Have you tried using a different ISA slot?
Another thing you could try is to use a different cable cuz you never know 😜
Same OS?

Yeah, I tried several ISA slots... the serial controllers are detected just fine, all 4 ports listed in the post table. I use DOS 6.22 and tried both CuteMS and mouse.com - no dice.

Win 3.11 won't detect a mouse either...

kanecvr wrote:

Yeah, I tried several ISA slots... the serial controllers are detected just fine, all 4 ports listed in the post table. I use DOS 6.22 and tried both CuteMS and mouse.com - no dice.

Win 3.11 won't detect a mouse either...

I haven't used a 386 board for ages, but could it be possible something in the BIOS is disabled?
Or maybe some conflict of some sort?
Does your board's BIOS have any features that can be disabled for testing purposes?

the board doesn't have on board anything, witch means it has no settings for COM ports or anything remotely related to mouse operation.

The isa multi I/O controller I'm using is working fine - fdd and HDD work perfecty, and I was able to use my mouse with said controller on another motherboard, so that's not to blame.

I'm thinking the problem might be related to the keyboard controller. When I got the board it was really damaged and dirty - broken traces, a missing capacitor, the works. I repaired the broken traces and added another capacitor in place of the missing one, but the KB controller was missing as well. I have no ideea what originally came with this board - probably an AMYKEY controller, but I don't have one of those on hand so I'm using an Opti controller I got off a dead 486 board. I also have a nec and an intel kb controller on hand, it seems to like the opti best.

Problem is, it doesn't work perfectly. If I press num lock, the light will not go on and the keypad will not work. Sometimes it makes the keyboard not work at all requiring me to reset the machine.

So - could it be related to the keyboard controller somehow?

P.S. Merry Christmas!

Have you tried using a different serial mouse?

You mentioned you tried your serial mouse on another computer, but you didn't say you used a different serial mouse on this computer.

I don't think the keyboard controller had anything to do with COM ports, but what you describe does sound similar to an IRQ conflict.

Years ago I was given a free 1200 baud internal modem, and I plugged it into my 286, but I could never get it to work. It wasn't until years later I discovered that the modem used the same IRQ as one of the COM ports on the serial card. I changed the serial card to COM1 and COM3 (which normally share an IRQ) and set the modem to COM2 (which uses a different IRQ) and it worked - only about 10 years too late as I had a 56k modem and Internet access by then.

If you have four serial ports in your machine across two controllers then that is definitely asking for trouble. You can certainly do this, but I'd wait until you have a working setup before trying it, because it can require some fiddling to get working.

Are the boards (when you use them one at a time) configured for the correct I/O and IRQ settings for COM1 and COM2?

Are you positive your motherboard doesn't have onboard serial ports that could be conflicting with the ISA card ones?

I had the second controller set up as com 3 and com 4, IRQ 10 and IRQ 12 - the first controller (the one with IDE uses IRQ 3 and IRQ 4, com 1 and com 2). The mouse works on other machines.

I plugged in the second controller thinking there was something wrong with the first, but it's not. The only other device in the machine is a Cirrus Logics ISA video card.... I don't see what it would conflict with since the board does not have on board I/O: http://stason.org/TULARC/pc/motherboards/S/SO … 86-SY-019C.html

Frankly I'm out of ideas...

To recap:
- mouse works on other PCs
- I/O card works correctly on a 486, mouse is detected fine
- Serial controller (2nd one) also works fine on other machines, mouse is detected right away
- motherboard does not have on board I/O whatsoever

Right now I removed the 2nd controller and only have com 1 (irq 3) and com 2 (irq 4) from the i/o card.

One other thing - the capacitor I had to replace was a 4uf (4.7?) tantalum cap, and I had to replace it with a 10uf cap since I had no 4uf cap available... it's right near the NPU socket - board will not post w/o it...

Interesting. I guess it's possible that the lines to the keyboard controller are problematic and the keyboard IRQ is somehow mixed up with IRQ3/4, but then if that was the case, the other COM ports on IRQ 10/12 should have worked.

I presume you tested on all four COM ports, correctly setting the IRQs in the mouse driver for COM3 and 4?

Do you have a null-modem cable? Can you connect one port to another machine and see if you can launch a terminal program and send data across the port?

Perhaps before that, you could try running a diagnostic program? Somewhere on this forum (sorry I forget where now) is diags.zip, a large archive of various types of diagnostic programs. There's an AMI diagnostic which can test an internal loopback feature of some (all?) serial ports, so that might be worth a try as well. That should certainly indicate whether there's any kind of hardware conflict.

Ok, got it working. I reset CMOS, removed the KB controller, turned the machine on and got Gate A20 error. Turned the machine back off, put the controller in the socket, booted up and mouse now works fine! (in DOS)

The board doesn't like this KB controller. It doesn't like the nec either - throws a KB interface error. It doesn't even post with the intel.... I need an Amikey or AMD controller most likely...

Still won't work with win3.11 - machine hangs after loading windows or will not take any input...

kanecvr wrote:

Ok, got it working. I reset CMOS, removed the KB controller, turned the machine on and got Gate A20 error. Turned the machine back off, put the controller in the socket, booted up and mouse now works fine! (in DOS)

The board doesn't like this KB controller. It doesn't like the nec either - throws a KB interface error. It doesn't even post with the intel.... I need an Amikey or AMD controller most likely...

Still won't work with win3.11 - machine hangs after loading windows or will not take any input...

Do you happen to have a correct KB controller in another (working) board you could borrow from it?

Problem is I don't know what the correct kb controller for this board is in the first place... 🙁 would any of you would happen to have a Soyo SY-019C?

What bios does it use? If its an AMI its needs an AMI keyboard controller.

Yup - AMI bios. It would work with and amikey kb controller but not necessarily. I have other ami bios boards (including 386) and none have an amikey kb controller - in fact my other boards use either Opti, AMD or NEC controllers...

I was encountering this on my Soyo 486 too, I tried about million drivers and finally found one that worked (well, sort of, it was using HUGE chunk of memory).

Yeah, mine doesn't work right. There's been work done on one of the ICs under the KB controller (the two installed in the middle of the KB controller socket) so I'm betting the controller that has been replaced was replaced with a similar but not identical one.

Regardless of what keyboard I plug in it will not work correctly. Plugging in one of my AT keyboard yields either a "keyboard error" or only half the kb works. Using a PS2 keyboard /w ps2 to at adapter kind of works, but the num pad only works in arrow key mode and pressing num lock makes the machine ignore all input.

My best bet would be a pic of under the KB controller and of said controller.

Mouse works now, KB still doesn't work right. And it hang as soon as it finishes loading win 3.11

i had problems getting my mouse to work about 1 month ago and it turned out that the serial cable from the host card to the port was wired incorrectly..
the only way to visually check this is to remove the black cap from the connector cable and push it back to expose the wiring scheme.. i had purposly replaced the cable with a new
one to make sure it was good not knowing that these cables could be wired differently inside the end connector