U.S. patent application number 16/655983 was filed with the patent office on 2021-04-22 for etalon based optical spectrometer.
The applicant listed for this patent is LightMachinery Inc.. Invention is credited to Jesse Dean, John H. Hunter, Hubert Jean-Ruel, Ian J. Miller, Edward S. Williams.
Application Number | 20210116302 16/655983 |
Document ID | / |
Family ID | 1000004421361 |
Filed Date | 2021-04-22 |
![](/patent/app/20210116302/US20210116302A1-20210422-D00000.png)
![](/patent/app/20210116302/US20210116302A1-20210422-D00001.png)
![](/patent/app/20210116302/US20210116302A1-20210422-D00002.png)
![](/patent/app/20210116302/US20210116302A1-20210422-D00003.png)
![](/patent/app/20210116302/US20210116302A1-20210422-D00004.png)
![](/patent/app/20210116302/US20210116302A1-20210422-D00005.png)
![](/patent/app/20210116302/US20210116302A1-20210422-D00006.png)
![](/patent/app/20210116302/US20210116302A1-20210422-D00007.png)
![](/patent/app/20210116302/US20210116302A1-20210422-D00008.png)
United States Patent
Application |
20210116302 |
Kind Code |
A1 |
Jean-Ruel; Hubert ; et
al. |
April 22, 2021 |
ETALON BASED OPTICAL SPECTROMETER
Abstract
Conventional etalon based spectrometers have either a limited
range of evaluation wavelengths or require continuous scanning of
the etalon. Conventional etalon based spectrometers also have
limited contrast between the peak transmission of a frequency on
resonance and the minimum transmission of a frequency off
resonance. An improved optical spectrometer includes a cylindrical
lens configured to converge the input beam of light in only one
direction, whereby the input beam of light is focused along a focal
line. Accordingly, a first etalon receives the input beam of light,
and transmits a series of sub-beams, each sub-beam transmitted at a
different angle from the normal, and each sub-beam including
multiple frequencies based on the FSR, whereby a secondary
dispersive element receives each sub-beam, and disperses each
sub-beam into individual frequencies. Ideally, a second etalon
receives and transmits the sub-beams with increased contrast, and a
second thickness of the second etalon is substantially identical to
the first thickness of the first etalon.
Inventors: |
Jean-Ruel; Hubert; (Ottawa,
CA) ; Dean; Jesse; (Nepean, CA) ; Hunter; John
H.; (Almonte, CA) ; Miller; Ian J.; (Ottawa,
CA) ; Williams; Edward S.; (Kanata, CA) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
LightMachinery Inc. |
Nepean |
|
CA |
|
|
Family ID: |
1000004421361 |
Appl. No.: |
16/655983 |
Filed: |
October 17, 2019 |
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G02B 5/20 20130101; G01J
3/18 20130101; G01J 3/26 20130101; G01J 3/14 20130101; G01J 3/0208
20130101 |
International
Class: |
G01J 3/26 20060101
G01J003/26; G01J 3/02 20060101 G01J003/02 |
Claims
1. An optical spectrometer for dispersing an input beam of
collimated light comprising a plurality of frequencies, comprising:
a cylindrical lens configured to converge the input beam of light
in only one direction, whereby the input beam of light is focused
along a focal line of the cylindrical lens and remains collimated
in a perpendicular direction; a first etalon comprising a first
spacer including a first thickness, a first reflective surface
including a first reflectivity, a second reflective surface
including a second reflectivity, a normal to the second reflective
surface, and a free spectral range (FSR), the first etalon
configured to receive the input beam of light, and transmit a
series of sub-beams, each sub-beam transmitted at a different angle
from the normal, and each sub-beam including multiple frequencies
based on the FSR; a second etalon configured to receive the series
sub-beams from the first etalon, and transmit the sub-beams with
increased contrast wherein the second etalon comprises a second
spacer with a second thickness, a third reflective surface
including a third reflectivity, and a fourth reflective surface
including a fourth reflectivity; and a secondary dispersive element
configured to receive each sub-beam from the second etalon, and
disperse each sub-beam into individual frequencies.
2. The optical spectrometer according to claim 1, further
comprising a camera configured to receive the individual
frequencies, and provide a measure of each of the individual
frequencies.
3. The optical spectrometer according to claim 1, wherein first
reflectivity is between 90% and 97% for a desired wavelength range,
and the second reflectivity is between 90% and 97% for the desired
wavelength range.
4. The optical spectrometer according to claim 1, wherein the focal
line of the cylindrical lens is between the first etalon and the
secondary dispersive element.
5. (canceled)
6. The optical spectrometer according to claim 1, wherein the third
reflectivity is between 90% and 97% for a desired wavelength range,
and the fourth reflectivity is between 90% and 97% for the desired
wavelength range.
7. The optical spectrometer according to claim 1, wherein the
second thickness of the second etalon is a substantially identical
to or an integer multiple of the first thickness of the first
etalon; whereby each sub-beam is transmitted by both the first
etalon and the second etalon at substantially a same angle, and
whereby each frequency is already sorted to angles of incidence
required for maximum transmission.
8. The optical spectrometer according to claim 7, wherein the
second thickness of the second etalon is within 50 nm of the first
etalon.
9. The optical spectrometer according to claim 1, wherein the focal
line of the cylindrical lens is in between the first etalon and the
second etalon.
10. The optical spectrometer according to claim 1, wherein the
first etalon comprises a virtually imaged phase array (VIPA).
