IDE hell

[Adcadet]

So I got a WD 1200JB for $90 through an Office Max special. Transfered the stuff from my old 800JB onto it, then removed the 800JB (for later inclusion into a new PC I'm building) and put the 1200JB in it's place in my case. That's when the problems start.

The machine won't boot. Windows will start to load, but the machine reboots during the splash screen. Booting in safe mode doesn't seem to help. I remove all PCI cards (other than my SCSI controller, which I boot off of) and disconect all IDE stuff, and it boots fine. I add back the PCI cards and it boots fine. I add back the IDE channel that's not connected to the new drive, and it's fine. I add back the IDE channel with the new drive, except I pull the molex power plug from the 1200JB, and the machine boots fine (even reads CDs in the CD-ROM drive, despite it being set as master and there is no slave, but whatever). If I put the old 800JB back in the machine reboots during the Windows splash screen. If I put my IBM HD in place of the WD drive, the machine works fine. When I unhook everything from the IDE channels but the 800JB or 1200JB the machine still reboots during the splash screen.

What on earth is going on? Could it be that I simply can't read WD's jumper settings? What is it that my DVD player will work just fine regardless of what else I have on the channel, despite being set as master? What could be causing this? Could I have somehow killed BOTH WD hard drives during my fiddling? Could it be a bad Windows install?

 

[mubs]

Well, well, I was replying and got locked out.

Adcadet, did you have a swap partition on the IDE drive(s)? If you did, hook it up, boot to PartitionMagic floppys and delete that partition. Nothing else makes sense to me now.

 

[Adcadet]

no, my swap files are on my "OS" and "Data" drive, both are SCSI and both are working perfectly. It's just this damn IDE crap.

 

[jtr1962]

You didn't have the drive data cable connected with the Molex power plug pulled, did you? That's what it sounds like from your post, anyway. That could possibly kill the drive(s) depending upon how the inputs are designed.

Because the machine works fine except when one of the WD drives are connected, I tend to think that's where the problem lies. I remember when older WD drives were run by themselves, you were supposed to leave the master jumper off. I'm not sure if the current ones are like that, but that could cause a problem like you're seeing. Why don't you try the WD drives in another machine just to make sure they still work? If so, then I'm fairly certain it comes down to a jumper or software problem, not a hardware problem.

I've been having a nagging problem with drives randomly shutting down so I can sympathize.

 

[Adcadet]

yeah, to disable the drive easily I would unplug the molex while leaving the data cable in ..... are you sure that can hurt drives? I've done that for a while and never had a problem (that I know of).

 

[Adcadet]

oh, and I'm hesitant to approach my wife about letting me mess with her PC. I'm fine killing my own drives, but I would feel horrible if I did anything to hers.

 

[jtr1962]

yeah, to disable the drive easily I would unplug the molex while leaving the data cable in ..... are you sure that can hurt drives? I've done that for a while and never had a problem (that I know of).

It may or may not-I just don't know the exact design of the data inputs. The problem is caused by the protection diodes on the inputs. If there is no power applied to the drive, then if a data input from the controller card goes high it will be clamped to about 0.7V by the protection diode, and a larger than normal current will flow. Fortunately, most outputs can't source a whole lot of current so in all likelihood nothing will happen. However, if many data lines are high then there could be enough current flowing through the protection diodes to heat up and damage the input IC, or even the current flowing in one input could blow a protection diode and short that input to ground permanently, rendering the drive useless. Of course, some inputs are configured with resistors to limit the current flow and prevent exactly that sort of thing. I doubt anybody outside of WD would know exactly which type of input their drives use.

Bottom line-what you did could cause a problem, but not necessarily. In all honesty, I would say it's not that likely unless the 5V line directly touches a drive input while the power plug is disconnected. That obviously can't happen unless it's done intentionally, so I'm leaning towards this being a jumper or software problem.

Do you have an old PC or M/B that you might be able to put those drives in to test? I understand your point about not wanting to screw up your wife's computer, but it needs to be established somehow whether or not the drives are bad.

 

[Adcadet]

no, I don't have any old PC here to test with. Though I did just order a new chip/mobo/ram for both my parents and wife. Perhaps when I build their PCs I'll see about my data drive. It sure would be nice to have that data, but I guess I can re-download my Lectures Online and MP3s. The important stuff (wedding pics, honeymoon pics, etc) were all backed up.

 

[Mercutio]

What did you use to affect the transfer? Ghost? WD's program?

 

[Mercutio]

Can you describe the exact jumper settings on your drives? eg. IDE0 M=WD80GB, S=DVD or whatever.

Have you tried them all in cable select mode?

 

[Adcadet]

I simply did a drag and drop to get the stuff over onto the new HD. Nothing fancy. Then I formatted the old HD, creating a 10 GB partition for an OS and the rest I used to store setup files, as this HD is going into my parents new computer.

 

[Adcadet]

I've tried so many cable settings by now....
I've been hesitant to use cable select as I'm not sure my motherboard (Tyan Tiger MP) or other drives supports it. I thought it would be wise to only hook up the questionable HDs and work from there.

