Reply 20 of 40, by Predator99
No Post Code? Tried another Slot? Any beep Code?
No Post Code? Tried another Slot? Any beep Code?
wrote:No Post Code? Tried another Slot? Any beep Code?
I think it actually completes POST. Beep codes are normal. I will cycle through the POST codes on the POST card this evening.
On a cold start, it stops at the WAIT.... as seen on the picture. I can then reboot it using CTRL-ALT-DEL (so it's not completely unresponsive). After this warm start, it will display Press <DEL> to enter setup (/bios?), but then it is completely unresponsive, CTRL-ALT-DEL does not respond (reset button will of course).
EDIT: found some time for a quick check. No POST codes on the POST card (tried multiple ISA slots)
EDIT2: been experimenting a bit with cache chips and cache jumpers, but no difference...
1982 to 2001
No POST codes shouldn't happen on this mobo - perhaps you have one of the address or data bit lines busted, that would also explain the hang. But do check if your test card is not set to XT class system (I haven't seen a modern ISA/PCI test card capable of that though).
Try booting without any memory sticks but speaker connected. If you don't get any beeps it's even more likely the address lines are suspect. D0-D7 on ISA should be OK if you get all that text displayed by the VGA card.
wrote:No POST codes shouldn't happen on this mobo - perhaps you have one of the address or data bit lines busted, that would also explain the hang. But do check if your test card is not set to XT class system (I haven't seen a modern ISA/PCI test card capable of that though).
Try booting without any memory sticks but speaker connected. If you don't get any beeps it's even more likely the address lines are suspect. D0-D7 on ISA should be OK if you get all that text displayed by the VGA card.
No option to set the POST card to XT class (and it works on another 386 board just fine). When I don't insert any (or the wrong amount) of SIMMs, I get a proper beep error.
1982 to 2001
Well at least there's that. I'd still check all the ISA address lines, all those should be connected to one of the big chips.
Well, since the kid was sleeping I could do a visual check of the board. No visible damage. I also reseated all socketed chips and reflowed some of the worse looking solder connections (esp. keyboard and in the battery corrosion area).
1982 to 2001
Visual inspection might not be enough. Use this pinout: https://upload.wikimedia.org/wikipedia/common … SA_Bus_pins.svg
Check all address lines (but above all A0-A11 on the 8-bit slot part) with a meter set to beep on continuity. All of them must be connected to the same big chip, it's not important which one exactly or what pin. Also, check all for shorts to ground and +5V while at it.
wrote:wrote:No Post Code? Tried another Slot? Any beep Code?
I think it actually completes POST. Beep codes are normal. I will cycle through the POST codes on the POST card this evening.
On a cold start, it stops at the WAIT.... as seen on the picture. I can then reboot it using CTRL-ALT-DEL (so it's not completely unresponsive). After this warm start, it will display Press <DEL> to enter setup (/bios?), but then it is completely unresponsive, CTRL-ALT-DEL does not respond (reset button will of course).
If this is the case then it be to do with the FPU / math co-processor.
Either put one in or disable it!
Currently, your system is set up for one and is looking but cannot find.
wrote:If this is the case then it be to do with the FPU / math co-processor. […]
wrote:wrote:No Post Code? Tried another Slot? Any beep Code?
I think it actually completes POST. Beep codes are normal. I will cycle through the POST codes on the POST card this evening.
On a cold start, it stops at the WAIT.... as seen on the picture. I can then reboot it using CTRL-ALT-DEL (so it's not completely unresponsive). After this warm start, it will display Press <DEL> to enter setup (/bios?), but then it is completely unresponsive, CTRL-ALT-DEL does not respond (reset button will of course).
If this is the case then it be to do with the FPU / math co-processor.
Either put one in or disable it!
Currently, your system is set up for one and is looking but cannot find.
There is no jumper for an FPU. I also don't have one available at the moment. I think someone earlier said it is autosensed.
1982 to 2001
wrote:Visual inspection might not be enough. Use this pinout: https://upload.wikimedia.org/wikipedia/common … SA_Bus_pins.svg
Check all address lines (but above all A0-A11 on the 8-bit slot part) with a meter set to beep on continuity. All of them must be connected to the same big chip, it's not important which one exactly or what pin. Also, check all for shorts to ground and +5V while at it.
I tried this for two channels, but could not find a connection. I will have to look for some datasheets of the chips.
1982 to 2001
Nah, if you can't find any connection at all then the chips are using the '244, '245 and '373 (which your board has plenty of) as buffers/latches and that complicates things. You're unlikely to debug that with just a meter at your disposal.
There is some battery corrosion though so check this area for any broken connections, weird things happen when the CMOS/RTC and keyboard controller are not powered properly.
wrote:Nah, if you can't find any connection at all then the chips are using the '244, '245 and '373 (which your board has plenty of) as buffers/latches and that complicates things. You're unlikely to debug that with just a meter at your disposal.