11. The optical spectrometer according to claim 10, wherein first
reflectivity is between 99% and 100% of a desired wavelength range,
and the second reflectivity is between 90% and 98% for the desired
wavelength range.
12. The optical spectrometer according to claim 10, wherein the
focal line of the cylindrical lens is near the third reflective
surface.
13. The optical spectrometer according to claim 1, wherein the
first etalon and the second etalon are tilted in a first direction
relative to an optical axis of the cylindrical lens by a same
amount to reduce effects of back reflection from the second etalon
on the first etalon.
14. The optical spectrometer according to claim 13, wherein the
first etalon and the second etalon are tilted by
0.1.degree.-2.degree. in the first direction in which the input
beam light is converging or diverging.
15. The optical spectrometer according to claim 14, wherein the
first etalon and the second etalon are tilted relative to each
other by 0.1.degree.-2.degree. in a second direction perpendicular
to the first direction to compensate for mismatch between the
second thickness and the first thickness.
16. The optical spectrometer according to claim 10, wherein the
VIPA comprises a first finesse; wherein the second etalon comprises
a second finesse; and wherein a ratio of the first finesse to the
second finesse is between 1.5:1 to 3:1.
17. The optical spectrometer according to claim 13, wherein an
angle between converging extreme rays of the input beam of light
after the cylindrical lens is less than a tilt of the first etalon,
whereby substantially all of the input beam of light reflects in a
same direction in the first etalon.
18. A method of manufacturing an optical spectrometer comprising
substantially identical first and second etalons comprising:
providing a spacer including a first side and a second side in
parallel to each other; coating the first side of the spacer to
form a first reflective surface; coating the second side of the
spacer to form a second reflective surface; cutting the spacer into
two sections forming the first etalon and the second etalon; and
fixing the first etalon and the second etalon substantially in
parallel to each other.
19. The method according to claim 18, further comprising
positioning a cylindrical lens so that a focal line of the
cylindrical lens is focused along a desired location between the
first etalon and the second etalon.
20. The method according to claim 19, further comprising tilting
the first etalon and the second etalon relative to an optical axis
of the cylindrical lens to reduce effects of back reflection from
the second etalon on the first etalon, and to compensate for
mismatch between a first thickness of the first etalon and a second
thickness of the second etalon.
21. The method according to claim 18, further comprising coating
the second side of the first etalon with an additional amount of
material to increase a reflectivity thereof, thereby forming the
first etalon into a VIPA with substantially a same thickness as the
second etalon.
Description
TECHNICAL FIELD
[0001] The present invention relates to a Fabry-Perot etalon based
optical spectrometer, and in particular to a high contrast optical
spectrometer utilizing multiple etalons.
BACKGROUND
[0002] Conventional Fabry-Perot etalons (etalon) are utilized as
optical filters, which receive an incident light beam including
constituent frequencies, and transmit evenly spaced light
frequencies as a function of the thickness of the etalon, i.e.
distance between parallel reflective surfaces, and the angle of
incidence. However, if the incident light beam is incident over a
range of angles, e.g. if the incoming light is converging or
diverging, then the full spectrum of the incident light beam may be
transmitted with each frequency at a different output angle, i.e.
from the etalon's normal, since each angle corresponds to a
different distance across the etalon and a different thickness.
Moreover, each angle will potentially represent multiple
constituent wavelengths separated by the corresponding free
spectral range (FSR) of the etalon at that angle.
[0003] Unfortunately, most etalon based spectrometers have either a
limited range of evaluation wavelengths, or require continuous
scanning of the etalon, either by varying the separation between
reflective surfaces or by varying the tilt of the etalon relative
to the incident light beam. Another common limitation of etalon
based spectrometers is limited contrast between the peak
transmission of a frequency on resonance and the minimum
transmission of a frequency off resonance.
[0004] One approach to increasing the etalon contrast is to pass
the incident light beam through a series of etalons with
substantially different thicknesses. Fine tuning of the
transmission of each etalon can be achieved by various tuning
means, e.g. temperature, angle of incidence or pressure tuning in
the case of an air spaced etalon. Unfortunately, this approach can
only yield very high contrast for a series of discrete frequencies
of light. To view a continuous spectrum with multiple etalons, the
etalons must all be tuned simultaneously, so that the each
wavelength of interest is transmitted through all of the etalons as
the etalons are tuned. Tuning can be through some combination of
pressure tuning, piezo scanning angle tuning or temperature tuning.
Additionally, the light incident on a series of etalons in this
configuration must be substantially collimated.
[0005] Another version of an etalon based spectrometer is too
illuminate an etalon with nominally collimated light. Multiple
frequencies of light are transmitted at each incident angle, spaced
according to the free spectral range of the etalon and the incident
angle. In order to separate the multiple wavelengths at each angle,
the etalon filtered light must be transmitted through a slit, after
which the different orders from the etalon may be split into bands
by a secondary dispersive element. FIG. 1 illustrates a plot of
transmission vs angle for two substantially different etalon
thickness, e.g. 2.0 mm thick and 2.1 mm thick, illustrating an
overlap, i.e. resultant transmission, of only a single
frequency.
[0006] An object of the present invention is to overcome the
shortcomings of the prior art by providing a spectrometer with high
resolution, broad wavelength coverage, no scanning required and
high speed acquisition of complete spectra.
