Last picture

Reseat all BIOS chips, sometimes they just lost cotacts. Carefully clean all contacts on isa/vlb groups.
PS. Attach photos!

That was my first try, forget to write here.

Pls, connect PC-speaker to SPKR pins and listen are there any sound out ?

Just a very short beep, not as a long what usually in bios

envagyok wrote on 2023-01-18, 14:19:

Just a very short beep, not as a long what usually in bios

So...

0) Connect LED to JP1
The LED connected to JP1 flashes some diagnostics right after power-up. The following shows the flash frequency of each fatal error condition:
- One short flash: No Cache DRAM installed
- Two short flashes: CPU error
- Three short flashes: SRAM error
-Four short flashes: Timer error

1) Check your motherboard BIOS for "If the system contains a 16-bit VGA adapter with a Fast address Decode option, disable that feature. Due to its wider address decoding range, this feature might cause some potential conflicts with the DC-6X0 or any other cards implemented with 8-bit memory mapped I/O."

2) Send screenshots of ALL BIOS screens of your system

3) Check SIMM slots for jammed contacts or "down-braked" pins

https://minuszerodegrees.net/manuals/Tekram/T … 80%20DC-690.pdf

Without cache ram installed 3 short flashes after and after looped
with cache ram installes 1 very short flash, at the moment i turn on the pc
Here is video from screen:
https://www.youtube.com/watch?v=sg6RgxF0Vok&a … 3%A1nHaj%C3%B3s

The Tekram controller answers to the IDE identify command only if it passed its internal POST. As you see two drives recognized, we can immediately conclude that none of the four fatal error conditions apply on your system. Furthermore, the "fast address decode" issue for the VGA card only applies for VGA cards that allow 16-bit BIOS access. Some ISA cards provide 16-bit BIOS access, but as the VGA BIOS is more often shadowed than not in 486 systems, nearly no VL VGA card claims 16-bit cycles in the C000-DFFF address space.

The BIOS on the Tekram controller is accessed from the ISA bus, whereas the IDE ports are accessed through the VL bus. Your issue looks like the communication between the ISA bus connector and the TRM680C is not working properly, maybe a trace is broken. The photo of the back side of your controller shows a scratch, which might be worth investigating. The scratch doesn't look that bad, but I'm not sure no trace is damaged. If you have a meter, you shoud check whether you get continuity to some TRM680C pins from all 20 ISA address pins (A12-A31), the low 8 ISA data pins (A2-A9) and the ISA SMEMR pin (B12). You could also check whether you get something on the ISA bus at C800:0, CC00:0, D000:0, D400:0, D800:0 or DC00:0. If you do, the kind of data you observe might give a clue to the type of the fault. Possibly a tool like CHECKIT or MSD also shows which address ranges in the upper memory area are occupied by ROMs, even if they don't seem to contain valid BIOS extensions.

Babasha wrote on 2023-01-18, 14:54:

The LED connected to JP1 flashes some diagnostics right after power-up. The following shows the flash frequency of each fatal er […]
Show full quote

The LED connected to JP1 flashes some diagnostics right after power-up. The following shows the flash frequency of each fatal error condition:
- One short flash: No Cache DRAM installed
- Two short flashes: CPU error
- Three short flashes: SRAM error
-Four short flashes: Timer error

It seems these codes are for the old hardware revisions that had dedicated SRAM for the CPU as well as cache DRAM. I took a peak at the DC680C firmware, and it doesn't contain the "one flash" logic any more. As observed later in this thread, missing Cache DRAM causes the "3 flashes" error code. That error code is in fact generated if the 64K working RAM for the 80286 does not respond properly. Obviously, the TRM680C chip is able to map the 80286 RAM address space into the DRAM, so no dedicated "Cache error" code is needed anymore.

mkarcher wrote on 2023-01-18, 19:18:

If you have a meter, you shoud check whether you get continuity to some TRM680C pins from all 20 ISA address pins (A12-A31), the low 8 ISA data pins (A2-A9) and the ISA SMEMR pin (B12). You could also check whether you get something on the ISA bus at C800:0, CC00:0, D000:0, D400:0, D800:0 or DC00:0. If you do, the kind of data you observe might give a clue to the type of the fault. Possibly a tool like CHECKIT or MSD also shows which address ranges in the upper memory area are occupied by ROMs, even if they don't seem to contain valid BIOS extensions.

I check it. Sratches isnt problem, no discontinuities.
A12-A31 and B12 is connected to Tekram Chip, A2-A9 is connected to Goldstar Floppy controller.
Here is a photo from checkit:

envagyok wrote on 2023-01-19, 08:53:

I check it. Sratches isnt problem, no discontinuities.
A12-A31 and B12 is connected to Tekram Chip, A2-A9 is connected to Goldstar Floppy controller.

