VOGONS


First post, by tsalat

User metadata
Rank Member
Rank
Member

Hi everyone,

I am struggling to get the ABit AB-AT3 running and I am stuck on a 13-12 code on the post card. The board came with not such a bad condition, some damage from the battery around the keyboard connector but no trace broken. Another damage was around the CPU, surprisingly it seems like most of the damage was taken by the pins connected to the ground - this was already handled by the guy who had the board before and has removed the oxidation, like a butcher but removed. One pin was detached on the CPU but I was able to solder it back and using the reference from theretroweb verify the connection.

The board indicates no RAM when no RAM is inserted. After inserting the 4x 30pin sticks and with or without the 72pin RAM, the board stops at 13-12 post card code.
I have checked all pins on the CPU, Chipset and so on, and all all solidly soldered. I have checked a few on the CPU around the oxidation and all are on the board and are having continuity. I had a board with similar behaviour and the reason for it was a detached DMA controller leg but here I am not able to find any so far. I have removed the cache, and as well tried a different one with the same result.

Q1: Any idea what could be blocking the booting process? Or what to try to verify?
Q2: Is there any cache tester that I could buy? or a way how to test them?

Pictures of the board are attached.
reference: https://theretroweb.com/motherboards/s/abit-ab-at3#docs

br, tomas

Reply 1 of 6, by nuno14272

User metadata
Rank Member
Rank
Member

1) is the ram sticks good ?
2) is a jumper to select 30pin or 72 pin ?
3) can be damage traces under de cpu.

1| 386DX40
2| P200mmx, Voodoo 1
3| PIII-450, Voodoo 3 3000

Reply 2 of 6, by tsalat

User metadata
Rank Member
Rank
Member
nuno14272 wrote on 2025-02-17, 14:42:

1) is the ram sticks good ?
2) is a jumper to select 30pin or 72 pin ?
3) can be damage traces under de cpu.

1| yes, they stick good in the slots, if I try to misalign them, I got a beep - forgot to mention, while 3x sticks are in, the board stops at 13-12 and beeps, with 4x sticks it stops at 13-12 but no deep
2| no jumper
3| yes, could be, I am using the reference image from retroweb with the 486 CPU to see what pin of the 386 CPU goes where under the CPU but so far all my measurements are showing continuity

edit: i took out the power connector as well but there were no indication of battery damage underneath.

Reply 3 of 6, by nuno14272

User metadata
Rank Member
Rank
Member

You nedd all the 4 rams inserted.

1) Are all the for ram the same value ? (amount of ram and nanoseconds)?
2) Disable the onboard cache and boot. several times the cache memory is corrupted.

1| 386DX40
2| P200mmx, Voodoo 1
3| PIII-450, Voodoo 3 3000

Reply 4 of 6, by tsalat

User metadata
Rank Member
Rank
Member
nuno14272 wrote on 2025-02-18, 11:36:

You nedd all the 4 rams inserted.

1) Are all the for ram the same value ? (amount of ram and nanoseconds)?
2) Disable the onboard cache and boot. several times the cache memory is corrupted.

1) all are the same value, 1MB. I have tried as well 4x 256kB, the same. I have as well tried Bank0 + Bank1/2 with 256 and 1MB configuration and 1MB and 4MB 72pin memory with the same result.
2) no jumper to do so, with no cache the symptoms are the same. I have checked the board yesterday, and especially the DIP sockets for the cache. It seems that there is some corrosion underneath, hence, will remove them to see the damage. will take some time to finish, heh.

Size Bank 0 --- Bank 1/2
1MB (4) 256K x 9 --- NONE
2MB (4) 256K x 9 --- (1) 256K x 36
4MB (4) 1M x 9 --- NONE
5MB (4) 1M x 9 --- (1) 256K x 36
8MB (4) 1M x 9 --- (1) 1M x 36
16MB (4) 4M x 9 --- NONE
17MB (4) 4M x 9 --- (1) 256K x 36
20MB (4) 4M x 9 --- (1) 1M x 36
32MB (4) 4M x 9 --- (1) 4M x 36
64MB (4) 16M x 9 --- NONE
192MB (4) 16M x 9 --- (1) 32M x 36

Reply 5 of 6, by nuno14272

User metadata
Rank Member
Rank
Member

You can just take the cache off, and the tag, and try to boot without nothing.. just to be shure the problem is not in the cache.

The left error code is the current. the board passed 12 and stoped at 13.

usually memory errors gives beeps.. connect the speaker.. (attention to ground and data pins). memory is ussually repeted smal beeps.

Error Code – 13

AMI (13)DMA controller#1,#2,interrupt controller#1,#2disa- bled. About to disable Video display and initialize port-B. Chipset initialize/auto memory detection about to begin. Replace first memory SIMM.(13)Chipset initialize/auto memory detection about to begin. Check first SIMM.(13) Interrupt vectors initialized.

AST (13)Initialize video.

Award (13)Test 8254 timer 0 channel 1. Test keyboard controller.

Compaq (13)Test timer 0.

Chips & Tech (13)LSL instruction failure.

Phoenix&Dell [Beep]=2-1-4 1st 64K RAM chip or data line failure-bit 3. Initialize PCI Bus Mastering devices

1| 386DX40
2| P200mmx, Voodoo 1
3| PIII-450, Voodoo 3 3000

Reply 6 of 6, by tsalat

User metadata
Rank Member
Rank
Member
nuno14272 wrote on 2025-02-19, 15:52:
You can just take the cache off, and the tag, and try to boot without nothing.. just to be shure the problem is not in the cache […]
Show full quote

You can just take the cache off, and the tag, and try to boot without nothing.. just to be shure the problem is not in the cache.

The left error code is the current. the board passed 12 and stoped at 13.

usually memory errors gives beeps.. connect the speaker.. (attention to ground and data pins). memory is ussually repeted smal beeps.

Error Code – 13

AMI (13)DMA controller#1,#2,interrupt controller#1,#2disa- bled. About to disable Video display and initialize port-B. Chipset initialize/auto memory detection about to begin. Replace first memory SIMM.(13)Chipset initialize/auto memory detection about to begin. Check first SIMM.(13) Interrupt vectors initialized.

AST (13)Initialize video.

Award (13)Test 8254 timer 0 channel 1. Test keyboard controller.

Compaq (13)Test timer 0.

Chips & Tech (13)LSL instruction failure.

Phoenix&Dell [Beep]=2-1-4 1st 64K RAM chip or data line failure-bit 3. Initialize PCI Bus Mastering devices

Hi and thank you for your answer,

I have tried already to remove the cache, exchange the cache, and remove it one by one as well until there was nothing in the DIP sockets but the behaviour is the same.
The beep code is only with no ram or 3x sticks, while fully occupied, there is no beep code at all. I have also tried to exchange the sticks, add pressure on them, and checked visually the RAM sockets - no issues. Also, I have remowed the power connector to see if the battery acid went below but no traces there - I hope nothing broken beneath the ram sockets.

I dont see any damage after removing the cache through the DIP socket as well - I want to avoid de-soldering it since it would be looking "different" but if there will be nothing to try later on, I will go for it.
The AMI description of the code is appreciated, I will check the DMA controller a bit more to see if all lines are connected, however, I am not able to find a datasheet for the ALI M1431 - anyone have something?

br, Tomas