Weird issue involving my ipac - please help! *SOLVED*

[leapinlew]

I'm having an issue with my PS2 ipac and computer that I can only describe as weird. 9 times out of 10 it doesn't register keystrokes and neither does the keyboard pass through once it gets in game. I'm able to get into the Bios using the F2 key, but once it boots into Mala... nothing. If I exit Mala (using a USB keyboard and mouse), windows doesn't report any errors. If I open the notepad - still nothing from the ipac or the pass through. Here is what I've done so far:

1. Hooked iPac to a different computer. No issues. Keyboard passthrough worked fine. Loaded winipac utility and keypresses registered as normal. For kicks, I restored all the defaults but I don't think it changed anything since it was a new ipac I just started using a few hours ago.  
2. Hooked ipac to my computer and it doesn't register keystrokes either through the CP or through a passthrough keyboard.
3. I started thinking it was a physical issue with my PS2 keyboard port on the mobo, so I changed the motherboard out with a brand new one (I had some spares). Same issue. It seems that something happens when windows loads up? I don't know... 1 out of 10 boots and it'll work as normal.

I'm running out of ideas. Any thoughts?

 

[Gatt]

I never thought I'd type this again after Windows XP released but...

This sounds to me like an IRQ issue.  Sounds to me like the introduction of the Ipac and/or just your PS2 port is somehow ending up in conflict with some other piece of hardware.

I suppose it's also possible it's a driver issue with the PS2 port.

Either way,  control panel -> Device Manager,  and start searching through there to see if there's an issue (The view tab will let you switch between device types and resources),  check the PS2 port entry under device types,  and then,  if necessary,  start tracing the IRQ's in resources to see if the PS2 port and something else are fighting over the same IRQ.  I really wouldn't be surprised if some sound-card-type-thing is grabbing the IRQ,  used to be that sound caused all kinds of problems with IRQ's and hardware.

Let us know.

*When the system powers on,  it starts with a limited set of resources,  when Windows boots it adds all kinds of extra hardware like sound and video cards,  so it's possible that system is hardwiring the PS2 port to IRQ X and then Windows is going "Ah hell,  who uses PS2 ports today?  We'll put the sound chip on that IRQ,  there's no way there'll be a problem!".  So you get a PS2 that works on power-on but fails when Windows boots.

[leapinlew]

I never thought I'd type this again after Windows XP released but...

This sounds to me like an IRQ issue.  Sounds to me like the introduction of the Ipac and/or just your PS2 port is somehow ending up in conflict with some other piece of hardware.

I suppose it's also possible it's a driver issue with the PS2 port.

Either way,  control panel -> Device Manager,  and start searching through there to see if there's an issue (The view tab will let you switch between device types and resources),  check the PS2 port entry under device types,  and then,  if necessary,  start tracing the IRQ's in resources to see if the PS2 port and something else are fighting over the same IRQ.  I really wouldn't be surprised if some sound-card-type-thing is grabbing the IRQ,  used to be that sound caused all kinds of problems with IRQ's and hardware.

Let us know.

*When the system powers on,  it starts with a limited set of resources,  when Windows boots it adds all kinds of extra hardware like sound and video cards,  so it's possible that system is hardwiring the PS2 port to IRQ X and then Windows is going "Ah hell,  who uses PS2 ports today?  We'll put the sound chip on that IRQ,  there's no way there'll be a problem!".  So you get a PS2 that works on power-on but fails when Windows boots.

I'm very familiar with IRQ issues. I did peek at the device manager, but didn't spend much time there. I will now. Thanks for your suggestion - sounds promising.

I also have a PCI add-in card to add PS2 ports. I'm going to drop that in and see how it goes. Thanks.

[leapinlew]

It's not an IRQ issue... this thing is driving me crazy!

I've tried the following:
Unplugged everything from the ipac except the pass through keyboard
A different mobo
PCI addin card with a PS2 port (keyboard works, but pass through on ipac doesn't)
A different PS2 cable
reloading the OS
A different ipac from a known working machine. (which really sucked to take apart)

I mean... what esle can I try?

I'm using a Intel Atom chip and Tiny XP. I've used this same setup before.

[leapinlew]

Another hint....

Remember how I said it works 1 out of 10 times? Well, I can get it to work everytime if I go into the BIOS and change a setting in the BIOS. For example, if I go into the bios and change anything it works.

I went to Intels site and got the latest bios upgrade. Upgraded it successfully, and it still has the same issue of not working unless I go into the bios and make a change. Does this make sense to anyone?

[Gatt]

Another hint....

Remember how I said it works 1 out of 10 times? Well, I can get it to work everytime if I go into the BIOS and change a setting in the BIOS. For example, if I go into the bios and change anything it works.

I went to Intels site and got the latest bios upgrade. Upgraded it successfully, and it still has the same issue of not working unless I go into the bios and make a change. Does this make sense to anyone?