SUMMARY OF THE INVENTION
[0007] Accordingly, the present invention relates to an optical
spectrometer for dispersing an input beam of collimated light
comprising a plurality of frequencies, comprising:
[0008] a cylindrical lens configured to converge the input beam of
light in only one direction, whereby the input beam of light is
focused along a focal line of the cylindrical lens and remains
collimated in a perpendicular direction;
[0009] a first etalon comprising a first spacer including a first
thickness, a first reflective surface including a first
reflectivity, a second reflective surface including a second
reflectivity, a normal to the second reflective surface, and a free
spectral range (FSR), the first etalon configured to receive the
input beam of light, and transmit a series of sub-beams, each
sub-beam transmitted at a different angle from the normal, and each
sub-beam including multiple frequencies based on the FSR;
[0010] a secondary dispersive element configured to receive each
sub-beam, and disperse each sub-beam into individual
frequencies.
[0011] Another aspect of the present disclosure relates to an
optical spectrometer further comprising a second etalon between the
first etalon and the secondary dispersive element configured to
receive the sub-beams and transmit the sub-beams with increased
contrast; wherein the second etalon comprises a second spacer with
a second thickness, a third reflective surface including a third
reflectivity, and a fourth reflective surface including a fourth
reflectivity.
[0012] Another feature of the present disclosure provides an
optical spectrometer, wherein the second thickness of the second
etalon is a substantially identical to or an integer multiple of
the first thickness of the first etalon; whereby each sub-beam is
transmitted by both the first etalon and the second etalon at a
same angle, and whereby each frequency is already sorted to angles
of incidence required for maximum transmission.
[0013] Another aspect of the present disclosure includes an optical
spectrometer, wherein the first etalon comprises a virtually imaged
phase array (VIPA).
[0014] Yet another aspect of the present disclosure relates to a
method of manufacturing an optical spectrometer comprising
substantially identical first and second etalons comprising:
[0015] providing a spacer including parallel first and second
sides;
[0016] coating the first side of the spacer to form a first
reflective surface;
[0017] coating the second side of the spacer to form a second
reflective surface;
[0018] cutting the spacer into two sections forming the first and
second etalons; and
[0019] fixing the first and second etalons substantially in
parallel to each other.
BRIEF DESCRIPTION OF THE DRAWINGS
[0020] The invention will be described in greater detail with
reference to the accompanying drawings which represent preferred
embodiments thereof, wherein:
[0021] FIG. 1 is a plot of etalon transmission vs etalon angle for
two different etalon thicknesses in a conventional multi-etalon
device;
[0022] FIG. 2 is a top view of an etalon spectrometer in accordance
with an exemplary embodiment of the present invention;
[0023] FIG. 3 is a side view of the device of FIG. 2;
[0024] FIG. 4 is an isometric view of a portion of the device of
FIG. 2;
[0025] FIG. 5 is a top view of a multi-etalon spectrometer in
accordance with an exemplary embodiment of the present
invention;
[0026] FIG. 6 is a side view of the device of FIG. 5;
[0027] FIG. 7 is a plot of etalon transmission vs etalon angle for
two identical etalon thicknesses in the multi-etalon device of FIG.
5;
[0028] FIGS. 8A to 8E illustrate a method of manufacturing the
device of FIGS. 5 and 6;
[0029] FIG. 9 is a top view of a multi-etalon spectrometer in
accordance with an exemplary embodiment of the present
invention;
[0030] FIG. 10 is a side view of the device of FIG. 9;
[0031] FIGS. 11A to 11E illustrate a method of manufacturing the
device of FIGS. 9 and 10.
DETAILED DESCRIPTION
[0032] While the present teachings are described in conjunction
with various embodiments and examples, it is not intended that the
present teachings be limited to such embodiments. On the contrary,
the present teachings encompass various alternatives and
equivalents, as will be appreciated by those of skill in the
art.
[0033] With reference to FIGS. 2 to 4, an exemplary embodiment of
an etalon spectrometer 1 includes a cylindrical lens 2 for focusing
the incoming, e.g. collimated, beam of light, including a plurality
of different frequencies, in only one direction along a focal line
3 onto an etalon 4, while the beam of light remains collimated in
the perpendicular direction. The etalon 4 is illuminated by a range
of angles in one axis only, as the light is collimated in the
perpendicular axis. The etalon 4 includes first and second
reflective surfaces, e.g. coatings, 6 and 7, separated by a spacer,
e.g. glass block, 8. The first and second reflective surfaces may
have a reflectivity of at least 70%, typically between 70% and
99.5%, preferably greater than 90% and more preferably between 90%
to 97% for all of the wavelengths in a desired wavelength range.
The focal line 3 of the cylindrical lens 2 may be in between the
reflective surfaces 6 and 7, e.g. in the middle of the spacer 8;
however, focusing the incoming beam of light along a line in
between the cylindrical lens 2 and the etalon 4 or on the far side
of the etalon 4 is within the scope of the invention. The preferred
location for the focal line 3 is somewhat behind, the etalon 4,
i.e. the opposite side of the etalon 4 from the cylindrical lens 2,
so as to minimize the area of the etalon 4 which contributes to the
transmitted signal. If the focal line 3 is far in front of the
etalon 4, the diverging beam will cover a larger area of the etalon
4, than if the beam is converging as it repeatedly bounces between
the etalon reflective surfaces 6 and 7. The ideal location of the
focal line 3 will depend on the etalon finesse, thickness, and the
tilt of the etalon 4. Greater etalon tilt will reduce the
significance of the location of the focal line 3 because the
walking of the beam becomes relatively more important. Higher
finesse will push the optimum position of the focal line 3 further
behind the etalon 4 because the total path through the etalon 4 is
increased.
