• Please review our updated Terms and Rules here

Supermicro P5STE Motherboard PS/2 Mouse Problem

Tincanalley

Experienced Member
Joined
Aug 26, 2019
Messages
176
Location
Los Angeles, CA
On a system with a PS/2 mouse, but just an AT Keyboard, is the mouse alone on a controller? Also, what is that controller? I have an odd mouse issue and had to switch to a serial mouse.

When I boot the system with a PS/2 mouse connected, all is well until I start using the mouse. It will work for a moment, then freak out. It will move all over the screen like hitting the hyperspace button on Defender. Not only will it do that, but it randomly selects icons and menus. They only way to stop it is using keystrokes to restart. It even does when mouse support for the BIOS setup is turned on and I enter the BIOS.

I have tried several mice, tried different header to PS/2 connectors.... I did have to replace the leaking battery in the area, but when this started happening, I never noticed the battery leaking. It was months later a leak developed. Still, I can't rule it out. I really want to get the mouse back on the port as the serial port won't allow me to use my KVM.

I'm attaching a pic of the board. One is a closeup of the battery and PS/2 mouse header and the other is a full view with battery removed (I found no evidence of trace damage). Maybe you can see something I missed, or help locate the controller IC.
 

Attachments

  • P5STE PS2 MB.jpg
    P5STE PS2 MB.jpg
    1.2 MB · Views: 9
  • P5STE PS2 Mouse.jpg
    P5STE PS2 Mouse.jpg
    840.5 KB · Views: 10
That type of setup is identical to the ones with ps/2 mouse and keyboard, just setup for an AT motherboard instead of ATX. Mouse and keyboard are on the same controller, an 8042 compatible integrated in the floppy/ide/serial/etc controller the National Semi PC87303 on that particular board.

You wouldn't happen to be using a serial to ps/2 adapter with that mouse? Those are only compatible with specific mice.
Should be able to find the pins on the controller from the datasheet here: https://pdf1.alldatasheet.com/datasheet-pdf/view/9310/NSC/PC87303.html to trace the pinout to the connector.
 
On the second pic, right above the negative terminal of the battery I think are two of the lines for the ps/2 mouse. Right above the silkscreen outline for the battery. Could just be the lighting but they look a bit iffy to me.
 
That type of setup is identical to the ones with ps/2 mouse and keyboard, just setup for an AT motherboard instead of ATX. Mouse and keyboard are on the same controller, an 8042 compatible integrated in the floppy/ide/serial/etc controller the National Semi PC87303 on that particular board.

You wouldn't happen to be using a serial to ps/2 adapter with that mouse? Those are only compatible with specific mice.
Should be able to find the pins on the controller from the datasheet here: https://pdf1.alldatasheet.com/datasheet-pdf/view/9310/NSC/PC87303.html to trace the pinout to the connector.
Yeah, lighting is making some traces seem like the coating is gone or wearing, but it's like that over the entire board. There were a couple traces that were kind of black and crusty, but when I cleaned them off, they still looked that way and tested good.

There is still some corrosion under the FB1 and 2, but I wasn't sure if they could handle the pressure of cleaning under them and didn't know if they are connected to the mouse circuits anyway.

The mouse I used to used is the same one I had used on the board since new in '95. It is the MS inport mouse with swappable connections (serial, bus and PS/2). The first thing I did when it acted up was try a compaq PS/2 mouse, MS PS/2 and Logitech PS/2. All of them did the exact same thing and went crazy while randomly clicking on things.

I will find the pins on the controller and check the traces. I know the 5V line and ground are working, so those are done.
 
As mentioned, there is some very obvious corrosion under the FB components. Those should be removed from the board and have the pads cleaned and traces inspected under them.

The black crusty traces also need to be repaired, just because you get continuity on them, doesn't mean they're good. They probably have altered resistance from the corrosion. I'd recommend either bypassing the bad section with a jumper wire, or using some Krud Kutter rust converter liquid on a cuetip to clean the trace and convert as much as possible back to copper and then still run a jumper wire.

You may also want to remove the nearby ICs and check under them as well to make sure there's not any more battery corrosion hidden.
 
Last edited:
As mentioned, there is some very obvious corrosion under the FB components. Those should be removed from the board and have the pads cleaned and traces inspected under them.

The black crusty traces also need to be repaired, just because you get continuity on them, doesn't mean they're good. They probably have altered resistance from the corrosion. I'd recommend either bypassing the bad section with a jumper wire, or using some Krud Kutter rust converter liquid on a cuetip to clean the trace and convert as much as possible back to copper and then still run a jumper wire.

You may also want to remove the nearby ICs and check under them as well to make sure there's not any more battery corrosion hidden.
I'm attaching a pic of the battery. In it you can the corrosion was limited to the board under the battery and the 2 FBs. I was told those were ferrite beads and not to push anything under them as they can easily break. That they wouldn't have a negative effect if removed as they are only noise suppressors.

That trace is clean as I can get it without scraping it, physically. I've used CLR, baking soda and IA. I will do a rundown of that trace to see what circuit it belongs to. If it is in the mouse circuit, I will do a jumper wire for testing.
 

Attachments

  • P5STE Battery Leak.jpg
    P5STE Battery Leak.jpg
    745.4 KB · Views: 5
Back
Top