Now that is really,  really,  weird.

Obviously,  the board isn't initializing the PS2 port on a standard boot,  but when you enter bios and change something,  it's forcing it to revalidate everything and initializing the port.  But why is this happening?


I *really* can't think of any reason why it might do that.  It's not an OS issue,  can't be,  it has to be an issue with the motherboard.  Perhaps even the chipset.  I'm not sure what motherboard you have though,  so I can't really do any research.  Sadly,  I don't think you're getting around this one.  

It actually sounds slightly familiar,  I had an Asus board that didn't like Razor mice,  if I didn't click the mouse through the whole boot process,  it wouldn't power it.  The only remaining advice I have...

1.  Tap a key through the whole boot process,  see if it's powering down the Ps2 port on boot and an input stream forces it to keep it alive.
2.  It's  a real reach,  and I think it's unlikely to work,  but check and see if your motherboard manufacturer has drivers for the board itself.  I really doubt this will solve the problem,  but it's really the only other thing I can think of.

Edit:

WAIT! Idea!

Power management!  Atoms are supposed to be lower power chips,  so their motherboards should be designed to cut power to anything they can.  One other thing I can think of,  try disabling power management(ACPI I think it's still called?) in the bios and XP's power management features.  It's within the realm of possibility that power management could be at fault.

[leapinlew]

Another hint....

Remember how I said it works 1 out of 10 times? Well, I can get it to work everytime if I go into the BIOS and change a setting in the BIOS. For example, if I go into the bios and change anything it works.

I went to Intels site and got the latest bios upgrade. Upgraded it successfully, and it still has the same issue of not working unless I go into the bios and make a change. Does this make sense to anyone?

Now that is really,  really,  weird.

Obviously,  the board isn't initializing the PS2 port on a standard boot,  but when you enter bios and change something,  it's forcing it to revalidate everything and initializing the port.  But why is this happening?


I *really* can't think of any reason why it might do that.  It's not an OS issue,  can't be,  it has to be an issue with the motherboard.  Perhaps even the chipset.  I'm not sure what motherboard you have though,  so I can't really do any research.  Sadly,  I don't think you're getting around this one.  

It actually sounds slightly familiar,  I had an Asus board that didn't like Razor mice,  if I didn't click the mouse through the whole boot process,  it wouldn't power it.  The only remaining advice I have...

1.  Tap a key through the whole boot process,  see if it's powering down the Ps2 port on boot and an input stream forces it to keep it alive.
2.  It's  a real reach,  and I think it's unlikely to work,  but check and see if your motherboard manufacturer has drivers for the board itself.  I really doubt this will solve the problem,  but it's really the only other thing I can think of.

Edit:

WAIT! Idea!

Power management!  Atoms are supposed to be lower power chips,  so their motherboards should be designed to cut power to anything they can.  One other thing I can think of,  try disabling power management(ACPI I think it's still called?) in the bios and XP's power management features.  It's within the realm of possibility that power management could be at fault.

GATT! Great idea - how'd you know? I didn't have ACPI in the BIOS to disable, but if I did - I bet it would've worked. I took your suggestion and disabled ACPI in Windows XP using these steps:

    *  Right click on My Computer
    * Choose Properties
    * Click the Hardware tab
    * Choose Device Manager
    * Double click on Computer. You will see Advanced Configuration And Power Interface (ACPI) PC
    * Double Click on the ACPI reference
    * Click the Driver tab
    * Click Update Driver
    * Click Next
    * Choose 'Display a list of the known drivers for this device so that I can choose a specific driver'
    * Click Next
    * Choose 'Show all hardware of this device class'
    * Select Standard PC
    * Click Next. Windows will display a warning message. Ignore it and click Yes i.e. you want to install the driver despite Microsoft's warning not to do so
    * Click Next. Windows will upgrade the driver. This takes a few seconds
    * Click Finish
    * Close the 'Upgrade Device Driver Wizard' window. Restart your computer when prompted
    * After restarting, return to the Device Manager and verify that your computer's configuration is now Standard PC and not ACPI

That resolved the issue. Weird eh? I burned up a lot of time on this issue. Thanks so much for your help. I've posted the solution here just in case someone searching later finds this thread.

[Gatt]

You're very welcome,  glad to be of service!

The Atom tipped me off.  I was thinking about it,  going into Bios reset something that permitted the PS2 port to be detected,  so the port must be good and something must've been shutting it off.  Which lead me to thinking about the architecture goals for the Atom,  low-power consumption.  ACPI was the root of my evil with my Razor Mouse,  the Asus board would power it down,  so it was possible that ACPI was to blame,  especially since the design goals for Atom is to have minimal power consumption.

Good idea to post this,  wasn't that long ago that someone else had a similiar issue.  I suspect as Atom's become more common the issue will be recurring.