That was one tempremental board...

That P3-533, is that a Katmai P3-533B (512kB L2) or Coppermine P3-533EB (256kB L2) 533?

In any event I was able to run it fine with a P3-1000EB So370 and via MS-6905 Master Slocket. Most of my issues with the board related to BIOS and ESCD. Did you try a CMOS clear *and* an ESCD reset (disconnect mains power, hold power button for >10 sec) after installing the P3-866EB? If not, that would be where to start.

Thanks for the tip, I will try to do what you suggest. By the way the PIII that works is 533EB.

Paar wrote:

Thanks for the tip, I will try to do what you suggest. By the way the PIII that works is 533EB.

That's useful, that means that whatever the problem is, it's not a Coppermine voltage issue.

If the CMOS and ESCD resets don't work out, it could be a CPU stepping issue. Possibly the P3-866EB is a late stepping (C0) and the current BIOS doens't know what to do with it. A BIOS upgrade might help in that case.

Success! I have followed your advice and did CMOS and ESCD reset. One more thing I did was that I used a heatsink with fan while powering the computer up. I don't know what helped but it is working now and I'm super happy. Is it possible for Coppermine CPU to get overheated right on POST screen with no cooler attached?

One interesting observation. I have two 866EBs from year 2000 that operate on 1.7V. Both work fine. The one that doesn't is from 2001 that operates on 1.75V. The CPU looks undamaged so I wonder if the small voltage spec difference can make it incompatible?

Paar wrote:

Is it possible for Coppermine CPU to get overheated right on POST screen with no cooler attached?

Not just possible, that's inevitable. Don't run Pentium IIIs without heatsink, you could damage them! 😵

I'm surprised that you didn't kill those CPUs instantly by not having a cooler on them.

Paar wrote:

One interesting observation. I have two 866EBs from year 2000 that operate on 1.7V. Both work fine. The one that doesn't is from 2001 that operates on 1.75V. The CPU looks undamaged so I wonder if the small voltage spec difference can make it incompatible?

The earlier versions of that cpu: SL49H, SL4MD, SL43J and SL4CB have BIOS identification code 0686h.
The last two versions: SL4ZJ and SL5DX have BIOS identification code 068Ah.
You may need to update your BIOS before the last two versions are recognized.

http://www.cpu-world.com/sspec/SL/SL4ZJ.html

cyclone3d wrote:

I'm surprised that you didn't kill those CPUs instantly by not having a cooler on them.

On old systems I usually just put the CPU in to see if the computer will boot up just to shut it down right afterwards. I had no idea PIII CPUs are so sensitive. It helped just to place a heatsink on top of it without applying any thermal paste. I'll surely remember that next time I will be testing something.

PARKE wrote:

The earlier versions of that cpu: SL49H, SL4MD, SL43J and SL4CB have BIOS identification code 0686h.
The last two versions: SL4ZJ and SL5DX have BIOS identification code 068Ah.
You may need to update your BIOS before the last two versions are recognized.

I already have the newest version of the BIOS and the CPU still doesn't work. The question is if I have fried it (unlikely as I didn't fry the other ones) or the BIOS just doesn't support. I have checked and the BIOS version 1.07 (which is the newest I could find) has been released at the end of year 2000. That probably explains why the board doesn't support the CPU.

Paar wrote:

[...]

On old systems I usually just put the CPU in to see if the computer will boot up just to shut it down right afterwards. I had no idea PIII CPUs are so sensitive. It helped just to place a heatsink on top of it without applying any thermal paste. I'll surely remember that next time I will be testing something.

Modern CPUs have sophisticated thermal regulation. Older ones don't. Back in the 486/Pentium days that didn't matter too much, but by P3-era it became critical. The P3 Coppermine has in internal thermal diode that registers temperature and shuts the CPU down if it gets too hot, but it's designed for slow increases in temperature that you get with heatsink but say too much dirt or other insufficient airflow. It's quite specifically not designed to cope with no heatsink at all.

Note that a P3 *may* survive this, an Athlon Thunderbird wil most definitely not give you a second chance:

https://youtu.be/UoXRHexGIok

Wow, that is one scary video. I guess I was lucky but probably just because I didn't go past POST screen (which didn't even turn up).