This is a picture of the assembled sensor. You see a cardanic mount on the
right side, the central corpus where the encoder is mounted inside and a
ring structure on the left side. This ring is a spring, as can be clearly
seen in the following picture:
Here you clearly see that the ring is not uniform, but has thinner and
thicker regions. Running diametrially from left to right is the extended
sensor tip from the encoder. This arrangement measures the deformation of
the ring, due to the varying distance of the two measurement points. This
becomes clear when you look at this picture:
This is one of the short distances to be monitored. You see the sensor at
the right, the mount on the wall at the left and a thread connecting both.
This thread is made from a special steel alloy that has almost no
temperature coefficient.
So, when the two points where the sensor is mounted and where the other end
of the thread is fixed, respectively, change their distance this is
transformed into a deformation of the ring (spring) which in turn is sensed
by the encoder. Why this complicated arrangement?
Because the thread needs to be under a constant and known tension. Therefore
we need the spring, although this translates into a non-linear behaviour.
But this is taken into account by the software. The changes, as seen by the
sensor, are an order of magnitude smaller than the real changes, due to the
thread being elastic, too. Of course this depends on the length of the wire;
a longer wire results in less change at the sensor, compared to a short one.
The distances to be monitored range from 1.6m to 19m, while the resolution (not accuracy, of course) is 1µm. This is due to the precise encoder from Heidenhain, shown to the right. These sensors are the size of a ball pen.