[0034] Typically, etalons are configured to receive a collimated
beam of light and transmit a single frequency or a series of
frequencies which are spaced apart by the etalon's free spectral
range (FSR). However, because the beam of light is converging or
diverging, e.g. in only one direction, due to the cylindrical lens
2, multiple incoming angles of light results in a series of
sub-beams with different frequencies being transmitted, since each
angle corresponds to a different distance across the etalon 4
Accordingly, each sub-beam is transmitted at a different angle from
the normal, and each sub-beam may include multiple frequencies
based on the FSR of the etalon 4. When multiple frequencies are
incident on the etalon 4, there is uncertainty in frequency of
nFSR. To resolve the nFSR uncertainty, a secondary dispersive
element 9 may be added before or after the etalon 4. Before the
etalon 4 is possible, but after is preferred because the area of
the etalon 4 is minimized. If the dispersive element 9 is first,
then the line focus becomes longer, and the etalon 4 becomes more
difficult to make. The secondary element 9 may be a dispersive
prism, a diffraction grating or a much thinner etalon which may be
able to resolve frequency differences at least as small as the FSR
of the etalon 4. The combination of the etalon 4 and the secondary
dispersive element 9 in principle resolves the spectrum of the
incoming beam of light with a resolution limited by the etalon
resolution, and a total bandwidth, i.e. frequency range, limited by
the performance range of the etalon's reflective surfaces 6 and 7,
and any other optical elements in the device, e.g. 10-500 times the
FSR range of the etalon 4. Practically, the secondary dispersive
element 9 may resolve at least 2-3 times smaller differences than
the FSR, ideally 4-6 times. More than 4-6 times tends to limit the
overall frequency range, and less than 4-6 times tends to produce
cross talk.
[0035] Ideally, the angle between the converging extreme rays after
the cylindrical lens 2, e.g. should be less than the tilt of the
etalon 4, e.g. whereby substantially all of the light (or at least
almost or substantially all) bounces, i.e. reflects, in the same
direction, e.g. upward, inside the etalon 4. Optimizing the tilt of
the etalon 4 in the invention is a balance between the area of the
etalon 4, the number of FSRs illuminated, and the linearity of the
frequency/angle response. The angular separation between orders
decreases as the angle of incidence at the etalon 4 increases.
Accordingly, if the etalon 4 is not tilted at all fewer different
FSRs are illuminated. However, as the etalon 4 is tilted more,
newly illuminated FSRs are added more quickly than the ones near
normal incidence go dark. An excessively large tilt will increase
the required etalon length (1.apprxeq.2t*tan(.theta.)*finesse), and
is also likely to produce larger phase and reflectivity differences
between s- and p-polarizations incident on the etalon 4. Too small
an etalon tilt will decrease the number of etalon FSRs that can be
illuminated for a given angle range which makes analysis of the
resulting spectrum more sensitive to alignment details.
Accordingly, the first etalon 4 may be tilted at an angle such that
the range of angles from the converging beam of light from the
cylindrical lens 2 is incident on the first etalon 4 from one half
of a full angle divergence to 1.5.times. the full angle
divergence.
[0036] In most cases, a diffraction grating is preferred as the
secondary dispersive element 9 because of its higher dispersion,
but a prism could be used. Because the light leaving the etalon 4
is collimated in one axis (to the extent that is possible based on
the extent of the illumination source), there is no need for a slit
between the etalon 4 and the secondary dispersive element 9. The
light leaving the secondary dispersive element 9 is thus encoded,
with each of the angles (.theta..sub.x, .theta..sub.y) in the x and
y planes corresponding to a unique light frequency. Typically, each
input frequency will be represented at multiple angle coordinates
corresponding to multiple orders of the etalon 4. The thickness of
the etalon 4 will correspond to some multiple of the transmitted
wavelength. The "order" of the transmitted light at wavelength
.lamda. is N=2nt cos(.theta.)/.lamda., where n is the refractive
index, t is the thickness of the etalon 4, and .theta. is the angle
of incidence. Accordingly, each wavelength will be transmitted at a
series of increasing angles corresponding to decreasing orders for
the etalon 4. A second, e.g. spherical, lens 11 may be used to
focus the dispersed light onto an array of photodetectors, e.g.
camera 12, e.g. a CCD, a CMOS, film, or other sensor array, and the
resulting pattern is evaluated to determine the spectral content of
the incident light, i.e. provide a measure of each of the
individual frequencies.
[0037] The version illustrated in FIGS. 2, 3 and 4 may generate
significant loss, e.g. if the finesse of the etalon 4 is 50, then
only 1/50th of the incident light will be transmitted, if the range
of angles from the cylindrical lens 2 corresponds to multiple FSRs
of the etalon 4. However, one key attribute of an etalon is the
discrimination between frequencies and angles. For example, for
frequency f1, and frequency f2, there may be different ratios of
the transmitted power of the two frequencies at the angle
corresponding to the peak transmission angle of frequency f1. These
ratios can be increased by raising the finesse of the etalon 4 to
higher and higher values. Practically, there are limits to the
finesse for an etalon with parallel planar reflective coatings,
e.g. 6 and 7. Depending on the working frequency range, the maximum
practical finesse is usually in the range of 50 to 300 for a planar
etalon. Etalons with curved mirrors may go much higher, but the
cylindrical focus is not appropriate then.
