Samsung DVD Drive

[CharlotteTheHarlot]

I see the IDE configuration you have now (jeez, I should have looked up the Abit BE-6 first ). I see a picture of it on this page. Is this the one?

When I noticed the eight IDE devices I figured resource sharing as a likely culprit. I noticed on that same page mentioned above that the author mentions this: "As with the BE6 before it, installation of the BE6-II passes without incident as long as some forethought is put to the installation of the ATA66 controller. The trick consists to refer to Appendix D in the users manual to proceed with the installation of the ATA66 controller drivers. Don't forget that the ATA66 controller integrated into this board shares the bus master designation for PCI slot 5, while PCI slot 3 shares an IRQ with that same controller. The driver for the ATA66 controller supports the sharing of the IRQ, but so must the peripheral device installed in PCI slot 3, otherwise you're in for a whole mess of annoyance..."

Since this arrangement is probably using/sharing 4 IRQ's and 12 I/O ranges (apologize if you already mentioned it) but could you describe what if anything is in the PCI slots? I agree with Drugwash that the problem lies somewhere in the HDD Controller configuration (as long as you are sure that there isn't some BIOS setting you missed).

I didn't see anything yet in a quick Google search but somewhere there must be someone that documented the steps of installing the Intel INF Chipset Utility (this gets installed first) and the Intel AA (if it applies to this board) or maybe Highpoint supplemental drivers (if they even exist).

When you get around to posting those registry entries, do a search for the string 'S222A' and list any that appear under the HKEY_LOCAL_MACHINE\Enum key. What I am thinking is that the two separate controllers use two different branches. Everything under \Enum matters.

P.S. as already mentioned (but hopefully not to late!), you need a registry backup especially when juggling HDD controllers. If you are not familiar with saving and restoring the configuration, please ask. But definitely make sure there is a recent copy (prior to this adventure) of both System.dat and User.dat that can be used in an emergency.

[Rivia]

Is this the one?

Yes !

"Don't forget that the ATA66 controller integrated into this board shares the bus master designation for PCI slot 5, while PCI slot 3 shares an IRQ with that same controller. The driver for the ATA66 controller supports the sharing of the IRQ, but so must the peripheral device installed in PCI slot 3, otherwise you're in for a whole mess of annoyance..."

OK, the user manual of the BE6 doesn't tell anything about that. Concerning the ATA/66, it only explains how to install the drivers. This could possibly account for a problem when I used the ATA/66 connector.

But could it when I put the drive on the secondary "classical" IDE ATA/33 connector ?

Since this arrangement is probably using/sharing 4 IRQ's and 12 I/O ranges (apologize if you already mentioned it) but could you describe what if anything is in the PCI slots?

I have 2 PCI slots occupied by a Sound Blaster card and a USB 2 card (based on a VIA chip).

Otherwise, my video card is a Matrox in the AGP slot and I also have a FM radio card in a ISA slot (yes, still working !).

I'm not familiar with those IRQs and I/O things, but I can copy you the informations displayed by MSInfo (some french terminology inside, I try to translate what could be not obvious) :

0	Horloge système [=System clock]
1	Logitech PS/2 Keyboard
2	Contrôleur d'interruptions programmable [=Programmable interrupt controler]
3	Port de communication (COM2)
4	Port de communication (COM1)
5	VIA PCI to USB Enhanced Host Controller
5	Matrox Millennium G400 DualHead - Français
5	IRQ Holder pour PCI Steering
6	Contrôleur de lecteur de disquette standard [=Floppy disk controler]
7	Port imprimante (LPT1)
8	Horloge système CMOS/temps réel
9	(libre) [=(Free)]
10	Creative SB Live! Value
10	Contrôleur VIA Tech 3038 PCI vers USB Universal Host
10	Contrôleur d'hôte PCI vers USB universel Intel 82371AB/EB
10	IRQ Holder pour PCI Steering
11	HPT366 Ultra DMA Controller
11	HPT366 Ultra DMA Controller
11	Contrôleur VIA Tech 3038 PCI vers USB Universal Host
11	IRQ Holder pour PCI Steering
12	Logitech-compatible Mouse PS/2
13	Coprocesseur arithmétique
14	Contrôleur IDE primaire (fifo double)
14	Contrôleur IDE standard double canal PCI
15	Contrôleur IDE secondaire (fifo double)
15	Contrôleur IDE standard double canal PCI

I agree with Drugwash that the problem lies somewhere in the HDD Controller configuration (as long as you are sure that there isn't some BIOS setting you missed).

