80 pin IDE cable and Primary slave disks make Win98SE very slow

[LarsHansen]

Hi,

mainboard: DFI AK-74 SC
chipset: Via KT133
northbridge: VT8363
southbridge: VT82C686A

BIOS rev.: 2001/03/22


when installing any hdd (tried maxtor and hitachi) using an 80 pin (BIOS calls it 80 conductor cable) IDE ATA100 cable as primary slave, Windows 98 SE gives a blue screen write error on drive D (the pr.sl.) when starting. Windows wasn't able to load any file from the disk and gave the same blue sceen error before it suddenly started with the start-up errors (occured with the maxtor). Using the Hitachi Windows start-up time was normal but as soon as the desktop appeared it became extremely slow. All other IDE channels (pr.ma.,sec.ma.,sec.sl.) are in use: hdd on pr.ma. and sec.sl.. DVD (burner) drive on sec.ma.. Jumpers and cable connections should be set correctly. Using a 40 pin IDE cable, there are no errors and the pr.sl. hdd works.
Any known reason for this behaviour? Jumpers? And what is lost using a 40 pin cable? bandwidth?

thanks for any enlightenment!

Lars

[Mr. Chris]

Lars, you are trying to run an 8 year old operating system 9 98SE ) with new components. I would not think it's going to work very well. I don't think you will lose a lot of performance if you go back to the 40 pin IDE cables. 80 pin cables are needed for ATA 100, 40 pin is for ATA 66. I'm not sure 98SE even will see ATA 100 as it was not an ATA standard back then. Maybe time to get XP? Chris.

[LarsHansen]

I got XP on my main machine :-) I'm building a secondary system using unused hardware. As it currently works using the 40 pin cable it's not a pressing problem. I was just curious (as it occurs with different hdds (back when I still used the system (I had to not use the Hitachi) and now again)); thanks!

[Stultify]

A 40 conductor IDE cable will only allow ATA 33 speeds, and probably a good solution for the VT82C686A "bug", though don't tell anyone who *thinks* they need ATA 100 speeds.

[Engine]

The 40 and 80 pin cables are NOT interchangeable. The devices that utilize each cable are built to different spec's.

[tcsenter]

The 40 and 80 pin cables are NOT interchangeable. The devices that utilize each cable are built to different spec's.

They are interchangeable as long as the active transfer mode does not exceed the cable's specs. The 80-conductor cable was actually designed for Ultra ATA/33, but Ultra ATA/33 was refined enough by market release to work with regular 40-conductor IDE/ATA cable, as long as the length did not exceed 18 inches. At least one major OEM (Dell, Gateway, HP, or IBM, I forget now) was using the 80-conductor cables with Ultra ATA/33 drives (UDMA-2), before UltraATA/66 drives were shipping en mass.

Possible causes of the behavior described could be:

- A defective 80-conductor cable (obvious)

- The motherboard is unstable with Ultra ATA/66 due to poor signal and/or electrical properties (PCB design flaw, bad caps between the IDE connector and Southbridge, or defective Southbridge)

- The BIOS has a bug with auto-detecting an Ultra ATA/100 or Ultra ATA/133 drive, since the 686A supports only up to Ultra ATA/66. Using a standard IDE cable causes the BIOS to automatically limit drives to a maximum transfer mode to Ultra ATA/33. I recall seeing this on several motherboards, including one from FIC where the BIOS couldn't cope with drives that reported a maximum supported transfer mode higher than the chipset supported. Solved by a BIOS update.


Windows 98SE should be fine with drives up to Ultra ATA/133. I've used numerous Ultra ATA/100/133 drives on Windows 98SE systems. However, there may be a patch or driver needed. Its been so long since I've setup a 98SE system that I've forgotten many of the patches and hotfixes that were needed.