VOGONS


First post, by gerwin

User metadata
Rank l33t
Rank
l33t

I have two socket 370 CPUs here:
- Pentium III 866MHz
- Celeron 850MHz
Both dead. One could say: the die died. Different motherboards fail to boot up with these.
Not such a big deal normally, but the celeron was a special model. The Celeron seems to emit more heat then normally when trying to boot with it. The Pentium III 866 was in my stash for years already. It should have been removed in working order back then. Not working anymore.
I wonder what happened, normally I would think these CPUs overheated at some point. Though this video seems to prove that Pentium III has a thermal protection: http://hackermaster.free.fr/ftp/thg/thg_video … cpu_cooling.zip
Elsewere I read that this protection is not always fast enough to save an overheating situation.

Anybody else have some anekdotes on this matter, preferably regarding the pentium III?

--> ISA Soundcard Overview // Doom MBF 2.04 // SetMul

Reply 1 of 18, by meljor

User metadata
Rank Oldbie
Rank
Oldbie

I have seen MANY cpu's trough the last 20+ years and the most common dead ones are athlon/athlon xp cpu's. I did have a slot1 intel p3 that had defective cache but otherwise i can't even remember seeing a defective Intel cpu.

Certainly not saying that these never die, just from my experience it doesn't seem to happen a lot?

asus tx97-e, 233mmx, voodoo1, s3 virge ,sb16
asus p5a, k6-3+ @ 550mhz, voodoo2 12mb sli, gf2 gts, awe32
asus p3b-f, p3-700, voodoo3 3500TV agp, awe64
asus tusl2-c, p3-S 1,4ghz, voodoo5 5500, live!
asus a7n8x DL, barton cpu, 6800ultra, Voodoo3 pci, audigy1

Reply 2 of 18, by deleted_Rc

User metadata
meljor wrote:

I have seen MANY cpu's trough the last 20+ years and the most common dead ones are athlon/athlon xp cpu's. I did have a slot1 intel p3 that had defective cache but otherwise i can't even remember seeing a defective Intel cpu.

Certainly not saying that these never die, just from my experience it doesn't seem to happen a lot?

isn't that due to the die usually cracking? never broke a CPU before, don't OC to much and be carefull with athlons....

Reply 3 of 18, by Errius

User metadata
Rank l33t
Rank
l33t

I lost a 366 MHz S370 Celeron for unknown reasons some years ago. It just stopped working. No OC or overheat that I was aware of. Worse, plugging this CPU into another board for testing purposes seems to have killed that board as well. (Original board was OK, but the CPU was running on a slotket which may have protected it.) CPU and second board were both trashed.

I have never had a Pentium fail

Is this too much voodoo?

Reply 4 of 18, by meljor

User metadata
Rank Oldbie
Rank
Oldbie
Richo wrote:
meljor wrote:

I have seen MANY cpu's trough the last 20+ years and the most common dead ones are athlon/athlon xp cpu's. I did have a slot1 intel p3 that had defective cache but otherwise i can't even remember seeing a defective Intel cpu.

Certainly not saying that these never die, just from my experience it doesn't seem to happen a lot?

isn't that due to the die usually cracking? never broke a CPU before, don't OC to much and be carefull with athlons....

Cracking can occur too but the dead Amd's came that way, i never killed one myself as far as i remember. The older versions didn't have thermal protection and a defective cooler was sometimes enough to kill them.

asus tx97-e, 233mmx, voodoo1, s3 virge ,sb16
asus p5a, k6-3+ @ 550mhz, voodoo2 12mb sli, gf2 gts, awe32
asus p3b-f, p3-700, voodoo3 3500TV agp, awe64
asus tusl2-c, p3-S 1,4ghz, voodoo5 5500, live!
asus a7n8x DL, barton cpu, 6800ultra, Voodoo3 pci, audigy1

Reply 5 of 18, by appiah4

User metadata
Rank l33t++
Rank
l33t++
meljor wrote:

Cracking can occur too but the dead Amd's came that way, i never killed one myself as far as i remember. The older versions didn't have thermal protection and a defective cooler was sometimes enough to kill them.

More often than not it was a clumsy first time attempt at mounting a heatsink onto a bare CPU die.

I know I killed at least one motherboard if not CPU..

Retronautics: A digital gallery of my retro computers, hardware and projects.

Reply 6 of 18, by gerwin

User metadata
Rank l33t
Rank
l33t
meljor wrote:

...but otherwise i can't even remember seeing a defective Intel cpu.

I also used to think these CPUs were pretty much indestructable compared to other hardware.

Never had a cracked die myself. But through all the swapping and testing, can I be 100% sure that the heatsink was always attached exactly flat to towards the die? and if not, then what?

--> ISA Soundcard Overview // Doom MBF 2.04 // SetMul

Reply 7 of 18, by Tetrium

User metadata
Rank l33t++
Rank
l33t++
Errius wrote:

I lost a 366 MHz S370 Celeron for unknown reasons some years ago. It just stopped working. No OC or overheat that I was aware of. Worse, plugging this CPU into another board for testing purposes seems to have killed that board as well. (Original board was OK, but the CPU was running on a slotket which may have protected it.) CPU and second board were both trashed.

I have never had a Pentium fail

I've not had this with a CPU, but I did have this happen with a DIMM and a harddrive. Both "infected" and killed any board they were put in and yes I killed several boards using either defective parts 😵
From the outside they looked perfectly fine.

I did kill at least one CPU, an Athlon XP (forgot TIM and there it went), but never had a P6 or any other Athlon die (or at least not verified).

Gerwin, what was so special about your Celeron? Don't tell me it was an ES? 😕

Whats missing in your collections?
My retro rigs (old topic)
Interesting Vogons threads (links to Vogonswiki)
Report spammers here!