[0038] With reference to FIGS. 5 and 6, a dual-etalon spectrometer
21 has the benefit of increasing the contrast of the spectrometer
21 for many incident frequencies simultaneously, while maintaining
comparable, e.g. substantially the same, throughput for all of the
frequencies. If a second etalon 24 with identical optical thickness
is placed in series with the first etalon 4, the transmission
function of the two etalons 4 and 24 may be identical as a function
of the transmission frequencies versus angles of incidence.
Accordingly, when the first etalon 4 is placed in a converging or
diverging beam, all of the incident frequencies are transmitted at
specific angles, as hereinbefore described with reference to FIGS.
2 and 3. For a conventional etalon with equal reflectivity mirrors
on both faces of the etalon, approximately 1/finesse of the
incident power will be transmitted assuming a uniform distribution
of angles for each incident frequency. When the light transmitted
through the first etalon 4 is incident on the second etalon 24,
each frequency is already sorted to the angles of incidence
required for maximum transmission, and the power loss is only
approximately 56% (corresponding to the increase in finesse), while
the increase in contrast is approximately as the square for all
frequencies, not just a single discrete frequency. Accordingly,
each sub-beam is transmitted by both the first etalon 4 and the
second etalon 24 at the same angle. If two different frequencies
are incident on the etalon at the same angle, the contrast is the
ratio between their transmitted intensities.
[0039] Another benefit of the dual etalon spectrometer is improved
control of scattered light. Scattered light is typically not
collimated, and instead of being transmitted in a single direction,
will be transmitted in a ring pattern at a constant angle. The
shape of each ring pattern is not affected by the cross disperser,
because it is a single frequency. The ring patterns from
frequencies separated by an etalon FSR or more will then tend to
overlap. If the second etalon 24 is perfectly parallel to the first
etalon 4, the scattered light will not be attenuated by the second
etalon 24. However, if the second etalon 24 is tilted relative to
the first etalon 4 in the horizontal plane (FIG. 6), the light in
the ring pattern will be strongly attenuated because the effective
thickness of the two etalons 4 and 24 will only be the same for
light which has zero horizontal angle.
[0040] If the first and second etalons 4 and 24 are not identical
in thickness, e.g. thickness1 is within 50 nm of thickness2, the
thicker etalon may be tilted in the horizontal plane (am), e.g.
0.1.degree.-10.degree., preferably 0.1.degree.-2.degree., so that
the thicknesses are equivalent. As an etalon is tilted, it's
apparent thickness decreases (t*cos(theta)). The physical distance
travelled through the etalon increases as the tilt is increased,
but the change in phase actually decreases when you evaluate the
phase in a plane perpendicular to an input ray.
[0041] So long as the thickness compensating tilt is purely
horizontal there is no theoretical limit to the thickness
difference which can be compensated. Both the first and second
etalons 4 and 24 may be tilted in opposite vertical directions
(.alpha..sub.v1 and .alpha..sub.v2), e.g. 0.1.degree.-2.degree., to
help prevent recycling or back reflection of the light
therebetween. Practically, if the thickness difference between the
first and second etalons 4 and 24 is large, e.g. greater than
approximately 1%, the performance of the optical coatings forming
the reflective surfaces 6, 7, 26 and 27 may start to introduce
differences between the transmission functions of the first and
second etalons 4 and 24. The condition for equivalence of the first
and second etalons 4 and 24 is:
cos(.alpha..sub.h1)*thickness1=cos(.alpha..sub.h2)*thickness2.
[0042] The first and second etalons 4 and 24 may also have a near
integer, e.g. a multiple N=2.times.-10.times., thickness ratio N,
i.e. the first thickness of the first etalon 4 or the second
thickness of the second etalon 24 may be an integer multiple N of
the other. This configuration has poorer performance for the orders
which are not transmitted by both the first and second etalons 4
and 24, but will still reduce cross talk and improve contrast.
[0043] Alternately, the tilt of the first and second etalons 4 and
24 may be adjusted so that
cos(.alpha..sub.h1)*thickness1=N*cos(au)*thickness2. Normally, this
configuration would only be used if the ratio of the etalon
thicknesses, thickness1/thickness2, is approximately equal, e.g.
within 1%, to N. In this configuration, the angular spacing between
transmitted orders of a given light frequency will be larger for
the thinner etalon than for the thicker etalon, but the orders that
are transmitted by the thinner etalon will still perfectly overlap
with transmitted angles for the thicker etalon.
[0044] The second etalon 24 includes first and second reflective
surfaces, e.g. coatings, 26 and 27, separated by a spacer, e.g.
glass block, 28, ideally with a width identical, e.g. less than 50
nm difference, to the first glass block 8. The first and second
reflective surfaces 26 and 27 may have a reflectivity of between
90% and 96% over a desired wavelength range. The first and second
reflective surfaces 26 and 27 of the second etalon 24 may be
identical, e.g. +/-1%, to the first and second reflective surfaces
6 and 7, to facilitate manufacture, but other reflectivity is
possible. The reflectivity of the first and second reflective
surfaces 26 and 27 may have a reflectivity of 80%-90% or even
70%-80%. In all cases the first and second reflectivities 26 and 27
may be as similar as possible to each other. The reflectivity
chosen will depend on a balance between the contrast, transmission
and cross talk of the complete spectrometer. Lower reflectivity
mirrors will reduce the maximum theoretical contrast, but because
of imperfections in practical etalons, optimum performance may be
reached with these lower reflectivity values. The focal line 23 of
the cylindrical lens 2 may be in between the reflective surfaces 6
and 7, e.g. in the middle of the spacer 8; however, focusing the
incoming beam of light along a line in between the cylindrical lens
2 and the first etalon 4 or on the far side of the first etalon 4,
e.g. in between the first etalon 4 and the second etalon 24, is
within the scope of the invention.
