atomic_fungus (atomic_fungus) wrote,

#354: Hard drive adventures

Today I installed the 160 GB Seagate I bought yesterday.

Installation was pretty painless. On the old P3 all I would have needed was a pair of hemostats for removing the drive jumper. That old case was well-designed. Hard drives just slide in, and a metal lever with a green plastic grip locks the drive in place.

Even though this new Gateway has a more typical case, where everything is held together with screws, it was still pretty easy. Two screws got the drive cage out; four screws hold the new drive in the cage, and SATA cables can only be plugged in one way.

The drive has a jumper on it which limits it to a 1.5 GB/second data rate, for compatibility with "older systems". I worried about that for a moment. Then I realized that any computer which has only XP and later drivers for its motherboard probably won't have trouble with 3.0 GB/second data transfer.

But I am an old-time computer technician. I cut my teeth on 286s, back when you still have to throw DIP switches on the motherboard to configure certain things. (That is, 1990.) When I first started fixing computers for a living, few machines had more than one hard drive in them. And, this was before IDE (what is now known as ATA or EIDE) became widespread, and only top-name machines had IDE drives--IBM and Compaq, mainly (and that shows how long ago that was: Compaq was a maker of good computers).

In 1990 it was highly common for a computer to have a hard drive which required a separate controller. There were two cables going from this controller: one was the cable which told the drive what to do, and the other was a data cable. The latter cable was an eight-bit parallel data cable.

Without the controller, the drive was utterly useless. There were no plugs or sockets on the motherboard for that. The controller was the brains of the hard drive--it contained the BIOS routines that the computer needed in order to run the hard drive, and when the computer requested a data function (such as "write the following data to block XXYYZZ") it was the drive controller which actually moved the drive's heads and told the drive what to write where.

So when you wanted to put more than one drive in a computer, you had to keep in mind what sort of controller you had and what sort of drive you had. You had to buy drives which the computer knew how to address--there were around 23-46 sets of drive parameters in the BIOS, and there were no "custom" drive types--and you had to know how to terminate drives.

Owing to the finicky nature of high-speed data circuits, there had to be terminating resistors at each end of the cable connecting the drive to the controller. If you had a Seagate ST-251 on your controller as Drive 0, and you wanted to add another drive, you had to remove the terminating resistors from it and hook the second plug of the cable onto it. The cable had a twist in it, and the socket past that twist--farthest from the controller, that is--was for Drive 0. The socket which was closer to the controller than the twist was for Drive 1.

Seagate had three SIP sockets on the underside of the drive's motherboard, and you removed these resistors from Drive 1. Once the hardware was installed, in some cases you even threw a DIP switch on the motherboard to tell it that you'd installed a second hard drive. Then you powered up the computer, went into the BIOS setup, and found the drive type (parameter list) which matched the one you'd installed. (The ST-251 was almost always Type 40, 860 cylinders, 6 heads, 16 sectors per track. At 512 bytes per sector that was 42 MB of formatted capacity. That's how common the ST-251 was; I still remember the parameters.)

But you weren't done yet. The first thing you did once the system finished booting--the primary OS of the day was DOS 3.3; it didn't take long--you ran a utility to low-level format the hard disk.

The low-level format basically put the data tracks on the drive platters. Before the low-level, it was utterly blank. (This was sometimes called a "physical format" because it was laying down the magnetic tracks on the platters.) In the case of the ST-251, which used a stepper motor to position the RW heads, they couldn't pre-format the thing at the factory because after all the shipping and handling, by the time it got to the consumer the registration of the drive would be off, and the customer would have to low-level format it anyway.

That was both good and bad. Stepper-motor drives were seen as somewhat inferior to voice-coil drives, because they were slower and had less thermal stability--but they were tanks. I used an ST-251 which had been dropped on its top cover, from a height of about five feet, and it worked perfectly and never lost a bit of data. (It howled like a banshee whenever it was on, but it worked.)

After the low-level format, then you ran the DOS FDISK utility to set up partitions on the drive. Partitions were how DOS dealt with hard drives. DOS 3.3 could only handle a drive of about 32 MB. If you had an ST-251 in your computer, that would leave about 10 MB of extra space, so you would set up an additional partition. (Some people split the drive into two equal partitions of 21 MB each.) FDISK set up the Master Boot Record (MBR) which let DOS "see" the hard drive and understand the drive's operating parameters.

Then you would use the DOS FORMAT utility to do the high-level, or "logical", format, which set up the sectors and FAT and directories which DOS needed in order to use the drive. After all that was done, you had a usable hard drive.

It took someone with some technical ability to do all this stuff. The commands were not very user-friendly and the hardware was complicated.

When IDE came into vogue, it eliminated all of the hardware stuff. There were no terminating resistors or DIP switches to worry about; all you needed to do was make sure that the drive "master/slave" settings were right, and that normally consisted of making sure a single jumper was either present or absent.

You still needed to low-level format the drive, and run FDISK, before you could run FORMAT and use the drive, but the lack of hardware BS made things easier for everyone. As long as the jumpers were set correctly it didn't matter which drive was where on the IDE bus, either, and the cable ran straight from controller port to the drive.

Also, BIOS finally got a "user-defined" drive type, which let you select--within reason--the drive parameters of your drive, so you were no longer confined to certain drives.

IDE stands for "integrated drive electronics" and it basically meant that the controller was now part of the drive's on-board electronics. The drive could be anything--MFM, RLL, SCSI, WTF--but as long as it had the 40-pin connector and obeyed the right commands, it was IDE and any IDE controller (really, an interface, not a "controller") would work with it.

I have never installed a hard drive without having to do a lot of this stuff to make it usable. Even with Windows ME, I still had to run FDISK and FORMAT and all the rest of it.

Even Macintoshes had a lot of stuff to do. They used SCSI exclusively for years, and that meant "terminating resistors"--Apple's drive setup program was a bit easier to use than DOS, but it was basically the same thing as DOS with a different appearance.

So what happened today, when I put the 160 GB into my machine running Windows Vista, was a bit disconcerting.

I installed the drive in the machine, and turned it on.


* * *

[about half an hour later]

Windows Vista installed drivers but I could not figure out how to get the stupid thing partitioned or formatted. Microsoft's web site was singularly unhelpful. There are no instructions on how to do either in the knowledge base.

Finally after playing around with the administrative software I figured it out. The drive is now formatting, even as I type this.

Besides all that, though, when I was installing the drive I unplugged the keyboard, because the cable was too short for me to lay the thing down. When I turned the computer on again, I forgot to plug the keyboard in. I got as far as starting to type a blog entry when I realized this.

...stupid keyboard would not work. I had to shut down the computer to get the computer to recognize that I'd plugged it in.

You know, the last PC-DOS-Intel-Windows machine which didn't have a hot-pluggable keyboard interface was the PC-XT. Morons.

  • #8259: Okay, that's a little better

    Flopped for about 20 min, had some ibuprofen and a shower; now I feel halfway functional. At least enough to eat dinner. Typing no longer hurts. This…

  • #8258: There is nothing left.

    I spent the last four and a half hours--with about a 20-minute respite--in motion. Pool is up. It's leaking, but I'm pretty sure I know where…

  • #8257: It really amuses me, in fact.

    He's right, this is their perennial response. "If we can't have abortions, then the men have to be sterilized." The theory is that the men must be…

  • Post a new comment


    default userpic

    Your reply will be screened

    Your IP address will be recorded 

    When you submit the form an invisible reCAPTCHA check will be performed.
    You must follow the Privacy Policy and Google Terms of use.