Alignment disks, and the creation thereof

[Fred Cisin]

On Mon, 9 Feb 2009, Mr Ian Primus wrote:
> I've been reading the now increasingly mistitled thread on alignment
> disks, and it's gotten me thinking. What IS required to make alignment
> disks? I've always aligned malfunctioning floppy drives "the wrong way",
> using a known good disk - or an original software disk. On the Commodore

It may well be adequate for you.
But, . . .   I PRODUCE "original software disks", although admittedly WAY
less than 20 years ago.  If you end up "aligning" a drive for best
readability using one of MY distribution disks (or even one of
MICROS~1's), then we end up with a RESPONSIBILITY (could prob'ly get
sued!) to make sure that they are properly/well aligned, not merely "good
enough to work in a "known good" drive" (which might be FINE for your
use).

Yikes!   People "aligning" drives to match MY drives??!?


> With today's modern machining technology, it should definitely be
> possible to create a system to write new alignment disks. Starting with
> a standard floppy drive, you could replace the stepper drive with
> something more precise and controllable. Then, you have to somehow
> accurately calibrate this whole thing. Is there a document out there
> that describes the track layout of a standard (say, 5 1/4" 48tpi, 40
> track) diskette? There has to be a spec, one that describes exactly the
> distance between tracks, the width of a track, and the distance from the
> center hub to the beginning of the first track. If you had that
> information, and precision measurement tools, one could set up and
> calibrate the drive to spec.

Yes, the specs are readily available.
It would not be a major feat of prototyping to build a drive that MAKES
alignment diskettes.


For RADIAL alignment, even I (grossly, totally incompetent as a machinist)
could create a CRUDE digital alignment disk with specs and Harbor Freight
tools!:

Take a drive with a narrow head (96TPI or 100TPI),  (The Micropolis lead
screw, if not too worn (lash!), seems ideal),
align it,  ACCURATELY!
add a dial indicator for more precision, maybe improve the clamping,
maybe change the tunnel erase?
record a few tracks at 48TPI,
Record a track very slightly further out than spec,
(you can manually rotate the lead screw a fixed amount)
record the next track very slightly further IN than spec,
record the next track further out from spec,
the next one further in than spec, etc.

Read some tracks.
Do you get more errors on odd numbered or even numbered tracks?


> Then, of course, there needs to be a controller for the thing, a
> computer interface to control the precise positioning of the head, as
> well as being able to read/write to the disk.

You can make a stand-alone controller for the task - called a "drive
exercisor" or "exOrcisor"
Or, use a PC.  Trivial software to convert a PC into a drive exercisor.

> It's possible. It's beyond the equipment and tools at my disposal, but
> it's definitely doable for a hobbyist who has access to a machine shop
> and good tools.

It actually wouldn't take much (see above).

> Anyone out there a machinist? :)
Although he won't admit to it, Tony is a better machinist than some
professionals that I have dealt with (I ended up doing my own valve seat
grinding, but sent out flywheel resurfacing, crank balancing, etc.)
It might be worthwhile to give Tony a 100TPI drive, and get him to make
some alignment disks.

--
Grumpy Ol' Fred     		cisin at xenosoft.com