[0045] Practically, the wavelength range of the coatings of the
reflective surfaces 6, 7, 26 and 27 may be a limiting factor. For
instances where the wavelength range of the spectrometer 1 or 21 is
small, e.g. less than 25% of an octave, the coating performance
won't likely be an issue. For example: if the camera 12 has 2000
horizontal pixels, and the spacing between stripes is 5 pixels,
there would be a total range of 400 FSRs. If 400 FSRs is less than
25% of an octave, the coatings 6, 7, 26 and 27 should not be an
issue. For overall ranges greater than 25% of an octave, coating
performance becomes significantly more challenging. This is
actually an argument in favour of using two or more etalons 4 and
24 because lower finesse etalons may still yield overall high
resolution, and increasing the bandwidth of the coatings is easier
if they have lower reflectivity.
[0046] FIG. 7, in contrast to FIG. 1, illustrates a plot of
transmission vs angle for two substantially identical etalon
thickness, e.g. spacers 8 and 28, 2.0 mm wide and within 10 nm of
each other, illustrating an overlap, i.e. resultant transmission,
of a plurality of frequency peaks.
[0047] With two or more etalons, e.g. first and second etalons 4
and 24, it may be necessary to avoid interference between the
reflections from the faces, e.g. reflective surfaces 7 and 26, of
the individual etalons 4 and 24, or simply to ensure that multiple
reflections between the etalons are not collected on the
spectrometer sensor array. This effect can be prevented by tilting
one or both of the first and second etalons 4 and 24 relative to
each other by a small amount, e.g. <15.degree.. This may be
accomplished by tilting the two etalons 4 and 24 by the same amount
in one plane, for example by .alpha..sub.v1 and
.alpha..sub.v2=2.degree.-10.degree. in the, e.g. vertical,
direction (FIG. 5) that the input beam is converging or diverging,
and/or by a 5-20 times smaller amount, e.g. by .alpha..sub.h1 and
.alpha..sub.h2.+-.0.1.degree.-2.degree., in the other,
perpendicular plane, e.g. horizontal plane (FIG. 6).
[0048] Alternatively, the horizontal tilt an (top view) may be on
the same order as the vertical tilt a, but then the relative tilt
of the first and second etalons 4 and 24 must be much better
matched to ensure accurate matching of the transmission peaks. If
the first and second etalons 4 and 24 are not identical in
thickness, e.g. but still less than 50 nm, then some degree of
compensation for the mismatch is possible by varying the relative
tilt slightly, e.g. .alpha..sub.h1.noteq..alpha..sub.h2, to ensure
the optical path lengths through the first and second etalons 4 and
24 are the same.
[0049] Alternatively, or in addition, the first and second etalons
4 and 24 may be positioned far enough apart, whereby light
reflected back from the first etalon 4 is not incident on the
second etalon 24.
[0050] With reference to FIGS. 8A to 8E, a method of manufacturing
identical first and second etalons 4 and 24 may start with forming
a block of material 31, e.g. glass, optically transmissive at the
desired wavelength range. Ideally, the block 31 includes a length
that is equal to or longer than the combined lengths of the first
and second etalons 4 and 24, with parallel sides perpendicular to
the length, e.g. a thickness within 50 nm, preferably less than 25
nm, and more preferably less than 10 nm, along the entire length
thereof. The second step may include coating one of the parallel
sides of the block 31 with a reflective coating 32 with the desired
reflectively of the first reflective surfaces 6 and 26, e.g.
greater than 70%, typically between 70% and 99.5%, preferably
greater than 90%, and more preferably between 90% to 97% for all of
the wavelengths in a desired wavelength range. Next, the opposite
side of the block 31 may be coated with a reflective coating 33
with the desired reflectively of the second reflective surfaces 7
and 27, e.g. greater than 70%, typically between 70% and 99.5%,
preferably greater than 90%, and more preferably between 90% to 97%
for all of the wavelengths in a desired wavelength range. Finally,
the coated block 31 is cut in half into two sections forming the
spacers 8 and 28, and the first and second etalons 4 and 24.
[0051] To simplify alignment, the first and second etalons 4 and 24
may then be assembled as a monolithic block via optical contacting
or other means, e.g. adhesive, with a frame 35 for fixing both of
the first and second etalons 4 and 24 in a relative position. The
frame 35 may include a first vertical tilt angle enabling both the
first etalon 4 and the second etalon 24 to be tilted at
substantially the same vertical angle .alpha..sub.v1 and
.alpha..sub.v2, as in FIG. 5, to limit back reflection from the
second etalon 24 onto the first etalon 4. One or more spacer wedges
36, including a small wedge angle .alpha..sub.w, e.g. 0.1.degree.
to 2.degree. in the horizontal plane may also be included with the
frame 35 between the first and second etalons 4 and 24 to enable
adjustment of the overall horizontal and vertical tilt angles
.alpha..sub.h1, .alpha..sub.h2, to effectively tune the thickness
of the second etalon 24. The entire two etalon assembly may then be
rotated or laterally translated so that the one or more spacer
wedges 36 may be aligned at the desired horizontal and vertical
tilt angles relative to the axis OA of the cylinder lens 2, as in
FIGS. 5 and 6, and then tilted by up to several degrees depending
on the thicknesses of the first and second etalons 4 and 24 to
ensure that each incident frequency is transmitted via at least two
different angles. With such monolithic configuration, a slight
thickness mismatch between the first and second etalons 4 and 24,
e.g. a few nm, may be compensated during alignment by rotating the
entire assembly effectively changing the angle of the spacer wedge
36 relative to the axis of the cylindrical lens 2 such as to make
am au. Three or more identical etalons may be placed in series in a
similar manner with a similar improvement in contrast performance,
albeit with more complicated alignment.
