Problems enabling DMA in Windows XP

[iq100]

This is my first post. If it does not belong here, my apologies, please copy elsewhere.

1. I read thread:
http://www.cdrlabs.com/phpBB/viewtopic. ... c&start=25

2. Thanks Hans and cfitz for patience and efforts, but even after deleting the key suggested above my LG 8482B still show PIO mode. I also tried deleting the LG drive from DM and the key. My configuration is:

Dell 1400SC, W2K, with only a Primary IDE channel on the motherboard.
Master is Liteon 48246 which does show Ultra DMA.
Slave is the LG 8482B.
There is also a Promise PCI IDE controller connected to two WD1200JB drives.
Fujitsu 18GB SCSI drive is connected to motherboards Ultra-160 SCSI.

No big problem for me running the LG in PIO mode and Liteon as DMA, but thought I would provide this data point to contribute to your noble attempts to get these things correct.

By the way, the Computer Management>removable storage>physical locations shows multiple occurences of the Liteon and LG drives, most with red X's. Is this normal? Is this just a kind of history of all the deletes and re-installs of these drives, in my attempt to get both to report DMA?

Keep up the good work,

iq100

[cfitz]

1. Did you try deleting the IDE channel altogether?

2. What are your BIOS settings, and what options are available to you in the drive settings?

3. It shouldn't make any difference, but you might want to try swapping the master-slave settings of the two optical drives.

4. The red X drives are simply the history of all the deletes and reinstalls as you suggest. You can delete them safely.

5. Yes, we would like to keep diagnoses questions in the other forums, and keep the FAQs clean with just the solutions. Perhaps a moderator will move this to the CD-RW forum.

cfitz

[dodecahedron]

Yes, we would like to keep diagnoses questions in the other forums, and keep the FAQs clean with just the solutions. Perhaps a moderator will move this to the CD-RW forum.

cfitz

done!

[UALOneKPlus]

cfitz does such a great job. I vote cfitz for Moderator!!!

[iq100]

cfitz posted: 1. Did you try deleting the IDE channel altogether?

I tried this before I read about deleting the key from W2K registry here. I tried deleting key from registry and the LG from DM before rebooting, but not the IDE channel, the key, and the LG before rebooting. If I get a chance I will try that tonight and report back. A special thanks for the info about the historical red X's. Sheeesh ... why doesn't BG get this into the help text. I checked the BIOS-nothing particular there. I believe that I previously had tried the LG as master and Liteon as slave and found that the LG became DMA while the Liteon kept reverting to PIO. So it appears that something wants to force this split mode use of the IDE channel. Could the problem be that the Liteon advertises UltaDMA while the LG can only do DMA and the IDE driver cannot mix this combo. Maybe DMA/DMA would work. Experiments will have to wait till I get some time.

Since the 'delete key' addition to this FAQ is recent, I'll continue to post feedback here and let moderator move when desired.

iq100

[Han]

Hmm, perhaps the IDE cable is defective. Try a different one; try both, 40- and 80-pin...

[ariranha]

cfitz does such a great job. I vote cfitz for Moderator!!!

me too

[cfitz]

Hmm, perhaps the IDE cable is defective. Try a different one; try both, 40- and 80-pin...

I agree with Han, particularly if you changed the positions of the drives when you changed the master/slave settings (maybe you are using cable select?) and the problems always stayed with the same connector.

cfitz

[iq100]

Yes, I am using cable select. Good call! When I get a chance I will try another IDE cable.

Here is another historical data point: When I originally tried to add the Liteon burner as a second CD device to the shipped LG configuration (both on the single motherboard IDE channel), I could not even manually open the LG device using the push button on its front panel (the red LED did go on like it was trying to open the tray, but did not have the power to do so)! Dell had no solution for this. Serendipity, when the push button began to work after I had swapped physical locations of the drives, I realized that I had used different power cables and power connectors when I swapped the physical locations because of power cable lengths. I then experimented with the 'bad' power cable, and found that when used to power the single SCSI drive that the computer failed to boot. Therefore I concluded that I had proved that one of the power cables was bad. It had two standard 4-pin molex connectors on it. When I used the other power cables, the failure to open tray problem disappeared. I report all this to agree that cables can be bad, and when time permits I will try and swap the data cable. Right now the W2K reported split mode of DMA/PIO does not cause me any operational problems, so I will have to put the experiment on the back burner.

If it does prove to be a data cable problem my hat will be off to you guys! Thanks,

iq100

[cfitz]

You are welcome. Let us know the results when you do get around to performing your experiments.

cfitz

[blakerwry]

This is a known problem I thought was moved to the permanent post section... it started occuring around the time Xp SP1 was released


deleting the IDE controller from device manager typically fixes the problem... upgrading the IDE driver (for example installing the VIA 4-in-1's) might also do the trick.

[dodecahedron]

This is a known problem I thought was moved to the permanent post section... it started occuring around the time Xp SP1 was released


deleting the IDE controller from device manager typically fixes the problem... upgrading the IDE driver (for example installing the VIA 4-in-1's) might also do the trick.

it's in the FAQ