[DEAD] Celeron 633 $17 @ Directron, free S/H

[AllPurposeNothing]

mindless1-

Switched out the cpus. Broke off the BSEL0 and VID2 pins on the 566 and wrapped the wire around the VID3 and VSS pins, put the CPU in and voila...it posted. BIOS showed it as a 850@100FSB.

Only problem is that thats about as far as it gets. In trying to load up WindowsMe, I never actually got to a workable point. Numerous errors would occur, from simple freezing on the partially loaded desktop, to inability to run program errors to system self-rebooting on start-up. So while it posts, its still unstable.

In taking the chip back out, somehow I broke the VSS pin once again. Just for the sake of argument, I got a new piece of wire, wrapped it around the VID0 and VID3 pins, but this time..it doesn't even post.

So my new questions are:

1. Is there likely a way to deal with the stability issue, or is this likely a voltage issue (too much or too little)?

2. Is there another option than the the VID0 and VID3 pins, given that they are not posting when tied together?

Thanks for your help.

 

[mindless1]

Originally posted by: AllPurposeNothing
mindless1-

Switched out the cpus. Broke off the BSEL0 and VID2 pins on the 566 and wrapped the wire around the VID3 and VSS pins, put the CPU in and voila...it posted. BIOS showed it as a 850@100FSB.

Only problem is that thats about as far as it gets. In trying to load up WindowsMe, I never actually got to a workable point. Numerous errors would occur, from simple freezing on the partially loaded desktop, to inability to run program errors to system self-rebooting on start-up. So while it posts, its still unstable.

In taking the chip back out, somehow I broke the VSS pin once again. Just for the sake of argument, I got a new piece of wire, wrapped it around the VID0 and VID3 pins, but this time..it doesn't even post.

So my new questions are:

1. Is there likely a way to deal with the stability issue, or is this likely a voltage issue (too much or too little)?

2. Is there another option than the the VID0 and VID3 pins, given that they are not posting when tied together?

Thanks for your help.

Breaking off the VID2 and wire-wrapping VID3 to VSS on a 1.5V default CPU will only result in 1.7V. Being a cB0, it most likely needs 1.8-1.85V. It is GOOD that it was running at all, even instabily, meaning the motherboard isn't going to be a problem, it's just that you needed more than 1.7V. Remember that each "mod" varies based on the default voltage it starts with. The reason that wire-wrapping VID0 and VID3 doesn't help now is because VID0 wasn't, isn't in use on the 1.5V default CPUs.

It is important at this point to be more gentle with the CPU pins... better to have the wire too loose and need to tighten it, than to bend the pin so it doesn't fit in the socket anymore and/or breaks off.

Looking again at Intel's VID chart we can study it and see that your Celery 566 started out with it's VID2 connected to VSS (ground), as evidenced by the "0" in that chart. You needed to break that pin off so it wouldn't be connected, turning the "0" into a "1" as is needed for 1.85V on the chart. Then you needed to wrap a wire around 4 pins, VID0, VID1, VID3, and VSS, making them all "0" as in the chart. A "0" in the chart means connected to ground, and for this exercise the VSS is ground.

However, since that nearest VSS pin is now broken off, you simply need to connect them to another VSS pin, instead. On the following linked picture I have greyed out the missing pins, just do as before, wrapping a wire between the marked pins, VID0, VID1, VID3, and VSS. Don't pull so hard on the wire, only a light contact is needed. Here is the pic.

 

[AllPurposeNothing]

mindless1-

Took some wire form the adapter and wrapped it as shown on the provided pic. Haven't installed the cpu yet, but yet another question:

The wire doesn't touch but is awfully close to touching both the TD0 and TD1 pins. I'm afraid that in installing the cpu, the wire may shift slightly. Is there any danger to the wire touching either or both of these pins on installation (other than a possible non-post)?

 

[mindless1]

AllPurposeNothing,
I advise using smallest wire possible, single-strand solid wire and wrapping it along the path shown in picture, in a "figure-8" around the pins (as much as possible given the multi-pin connection you're making). I cannot guarantee that it won't shift nor that if it touched those pins, it wouldn't be a problem or cause damage, I just don't know. If you think the wire needs redone, redo it. If you think you need smaller wire, PM me your address and I'll mail you some. Just be careful, take your time, and it should work.

 

[AllPurposeNothing]

Well, if you don't think there's any inherent danger to the cpu, I'll just try installing it. If it posts and is stable, great!

If not, I may have to take you up on the piece of wire offer.

If it should post and be stable, I'll tackle the 633 and see what happens. At least, I'll know I have an 850 if nothing else.

 

[kctopitz]

got the slocket today, left it set for 66MHz FSB to make sure the thing ran at 633 first, of course. installed the cpu and hsf, popped it in, powered on:

no video and (oddly) the system fan doesn't run when this thing is plugged in.

bad chip? bad slocket? only on arrival did i discover that the slocket i ordered is the Super Slocket III - the very one i was trying to avoid.

again, thus far i have done *no* mods to anything. i even tried switching the slocket to Auto FSB just in case that might help, but no luck. the board is an Intel SE440BX in a Dell R400, and from the info on (http://www.roberthancock.com/dell/xpsrproc.htm) it seems these systems have a good history of being able to take a processor well faster than this.