[0052] The lens 2 may then be positioned so that the focal line 23
is focused along the desired location, e.g. between the first and
second etalons 4 and 24, as in FIG. 5. For spectrometer
applications, the secondary dispersive element 9, the second lens
11 and the camera 12 are positioned in suitable locations to
capture the transmitted light, see FIG. 5.
[0053] With reference to FIGS. 9 and 10, a spectrometer 71 includes
a virtually imaged phase array (VIPA) 74, such as the ones
disclosed in U.S. Pat. No. 10,162,186 issued Dec. 25, 2018 in the
name of Hunter et al, which is incorporated herein by reference, as
the first etalon. The VIPA 74 is a special type of etalon, also
with parallel, partially reflective surfaces, but with a first
highly reflective surface 76, e.g. between 99% and 100% reflective
at desired frequencies, and a second partially reflective surface
77 having less reflectivity, e.g. 90-98%, preferably 95%-98%
reflective on opposite parallel sides of a transparent spacer 78,
as hereinbefore discussed. The first reflective surface 76 includes
an input window 75, which may include an antireflection (AR)
coating, for receiving the light along a focal line 73. Light is
typically focused into the VIPA 74 with the cylindrical lens 2
which results in high divergence in one direction, e.g. across the
width of the VIPA 74, and low divergence in the other direction,
e.g. along the focal line 73 and into the input window 75. Unlike
the two etalon case above, the VIPA 74 should be positioned near
the focus, i.e. focal line 73, of the cylindrical lens 2 in order
to efficiently couple light into the VIPA 74, because the first
reflection of the incoming light from the partial reflective
surface 77 of the VIPA 74 must be offset from the AR coated input
window 75 sufficiently to reflect from the high reflective surface
76 of the VIPA 74. For optimal coupling, the focal line 73 should
be close to the partially reflecting surface 77 of the VIPA 74. The
light exiting the VIPA 74 is dispersed in the high divergence
direction with each frequency at a different angle, but as with the
etalon 4, multiple frequencies separated by the FSR appear at the
same angle. As with the aforementioned embodiments, the cross
dispersive element 9 is required to enable identification of each
incident wavelength. When the first etalon is the VIPA 74 there is
much more transmitted power, typically an improvement compared to a
symmetric etalon on the order of the finesse. However, this
increase in transmitted power is only available if the input light
can be focused into a focal line 73 that is narrow (<100
.mu.m).
[0054] In a preferred embodiment in which the VIPA 74 is followed
by the conventional second etalon 24, the thickness of the VIPA 74
is identical to the thickness of the second etalon 24; however, the
increase in throughput is so large that there are advantages to
their combination even if the thicknesses of the second etalon 24
and the VIPA 74 are not identical. As the thickness difference
between the etalons 24 and 74 increases, the range of angles or the
range of frequencies where there is a good match between the two
etalons decreases, i.e. smaller relative shift than the full width
at half maximum (FWHM) of the etalons. The horizontal tilt of the
second etalon 24 may be adjusted to match the vertical angular
transmission function to that of the VIPA 74. In the case where a
small range of frequencies is of interest, e.g. corresponding to
about 1 FSR, matching over a single FSR could be sufficient.
Moreover, there may be applications for the combination of the VIPA
74 and the second etalon 24 other than a spectrometer, that are
within the scope of the invention.
[0055] In the case of the VIPA 74 followed by the conventional
second etalon 24, the finesse of the second etalon 24 need not be
as high as the finesse of the first etalon 4 or VIPA 74 to still
have strong contrast and signal to noise benefits. A lower finesse
for the second etalon 24 will result in higher overall light
throughput, but lower contrast improvement.
[0056] There may be practical limits to the increase in contrast
achieved by adding a second etalon 24. For example, if the first
etalon 4 or VIPA 74 has a finesse of 70, adding a second etalon 24
of a finesse 70 may not provide significant contrast improvement
compared to adding a second etalon 24 with a finesse of 40-50. In
general, the choice of the finesse of the second etalon 24 will be
a balance between increasing resolution, higher contrast, and
reduced peak transmission. The reduced peak transmission is an
inevitable effect of higher resolution, but in many cases, the
primary purpose of the second etalon 24 is increased contrast not
resolution. Accordingly, there will be a range of finesse values
which may be chosen for the second etalon 24, e.g. in the range of
20-60, compared to the finesse of the first etalon 4 or 74, e.g.
50-100. In some cases a finesse as low as 10 for the second etalon
24 could still improve contrast, but in most cases it would be
better to lower the finesse of the first etalon 4 or 74 and raise
the finesse of the second etalon 24 rather than have such a big
ratio between the two finesses. Accordingly, the ratio of the
finesse of the first etalon or VIPA 74 to the finesse of the second
etalon 24 is preferably in the range of 1:1 to 4:1, ideally 1.5:1
to 3:1.
