• Please review our updated Terms and Rules here

Most uncommon 486 chips

Chuck(G) said:
I've seen AM486DX5-133 chips. They've got to be uncommon.
Click to expand...

The early production AMD chips said both 486DX5 and 5x86 on them. Perhaps a pre-production test/sample run said just 486DX5 on it, but I've never seen any.

I wonder who was the first to officially call their chip the 5x86: either AMD or Cyrix. (Remember, the Cyrix 5x86 was initially called the "M1SC".) Perhaps one company renamed their chip to 5x86 in response to hearing that the other company was going to use that name.
 
as for the 486-DX-50 being fast for VLB, alot of the boards i have allow you to decouple the VLB from the main clock speed, you can set the VLB @ 33, and keep the rest chugging away at 50. At least that's what the jumpers on the board indicate. Unfortunately my only pc that would allow you to run EVERYTHING @ 50 properly just went tits-up, my Digital DECpc, the CPU card got zapped some how, there is a rather nice, shapely, hole in it that was proceeded by a rather large bang from the SX-25 so i think i may have been the cause of it's demise :(
 
I don't think your board is decoupling the local bus. Most likely the jumpers say something like ">33MHz" or "<=33MHz". What the jumper is actually doing is adding a wait state.
 
VLB BUS speed is the same as the processor FSB, that is the reason PCI took over (PCI slots are 33mhz and the CPU bus is a different speed).
 
On some 486 implementations of PCI it was FSB-based as well, with divider settings and the like. Sometimes your PCI could end up at 28mhz or so, sometimes 35mhz, etc.
 
The PCI 486 boards are the best for the fastest 486 systems, but the VLB or EISA ones are the most interesting to me.
 
I have an Intel 487SX/2 50mhz CPU... It was used on the 487 socket of a 486SX 25mhz system to add both co-processor and DX2 internal clock-doubling, the SX 25 CPU being disabled, of course.

Not sure how rare that is, but I haven't seen very many.

I also have a DX50 somewhere.
 
I've never seen a 487 chip - likely that was made early on, back when clock multipliers were just starting to take off. 487 sockets (also called "Upgrade" sockets) can be used for all kinds of things in later boards, including replacing the onboard soldered CPU on many boards with the appropriate jumper setting. AFAIK the 487/Upgrade socket is just a Socket 1 that may or may not be wired differently depending on board design and jumpers.

Can we get a pic of that chip?
 
Raven said:
I've never seen a 487 chip - likely that was made early on, back when clock multipliers were just starting to take off. 487 sockets (also called "Upgrade" sockets) can be used for all kinds of things in later boards, including replacing the onboard soldered CPU on many boards with the appropriate jumper setting. AFAIK the 487/Upgrade socket is just a Socket 1 that may or may not be wired differently depending on board design and jumpers.

Can we get a pic of that chip?
Click to expand...

Do I sense a tad bit of doubt in your tone there, Raven....? ;)

The 487SX and SX/2 chips were full 486DX chips and disabled the existing 486SX chip. The rarity I perceive in the one I have is that it's a 487SX/2. I've seen many 487SX chips, but not many (if any) SX/2's.

EDIT: I will try and get the glued heatsink off the chip so that I can take a picture of it. Any advice on how to do this without causing damage to the ceramic surface? I don't have the greatest dexterity for this type of thing.
 
No, I believe you fully - I know they exist and seemed to remember that they might be something along the lines you just mentioned (I was thinking that they had the core part disabled but were a normal DX chip, just the FPU enabled, but I had it kinda sideways in my mind it seems).

What it sounds like to me is that later in production they just stopped calling them 487s and started calling the socket an Upgrade socket since you can really put any 486 in there and override the original. That would make them somewhat rare, as anything marked 487 would be an early revision of a 486DX or DX2 marked 487SX or SX2.

As for removing heatsinks - twist the heatsink if you can to loosen it, otherwise use a flathead screwdriver or chip pry tool to try to pry it up (carefully). As long as you don't lose control of the screwdriver (which CAN happen when prying things, don't laugh) you can't really damage the ceramic top of the chip very easily.
 
Raven said:
No, I believe you fully - I know they exist and seemed to remember that they might be something along the lines you just mentioned (I was thinking that they had the core part disabled but were a normal DX chip, just the FPU enabled, but I had it kinda sideways in my mind it seems).

