VOGONS


First post, by brian.t

User metadata
Rank Newbie
Rank
Newbie

I have an FIC PA-2013 (Revision 2.1, 1MB cache, new capacitors) with a K6-III-450. It will not boot with "External Cache Enabled" (goes to the blinking underscore in the top left corner after drive detection). Works fine when I disable the external cache. Wondering the best way to troubleshoot if I want to get the cache working. I verified the jumper settings are correct. I've set the BIOS to defaults. Wondering if I need to replace the four cache chips or if there's something else. Thanks

Reply 1 of 15, by dominusprog

User metadata
Rank Oldbie
Rank
Oldbie

Have you checked the chips legs? Also you should check the TAG chip for broken solder joints.

Duke_2600.png
A-Trend ATC-1020 V1.1 ❇ Cyrix 6x86 150+ @ 120MHz ❇ 32MiB EDO RAM (8MiBx4) ❇ A-Trend S3 Trio64V2 2MiB
Creative AWE64 Value ❇ 8.4GiB Quantum Fireball ❇ Win95 OSR2 Plus!

Reply 2 of 15, by Many Bothans

User metadata
Rank Newbie
Rank
Newbie

Yes, check that tag sram... there are five chips.

  • Zenith Z386SX-20, 8MB FPM, Video 7 1024i, Unhoused
  • AOpen AP43, Am5x86-133@160, 1MB L2, 128MB FPM, Stealth III S540 32MB Savage4, SB32
  • ITX-Llama, 3Dfx V3
  • Asus CUV4X-E, P3-933, 512MB PC133, Hercules 3D Prophet II MX 32MB, SB Live!

Reply 3 of 15, by brian.t

User metadata
Rank Newbie
Rank
Newbie
dominusprog wrote on 2024-10-08, 15:39:

Have you checked the chips legs? Also you should check the TAG chip for broken solder joints.

The SRAM tag seems ok, however the pins are j shaped and the solder is underneath. Should I try to reflow the tag with hot air, or replace it from an identical donor motherboard?

Reply 4 of 15, by brian.t

User metadata
Rank Newbie
Rank
Newbie
brian.t wrote on 2024-10-09, 12:19:
dominusprog wrote on 2024-10-08, 15:39:

Have you checked the chips legs? Also you should check the TAG chip for broken solder joints.

The SRAM tag seems ok, however the pins are j shaped and the solder is underneath. Should I try to reflow the tag with hot air, or replace it from an identical donor motherboard?

Also would there be any value in changing the bus speed from 100MHz to like 66MHz to see if it's an access time issue?

Reply 5 of 15, by dominusprog

User metadata
Rank Oldbie
Rank
Oldbie
brian.t wrote on 2024-10-09, 12:19:
dominusprog wrote on 2024-10-08, 15:39:

Have you checked the chips legs? Also you should check the TAG chip for broken solder joints.

The SRAM tag seems ok, however the pins are j shaped and the solder is underneath. Should I try to reflow the tag with hot air, or replace it from an identical donor motherboard?

Reflowing the solder is a good idea, but make sure that you're applying enough flux to the pins. For the bus, it should run at 100MHz without any issues.

Duke_2600.png
A-Trend ATC-1020 V1.1 ❇ Cyrix 6x86 150+ @ 120MHz ❇ 32MiB EDO RAM (8MiBx4) ❇ A-Trend S3 Trio64V2 2MiB
Creative AWE64 Value ❇ 8.4GiB Quantum Fireball ❇ Win95 OSR2 Plus!

Reply 6 of 15, by ratfink

User metadata
Rank Oldbie
Rank
Oldbie

There was a thread about this before by the look of it - same mobo and cpu (though not specific on versions)
to disable Cache or not to
Suggestion there that the board may not be able to operate the cache as L3.

Reply 7 of 15, by brian.t

User metadata
Rank Newbie
Rank
Newbie
dominusprog wrote on 2024-10-09, 15:42:

Reflowing the solder is a good idea, but make sure that you're applying enough flux to the pins. For the bus, it should run at 100MHz without any issues.

I managed to order a couple 8ns replacement SRAM tags (IS61C256AH-8JR). I'm going to try replacing the tag chip and report back once I do. Thanks

Reply 8 of 15, by brian.t

User metadata
Rank Newbie
Rank
Newbie
ratfink wrote on 2024-10-09, 16:50:

There was a thread about this before by the look of it - same mobo and cpu (though not specific on versions)
to disable Cache or not to
Suggestion there that the board may not be able to operate the cache as L3.

Thanks for the thread link. Good info. I do recall running this processor on this board, but it was over 20 years ago 😀.

