Tarbell is making me insane

Chuck Guzis cclist at sydex.com
Tue Nov 13 15:55:01 CST 2007


On 13 Nov 2007 at 11:34, Grant Stockly wrote:

> Also, are DD disks the same as Single Sided 3.5" 360/400k 
> disks?  Except for the double sided part?

Very close--I've got 3M's published specs on all of these, if anyone 
cares.  And 3.5 DSHD media is fairly close to DS2D; the coercivity is 
somewhat higher, requiring increased write currents, but not as 
drastic as the difference between 5.25" DSDD and DSHD.
  
> When copying from drive A to drive B it has an error at the end of 
> the disk.  Is this where a data rate issue would be the worst?

> DISK WRITE ERROR: =*.*
> 
> A: R/O, SPACE: 0K
> B: R/W, SPACE: 11K

Of course--for at least two reasons.  The first is that the linear 
path is shorter; you're packing more bits per unit length.  The 
second is that the surface velocity with respect to the head is 
lower.  Since the induced write current is a function of the square 
of the velocity (another one of those v**2 laws that are so common in 
physics), the signal is weaker.

The methods of getting around this are several.  One can vary the 
rotational speed of the drive so the linear velocity remains 
relatively constant (Apple Lisa and the Victor 9000 come to mind).  
The upshot here is that outer tracks will hold more than inner ones, 
so a translation layer is necessary in software when attempting to 
determine the address of a block.

One can also vary the speed of the bit clock, slowing it as the 
center of the disk is approached.  This doesn't elminate the problem 
of a lower read signal on the inner tracks, but it helps.

One can reduce the write current on the inner tracks (8" drives do 
this on receipt of the TG43 signal) to make the transitions a bit 
cleaner (oxide coated media has a tendency to "blur" or "shift" the 
edges of domains if they're close together).  

Finally, one can play some tricks, varying the timing of bit 
transitions during writing depending on the pattern of adjacent bits 
to make things look better when reading--in other words, if two bits 
are close together, one can make the transition of the first one a 
bit earlier or the transition of the second one a bit later.  This is 
mostly done with MFM recording, not FM.

I don't know enough about your controller to diagnose things, but 
perhaps this will help a bit.

Cheers,
Chuck




More information about the cctech mailing list