Unusual track pad problem

[joydivision]

Got an old Acer which has hardly been used, is FAT32 and is very slow (HD is ok). However when it boots up the track pad is very slow to respond but is faster in safe mode, I have tried Kaspersky rescue CD and the trackpad doesn't work at all in that.

It does work ok in safe mode though. I try a Ubuntu boot CD before writing it off, but does this not seem a bit strange?

 

[mraikes]

It works OK in safe mode? Then I'd suspect there's something eating up the CPU cycles or RAM in normal mode and is interfering with the pad.

 

[joydivision]

Will try it Ubuntu, wouldn't say the trackpad is exactly smooth in safe mode, but its much smoother than it is in normal mode.

 

[joydivision]

Works fine with Ubuntu, so I am just going to N&P it, I suspect it has some sort of rootkit on it at the moment. Client just wants the job done as cheaply as possible, so in this case N&P makes sense given how slow it is. It is FAT32 too which doesn't help.

 

[KompuKare]

Don't forget that it might be the infamous Microsoft APM fault where when the battery is poor, it spends lots of cycles waiting for a response (Microsoft have allegedly fixed it a few times...).

With process explorer see if hardware interrupts go high - if yes try disabling APM in dev manager

 

[joydivision]

At the moment it is so slow I can't even get to process explorer, but I shall try disabling the services via safe mode. Thanks The battery is dead,.

 

[Fixedathome.com]

Out of interest does a mouse suffer the same problems if you connect one?

 

[joydivision]

No the mouse works perfectly fine, even in XP normal mode, which is why I have got confused. XP normal mode crashes though as soon as it loads explorer, but the RAM and HD test fine. Which is why I am suspecting a faulty file system or rootkit.

 

[joydivision]

An update, you were right about the battery. I removed it and it works perfectly fine.

First time I've seen this problem

 

[KompuKare]

An update, you were right about the battery. I removed it and it works perfectly fine.

First time I've seen this problem

I've only seen it the once myself but it was such a weird diagnose that I keep jumping to that conclusion first now...

I know that disabling the APM device works, but don't recall if switching to ACPI works (I think the laptop where I came across it might not have had any such option in the BIOS). As I mentioned before, Microsoft have 'fixed' this problem a few times for XP and they have some KBs about it somewhere on their site. Mind you, AFAIR it was an Acer in my case too so maybe it's a combination of poorly programmed interrupt plus a bad implementation at the hardware level

 

[joydivision]

There is no APM option in the power management, so I can't seem to think of a solution. The battery forms part of the chassis too, so its unstable with the battery removed.

A new battery is an obvious choice but my client says he never needs to use the battery.

 

[KompuKare]

There is no APM option in the power management, so I can't seem to think of a solution. The battery forms part of the chassis too, so its unstable with the battery removed.

I'm sure I just disabled the relevant device in device manager, question being which was the 'relevant' one: try checking under batteries or System devices and disabling any you find until the trackpad works.

The other solution might be some thin insulation tape over the contacts of the battery: you'd have the physical chassis prop (silly Acer design) but no actual contact.

Somewhat off topic: While Process Explorer is gold compared to TaskMan it would really be nice to have something more useful to go on than 'Interrupts', especially for some of the conflicts with services I've seen. In those cases high CPU from an instance of svchost.exe is not too useful since most of those instances can themselves host dozens of services and threads.

 

[joydivision]

Sorted it, disabled the battery in device manager and it runs like a dream now. I would never have imagined that a dead battery would cause Windows not to work.

I couldn't even load process explorer to see what was going on, it was that slow.