There is some battery corrosion though so check this area for any broken connections, weird things happen when the CMOS/RTC and keyboard controller are not powered properly.
Yeah suspected something like that. I can find the IRQ lines going to/from the SiS85C206. I have checked all the traces that seem affected by corrosion (it's not really bad fortunately). The multimeter says everything is connected.
1982 to 2001
...
...
... trust me... it's the missing FPU.
Bios or jumper, disable it... and you will pass this "Wait".
wrote:... […]
...
...
... trust me... it's the missing FPU.
Bios or jumper, disable it... and you will pass this "Wait".
Yes, can confirm this. I remember to also solved such issue with the FPU jumper in the past. Or the missing drive controller.
Dont think the board is damaged, otherwise you wouldnt have got such far.
Did you test your POST-card in another board?
wrote:... […]
...
...
... trust me... it's the missing FPU.
Bios or jumper, disable it... and you will pass this "Wait".
I have this as a top option, but I don't have an FPU, there is no jumper and I can't get in the BIOS. It would be a pretty weird design that the CMOS default is FPU enabled/needed?
The POST card works fine in many other boards.
1982 to 2001
wrote:wrote:... […]
...
...
... trust me... it's the missing FPU.
Bios or jumper, disable it... and you will pass this "Wait".
there is no jumper and I can't get in the BIOS. It would be a pretty weird design that the CMOS default is FPU enabled/needed?
If FPU is auto-sensed... something I've never really seen before, and without an FPU, the Bios is the only option.
To get into the Bios, you will have to.... Re-set / Clear it with the jumper, every time you get a..."Wait".
Once you Reset / Clear the Bios with jumper, and reboot the system... you have to be very quick F1, F1, F1, F1... or whatever the key to enter the Bios, before it loads the rest of POST... and of course... "Wait"!
If there is nooooooooo way whatsoever into the Bios, then the only option is to install an FPU.
wrote:If FPU is auto-sensed... something I've never really seen before, and without an FPU, the Bios is the only option. […]
wrote:wrote:... […]
...
...
... trust me... it's the missing FPU.
Bios or jumper, disable it... and you will pass this "Wait".
there is no jumper and I can't get in the BIOS. It would be a pretty weird design that the CMOS default is FPU enabled/needed?
If FPU is auto-sensed... something I've never really seen before, and without an FPU, the Bios is the only option.
To get into the Bios, you will have to.... Re-set / Clear it with the jumper, every time you get a..."Wait".
Once you Reset / Clear the Bios with jumper, and reboot the system... you have to be very quick F1, F1, F1, F1... or whatever the key to enter the Bios, before it loads the rest of POST... and of course... "Wait"!
If there is nooooooooo way whatsoever into the Bios, then the only option is to install an FPU.
I will try this a few times, it might work. Would still be really hard to work with this board but who knows 😀
I might be able to source an FPU, might take some time though 😀
1982 to 2001
wrote:<snip> If FPU is auto-sensed... something I've never really seen before, and without an FPU, the Bios is the only option. […]
<snip>
If FPU is auto-sensed... something I've never really seen before, and without an FPU, the Bios is the only option.To get into the Bios, you will have to.... Re-set / Clear it with the jumper, every time you get a..."Wait".
Once you Reset / Clear the Bios with jumper, and reboot the system... you have to be very quick F1, F1, F1, F1... or whatever the key to enter the Bios, before it loads the rest of POST... and of course... "Wait"!
If there is nooooooooo way whatsoever into the Bios, then the only option is to install an FPU.
I tried this a few times, but could not get into the BIOS.
I also let it sit at the Wait prompt for over 30 minutes, but no result.
Time to arrange an FPU...
1982 to 2001
Tested with a borrowed FPU, but this made no difference.
Also tried with a different POST card (a proper old one instead of an eBay one). This gave me two POST codes!
At cold boot: 85
After CTRL-ALT-DEL: 4B
I also dumped the BIOS contents and compared them to a copy I found online. They are identical.
1982 to 2001
Another update! And finally the board seems to be working completely! Today I decided to look at the board again and to do some more de-soldering and analysis. Since I still suspected the problem to be related to the (little) acid damage near the keyboard connector, I removed the keyboard connector and power connectors. There was some acid staining, but everything was checked with the multi-meter and everything had continuity.
After that I did a thorough visual inspection, and something underneath one of the sockets with the previously removed logic chips caught my eye. I removed the socket and:
Luckily I have been watching a few RetroRepairs videos in YouTube so I knew what to do:
And after reassembly she seems to work a lot better!
So happy to (hopefully) finally close this repair project.
1982 to 2001