What it sounds like to me is that later in production they just stopped calling them 487s and started calling the socket an Upgrade socket since you can really put any 486 in there and override the original. That would make them somewhat rare, as anything marked 487 would be an early revision of a 486DX or DX2 marked 487SX or SX2.

As for removing heatsinks - twist the heatsink if you can to loosen it, otherwise use a flathead screwdriver or chip pry tool to try to pry it up (carefully). As long as you don't lose control of the screwdriver (which CAN happen when prying things, don't laugh) you can't really damage the ceramic top of the chip very easily.
Click to expand...

This thing is on there good... But a quick search of the IBM part number (it's an Intel chip but was sold as an upgrade by IBM...not an IBM 486SLC2 or the like) revealed that what you're saying is more accurate (FRU92F0147...http://www.chipdb.org/img-intel-a80486dx2-50-ibm-fru92f0147-2772.htm). It appears this is an actual labelled 486DX/2 50. The "upgrade" socket allowed for a standard 486DX or DX/2 to replace the 486SX, as you stated. What a weird thing to do, seeing how it would have been cheaper to build single-socket machines, but I guess they intended that socket to be used for non-overdrive processors, so it could automatically handle voltage differences. If this is the case with the one I have, there goes the delusion of rarity.

I still want to get the heatsink off to confirm this, but they did a great job of having this thing flush with the ceramic surface. Maybe if I apply a little heat with a hair dryer or something...
 
EddieDX4 said:
I still want to get the heatsink off to confirm this, but they did a great job of having this thing flush with the ceramic surface. Maybe if I apply a little heat with a hair dryer or something...
Click to expand...

Try the opposite. Stick it in the freezer overnight, then give the heatsink a twist. If it's held on with a superglue-like substance, cold temps make it brittle.
 
Dave Farquhar said:
Try the opposite. Stick it in the freezer overnight, then give the heatsink a twist. If it's held on with a superglue-like substance, cold temps make it brittle.
Click to expand...

This makes sense, when you realize a heatsink needed to stay on under heat, and the ceramic packaging of the 486 was designed for a 135 degree Fahrenheit normal operating temperature...
 
You have a 487SX/2? Do you have a photo? I'm sure the guys over at CPUworld would like to see a 487SX/2. I know I would.

I recently sold my 487SX. Kind of uncommon I guess, but not very interesting.
 
if its like either of these.. dont try & remove the heatsink, as it is worth more intact

49G2988 80486DX2-50
ibm__overdrive_234.jpg


49G2989 80486DX2-66
IBM_FRU92F2582_49G2989_DX2-66_cf.jpg
 
here is the bottom of one... shows it has the extra pin to disable the SX processor.
IBM didnt call them an Overdrive, but they did use overdrive technology

IBM49G22988-FRU92F0147_B.jpg


on a side note the IBMs were made from these Intel parts
for 25MHz bus 49G2988 >> ODP486SX-25 B1 stepping
for 33MHz bus 49G2989 >> ODP486SX-33 / ODP486DX-33 B1 stepping
 
Last edited:
EddieDX4 said:
This thing is on there good... But a quick search of the IBM part number (it's an Intel chip but was sold as an upgrade by IBM...not an IBM 486SLC2 or the like) revealed that what you're saying is more accurate (FRU92F0147...http://www.chipdb.org/img-intel-a80486dx2-50-ibm-fru92f0147-2772.htm). It appears this is an actual labelled 486DX/2 50. The "upgrade" socket allowed for a standard 486DX or DX/2 to replace the 486SX, as you stated. What a weird thing to do, seeing how it would have been cheaper to build single-socket machines, but I guess they intended that socket to be used for non-overdrive processors, so it could automatically handle voltage differences. If this is the case with the one I have, there goes the delusion of rarity.
Click to expand...

Told you I know my 486s, heh. ;)

Even if the chip underneath is just a standard one, the whole thing repackaged as a 487SX/2 makes it rare.

Cool chip either way, just not very useful in this age when alternatives abound for these sockets (i.e., I have a 25SX soldered onto my Presario 425, but I dropped an Evergreen 586 into the socket instead of something along the lines of the 487SX/2 - far greater speed increase).

On the bright side, you can likely drop that into any ordinary system and use it instead of having to find a special system that requires a "487" upgrade (unless that extra pin prevents it from being used in non-upgrade sockets, but I don't believe so). :D
 
You must log in or register to reply here.
Back
Top