vibration, interferometry and gratings
Posted by
Elliot Burke
on 2000-03-21 07:30:59 UTC
Vibration can be a big problem with a sensitive detection system, here is an
example.
As a numericaly example assume a machine was taking a heavy cut and the
stage was vibrating with a 0.004" amplitude, (0.008" PV), at a 60 Hz rate
(3600 rpm cutter). The equation of motion in mm is 0.1 sin(2 pi *60 *t).
The velocity is 2 pi *60* 0.1 cos(2 pi *60*t), peak velocity is 2 pi*60*0.1
mm/sec = 37.7 mm/sec. The fringe spacing in interferometry is about 0.3
micron, or 0.0003 mm, thus the fringe frequency would be 37.7/0.0003 = 125
kHz.
Interferometer frequently have bandwidths higher than this, the HP
interferometer had about 1 MHz bandwidth, if memeory serves.
Crossed gratings was the original way that quadrature signals were obtained
from moire gratings. It works OK. Modern systems use a special readout
grating, so that a compact detector array can be used.
I don't know of a cheap source of gratings except having them printed on a
linotronic machine. These are, as others have noted, of uncertain accuracy.
But they are very entertaining to look at.
The trick to be able to get moire patterns of high contrast on a detector is
to use collimated or nearly collimated light. A narrow angle
("ultrabright") led will do OK if your pattern frequency isn't too high.
Many commercial readouts use special collimated leds, these are cans with an
led die and a lens.
Elliot Burke
example.
As a numericaly example assume a machine was taking a heavy cut and the
stage was vibrating with a 0.004" amplitude, (0.008" PV), at a 60 Hz rate
(3600 rpm cutter). The equation of motion in mm is 0.1 sin(2 pi *60 *t).
The velocity is 2 pi *60* 0.1 cos(2 pi *60*t), peak velocity is 2 pi*60*0.1
mm/sec = 37.7 mm/sec. The fringe spacing in interferometry is about 0.3
micron, or 0.0003 mm, thus the fringe frequency would be 37.7/0.0003 = 125
kHz.
Interferometer frequently have bandwidths higher than this, the HP
interferometer had about 1 MHz bandwidth, if memeory serves.
Crossed gratings was the original way that quadrature signals were obtained
from moire gratings. It works OK. Modern systems use a special readout
grating, so that a compact detector array can be used.
I don't know of a cheap source of gratings except having them printed on a
linotronic machine. These are, as others have noted, of uncertain accuracy.
But they are very entertaining to look at.
The trick to be able to get moire patterns of high contrast on a detector is
to use collimated or nearly collimated light. A narrow angle
("ultrabright") led will do OK if your pattern frequency isn't too high.
Many commercial readouts use special collimated leds, these are cans with an
led die and a lens.
Elliot Burke