PDA

View Full Version : PS/2 Mouse Locks Keyboard



cr1901
February 28th, 2014, 02:28 PM
I have a 1998 motherboard- BCM SQ575- that I use for testing old DOS hardware, and while it's a fine motherboard for my purposes, it bugs me that I've never been able to fix certain issues with it- hard disk limited to 32GB (bug in AWARDBIOS, even after a BIOS Flash from v4 to v6 for this motherboard), K6-2 Max Speed (requires a special jumper setting for the clock multiplier), figuring out which IRQs are assigned to where :(.

One of the most irritating things for me however, is the fact that the PS/2 mouse port doesn't work properly. More than that, if I even move the PS/2 mouse the slightest after the mouse driver has loaded, the keyboard and mouse will both lock up, preventing further input. This happens in both DOS and Windows 98! Tuning the PC up recently reminded me of this issue. This started happening at least two years ago, and it occurs even when there is no extra hardware besides a video card present. I recall using the mouse successfully after the flash, which was over 3 years ago. I seem to even recall the mouse working at the Windows 98 install screen just before the issues started. But this was all a while ago, so it's very possible my mind is slipping :P. More importantly- right now, it just doesn't work.

Before I try flashing my BIOS again as a last resort, I'd like to figure out if there is a different underlying issue causing the PS/2 mouse to fail- hardware failure? Bad initialization settings written to flash somewhere? Unfortunately, I have no clue how to go about debugging this problem- I just know that the keyboard will lock up when either of the following two things happens:

Mouse driver is loaded in DOS, and mouse is moved even slightly. This ALSO happens in Windows 98, minus loading the driver. So I don't think it's a mouse driver issue.
Load a USB driver in DOS (USBUHCI.COM), and then load Bret Johnson's USBMOUSE.COM driver. The driver resets the PS/2 mouse before installing. Control returns to COMMAND.COM, but the keyboard is locked (can't test the mouse at this point :P). USB mass storage devices do work in DOS on my machine, albeit slowly of course XD.


The latter makes me think it could very well be hardware trouble, but I'll have to look at Bret's USBMOUSE source to know for sure.

Given the information above, could someone possibly point me in a direction or towards a utility that could help me debug this problem? It's obvious the keyboard isn't processing keys if this happened (i.e. IRQ 1 got screwed), but the question for me is: why? I suppose one advantage to this is that the IRQ is available for my PCI network card now...

Some additional information- the keyboard is NOT a PS/2 keyboard, but a regular AT keyboard. Though they are the same electrically, I do not know if the motherboard will look in different I/O ports for keys for a PS/2 keyboard vs an AT keyboard, or if the AT/PS/2 keyboard has any relation I/O port-wise or chipset-wise to the PS/2 mouse.

SomeGuy
February 28th, 2014, 03:27 PM
That sounds similar to a problem I was having with an interrupt conflict on a P200mhz board. In my case though, it was because the system was loaded with stuff - a SCSI card and unused sound card IDE were hogging interrupts. One removed SCSI card later, the sound card IDE (no way to disable that) got out of the way of the PS/2 port and let everything work.

Not sure what could be causing it if there are no installed cards! You might try clearing the CMOS. Some BIOSes also let you reset the stored PNP information.

If you can get Windows 9x to boot without the mouse, you can view resources in device manager by interrupt and see if there is anything else on IRQ12.

krebizfan
February 28th, 2014, 04:05 PM
I would check for a short across the mouse/keyboard traces. Not likely, but doesn't take much. Also double check the cabling just in case it was off by one pin and mucking with the ground.

You may have to forget about the PS/2 port and just use the USB mouse or switch over to a serial mouse.

cr1901
February 28th, 2014, 04:14 PM
I would check for a short across the mouse/keyboard traces. Not likely, but doesn't take much. Also double check the cabling just in case it was off by one pin and mucking with the ground.

You may have to forget about the PS/2 port and just use the USB mouse or switch over to a serial mouse.


Load a USB driver in DOS (USBUHCI.COM), and then load Bret Johnson's USBMOUSE.COM driver. The driver resets the PS/2 mouse before installing. Control returns to COMMAND.COM, but the keyboard is locked (can't test the mouse at this point :P). USB mass storage devices do work in DOS on my machine, albeit slowly of course XD.
Currently I use a serial mouse- this isn't a big deal to me, other than IRQ sharing between COM2 and COM4 breaks the mouse AND COM port in Windows XP (works fine in Windows 98- and kernelEx isn't enough for my needs). I have to disable COM4 while it's being used in Windows XP- and yes, I DO have 4 serial ports on this machine for vintage hardware, as well as two parallel ports! For those who didn't think serial mice worked in XP- well, they do :P.

The problem in and of itself and not being able to figure out what locks the keyboard is what bothers me. Additionally, I find it interesting that DOS USB mouse drivers are also broken. Whatever software routine or I/O polling locks the keyboard when the PS/2 mouse is simply moved, also locks the keyboard when the PS/2 port is simply reset.