one more question - which models/brands of crystal oscillators for overklocking of a 386-CPU over 40 MHz (maybe 45-50 MHz) are recommended?

I took a look at page 1 of this thread and followed the links for some 386-systems, but could not find any info.

Never seen 386 cpus packaged like that. Are they functional ?

As for your second question - all crystals work as advertised, except in one instance of extreme overclock.
What are you after ? : )

pshipkov wrote on 2024-04-12, 04:45:

Never seen 386 cpus packaged like that. Are they functional ?

As for your second question - all crystals work as advertised, except in one instance of extreme overclock.
What are you after ? : )

The CPUs are still not tested, as I am affraid, they are maybe not 386 (with risk to damage my mainboards).
Maybe I should try them anyway (as I have a very corroded 386-board here still booting/posting - no matter if it goes even more damaged by those IBM-chips, or not)

About the crystals - I would be thankful to know, which models do you (or other people here) recommend for overclockng a "standard" AMD 386DX-40 to 45/50 MHz (if lucky).

Thereafter I can have a look where to buy them (at this time I do not have any crystals for overclocking above 40 MHz).

Next update: both IBM-CPUs are 386!
Tested with a FIC-386SC, Rev. A3. Both work at 33 MHz, no one work at 40 MHz. Identified by CheckCPU as "standard 386"; identified by Speedsys as 80386DX (not sure about the correctness of this, maybe they are DX, or maybe SX).
In generally, they seems to be "nothing special".
Another test with a FIC-386-HC-HG, Rev. B, was not able to start both IBMs. Sadly, after that, no one more boot with every other 386-CPU is possible on this board (the IBMs still boot on the 386SC, so maybe the HC-HG has another problem).

Next thing is: few hours ago somebody brought me a desoldered IBM-chip in more or less bad condition, no idea is it working or not. Fortunately, all pins are present.
Is this maybe an IBM Blue Lightning BL2-BL3? Can the original frequency be recognized by the imprint?

Here Creating a voltage regulated 386 BL3 module from existing parts and here Re: 3 (+3 more) retro battle stations they are shown similar chips.
Compared to an adapter PGA-132 to QFP-132 with AMD-386DX-40, this IBM-chip is smaller, so it maybe can not be soldered instead of the AMD-CPU on the same adapter. And, of course, the number of the pins is different.
Are they other (smaller) adapters ecpecially for this IBM-chip on the market?

gonzo wrote on 2024-04-12, 17:50:

Next update: both IBM-CPUs are 386! Tested with a FIC-386SC, Rev. A3. Both work at 33 MHz, no one work at 40 MHz. Identified by […]
Show full quote

Next update: both IBM-CPUs are 386!
Tested with a FIC-386SC, Rev. A3. Both work at 33 MHz, no one work at 40 MHz. Identified by CheckCPU as "standard 386"; identified by Speedsys as 80386DX (not sure about the correctness of this, maybe they are DX, or maybe SX).
In generally, they seems to be "nothing special".
Another test with a FIC-386-HC-HG, Rev. B, was not able to start both IBMs. Sadly, after that, no one more boot with every other 386-CPU is possible on this board (the IBMs still boot on the 386SC, so maybe the HC-HG has another problem).

Next thing is: few hours ago somebody brought me a desoldered IBM-chip in more or less bad condition, no idea is it working or not. Fortunately, all pins are present.
Is this maybe an IBM Blue Lightning BL2-BL3? Can the original frequency be recognized by the imprint?

Here Creating a voltage regulated 386 BL3 module from existing parts and here Re: 3 (+3 more) retro battle stations they are shown similar chips.
Compared to an adapter PGA-132 to QFP-132 with AMD-386DX-40, this IBM-chip is smaller, so it maybe can not be soldered instead of the AMD-CPU on the same adapter. And, of course, the number of the pins is different.
Are they other (smaller) adapters ecpecially for this IBM-chip on the market?

Yes, the online search confirmed these are some early 386DX models.
Your tests confirm they don't scale very well, which i kind of expected.

---

There are 386 DX and SX variants. I am not going to bring the well know details about them here - you can check online - plenty of info on the subject.
DX variant has 132 pins.
SX variant has 100 pins.

Later on improved processor models came to market as upgrades to the standard 386SX/DX CPUs.
DLC -> DX
SLC -> SX
The processors you show are SLC variants - pin incompatible with DX/DLC.

For them you can try motherboards like Alaris Leopard, or Matra 486SLC2.

deleted (double post)

deleted (double post)

pshipkov wrote on 2024-04-12, 18:52:

DX variant has 132 pins. SX variant has 100 pins. The processors you show are SLC variants - pin incompatible with DX/DLC.

That explains a lot for me, thank you.

About the crystal oscillators for overklocking of a 386-CPU: I just found some info for one of your systems here Re: 3 (+3 more) retro battle stations, it is a crystal from ecliptek shown in the picture. If you know other manufacturers, please let me know, thanks.

Forgot to confirm your question about crystal oscillators.
Ebay is littered by them. Brands dont matter. Search for “crystal oscillator”. Can include additional keywords like “dip14” or “full can”.
Depends on how far you anticipate to go, consider buying a clock generator. They are more expensive, but even few crystal oscillators will build up to the same sum of $, wile leaving you with far fewer options.

pshipkov wrote on 2024-04-13, 08:15:

Forgot to confirm your question about crystal oscillators.
Ebay is littered by them. Brands dont matter. Search for “crystal oscillator”. Can include additional keywords like “dip14” or “full can”.
Depends on how far you anticipate to go, consider buying a clock generator. They are more expensive, but even few crystal oscillators will build up to the same sum of $, wile leaving you with far fewer options.

Thanks. Is the correct voltage 5 Volt?

Yes.

gonzo wrote on 2024-04-11, 15:48:

Are these two chips maybe 386-CPUs from IBM?
Should I risk to open/take-down the aluminium covers of them without damage of the cores?

Sadly I can not give any additional infos about them, as they are from "a friend of a friend of a friend"...

Did somebody crack open a few ps/2s to get these? I’m surprised they worked at 40mhz

I didn't see frequencies associated with these CPUs, but i assume they are early/slow models, still reaching 40MHz, at least getting lights on screen, is doable for even 25MHz rated silicon. Do you know more about them ?

Anonymous Coward wrote on 2024-04-15, 03:08:

I’m surprised they worked at 40mhz

No, they did not - only 33 MHz.
See my posting above "Both work at 33 MHz, no one work at 40 MHz"

pshipkov wrote on 2024-04-15, 19:12:

Do you know more about them ?

I don't know too much, except that CPU collectors like them. As far as I know they were only manufactured by IBM and only used in their own hardware. Functionally they should be the same as their intel counterparts.