The BIOS is an Award one, and there not very much parameters about IDE to play with.

In fact, for each of the four ATA/33 channels, I can choose the PIO mode (0 to 4 and Auto) and the ultra DMA mode (Disabled or Auto). I left everything on Auto.

However, and despite all my unsuccessful attempts so far to activate the DMA, I decided to finally burn a DVD (not only simulate) in order to see if it could work. And... it worked ! The burner needed a little more than half an hour to write about 3,9 GB of data, but it did write them.

A bit of consolation in this cruel world...

[dencorso]

Of course it'll work, but much slower, since you've not got DMA. What CharlotteTheHarlot is saying amounts to this: there are 5 PCI slots and you're using just two cards, so move both cards out of their present slots to any of the now unused slots, taking care to let both the presently used slots unused, and see whether this solves your issue.

Edited July 1, 2009 by dencorso

[Drugwash]

Apologies for not making it in time for the promised test - it's half past midnight and I just got home, tired and a bit drunk. Will have to postpone it for tomorrow... hopefully.

[CharlotteTheHarlot]

I just looked at some references and see that there is a possibility of conflicts arising from real-mode drivers loading in Autoexec.bat and Config.sys that can disrupt the use of opticals in Windows protected mode. You should check those two files in the root directory (if they exist) and see if there is anything within that is suspect (oak, atapi, mscdex). In theory, this could affect one optical drive but another might be just fine.

You didn't mention your ability to restore the registry from system.dat backups. Please tell us you have backups and know how to restore it from an increasingly likely FUBAR disaster. Swapping HDD controllers, INF chipset software, and moving PCI cards is not for the ill-prepared!

The MSINFO you supplied is definitely one for the record books, the only thing missing is the kitchen sink! It appears to me that the Highpoint implementation is non-standard since it does not show up as two discrete HDD controllers on their own IRQ's (unless the two so-called HPT366 Ultra DMA Controller is an alias, in which case they appear to share IRQ 11 ouch! lucky the secondary appears to be empty).

Re-reading this thread and your replies tells me that this DVD would not work in UDMA-2 on the 33 channel(s) or in UDMA-4 on the 66 channel(s). The latter is troubling because the correct cable is there and a drive is working in UDMA-4 as it should.

If you are game, I have an idea. After it is done I will give you a registry patch to try. There is some prep first...

1) remove the DVD burner

2) boot into to Safe Mode

3) Device Manager > System (delete all occurences of the Samsung)

4) shutdown, install DVD on 80-wire slave to the HDD (jumper!)

5) boot normal, let it detect, reboot

6) export two registry branches (place in zip) ...

[HKEY_LOCAL_MACHINE\Enum]

[HKEY_LOCAL_MACHINE\System\CurrentControlSet\Control\Asd]

7) get them to me somehow: maybe try attaching ZIP in forum (not sure if possible), I estimate those would be like 100 kb as text, probably 10 kb as ZIP. Text is probably too big for CODEBOX (does anyone here know?). If all else fails, PM me and we'll do an email.

Here is what I can definitely do (given that data). I can determine which entries the Samsung is being loaded from and tailor the registry branch and the Enumerate/Start key the way it should be (if needed! it might actually be correct). The resultant registry patch plus a reboot will pretty much tell us if this is a Windows problem or a hardware problem. In other words whether the drive should be replaced. The reason I can say this is because of that HDD sitting in the master position with UDMA-4 working. This drive as slave with the correct registry info should work (again barring all the previous things we verified: jumpers, BIOS, cables, and don't forget to look at Autoexec.bat and Config.sys).

P.S. how much time before the new drive replace/return RMA elapses? FYI, you can short circuit this adventure by stuffing the DVD drive into a WinXP (or newer) machine. With no DMA I would suggest with high confidence to RMA the thing and start over.

