fsinan wrote on 2024-04-21, 08:02:

I bought this board and found out that the cache is not working. It seems there is 256K cache on board but was not working accor […]
Show full quote

I bought this board and found out that the cache is not working. It seems there is 256K cache on board but was not working according to cache tests.

Cache chips are fake and fake cache bios is loaded on board. Bought a set of cache chips. Tried out 128K config and 512K config with UMC256K srams and ISSI 1024K srams. Tried out 20 chips 10 from each other and they did not work how many combinations I tried, at that time I did not have a TL688 or T48 programmer to test the chips.

System does not boot into DOS while external cache is enable even with slowest settings of cache. When cache disabled, it is OK.

Then I bought a T48 and tested all srams, all new chips that I bought were tested OK. The ones came with board were all fake as we understood before from the printings and fake cache bios which is listed on the retroweb too.

Then I thought maybe if I use the normal bios, things may work and program a new bios chip. Installed it, again failed with external cache enabled.

Then I borrowed a set of working cache from my other motherboard with same -15 speed and tested, again failed. I tested my new cache chips on that board and they worked.

It seems new cache chips are ok and bios is new but this ST1A-JK-042E Rev.B mobo is not working with cache even with slowest settings.

Any ideas? Might be a hidden feature to set in bios or a hw problem?

That does not sound good.
I guess it is a hardware problem. Since the board was shipped with fake cache chips, for sure the producer of the board did not test the board with real cache.
On this picture
https://theretroweb.com/motherboards/s/u-boar … rev-semicolon-b
there are resistors on the non equipped F244 place (between first and second PCI slot)
No clue if this has any relevance.
Maybe an other Vogons user owns such a board, too and can tell.

CoffeeOne wrote on 2024-04-22, 17:44:

That does not sound good. I guess it is a hardware problem. Since the board was shipped with fake cache chips, for sure the prod […]
Show full quote

fsinan wrote on 2024-04-21, 08:02:

I bought this board and found out that the cache is not working. It seems there is 256K cache on board but was not working accor […]
Show full quote

I bought this board and found out that the cache is not working. It seems there is 256K cache on board but was not working according to cache tests.

Cache chips are fake and fake cache bios is loaded on board. Bought a set of cache chips. Tried out 128K config and 512K config with UMC256K srams and ISSI 1024K srams. Tried out 20 chips 10 from each other and they did not work how many combinations I tried, at that time I did not have a TL688 or T48 programmer to test the chips.

System does not boot into DOS while external cache is enable even with slowest settings of cache. When cache disabled, it is OK.

Then I bought a T48 and tested all srams, all new chips that I bought were tested OK. The ones came with board were all fake as we understood before from the printings and fake cache bios which is listed on the retroweb too.

Then I thought maybe if I use the normal bios, things may work and program a new bios chip. Installed it, again failed with external cache enabled.

Then I borrowed a set of working cache from my other motherboard with same -15 speed and tested, again failed. I tested my new cache chips on that board and they worked.

It seems new cache chips are ok and bios is new but this ST1A-JK-042E Rev.B mobo is not working with cache even with slowest settings.

Any ideas? Might be a hidden feature to set in bios or a hw problem?

That does not sound good.
I guess it is a hardware problem. Since the board was shipped with fake cache chips, for sure the producer of the board did not test the board with real cache.
On this picture
https://theretroweb.com/motherboards/s/u-boar … rev-semicolon-b
there are resistors on the non equipped F244 place (between first and second PCI slot)
No clue if this has any relevance.
Maybe an other Vogons user owns such a board, too and can tell.

This is my board:
https://theretroweb.com/motherboard/image/img … c2909683688.jpg

fsinan wrote on 2024-04-22, 18:09:

CoffeeOne wrote on 2024-04-22, 17:44:

That does not sound good. I guess it is a hardware problem. Since the board was shipped with fake cache chips, for sure the prod […]
Show full quote

fsinan wrote on 2024-04-21, 08:02:

I bought this board and found out that the cache is not working. It seems there is 256K cache on board but was not working accor […]
Show full quote

I bought this board and found out that the cache is not working. It seems there is 256K cache on board but was not working according to cache tests.

Cache chips are fake and fake cache bios is loaded on board. Bought a set of cache chips. Tried out 128K config and 512K config with UMC256K srams and ISSI 1024K srams. Tried out 20 chips 10 from each other and they did not work how many combinations I tried, at that time I did not have a TL688 or T48 programmer to test the chips.

System does not boot into DOS while external cache is enable even with slowest settings of cache. When cache disabled, it is OK.

Then I bought a T48 and tested all srams, all new chips that I bought were tested OK. The ones came with board were all fake as we understood before from the printings and fake cache bios which is listed on the retroweb too.

Then I thought maybe if I use the normal bios, things may work and program a new bios chip. Installed it, again failed with external cache enabled.

Then I borrowed a set of working cache from my other motherboard with same -15 speed and tested, again failed. I tested my new cache chips on that board and they worked.

It seems new cache chips are ok and bios is new but this ST1A-JK-042E Rev.B mobo is not working with cache even with slowest settings.

Any ideas? Might be a hidden feature to set in bios or a hw problem?

That does not sound good.
I guess it is a hardware problem. Since the board was shipped with fake cache chips, for sure the producer of the board did not test the board with real cache.
On this picture
https://theretroweb.com/motherboards/s/u-boar … rev-semicolon-b
there are resistors on the non equipped F244 place (between first and second PCI slot)
No clue if this has any relevance.
Maybe an other Vogons user owns such a board, too and can tell.

This is my board:
https://theretroweb.com/motherboard/image/img … c2909683688.jpg

Yes. It does not have those resistors over the missing F244.

Almost identical, the resistors, a capacitor, and U16 (no idea what that is) are present on one and not the other.

fsinan, is the photo of your board with the fake cache chips present ?

weedeewee wrote on 2024-04-22, 18:23:

Almost identical, the resistors, a capacitor, and U16 (no idea what that is)

https://www.ti.com/lit/ds/symlink/lp2951.pdf
something is powered by this regulator, no idea what tho

weedeewee wrote on 2024-04-22, 18:23:

Almost identical, the resistors, a capacitor, and U16 (no idea what that is) are present on one and not the other.

fsinan, is the photo of your board with the fake cache chips present ?

https://theretroweb.com/motherboard/image/img … c2909683688.jpg

fsinan wrote on 2024-04-23, 19:28:

weedeewee wrote on 2024-04-22, 18:23:

Almost identical, the resistors, a capacitor, and U16 (no idea what that is) are present on one and not the other.

fsinan, is the photo of your board with the fake cache chips present ?

https://theretroweb.com/motherboard/image/img … c2909683688.jpg

apparently this link got mangled. But I get that it's the same one as the one you previously posted,
So I'm going for Yes. that photo was the board with the fake cache chips present.