrote:

If you wanna run a higher bus speed, I'd suggest looking at a 1.2ghz Celly. 1.6ghz would be much more attainable, still probably need a minor voltage bump, but a quick search tells me a lot of people had luck with 1.65 or 1.7v.

I've just spotted a 1.2ghz (100 FSB) Celly on ebay, with a CPU code of SL68P. I notice that cpu-world.com says it's just 1.5v:

http://www.cpu-world.com/sspec/SL/SL68P.html

In terms of successfuly OC'ing, is this a good or a bad thing? I appreciate your comment about 1.65v and 1.7v OC'ing, but I wondered if a 1.5v CPU would be even better? This is a complete guess on my part. If I'm wrong, I'll simply hunt about on ebay for either a 1.65v or a 1.7v 1.2Ghz Celeron. Thanks a lot.

There are better steppings. Try getting one with the 'B' in the stepping code.

Reply 7 of 14, by retro games 100

User metadata
Rank l33t
Rank
l33t
Tetrium wrote:
retro games 100 wrote:

There are better steppings. Try getting one with the 'B' in the stepping code.

Thanks. BTW, I seem to be having some power related problems with my mobo, after my unsuccessful OC'ing experiments. If I switch off power, then wait a bit, then switch on power at the wall mains plug, the PSU works immediately and the Powerleap's green light switches on. I get no activity from the mobo, and also no POST. I then have to hold down the mobo's power button for about 4 seconds, at which point the power is cut off. I can then switch on power to the mobo using the mobo's power button, and I finally get a successful POST. This process keeps repeating itself after I switch off the PC. Any ideas please people? Thanks a lot.

Reply 8 of 14, by retro games 100

User metadata
Rank l33t
Rank
l33t

Solved: I have solved the issue regarding the power not switching on and off correctly. For testing purposes, I always use two PSUs. I find having two PSUs useful, because you've got double the amount of connectors, and it's sometimes useful to spread the power load over two PSUs, for extra stability. Initially, I had the Powerleap's power cable plugged in to the PSU which is not connected to the mobo, but "artificially" switched on using the "paper clip method". The paper clip method is where you stick a paperclip in to two of the PSU power connector plug's holes, to force it to power itself on, even if it is not connected to a mobo. I solved the mobo's power problem by disconnecting the Powerleap's power cable from this PSU, and attaching it to the other PSU: the one which is connected to the mobo. Now, the PC doesn't appear to get "upset" that I have arranged power distribution in this manner, and the PC behaves itself correctly, when I switch it on and switch it off. Thank you very much for reading this dull message.

Reply 9 of 14, by Old Thrashbarg

User metadata
Rank Oldbie
Rank
Oldbie

I appreciate your comment about 1.65v and 1.7v OC'ing, but I wondered if a 1.5v CPU would be even better? This is a complete guess on my part. If I'm wrong, I'll simply hunt about on ebay for either a 1.65v or a 1.7v 1.2Ghz Celeron.

I meant that you'd probably have to bump the voltage up from the ~1.5V stock to somewhere in the range of 1.65 or 1.7V to get it stable at 1.6ghz. You may be able to get by with less, depending on the chip, but even 1.7V is not too bad on that architecture, as long as you're not using a cheesy cooler.

Reply 10 of 14, by Tetrium

User metadata
Rank l33t++
Rank
l33t++
retro games 100 wrote:

I've just spotted a 1.2ghz (100 FSB) Celly on ebay, with a CPU code of SL68P. I notice that cpu-world.com says it's just 1.5v

I forgot to reply to this. It's lower voltage then the 1Ghz parts because it's a newer fabrication process. The 1.2 is a Tualatin and the 1.0 a Coppermine.
At 1Ghz the Coppermine is already close to it's maximum limit while the Tualatin could clock to 1.4Ghz and above with relative ease. The 'B' stepping should be even slightly better as it's newer and Tualarons should clock a lil higher then proper Tualatins since Tualeron cache is artificially slowed a bit to make it more crippled compared to the (Tualatin) Pentium 3 line.

Reply 11 of 14, by Mau1wurf1977

User metadata
Rank l33t++
Rank
l33t++

Retro Games 100, just one more thing.

My cpu actually has a FSB of 133. So the CPU doesn't get overclocked. But I need to overclock the FSB for the cpu to run at its full speed.

I guess that's what I meant with "every 440BX board should do 133 FSB". I didn't mean every 100 FSB cpu will be able to do the same thing 😁

My website with reviews, demos, drivers, tutorials and more...
My YouTube channel

Reply 13 of 14, by retro games 100

User metadata
Rank l33t
Rank
l33t
Old Thrashbarg wrote:

I appreciate your comment about 1.65v and 1.7v OC'ing, but I wondered if a 1.5v CPU would be even better? This is a complete guess on my part. If I'm wrong, I'll simply hunt about on ebay for either a 1.65v or a 1.7v 1.2Ghz Celeron.

I meant that you'd probably have to bump the voltage up from the ~1.5V stock to somewhere in the range of 1.65 or 1.7V to get it stable at 1.6ghz. You may be able to get by with less, depending on the chip, but even 1.7V is not too bad on that architecture, as long as you're not using a cheesy cooler.

Success! I have successfully over clocked my Asus P2B board to 133 FSB, with a Celeron 1200 GHz (100 FSB) processor. Its real clock speed is now running at 1595 MHz. I have more details in this thread, at the bottom of page 2.