Is the mouse working on any system? Often mice cables get damaged. Run CheckIt serial ports test without loopback and see how it goes. If you can make a loopback connector it would be even greater since it will fully test the RS232 port.

Yes, tested many good working mice.

So this is really puzzling to put it mildly...after more tests, in fact with the ISA controller the mouse is detected as Mouse Systems Mode by ctmouse (alternatively, Logitech drivers 6.50 also finds it fine) and works fine in CheckIt Pro, FastTracker 2 and Space Quest 3. Went back to one VLB controller (QD6580 - only change in the whole config, everything is the same but swap the ISA IO to VLB IO) and the mouse is also detected as Mouse Systems Mode by ctmouse but Logitech doesn't find it, and it doesn't work at all in anything. Tested another VLB controller (Prime 2C) and same thing. Ran the whole CheckIt Pro 1.12 Serial test with the VLB and everything passes fine 🙁 no loopback.

I have a female DB9 connector I could build a loopback connector from, but googling for it gave me different wirings so would appreciate a confirmation of the proper one please.

This is really turning me crazy. This is too ridiculous and feels like I must be missing something obvious.

Ctmouse will report a "mouse systems" mouse even when nothing is plugged in to the serial port at all

Isn’t an « enable PS/2 mouse » jumper set on your mainboard, which would disable a serial mouse support or create a conflict of some sort? It shouldn’t…and on 2 different mainboard it’s very strange.

maxtherabbit wrote on 2024-01-20, 22:04:

Ctmouse will report a "mouse systems" mouse even when nothing is plugged in to the serial port at all

Indeed and why I initially thought it wasn't working even with the ISA IO card... used to the "Logitech" mode coming up when all is fine.
No it's not a jumper, and the Soyo 025K motherboard I'm focusing on right now I've definitely used for a very long time with a VLB IO, serial mouse and ctmouse 1.9 in the past...

Is the serial port on the vlb card or on a breakout cable? I have seen different pinouts being used for these, perhaps that is the problem. Otherwise there might be a problem with the 232 line driver or receiver on it, since it passes internal tests.

Check if checkit 3.0 tests sucessfully IRQ line assigned to your serial port. Which IRQ is that?

I'm pretty curious of how you've ended up in your testings as I've got the exact same problem: just swapped in a VLB I/O card and ctmouse (or even ms mouse.com for that matters) is not capable anymore to find my (perfectly working) mouse. If I switch back to my 16bit I/O it works just fine(Prime2C chip based)
Setup has 2 VLB cards, a Trident TGUI9400CXI in vlb slot 1, and I/O Astar VLMIO 1.4 (based on UMC 82C865F) in slot 0 (closest to CPU). Note that this placement seems the only one working properly, otherwise one card fails to work.
Set wait state 1 and <=33MHz on my mobo, already tried disabling any other jumper on the I/O card, but it seems just like it fails making the mouse work.
Additional info:
- Hwinfo16 sees correctly the IRQ from COMs
- Even when switching back to a 16 bit ISA only graphic card ctmouse fails finding the mouse
- Already tried a different ribbon cable for that matters

are you sure serial ports work on VLB card in the first place? Connect pins 2 and 3 together, open terminal program and type something to see if data is coming back, or even without linking plug and move mouse - there should be some garbage in serial terminal. Finally test if voltage generated by Serial Port is sufficient, maybe -12V on ISA slot is missing and card uses that?

To make a serial loopback cable, connect pins:

RS232 Signals:
TransmitData <> ReceiveData
RequestToSend <> ClearToSend
DataTeminalReady <> DataSetReady

DB9: 2<>3 7<>8 4<>6

DB25: 2<>3 4<>5 6<>20

You might also want to grab my SDT (SerialDebugTerminal)
This is a little terminal programs (SDT.COM runs under DOS)
which will let you:

Transmit Data characters
Raise/lower RTS & DTR (independantly)
See Receiveed Data characters
See current state of CTS & DSR

This with a loopback connector would be a good way to confirm that a serual
connection is working.

Easy place to get it is MySite->Downloads->DDLINK
(This is a file transfer tool, but I included SDT.COM as a way to help
test null-moden serial cables should you have to build them)

Dave ::: https://dunfield.themindfactory.com ::: "Daves Old Computers"->Personal

Thanks for the tips guys.

I've just tested with both a loopback DB9 and the serial mouse attached with CheckIt PRO 1.11 and the serial just passes all tests (when loopback dummy plug is in, modem tests fails, of course).

I've tested succesfully both by booting my environment, and from a vanilla DOS boot floppy environment, so I'd rule out env as a failure per se.

I'm will now test it with your SDT software, Dave.

Well, first thing first, I was not able to use SDT. I couldn't understand how to make the terminal actually work.

I've ended up testing COM3/4 instead of COM1/2, disabling LPT+GAME+COM2/4. Tried ctmoues 1.9 (stable) instead of latest beta. I've also tried unplugging any card, but the I/O, floppy, and HDD, and using a basic ISA 16 bit graphic card. None of these worked.

I've put my previous I/O card (Goldstar Prime 2C MKII) enabling it to just handle COMs and disabling the Astral VMLIO COMs. CTmouse is now happily finding my mouse back in Logitech mode, and it actually works.

I might just give another testing trial by putting the VLB Astral I/O card into a different slot and see if it ends up working.

Update: this VLB COMs just don't seem to be working. Swapped it in any VLB slot, even with various HW cards removal, nothing changes.

