On the Emulation of TU58s

[tony duell]

> >
> >> And yet another (possibly the most common one on computers) is to have a
> >> small drive wheel that pulls the tape at constant speed across the
> >> heads, and then have some other construction that drives the tape reels
> >> depending on tape tension or length. Think vacuum columns or spring
> >> loaded arms.
> >
> > That is closely related to the capstan and weakly-driven take up spool I think.
> 
> I would disagree. There is no slipping clutch. Instead, the motors (on

True. I meant it was a system where the tape speed was determined by the
capstan, and the reels are driven to keep the tape round up, not all over the 
machine room floor :-)

> The actual tape movement as such, is all done by the small wheel next to
> the head, which just runs the tape past the head.

Which is essentailly the same as the capstan in an audio tape recorder, albeit the computer
drive doesn't have a pinch roller

> I've been wondering if there is some sensor of tape tension/pressure at
> the heads, and this will allow the drive to figure out how much
> faster/slower the reels must run, relative to each other, in order to
> keep the tape tensioned. Then you can figure out tape speed across the
> heads (if you care) by just observing flux changes.

Only if there is something on the tape. These computer tape drives could surely record on a 
totally blank tape and get the right number of bits per inch. So the thing can't use the data rate
at the head as a speed measurement.

> > One of my VHS video recorders does something similar to work out how much recording space is
> > left on the cassette.
> 
> That can't be very precise... :-)

It always underestimates the remaining tape (so that if it says there is 1 hour left, you can definitely fit
a 1 hour TV programme on there). But it is suprisingly good. 

-tony