VOGONS


First post, by majestyk

User metadata
Rank Oldbie
Rank
Oldbie

Years ago I found a FIC PAC-2003 LPX mainboard in poor condition. I cleaned it, replaced some parts, upgraded VIDEO-RAM and installed the latest BIOS.
It worked fine with a Pentium MMX 233 (AMD K6-2 CPUs overload the onboard VRM with its tiny heatsink).

The attachment PAC_2003_1a.jpg is no longer available

Onboard there´s 256K PB-SRAM and I thought why not upgrade for 512K by populating the COAST (CELP) socket with another 256K?
In most cases this is a no-brainer, so I set the jumpers "SRAM1" and "SRAM2" accordingly. The jumpers allow to hook A13 and A14 of the TAG-RAM to either 5V or to the chipset / CPU for 512K or even 1M cache. The VIA "Apollo VP" chipset supports both.

After setting the jumpers and inserting a COAST module I was diasappointed when the POST screen reported "none" instead of 512K cache.
First I suspected the CELP slot´s contacts to be corroded / dirty after decades without a module, but cleaning wouldn´t help.
Next I tried dozens of different COAST modules and all possible jumper settings t0 no avail. I also tried an original FIC module just in case they did some proprietary configuration here (which wouldn´t be "FIC-like" btw). Unless staying with the 256K onboard configuration everything else turned out "none".
Interrupted traces, vias, chipset pins etc. could also be ruled out safely.
I then decided to put this mainboard aside - until today.

First I measured all the pins of the CELP slot and found out the clock connector was "nc"! Following the trace from the slot to the clock-generator finally revealed the culprit: The resistor between the clock-gen. output and the trace to the CELP slot had not been populated. Shame on the FIC QC guys...

The attachment PAC_2003_2a.jpg is no longer available

After soldering a 10R resistor 512K L2 cache are fully operable now.

The attachment PAC_2003_3a.jpg is no longer available
The attachment PAC_2003_4.JPG is no longer available

Reply 1 of 2, by rasz_pl

User metadata
Rank l33t
Rank
l33t

Nice one. Theretroweb has terrible picture, cant tell if darker blob is a component of empty space. Ebay has great photos with populated resistor
https://i.ebayimg.com/images/g/9wUAAOSwxD5lTYsi/s-l1600.webp
https://i.ebayimg.com/images/g/YvAAAOSwnGJlTYsw/s-l1600.webp

Kinda weird omission. Now I wonder when Optical Inspection was widely adopted.

https://github.com/raszpl/sigrok-disk FM/MFM/RLL decoder
https://github.com/raszpl/FIC-486-GAC-2-Cache-Module (AT&T Globalyst)
https://github.com/raszpl/386RC-16 ram board
https://github.com/raszpl/440BX Reference Design adapted to Kicad

Reply 2 of 2, by MagefromAntares

User metadata
Rank Member
Rank
Member

Hi,

Nice reminder that sometimes a small component that is missing or damaged that causes annoying issues.

About the K6-2, if the VRM can be set to 2 volts core then the Mobile versions of K6-2 can be tried which has lower TDP and such lower wattage requirements, I'm not familiar with this specific motherboard so I don't know if that is a valid setting for this specific board however.

EDIT: Some mobile versions of the K6-2 has 1.8 volt core not 2.0 volt core, so if you try the Mobile versions of K6-2s check which one is used by it, don't do an accidental over-volt of the K6-2 😀.

"A process cannot be understood by stopping it. Understanding must move with the flow of the process, must join it and flow with it." - Dune