Also, A31 to B22 is connected to the Goldstar floppy controller, and those connections are obviously OK, as the floppy controller works. Furthermore, A2-A9 is also connected to the Tekram chip. Those traces look fine.

envagyok wrote on 2023-01-19, 08:53:

Here is a photo from checkit:

That photo shows that the card does not seem to respond to ROM cycles at any address. The only lines required for the TRM680C to detect ROM requests are A12-A31 (address lines) and B12 (memory read request), and you checked those. Even if one of the data lines (A2-A9) were broken, you should see something in CheckIt. If the design of the DC-680C is similar to the DC-680T (the earlier variant, which I own), there is a small stub ROM integrated into the TRM680C chip, which is enabled under firmware control (for 2.x firmware), whereas 1.x firmware required an external host ROM. The 680C still has the solder positions for a ROM chip and an 8-bit buffer chip near the GoldStar, but as it never shipped with 1.x firmware, these components were not populated.

If the firmware was corrupted, the integrated stub ROM is possibly not enabled, but we know that the firmware works and passes its internal checksum test, because we observed correct behaviour with and without cache RAM, so it is highly unlikely that the firmware is corrupted. So I am out of ideas why your computers do not see the ROM.

Oh… where the screenshots of all BIOS settings?

Where stored the hdd settings?
Can i delete without entering tekram bios?
I think maybe corrupted some data, and that's the reason. The chs settings read from hwinfo is wrong too, one of hdd had only 0 heads there, that's impossible.

Or maybe bios chip has corrupted?

Sadly i don't have spare controller, i can't test swapping bios ic's

envagyok wrote on 2023-01-20, 05:52:

Where stored the hdd settings? Can i delete without entering tekram bios? I think maybe corrupted some data, and that's the reas […]
Show full quote

Where stored the hdd settings?
Can i delete without entering tekram bios?
I think maybe corrupted some data, and that's the reason. The chs settings read from hwinfo is wrong too, one of hdd had only 0 heads there, that's impossible.

Or maybe bios chip has corrupted?

Sadly i don't have spare controller, i can't test swapping bios ic's

You dont need spare card.
Just flash on programmer BIOS image - http://files.mpoli.fi/hardware/HDD/TEKMAR/

Babasha wrote on 2023-01-20, 06:28:

envagyok wrote on 2023-01-20, 05:52:

Where stored the hdd settings? Can i delete without entering tekram bios? I think maybe corrupted some data, and that's the reas […]
Show full quote

Where stored the hdd settings?
Can i delete without entering tekram bios?
I think maybe corrupted some data, and that's the reason. The chs settings read from hwinfo is wrong too, one of hdd had only 0 heads there, that's impossible.

Or maybe bios chip has corrupted?

Sadly i don't have spare controller, i can't test swapping bios ic's

You dont need spare card.
Just flash on programmer BIOS image - http://files.mpoli.fi/hardware/HDD/TEKMAR/

Sadly i don't have flasher.
Has any modern flash chip compatible with old socket, and voltages?

And settings where stored, how can i delete?

envagyok wrote on 2023-01-20, 05:52:

Where stored the hdd settings?

It is stored in a 93c46 EEPROM (or something compatible to it). As I don't find an 8-pin DIP or SOIC IC on the pictures of the DC-680C card in the internet, I guess that chip is likely mounted below the socketed ROM chip. Maybe the EEPROM is no longer a dedicated chip, but integrated into the TRM680C.

envagyok wrote on 2023-01-20, 05:52:

Can i delete without entering tekram bios?

The EEPROM is only accessible from the 80286 processor on the Tekram card. So you would need to send a "delete the EEPROM" command to the Tekram firmware. There is no simple delete command, but there is an EEPROM write command. I could tell you how to write a program that reprograms the EEPROM with contents with a bad checksum, so the Controller will fall back to a valid default configuration, but I have an even better idea:

You already got a link to the tekram Firmware, which includes the host BIOS, it's in BDC680CR.ZIP in the file DC680C.FRM. You can load it into conventional RAM using DEBUG, and jump into the setup routine like this:

1C:\>DEBUG
2-nDC680C.FRM
3-l5000:0
410000 bvtes loaded to 5000:0000
5-g5000:1939

Or maybe bios chip has corrupted?

The combined Firmware/BIOS chip might be corrupted. The firmware calculates a checksum over the ROM during POST, but ignores the result (I just re-checked it). Damage to the BIOS/Firmware ROM is not very likely, though.

I try it, but not see anything, only cursor blinks

envagyok wrote on 2023-01-21, 12:53:

I try it, but not see anything, only cursor blinks

Ouch, sorry for your inconvenience. I messed up the DEBUG syntax. The correct command in the last line would be "g=5000:1939". Note the extra equal sign, which I forgot.

Thank you, its working.

When i enter to settings, there is every drive empty, i try filled one connected drive's info.
Then restart the pc, run hwinfo, and there again the old drive settings with 2 drive what i photo from first 2 message this thread.
I try run again setup, i see again every drive empty 🙁 (but with hwinfo there is a 2 old drive sewttings)
Sadly not store the settings, and remain the old settings, and no see f2/f6 login screen at start.

@mkarcher
"The EEPROM is only accessible from the 80286 processor on the Tekram card. So you would need to send a "delete the EEPROM" command to the Tekram firmware. There is no simple delete command, but there is an EEPROM write command. I could tell you how to write a program that reprograms the EEPROM with contents with a bad checksum, so the Controller will fall back to a valid default configuration"

its not to difficult?