Earlier this year, in my work as a system administrator, I was assigned to investigate why approximately 40 co-workers (out of 1700 using a PC or Mac) were having USB optical mouse issues. The symptoms were generally the same. While using the computer, the user's USB optical mouse would simply stop working. If they flipped the mouse over, it was as though the mouse was powered off. The laser would appear to have shut off. If they unplugged and re-plugged the mouse, it would nearly always come back to life. Only once in a while was it necessary to reboot the machine. Once in a great while, the USB keyboard would stop working as well. Again, unplugging and re-plugging seemed to fix it.

When I went through my first round of problems early in the year, a lengthy investigation proved that all of my 40 cases could be solved by performing the following steps:

1. Update the system BIOS from the manufacturer's web site. Sometimes this corrected the issue and no more work was needed.
2. Update the motherboard chipset drivers from the manufacturer's web site. Sometimes this corrected the issue and no more work was needed.
3. Update the mouse drivers from the mouse manufacturer's web site. Again, sometimes this corrected the issue and no more work was needed.
4. Replace the mouse with another mouse from stock.
5. Replace the keyboard with another keyboard from stock.
6. Install a self-powered USB hub between the computer and the mouse/keyboard.

The above 6 remedies seemed to take care of all 40 users.

Then, about 3 weeks ago, the problem reared its ugly head again. This time, the above 6 steps didn't work. People who were run through all 6 of them still had the problem. Our help desk bumped the problem back to me to resolve, since I'd handled the one earlier in the year.

This time I went through a pretty thorough search on the web, Microsoft's knowledgebase, Dell's knowledgebase, etc. What I learned was interesting. When I searched for mouse failures as a whole, probably 99% of the time it was an optical mouse that failed. Moreover, it was almost always a USB optical mouse that failed. This was true on Mac OS X, Linux, Windows, and one or two other lesser-known operating systems. The brand of mouse didn't seem to matter, either. I read about the problem existing with every major brand and even some generic types. It made me wonder about the reliability of USB optical mice.

As a troubleshooting step, I suggested that we identify 1-2 of our users who were having the most "mouse failures". We swapped them with some new mechanical (i.e., old "ball style") mice from stock. The problems vanished for those users. As a follow-up step, I attacked the users' former optical mice to my system. Sure enough, I began seeing the problem myself.

An interesting observation occurred when we had one user who wanted to attach both an older mechanical mouse and an optical mouse to his Windows XP Pro system at the same time. He found that his optical mouse would fail, while the mechanical one kept working normally. (And again, when I swapped his optical mouse for mine, I began to see the problem myself.)

Another interesting observation was made after we installed Microsoft Windows XP Hotfix 914015 and 918365 on the affected systems. Occasionally the mouse would stop working as before, but within a couple of seconds it tended to come back to life on its own without any effort. (Please do not ask me for these hotfixes. If Microsoft won't provide them to you, I can't help as it would be illegal to do so.)

So it appears to me at the moment that we may have two separate issues here. The first issue is that there is a bug in the Windows XP Pro USB stack that causes a mouse to go offline if it asks to have the USB controller reset due to an error. The Microsoft hotfix takes care of that. The other issue is that we seem to have quite a few defective USB mice on-hand. In fact, I pulled one randomly out of stock and tested it, and it turned out to be defective. I confirmed this by attaching it to a Mac OS X system (in addition to a Windows XP system) and finding the same result, the mouse quit working.

I have a suspicion that the reliability of USB optical mice is not as high as we all might think. Just gauging from the numbers I'm seeing at our site, as many as 3% of the optical mice in use are in fact defective. Whether this is just the failure rate we should expect or whether it's symptomatic of a widespread defect in USB optical mice, I don't know.

In any case, it's caused me to add a few items to the above list of 6 steps:

7. Apply Microsoft Hotfix 914015 (if it's a Windows XP system).
8. Apply Microsoft Hotfix 918365 (if it's a Windows XP system).
9. Test the user's mouse on a system that isn't currently experiencing the mouse issue and give the user a mouse that hasn't been shown so far to have the problem.
10. Swap the optical mouse with a mechanical (roller ball type) mouse.
11. If the system and mouse support it, try installing the mouse on a PS/2 port instead of USB. We didn't see these issues with PS/2 mice on any platform.

Those steps, so far, seem to eliminate the problem in all cases. Perhaps they'll help you if you're experiencing a problem with your USB mouse.