Scor­pio News

Dave Hunt Begins

My bit in the last issue about making AT High Density drives format 730K normal double density disks has caused quit a lot of fun and burrowing into the undocumented realms of IBMs, clones, MS-DOS PC-DOS. Incidentally, the last article referred to 720K disks, no, the disks haven’t grown an extra 10K in the space of three months, just that I got it wrong. I bad done those experiments some time before I wrote it up and at the time I simply said two 360 is 720 and there you are! The actual size is 731K, so from now on, they’ll be referred to as 730K disks in sympathy with the MS-DOS habit of rounding down to the nearest 10K.

Firstly, the Drivparm statement only applies to MS-DOS 3.2 and later, Microsoft have provided a for various disk parameters and the syntax is similar to that given for Drivparm in the last issue. Examples of the use of DRIVER.SYS are given in John’s letter below. Secondly, what follows only applies to AT type computers fitted with IBM AT type disk controllers and one or more High Density drives. The 730K disk trick will work with most other IBMs and clones which are NOT AT types.

Now we (the collective is John and myself, with phone calls to Paul and Brian) think we’ve got to the bottom of it and we’re fairly sure who’s to blame. It all comes down to the IBM spec. for the controller card for the IBM AT, and clone manufacturers being in the business of cloning Big Blue, clone away, warts and all. Put simply, the IBM controller card is physically incapable of formatting a disk with 80 tracks and in normal density.

For starters, I suppose I’d better explain a bit about High Density drives and disks. The drive itself looks and feels, and to all intents and purposes IS, a fairly normal 5¼″ drive. The only real difference is pin 2 on the drive connector which for a normal drive is marked ‘Reserved’. Actually, on a normal drive it isn’t usually connected to anything, it just sits there. The HD drive uses this pin to switch it into HD mods. Put a logical 1 on the pin and two things happen; firstly, the drive increases rotational speed from 300 r.p.m. to 360 r.p.m., secondly, the electronics switch into a mode suitable for use at double the normal data rate, 500K bits instead of the usual 250K bit/s. In other words, it now behaves like an 8″ drive, only smaller.

The problem is the data packing density on the disk, related to the disk speed. Obviously, the angular velocity of an 8″ disk is a lot higher (the tracks are near the outer edge of the disk as is the case for 5¼″ disks) than the smaller disk. When it comes to 5¼″ disks, the normal disk can’t cope, the packing density is too high to make it reliable. Two approaches are made to overcome this, the first is to increase the angular velocity by jacking up the speed from 300 r.p.m. to 360 r.p.m., the second is to use disks coated with a magnetic pudding of higher coercivity, we call these HD disks. Higher coercivity magnetic coatings means lower head currents which don’t saturate the disks to the that is on normal disks, so it means that HD disks format unreliable (if at all) when used in place of ‘normal’ disks.

So back to the IBM controller card. Why IBM? Because that’s the one I’ve got the circuit diagrams for. Now, apart from s real genuine Big Blue AT at work, the rest of us use an assortment of clones with an amazing variety of controller cards, and despite the different chippery on the controller cards, they share two things in common. They were intended for AT clone computers and they are all capable of running two floppy disks (not more) and two hard disks.

Now there are three possible permutations covered by the combinations of drives we could fix.

80 track HD drive should be able to do:
a) 1.2M HD disks (80 tracks 15 sectors)
b) 730K Normal density (80 tracks 9 sectors)
c) 360K Normal density (40 tracks 9 sectors) double step
80 track ’Normal’ drive should be able to do:
b) 730K Normal density (80 tracks 9 sectors)
c) 360K Normal density (40 tracks 9 sectors) double step
40 track ’Normal’ drive should be able to do:
c) 360K Normal density (40 tracks 9 sectors) double step

(There are other perms. as well, all should be capable of the older 8 sector format in both double and single sided, but as far as I’m concerned, these don’t count.)