Reply 9 of 15, by dominusprog

User metadata
Rank Oldbie
Rank
Oldbie
brian.t wrote on 2024-10-10, 12:32:
dominusprog wrote on 2024-10-09, 15:42:

Reflowing the solder is a good idea, but make sure that you're applying enough flux to the pins. For the bus, it should run at 100MHz without any issues.

I managed to order a couple 8ns replacement SRAM tags (IS61C256AH-8JR). I'm going to try replacing the tag chip and report back once I do. Thanks

Don't forget to use Kapton tape to isolate the surrounding area.

Duke_2600.png
A-Trend ATC-1020 V1.1 ❇ Cyrix 6x86 150+ @ 120MHz ❇ 32MiB EDO RAM (8MiBx4) ❇ A-Trend S3 Trio64V2 2MiB
Creative AWE64 Value ❇ 8.4GiB Quantum Fireball ❇ Win95 OSR2 Plus!

Reply 10 of 15, by brian.t

User metadata
Rank Newbie
Rank
Newbie
dominusprog wrote on 2024-10-10, 14:55:

Don't forget to use Kapton tape to isolate the surrounding area.

I replaced the tag with the ISSI chip. Appears to have the same problem. I checked all tag pins for solder bridges/shorts and made sure each pin had a good solder connection.

Reply 11 of 15, by Horun

User metadata
Rank l33t++
Rank
l33t++
ratfink wrote on 2024-10-09, 16:50:

There was a thread about this before by the look of it - same mobo and cpu (though not specific on versions)
to disable Cache or not to
Suggestion there that the board may not be able to operate the cache as L3.

Yes, good info ! probably a bios issue.
With K6-III and onboard 256k cache an external cache is really not needed (just double caches the data iirc) which on most boards gives no real improvement. Be like adding external cache to a P.Pro. Or am I missing something...

Hate posting a reply and then have to edit it because it made no sense 😁 First computer was an IBM 3270 workstation with CGA monitor. Stuff: https://archive.org/details/@horun

Reply 12 of 15, by dominusprog

User metadata
Rank Oldbie
Rank
Oldbie
brian.t wrote on 2024-10-18, 02:22:
dominusprog wrote on 2024-10-10, 14:55:

Don't forget to use Kapton tape to isolate the surrounding area.

I replaced the tag with the ISSI chip. Appears to have the same problem. I checked all tag pins for solder bridges/shorts and made sure each pin had a good solder connection.

It looks fine to me, what is the problem?

Duke_2600.png
A-Trend ATC-1020 V1.1 ❇ Cyrix 6x86 150+ @ 120MHz ❇ 32MiB EDO RAM (8MiBx4) ❇ A-Trend S3 Trio64V2 2MiB
Creative AWE64 Value ❇ 8.4GiB Quantum Fireball ❇ Win95 OSR2 Plus!

Reply 13 of 15, by brian.t

User metadata
Rank Newbie
Rank
Newbie
dominusprog wrote on 2024-10-18, 09:39:

It looks fine to me, what is the problem?

With the external cache enabled, the bios goes to the blinking cursor. I'm going to try to put some of the tag pins on the logic analyzer or oscilloscope and see if it's attempting to access the tag when the bios tries to test the cache.

Reply 14 of 15, by dominusprog

User metadata
Rank Oldbie
Rank
Oldbie

Have you ever tried another CPU? Like @Horu mentioned above, the K6-III has an 256KiB on-die L2 cache.

TriLevel Cache
The original K6-2 had a 64 KB primary cache and a much larger amount of motherboard-mounted cache (usually 512 KB or 1024 KB but varying depending on the choice of motherboard). The K6-III, with its 256 KB on-die secondary cache, re-purposed the variable-size external cache on the motherboard as the L3 cache. This scheme was termed "TriLevel Cache" by AMD. The L3 cache has a capacity of up to 2 MB.

Duke_2600.png
A-Trend ATC-1020 V1.1 ❇ Cyrix 6x86 150+ @ 120MHz ❇ 32MiB EDO RAM (8MiBx4) ❇ A-Trend S3 Trio64V2 2MiB
Creative AWE64 Value ❇ 8.4GiB Quantum Fireball ❇ Win95 OSR2 Plus!

Reply 15 of 15, by brian.t

User metadata
Rank Newbie
Rank
Newbie
dominusprog wrote on 2024-10-18, 13:26:

Have you ever tried another CPU? Like @Horu mentioned above, the K6-III has an 256KiB on-die L2 cache.

That is a good point. I will try a K6-2 first.