Reply 8 of 18, by gerwin

User metadata
Rank l33t
Rank
l33t

You guessed it: ES. Just received it and never got it to boot. DOA I suppose.

I almost killed an Athlon XP once, which was turned on without heatsink. I was really quick to remove the power, and it was still OK. Phew. At that time it would matter more to me in case it was broken. Now, not so much.

--> ISA Soundcard Overview // Doom MBF 2.04 // SetMul

Reply 9 of 18, by Tetrium

User metadata
Rank l33t++
Rank
l33t++

I've had a Thunderbird 900 get lucky several times, including no TIM and accidental overclocks (from 900MHz to 1200MHz).

My own ES may have fallen victim to an accidental overclock but that had to do with the CPU multipliers being different from the multipliers the CPU actually (so it booted with something totally else compared to what I set in the BIOS).

I may have an ES 1GHz Coppermine, but I never actually tested it (I actually didn't even remember I had it till I stumbled upon it recently). Don't even know if it's got unlocked multipliers or not.

I was actually very hesitant to try the ES again since I saw no point in frying it for good just because I did not know what I was doing...

Whats missing in your collections?
My retro rigs (old topic)
Interesting Vogons threads (links to Vogonswiki)
Report spammers here!

Reply 10 of 18, by gerwin

User metadata
Rank l33t
Rank
l33t
Tetrium wrote:

I was actually very hesitant to try the ES again since I saw no point in frying it for good just because I did not know what I was doing...

That is a nice chip for sure. A good way to test it, is to take a motherboard with jumper configuration, then set the FSB to the lowest setting of 66 or even 50MHz. That way, even if the multi turns out to be rather high, the net speed is still safe.

--> ISA Soundcard Overview // Doom MBF 2.04 // SetMul

Reply 11 of 18, by Tetrium

User metadata
Rank l33t++
Rank
l33t++
gerwin wrote:
Tetrium wrote:

I was actually very hesitant to try the ES again since I saw no point in frying it for good just because I did not know what I was doing...

That is a nice chip for sure. A good way to test it, is to take a motherboard with jumper configuration, then set the FSB to the lowest setting of 66 or even 50MHz. That way, even if the multi turns out to be rather high, the net speed is still safe.

It's a pointless chip if I can't even use it and it's totally impractical if I don't even know how to set it. If I don't use its free multiplier then I might as well use a regular Tualatin.

That was what I was doing actually, except I was using the BIOS as I didn't consider a jumpered board would make the PCU behave differently compared to multiplier settings set from within BIOS (and it was faster that way).

But it's rather pointless if I have to figure out all the multiplier settings that way. Iirc some of the multipliers were totally different (much higher) compared to the multiplier I had set.

How the heck did you get your ES running gerwin? Didn't you run into the same issues?

Whats missing in your collections?
My retro rigs (old topic)
Interesting Vogons threads (links to Vogonswiki)
Report spammers here!

Reply 12 of 18, by gerwin

User metadata
Rank l33t
Rank
l33t

Write down a table of all 16 different settings of 4 multiplier jumpers/dip-switches. Try each setting, and note each result behind the setting. In case the BIOS shows a weird speed for the setting: ignore it. Instead do a speed measurement with SetMul or similar, and divide the total measured speed by the FSB speed that was configured, to get the real multi. A few jumper settings will not boot, so then note it as such.

The BIOS tries to be helpful, but it regularly fails to set or diagnose the odd speeds right. Just ignore it. This is a CPU matter, not a BIOS matter.

--> ISA Soundcard Overview // Doom MBF 2.04 // SetMul

Reply 13 of 18, by gerwin

User metadata
Rank l33t
Rank
l33t

Here is a burning Pentium III 450MHz Katmai:
https://www.youtube.com/watch?v=5umDJhIfrt0
I am not sure what he means with 'elapsed time 1:36'. Did he skip more then a minute to shorten the video, and start running again when smoke appears + system crashes? In that case it can survive for quite a while.

--> ISA Soundcard Overview // Doom MBF 2.04 // SetMul

Reply 15 of 18, by gerwin

User metadata
Rank l33t
Rank
l33t

ESD is good candidate indeed. The broken celeron was shipped with all legs pushed in a block of ordinary white Styrofoam which was wrapped in aluminium foil.

--> ISA Soundcard Overview // Doom MBF 2.04 // SetMul

Reply 16 of 18, by appiah4

User metadata
Rank l33t++
Rank
l33t++

In my 30 years of handling computers and hardware I never witnessed anything die from ESD ever. To me it might as well be a myth..

Retronautics: A digital gallery of my retro computers, hardware and projects.

Reply 17 of 18, by SSTV2

User metadata
Rank Oldbie
Rank
Oldbie
appiah4 wrote:

In my 30 years of handling computers and hardware I never witnessed anything die from ESD ever. To me it might as well be a myth..

You must be a poor conductor then 🤣

Personally, i've killed 2 CPUs with ESD, one got killed while i was fiddling with a socket 7 system (just took CPU out, reapplied thermal paste and put it back in), another died when i was transporting it in a pants pocket in my adolescence *facepalm*.

Reply 18 of 18, by Errius

User metadata
Rank l33t
Rank
l33t
appiah4 wrote:

In my 30 years of handling computers and hardware I never witnessed anything die from ESD ever. To me it might as well be a myth..

This is my experience as well. It probably depends a lot on the electrical characteristics of your house/workshop though.

I remember a house where I would get minor electric shocks when handling electrical equipment while standing barefoot on the floor. This was a bare stone floor on the ground floor of the building. The problem disappeared in places where the floor was made of wood or when I wore shoes.

Is this too much voodoo?