Stephen Goodman was wondering about a single finger tracking system
tracking torn paper when it wants to go to the right.

The left finger, I have found, tends to monitor what the industry
considered to be the guide edge of the roll as it went through the
perforator.  That's the edge that never changes in relationship to the
paper swelling or shrinking.  So when the left paper edge is right on
the finger line but not actuating it, the first holes in the paper will
be exactly over their trackerbar holes.

Some perforators nibbled (trimmed) both edges of the paper and some
just on one edge, by the way.  The criteria as to which way a roll
swells or shrinks was just the industry definition, as I understand it.
The first note hole on either edge would be almost perfect with its
edge.  But 90% of the time, tracking the left edge of the paper
produces the best normal roll tracking because this was the edge they
designated as the guide edge, and perforated to this edge.

When it comes to torn and stretched rolls, the single finger tracker
works fairly well when its a mechanical tracker in contrast to an air
tracker, because it is then set at a central position inboard of the
ideal paper edge, and so it monitors both the in and the out movements
of the paper.

Air trackers cannot do that, and can monitor only the "in" movement,
against the finger.  So Themodists with single finger tracking systems
are out of luck when the paper moves to the right.

The way I then adjust them to catch both "iny" and "outy" is to set the
finger such that a correctly tracking roll is just opening the finger
all the time, and the tracker bleed is large enough to have a bit of
proportional characteristic with the resistance of the heavy napped
leather I try to utilize for the finger pad.  When the pad nap "cones"
into its nipple hole, then it becomes very slightly proportional
relative to bellows speed, which equates to percent bellows travel/inch
roll travel.

That allows the finger to be in constant touch with the paper and
always correcting it just a bit.  The correct path then tends to
subtract a bit from the overall width of correction available, but
divides it between left and right tracking.

Craig Brougher