[Drugwash]

As promised, with apologies for the delay:

First attempt at getting DMA enabled failed on a Win98SE2ME AMD K6-II machine with an ALi chipset. However, it also failed for the Plextor PX-54TA CD-ROM that I already had mounted on that machine and I do remember that particular Windows installation has been quirky in regard to DMA back when it was run on its original mobo. So I'd file this test as inconclusive.

Second test was performed on a Compaq Deskpro EN400 under Win95 v4.01.971 B. As can be seen in the attached screenshot, the IDE controller is the very same as Rivia's, the only difference being that my DVD-RW has firmware revision SB00 as I had not yet searched for an upgrade (and well... if it ain't broke, don't fix it )

DMA got enabled by itself as soon as I mounted the drive and it appears I do have the Intel Bus Master driver installed. So despite this isn't a Win98SE system, I'd say the drive should work fine with DMA enabled. The conclusion I would draw is either (in order):

1. Your IDE driver is bad/badly installed.

2. There is hardware incompatibility between the DVD drive and the IDE controller(s) on your mobo.

3. The SB01 firmware revision breaks DMA (only in certain configurations, maybe)

4. The drive is defective

[Rivia]

Hi, I thank you again for the time you took in helping me. :

Charlotte :

I won't have very much time in a near future to change my internal connections again and make new tests with the registry and so on... I'm not the only person who is using this PC, so it has to be operational from time to time...

If I a propitious period occurs (maybe during holidays), I will try what you suggest and give you reports.

Drugwash :

Interesting... the only difference here seems to be the firmware version.

One thing I could perhaps try would be a downgrade of the drive, if I can put a hand on a SB00 version ?

PS : you apparently use some very nice wallpapers, don't you ?

[Drugwash]

Heh.. sorry, that wallpaper is not work-safe in its full version but I hope the small visible area would not create problems to anyone here; I took that screenshot in a hurry.

I see firmware revision SB00 is available for download at the Samsung site. You may try it, who knows... Interesting to know if that software works in Win9x since it's advertised as Win2000/XP/Vista only.

[Rivia]

I downloaded the SB00 version and installed it. I was a bit anxious to proceed, but it appeared that the Samsung flashing utility works fine with W98SE.

Unfortunately, that brought no changes for the DMA.

I put the SB01 back, just in case.

PS : there's no offense for me about your wallpaper. Simply, a French man notices that kind of details...

[Drugwash]

Haha! Us europeans would never be offended by the beauty of the human body.

Sorry to hear you still couldn't get that drive to work. The ultimate test would be to stick it into another machine (preferrably with a completely different chipset) and see if DMA works there - this would rule out the defective state of the drive and we'd be left with either bad IDE driver or hardware incompatibility (bad chipset/IDE implementation, since it worked here with the very same chipset).

In the mean time, have a nice week-end!

[jaclaz]

Haha! Us europeans would never be offended by the beauty of the human body.

Hey Drugwash , you understand that "US europeans" can be read differently from what you meant, do you?

A whole new breed of people, still with short sleeved shirts and ties, but more open minded when it comes to nudity?

jaclaz

[submix8c]

Ok... silly suggestion - have you tried going into "Safe Mode" and removing the HDD's and the CD/DVD drives then rebooting into "Normal Mode"?

[sp193]

Hey, do you have the Intel Application Accelerator (Or a replacement driver from VIA if your chipset is from VIA)? If you do, you should not have the "DMA" checkbox there (Mine doesn't have it).

If you have the Intel Application Accelerator installed:

Check under the "Secondary Channel" (Your drive is plugged there right?), and tell us the following:

PIO Mode Support

What are the supported DMA SW Modes?

What are the supported DMA MW Modes?

UDMA Mode Support?

Current Transfer Mode.

Transfer Mode Limit.

Cable Type (Host).

Or even better, send us a report generated by Intel Application Accelerator.

Also, you could reinstall those drivers (Chipset drivers first, then any 3rd party IDE bus drivers like the Intel Application Accelerator)