[0057] With reference to FIGS. 11A to 11E, a method of
manufacturing identical first and second etalons 74 and 24 may
start with forming a block of material 91, e.g. glass, optically
transmissive at the desired wavelength range. Ideally, the block 91
includes a length that is equal to or longer than the combined
lengths of the first and second etalons 74 and 24, with parallel
sides perpendicular to the length, e.g. a thickness within 50 nm,
preferably less than 25 nm, and more preferably less than 10 nm,
along the entire length thereof. The second step may include
coating one of the parallel sides of the block 91 with a reflective
coating 92 with the desired reflectively of the second reflective
surfaces 27 and 77, e.g. between 90% and 98%, preferably 94% to 96%
for all of the wavelengths in a desired wavelength range. Next, the
opposite side of the block 91 may be coated with a reflective
coating 93 with the desired reflectively of the first reflective
surface 26, e.g. between 90% and 98%, preferably 94% to 96% for all
of the wavelengths in a desired wavelength range. Then the bottom
half of one side of the block 91 is subjected to further coating
treatments to raise the reflectivity up to the desired
reflectivity, e.g. between 99% and 100% for the first highly
reflective surface 76 of the VIPA 74 Finally, the coated block 91
may be cut in half into two sections forming the spacers 28 and 78,
and the first and second etalons 74 and 24. The input window 75 for
the VIPA 74 may then be fabricated, as in any one of the
embodiments disclosed in U.S. Pat. No. 10,162,186.
[0058] To simplify alignment, the first and second etalons 74
(VIPA) and 24 may then be assembled as a monolithic block via
optical contacting or other means, e.g. adhesive, with a frame 95
for fixing both of the first and second etalons 74 (VIPA) and 24 in
a relative position. The frame 95 may include a first vertical tilt
angle enabling both the VIPA 74 and the second etalon 24 to be
tilted at substantially the same vertical angle .alpha..sub.v1 and
.alpha..sub.v2, as in FIG. 9, to limit back reflection from the
second etalon 24 onto the first etalon 74. One or more spacer
wedges 96, including a small wedge angle .alpha..sub.w, e.g.
0.1.degree. to 2.degree. in the horizontal plane may also be
included with the frame 35 between the first and second etalons 4
and 24 to enable adjustment of the overall horizontal and vertical
tilt angles .alpha..sub.h1, .alpha..sub.h2, as in FIG. 10, to
effectively tune the thickness of the second etalon 24. The entire
two etalon assembly is then rotated so that one or more spacer
wedges 96 are aligned parallel to the axis of the cylinder lens 2,
as in FIG. 9, and then tilted by several degrees depending on the
thicknesses of the first and second etalons 74 and 24 to ensure
that each incident frequency is transmitted via at least two
different angles. With such monolithic configuration, a slight
thickness mismatch between the first and second etalons 74 and 24,
e.g. a few nm, may be compensated during alignment by rotating the
entire assembly, effectively changing the angle of the spacer wedge
96 relative to the axis of the cylindrical lens 2, such as to make
.alpha..sub.h1.noteq..alpha..sub.h2.
[0059] The vertical tilt angles .alpha..sub.v1 and .alpha..sub.v2
in the case of the VIPA 74 and the etalon 24 are essential to
coupling the light into the VIPA 74. In the case of the first and
second etalons 4 and 24, the vertical tilt angles .alpha..sub.v1
and .alpha..sub.v2 are selected more to simplify the analysis of
the spot pattern on the camera 12 than for back reflection. For a
given range of input angles determined by the cylinder lens 11,
increasing the vertical tilt angles .alpha..sub.v1 and
.alpha..sub.v2 means that more interference orders are illuminated,
but so long as sufficient orders are illuminated, e.g. 3 or 4, the
exact tilt angles .alpha..sub.v1 and .alpha..sub.v2 are not
critical. The important part of the vertical tilt angles
.alpha..sub.v1 and .alpha..sub.v2 is that they are the same for
both etalons 4 and 24, so that the wavelength/angle mapping is the
same. Accordingly, the frames 35 and 95, and the spacer wedges 36
and 96 typically just need to hold the two etalon components
parallel. The assembly, e.g. the block 31 or 91, of the first
etalon 4 or VIPA 74 and second etalon 24 could include an
approximate vertical tilt angle, but it's not necessary, and the
overall vertical tilt angles .alpha..sub.v1 and .alpha..sub.v2
could be provided by the mounting of the assembly on the frame 35
or 95. As an example, for 1 pm resolution spectrometers, the
grating 9 with the highest dispersion possible is used, which
dictates the stripe spacing. Then the focal length of the imaging
lens 11 is selected based on the pixel size and the target of -7-15
pixels stripe spacing. Accordingly, the first etalon 4 is tilted,
such as to get the desired number of orders on the camera 12, e.g.
3-5.
[0060] Three or more identical etalons, one of which may be a VIPA,
may be placed in series in a similar manner with a similar
improvement in contrast performance, albeit with more complicated
alignment.
[0061] The lens 2 may then be positioned so that the focal line 73
is focused along the input window 75, as in FIG. 9. For
spectrometer applications, the secondary dispersive element 9, the
second lens 11 and the camera 12 are positioned in suitable
locations to capture the transmitted light.
[0062] The foregoing description of one or more embodiments of the
invention has been presented for the purposes of illustration and
description. It is not intended to be exhaustive or to limit the
invention to the precise form disclosed. Many modifications and
variations are possible in light of the above teaching. It is
intended that the scope of the invention be limited not by this
detailed description, but rather by the claims appended hereto.
* * * * *