 

[mindless1]

kctopitz,
I don't know for certain why yours isn't working but I'd try the following:

- Set slotket to 1.8V
- Try newest BIOS, then 2nd, 3rd, 4th newest BIOS. My memory is vague about exactly what BIOS is needed.
- See a pic I provided earlier about modding a slotket, removing the RSV pin from CPU and soldering two pins on slotket. slotket mod

 

[kctopitz]

ahh, how i love computers sometimes...

switched to the previous bios, the new processor didn't work - same thing as before. put the old processor back in, it POSTS but then goes to a black screen. can't boot to the hard drive or even a floppy to go back to the newer bios.

so, for the time being i'm just working on getting my old system back up and running. if i set the board to recover mode i can get a floppy to boot, but it's not taking the new bios. could be that it doesn't work will Dell's and i have to switch to the intel bios, but once i do that i don't think i can go back so i'm trying to avoid that, especially as the intel bios disallows any processor over 450.

yeah, aren't these problems fun.

 

[mindless1]

I just checked a file archive, and the only notes and bios I have for a SE440BX is for the "SE440BX-2" board. For THAT board, Intel bios revisions prior to "13" will allow use of Coppermine CPU.

I don't remember how to tell a BX-2 from earlier or later BX though, but the BIOS ID string for what I have is "4S4EB2X0.86A.0019.P12" The "P12" is the revision number, and "86A" is Intel's vendor ID (substitute Dell's vendor ID for a Dell BIOS). So, if it's appropriate for your board the "4S4EB2X0" part, at least, should match the BIOS ID string of what you were already running.

As for going back and forth between Intel and Dell BIOS, if you do that recovery-mode thing where you move the jumper, I "thought" you could go back from either, to the other bios. I could be wrong though, time erodes my memory of it.

 

[mindless1]

 

[Lucky777a]

Can I run 1.1Ghz Celery FC-PGA2 on Asus P3B-F Mobo?

 

[kctopitz]

EDIT: got it going with an intel bios. ahh the joy of messing up your system and getting it back to normal. some reading on that Super-Slocket III suggests that it can be very problematic with any FC-PGA chip.

i am now unable to go back to the Dell bios though. can't do a bios recovery with Dell bios files because they lack the .rec file needed to do it - unless someone knows where i can find that?

i think i might look into a cheap fc-pga motherboard to try running this chip.

 

[VirtualLarry]

Originally posted by: kctopitz
EDIT: got it going with an intel bios. ahh the joy of messing up your system and getting it back to normal. some reading on that Super-Slocket III suggests that it can be very problematic with any FC-PGA chip.

Hmm, interesting. I haven't jumped on this deal yet, but I was thinking about it, I have a slot-1 OEM Motorola 440BX-based mobo, and a Super Slotket-III (ver 1.0) sitting here brand new. If it doesn't work with this slotket, I might reconsider. I don't believe that this slotket has any voltage-adjustment jumpers.

 

[DJMiX]

Anyone happen to have a link to intel's 440bx bios files?

 

[kctopitz]

Intel SE440BX Bios

 

[DJMiX]

Much Thanx, Fixed link to bios

 

[AllPurposeNothing]

An update-

With mindless1's help (both advice and some wire), I was able to successfully overclock the 566 Celeron chip to 850 mHz.

Not so lucky with the 633 chip, however. While it posts at 950 @ 1.85v, it is far from stable. Also, made a rather severe mistake in pin pulling (should remind myself to do this during the day and not the wee hours of the night), but it doesn't seem to have had a noticable effect on use of the CPU (at least, so far).

Will see what we can do with the 633. If it just won't hit 950 on my board, I'll offer the cpu for adoption. mindless1 gets first dibs, but if he passes on it, than the next taker will get it.

Stay tuned...

 

[Flatpicker]

AllPurposeNothing
Member

If mindless passes, I promise to give it a good home.
Please PM or email me.

 

[mindless1]

AllPurposeNothing,
It is good to hear that it works for you. You might want to run a lot of stability tests on the 566@850 CPU before giving away the 633 though, just in case it isn't completely stable then the 633 would still be your faster CPU.

I don't need the 633 though, but thanks for the offer. Looks like Flatpicker was in the right place at the right time.

 

[AllPurposeNothing]

mindless1

Is there a piece of software designed to run stability tests? I'm not exactly sure how to do a clinical test to test stability.

The fact that it didn't crash on me was good news to me.


Flatpicker

After running the stability tests run on the 566@850, assuming all goes well, I will PM you and make arrangements for the transfer of the cpu. Glad you can use it.

 

[mindless1]

There are lots of ways to test stability, do a Google search for a few of the following:

CPUBurn
Prime95
Folding@Home
SuperPI
3Dmark2001 3DMark2003 3DMark2000 3DMark99Max (one of those, for your older system probably 3DMark99Max is most appropriate)

Zip some very large files and CRC check them against the originals (if you don't know how to, just skip this or any others above).

Many other benchmarks or game demos can run looped so they don't stop unless they crash. Be sure to run one of the first three (CPUBurn, Prime95 or Folding@Home) as they'll have a greater chance of raising CPU temp, which can contribute to instability. Any test should run for a minimum of 30 mintues continuous but longer, as long as practical in the situation, is better.

One important point though is that if, since you hadn't ran any stability tests BEFORE when the CPU wasn't overclocked, you don't have a "baseline" of stability to compare against... if it crashes you can't be certain that it wouldn't have crashed even with the CPU non-overclocked, but if it doesn't crash at all, that won't matter.

 

[forkd]

Can a socket A cooler be used on these.....I'm an AMD guy taking my first dive in intel

 

[Greg03]

I need one of these for my great great grandfathers' great uncle.

 

[mindless1]

forkd, yes a socket A 'sink will work but consider the space available on the motherboard, if it fits.

 

[minidanas]

Aopen MX3WE2 MATX S370 Motherboard - 39.95

Does anyone know what's the form factor on that mobo? MATX could be either micro- or mini-ATX. Will it fit standard ATX case?