 

[Adcadet]

great, and now this AM my system is not booting even without ANY IDE devices plugged in. I reset the motherboard and it seems back to "normal."

I boot into Windows with no IDE, and during the splash screen the machine reboots. Next time it gets into Windows, but after I log on the screen goes all crazy. When I reboot the BIOS screen is all funky. OK, so my video card may be crapping out on me. Could this be causing my IDE problems?

Video card = ATi Radeon 8500 128 MB hoooked up to two 19" monitors.

 

[Adcadet]

I start up the computer from a dead stop, and during the Windows splash screen one of the monitors flickers, then they both go blank.

 

[Mercutio]

Cable select is a standard, AD. It certainly should work.

 

[Adcadet]

hmm....who says I've ever had HD problems? My complaint has always been that "the screen goes blank." Sure, I've gotten the machine to boot a time or two without any IDE stuff hooked up, but I never ran the machine for much more than enough to get past the log on.

commnets? Am I just hoping?

So, the next question is, is this a video card problem, or a motherboard problem?

 

[Adcadet]

Cable select is a standard, AD. It certainly should work.

Oh. I guess I just never got past 1998 when I was working with computers that probably did not support it.

 

[Adcadet]

anybody know what the warranty length of retail ATi stuff is? My card is 1 year 7 months old.

 

[P5-133XL]

Cable select is a standard, AD. It certainly should work.

Oh. I guess I just never got past 1998 when I was working with computers that probably did not support it.

In 1998 it was probably supported, but you needed a special cable to make it work: CS has always been an option on IDE.

The special cable requirement disappeared with the advent of 80 pin cables and (ATA-66)

 

[Adcadet]

ah. Back on my first PC I built in 1998 or 1999 I distinctly remember using 40 pin cables for the CD-ROM.

 

[Adcadet]

So Mark - what do you think about my bad graphics card hypothesis? Any comment you wish to share?

 

[Handruin]

anybody know what the warranty length of retail ATi stuff is? My card is 1 year 7 months old.

My ATI retail says 3 year limited warranty. I'll guess that your's should be similar if not longer. I thought in the past ATI retail cards were 5 years, but it seems they have changed since mine says 3 years.

 

[Adcadet]

good to know. My wife has offered her computer. So I think later tonight I'll put her graphics card in my computer and pray that it works.

 

[P5-133XL]

So Mark - what do you think about my bad graphics card hypothesis? Any comment you wish to share?

Possible but a device conflict is more likely.

Q1. - for your IDE are you using the MB IDE controllers or an add-on card?

Look at your MB documentation and find out what PCI slot shares IRQ's with the AGP port. If you have a card there, then move it to another slot. Preferably, leaving the shared slot empty: You could be running into a IRQ/Port/Mem conflict with the AGP slot. Also clear your ESCD. That is where the MB keeps track of what it has assigned to what card and clearing it will cause it to re-assign everything.

Every time you change a card the MB dynamicly reassigns IRQ's, Ports, and bios mem-locations from the choices that the cards allow. It may be that the process of moving cards around the MB has created a conflict between two devices.

Also be careful about where you put your SCSI adapter relative to you IDE controller. Windows is very picky about device order in its boot.ini file. The boot.ini file points to the location to boot from by specifying the controller by sequence number (i.e. first controller, second controller, ...) Typically the controller number is in PCI slot order (but not necessarily, especially with PCI slots that share IRQ's). If you are using the internal PCI controllers then you need to look in the MB manuel to find out the PCI slot that they share IRQ's to find their seq. number relative to the SCSI controllers.

Make sure that your WD1200JB is properly jumpered. Specificly if it is the only device on the channel it must be jumpered as "single" - Not master or slave. If there are additional devices on the channel then make sure it is again properly jumperd. WD drives are very picky about their jumper settings.


Sorry about the complexity of my answer, but unfortunately the more the MB/OS automates stuff, the more complex the answer becomes when the automation goes wrong. Hopefully, what I said makes sense to you, is understandable, and helpful. If you have further questions

 

[Adcadet]

so after gutting the turkey that my father in law got us (wifey and me), I decided to give things one more shot. No IDE stuff hooked up. Started it up, screen stayed black. Rebooted a few times, and finally got a BIOS screen that was garbled. Later tonight I'll try my wife's card, and assuming hers works I'll call ATi.

 

[Adcadet]

threw in my wife's video card, and Windows started up and asked to be reactivated. So it looks like Windows is back to its nagging self, so it must be a bad video card. RMA request placed.

 

[sechs]

anybody know what the warranty length of retail ATi stuff is? My card is 1 year 7 months old.

My ATI retail says 3 year limited warranty. I'll guess that your's should be similar if not longer. I thought in the past ATI retail cards were 5 years, but it seems they have changed since mine says 3 years.

FYI, older cards are 5 years, newer ones are 3 -- either way you're fine.

 

[Adcadet]

I installed my wife's video card, and hooked up the new HD and Windows sat at the splash screen. I haven't had time to do anything else with the machine, so I don't really know what the deal is. Hopefully it's just Windows freaking out over the hardware changes. A new Windows install never hurt anybody (too much).