Alaris Cougar II Motherboard

Best to replace it, if the anode and cathode leads are pulled that far out, the internals are going to be damaged.

I always thought 386 boards with VLB were weird, the CPU doesn't have near enough grunt to drive them at a respectable speed. Fortunately it looks like you can put a 486 upgrade in and make the board a whole lot faster.
 
I looked up that blue lightning CPU and it looks pretty neat, cool board for a collector. I need to build an old 3.11 system but this is definitely not the one lol too much.
 
The weird CPUs and upgrade interposers were pretty interesting, more from a "why was this even made" perspective from me.

There exists a 386 to 486 interposer board which allows you to run an AMD Am5x86-133 on a 386 board, I believe it was called "Rev to 486" or some such. You could run the bus speed at 40 MHz and overclock it to 160 MHz for a bizarre 386 board running at 160 MHz.
 
RazorWind said:
Is that two AGP slots?
Click to expand...

As Bigbacon said, they're VLB slots.

VLB was a stopgap solution created out of the need for increasingly bandwidth heavy cards such as disk controllers and video cards. ISA was very bandwidth limited, operating at around 8 MHz (there was no defined speed, so this varied from board to board) and having a shared bandwidth between all cards of only about 33 MB/s. IBM had tried to replace ISA with their own MCA architecture in 1987, but it was very unpopular due to being proprietary and requiring licensing fees and royalties to be paid to IBM for use of their bus design.

In order to be as simple as possible and gain industry support, VLB was designed to be a processor direct slot where pins of the host CPU were brought out to the slot (mostly address and data lines) so the card could communicate directly with the CPU while the ISA bus was used for control and addressing. These design choices caused a number of headaches though, with the first being that since the CPU was effectively driving the slot directly from its address and data pins. This meant that you couldn't have more than around 3 slots in the system due to the electrical load placed on the CPU by the cards. Bus glitches between cards was a problem, which could lead to system instability and data loss if something like a disk controller conflicted with a video card.

The second headache is since VLB was based on the 486 bus architecture, it was not cross platform compatible (with the exception of the 386 since they had a similar bus architecture.) This is why the slot effectively died when the Pentium was introduced in 1993, there was no easy way to interface with the new bus. There were a few Pentium boards with VLB slots, but interest in the design quickly waned with the introduction of the PCI bus which was far simpler and had similar bandwidth.

There was yet another bus at the same time called EISA designed to compete directly with MCA, but it was more expensive to implement and had a lower bus bandwidth than VLB. The slots looked identical to ISA slots, with the only discerning difference being the slots were brown to indicate EISA. The slot was deeper than a normal ISA slot and had a second row of pins.

EISA-BT-747-BusTek-SCSI-Controller-ID531-531_3.jpg
 
Yep, VLB and EISA were precursors to PCI and really only served as stop-gap solutions to ISA limitations. When the Pentium launched, basically so did PCI. If you bought a Pentium system, you were pretty much guaranteed to have a few PCI slots (and some ISA slots as well). There were some 486 motherboards retrofitted with PCI, but they were expensive, and odd ducks because even the 486 DX4 100Mhz couldn't really benefit much from them. 486 was simply on its way out by the time PCI rolled around.

To summarize what GiGaByTe just said, during the 486 era the ever increasing demand for bandwidth outgrew the original ISA bus design, so a few solutions came out; however, there never was a universal solution agreed upon. For OEMs, this was pretty much a mess because it meant the end customer had to do their homework and shop for a video card or disk controller that used the correct bus interface since none of these solutions were interchangeable with each other.

The best thing PCI did was not try to form any backwards compatibility with this slop. It was the 32-bit bus solution everyone wanted. And the rest is history.
 
Last edited:
You must log in or register to reply here.
Back
Top