I've ended up putting back my Goldstar Prime 2C MKII card for using COMs, and jumper the whole setup for having COM1, COM2, COM3, COM4, LPT1, LPT2 coming from both cards. Mouse works fine now. Testing it as a whole with CKPRO, let's finger cross.

fix_metal wrote on 2024-12-22, 19:01:

Update: this VLB COMs just don't seem to be working. Swapped it in any VLB slot, even with various HW cards removal, nothing changes.

I've ended up putting back my Goldstar Prime 2C MKII card for using COMs, and jumper the whole setup for having COM1, COM2, COM3, COM4, LPT1, LPT2 coming from both cards. Mouse works fine now. Testing it as a whole with CKPRO, let's finger cross.

Could you post a photo or multiple photos of your VLB card, here or on some image hosting site like imgur, so we can clearly see the jumpers and connections, and post the link to them here.

fix_metal wrote on 2024-12-22, 17:56:

Well, first thing first, I was not able to use SDT. I couldn't understand how to make the terminal actually work...

It's got some more advanced/capable features to help test more complex
things, but just verifying the serial port works is pretty simple.

From a DOS command line:

> SDT p=1-4

Defaults to COM1 but you can pick the serial port with P=1 .. P=4

You will see a 24x80 terminal screen, with a single line across the bottom
showing the available function keys and the status of the serial signals:

1:dtr 2:rts 3:Asc/Asc 4:9600-None-8-1 5:Cls 6:Review 9:Help cd,ri,dsr,cts

dtr,rts,cd,ri,dsr,cts are the RS-232 signsls
- Show in lower case if signal is de-asserted
- Show in upper case if signal is asserted
DTR,RTS = output sgnals: DataTerminalReady,RequestToSend
CR,RI,DSR,CTS = inputs: CarrierDetect,RingIndicator,DataSetReady,ClearToSend

Press:
F1 toggle DTR
F2 toggle RTS
F3 toggle Ascii/Hex rx/tx modes
F4 presents a menu to set COM: speed,parity,data-bits,stop-bits
F5 Clear the screen
F6 review previously received data
F9 some help screens
F10 exit
The default values of: 3:Asc/Asc 4:9600-None-8-1 are good for simple tests.

Any other typewriter key will transmit the corresponding ASCII character via
the serial port.

Any characters which are received via the serial port will be shown in the
main 80x24 area.

So, with the loopback attached:
Type 'ABCD' - you should see 'ABCD' as received characters.

Press F1 to toggle DTR - you should see DSR toggle with it.
Press F2 to toggle RTS - you should see CTS toggle with it.

Press F10 to exit!

Dave ::: https://dunfield.themindfactory.com ::: "Daves Old Computers"->Personal

Sure thing, here they are:

• Astral VLMIO V1.4 Front (ISA + VLB)
• Astral VLMIO V1.4 Back (ISA + VLB)
• Goldstar Prime 2C MKII Front
• Goldstar Prime 2C MKII Back

I've used the [0] for identifying the Goldstar - I guess mine is MKII as it differs somehow from the MK III shown at [0]

Same applies for the VLB card: It's very similar to [1], but mine doesn't have a secondary IDE channel, which AFAICT should come with the v1.4. Mine is more likely a late V1.3 release with V1.4 printed on it [2]

[0] https://theretroweb.com/expansioncards/s/gold … -prime-2c-mkiii
[1] https://theretroweb.com/expansioncards/s/astral-vlmio-v1-4
[2] https://theretroweb.com/expansioncards/s/astral-vlmio-v1-3

You need to measure voltages. Here is similar 8 page saga and it turned out -12V wasnt reliably going to the card Re: serial mouse not detected on a 386 motherboard

On your Astral UMC UM8663 has build in voltage level translators for RS232. Those can go bad. Loopback test might work OK with bad voltage translators.

Your card uses both -12V and +12V to work correctly. -12V is on seventh pin on ISA slot bottom side, +12 is on pin 9. Both are wired up and lead to jumpers soldered in place of DIODES (cheap bastards) with D1 (-12V) and D2 (+12V) over them. Start there. Next step would be measuring pins on serial port and checking if levels are high/low enough.

fix_metal wrote on 2024-12-21, 17:01:

I'm pretty curious of how you've ended up in your testings as I've got the exact same problem: just swapped in a VLB I/O card and ctmouse (or even ms mouse.com for that matters) is not capable anymore to find my (perfectly working) mouse. If I switch back to my 16bit I/O it works just fine(Prime2C chip based)
(...)

Honestly I don't even remember...but I didn't "fix" it that's for sure, probably worked around it! The latest developments here for the past few days give me hope of getting to the bottom of it if I come across it one day 😀

rasz_pl wrote on 2024-12-23, 04:28:

You need to measure voltages. Here is similar 8 page saga and it turned out -12V wasnt reliably going to the card Re: serial mouse not detected on a 386 motherboard

On your Astral UMC UM8663 has build in voltage level translators for RS232. Those can go bad. Loopback test might work OK with bad voltage translators.

Your card uses both -12V and +12V to work correctly. -12V is on seventh pin on ISA slot bottom side, +12 is on pin 9. Both are wired up and lead to jumpers soldered in place of DIODES (cheap bastards) with D1 (-12V) and D2 (+12V) over them. Start there. Next step would be measuring pins on serial port and checking if levels are high/low enough.

Makes sense. I'll give it a further try.