Ultra DMA 5 Hard Drive in PIO Only Mode

[Charles Elliott]

My problem is that one of the three hard drives (IDE Secondary Channel
Master) is detected by the O/S (Win XP w/SP2) as PIO Mode 4 instead of Ultra
DMA 5. I originally noticed this when Nero tested all the logical hard
drives in the system to determine where to put its cache, and found the
logical drives on this hard disk to transfer data about 1/20th as fast as
the other logical hard disks (about 2.5 versus 50 MBps, as I recall).

This machine is tri-boot: Win NT 4.6a, Win2K, and WinXP. Over the weekend
I had to boot into Win2K to run a backup program. Just on a hunch, I
checked with Win2K's device manager and the drive was listed as Ultra DMA 5.
I suspected this was so because I had taken the drive to another computer
and the BIOS splash screen on that machine indicated the drive had been
detected as Ultra DMA 5.

So the problem appears to be that WinXP w/SP2 is detecting the drive as PIO
Mode 4 but other O/Ss on the same machine and other computers detect it as
Ultra DMA mode 5. I tried changing the drop-down box entry in device
Manager (Advanced settings) from "DMA if available" to "PIO only" and back.
That worked for the CD ROM drive on the same channel, but not for the hard
drive. The cable is brand new. I also tried that same experiment with the CD
ROM drive out of the system, but no change.

WinXP will downgrade a drive from Ultra DMA to PIO Mode if it detects errors
on the drive (according to
http://www.microsoft.com/whdc/device/storage/IDE-DMA.mspx), but when it does
it logs the errors in Event Log. My event logs show no errors.

I feel certain the problem is in WinXP. I think it began when I used that
channel to test a DVD burner to see if it worked. Before I tested the
burner, I am almost sure the hard drive was OK; now WinXP will not detect
the hard drive as Ultra DMA 5. But it is not the drive; obviously the
drive is OK. Do you know what the problem could be?



Charles Elliott

 

[Ground Cover]

Depending how it is currently set, distinctly, deliberately set the drive to
PIO.

Reboot

Go back and set it to DMA

Reboot

Did that fix it??

 

[Jim]

My problem is that one of the three hard drives (IDE Secondary Channel
Master) is detected by the O/S (Win XP w/SP2) as PIO Mode 4 instead of
Ultra DMA 5. I originally noticed this when Nero tested all the logical
hard drives in the system to determine where to put its cache, and found
the logical drives on this hard disk to transfer data about 1/20th as fast
as the other logical hard disks (about 2.5 versus 50 MBps, as I recall).

This machine is tri-boot: Win NT 4.6a, Win2K, and WinXP. Over the
weekend I had to boot into Win2K to run a backup program. Just on a
hunch, I checked with Win2K's device manager and the drive was listed as
Ultra DMA 5. I suspected this was so because I had taken the drive to
another computer and the BIOS splash screen on that machine indicated the
drive had been detected as Ultra DMA 5.

So the problem appears to be that WinXP w/SP2 is detecting the drive as
PIO Mode 4 but other O/Ss on the same machine and other computers detect
it as Ultra DMA mode 5. I tried changing the drop-down box entry in
device Manager (Advanced settings) from "DMA if available" to "PIO only"
and back. That worked for the CD ROM drive on the same channel, but not
for the hard drive. The cable is brand new. I also tried that same
experiment with the CD ROM drive out of the system, but no change.

WinXP will downgrade a drive from Ultra DMA to PIO Mode if it detects
errors on the drive (according to
http://www.microsoft.com/whdc/device/storage/IDE-DMA.mspx), but when it
does it logs the errors in Event Log. My event logs show no errors.

I feel certain the problem is in WinXP. I think it began when I used
that channel to test a DVD burner to see if it worked. Before I tested
the burner, I am almost sure the hard drive was OK; now WinXP will not
detect the hard drive as Ultra DMA 5. But it is not the drive; obviously
the drive is OK. Do you know what the problem could be?



Charles Elliott

No, your problem is caused by excessive errors. XP takes note of the errors
and sets the driver to PIO mode. It never reverts to DMA by itself.

The fact that the device works in DMA mode on a Windows2000 system may
merely mean that it isn't generating errors at the moment.

A way to proceed is to remove the IDE device driver. Then reboot. On
rebooting, XP will copy a new version of the driver from the i386 folder.
This new driver does not contain the record of prior errors; hence DMA will
be restored.

This is the method that I used when presented with the same problem.
Jim

 

[Martin S.]

You could also try uninstalling that drive channel and restarting the
computer. You don't say what IDE controller you have, if it's Intel have you
installed the chipset drivers? If it's a VIA chipset have you installed the
4 in 1 drivers?

 

[Charles Elliott]

Ta-Da. That worked. Ta-Da. Thanks for your help.

CHE

 

[Charles Elliott]

I uninstalled the Secondary Channel controller and let WinXP re-install it.
That worked. I have an Intel chipset, and, yes, the driver is installed.
Thanks for your help.

CHE

 

[Jim]

Ta-Da. That worked. Ta-Da. Thanks for your help.

CHE

Glad to see that it helped. I certainly can't claim to be the originator of
this particular fix.
Jim

 

[MAP]

No, your problem is caused by excessive errors. XP takes note of the
errors and sets the driver to PIO mode. It never reverts to DMA by
itself.

The fact that the device works in DMA mode on a Windows2000 system may
merely mean that it isn't generating errors at the moment.

A way to proceed is to remove the IDE device driver. Then reboot. On
rebooting, XP will copy a new version of the driver from the i386
folder. This new driver does not contain the record of prior errors;
hence DMA will be restored.

This is the method that I used when presented with the same problem.
Jim

I know this problem still exits but this "was" supposed to be fixed with
SP1.
So much for M/S