See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/253342601

Design and analysis of a compact wide-field unobscured zoom mirror system

[3129-09]

Article in Proceedings of SPIE - The International Society for Optical Engineering · September 1997
DOI: 10.1117/12.279094

CITATIONS READS

5 202

2 authors, including:

R. Barry Johnson
Alabama Agricultural and Mechanical University
128 PUBLICATIONS 1,031 CITATIONS

SEE PROFILE

All content following this page was uploaded by R. Barry Johnson on 25 October 2015.

The user has requested enhancement of the downloaded file.

 DESIGN AND ANALYSIS OF A COMPACT, WIDE FIELD,
UNOBSCUIRED ZOOM MIRROR SYSTEM
Allen Mann and R. Barry Johnson
Optical E.T.C., Inc.
3077-K Leeman Ferry Road
Huntsville, AL 35801

ABSTRACT

The evolution ofan unobscured all-reflective zoom optical system for utilization in the infrared
spectrum Is presented. The objective is to develop a system which has a flat Image surface, wide field-
of-view, spatially remote entrance pupil, 2:1 zoom range, and Is spatially compact. The optical
system comprises three aspheric mirrors sharing a common optical axis where the primary mirror Is
spatially fixed with respect to the entrance pupil, and secondary and tertiary mirrors move during
zm. The field-of-view ranges from 2.2 degs by 2.2 degs to 4.4 degs by 4.4 degs. The focal ratio
varies from F/4 to F/S. The Inherent characteristics ofthIs type optical system are discussed, as are
design methods to control aberrations, distortion and anamorphic error over the zoom range. The
baseline design configuration Is presented along with MTF performance data. The results of
tolerance sensitivity analysis are also discussed.

KEYWORDS: Zoom optics, collimator, reflective optical system, lens design, telescope.

1. INFRODUCTION

All-reflective telescopes have found significant utilization In the astronomical community,

governmenl and Industry. Until recenfly, the concept of an all-reflective zoom telescope, parficularly
having an unobscured pupil, with meaningful capabilifies was relatively unexplored1'2'3'4'5. A wide
field-of-view, unobscured, all-reflective zoom system was specified for inclusion In an Infrared test set
to project a large-am thermal pixel array. The optical system Is required to have a remote exit pupil
to match the entrance pupil ofihe unit under test The evolution ofan appropriate three-mirror,
unobscured, all-reflective zoom optical system for this purpose and utilization in the infrared
spectrum Is presented in the following sections. The system Is based upon earlier investigations
with the objective of developing a system which has a fiat Image surface, wide field-of-view, spatially
remote entrance pupil, 2:1 zeem range, and Is spatially compact

The zoom telescope configurations discussed in this paper are based upon the previously reported
low-F/#, wide field-of-view (FOV), unobscured, eccentric pupil, off-center field, flat-field three-mirror
telescopes6. The general configuration of these telescopes features a remote pupil (aperture stop)
located in front of the telescope by a considerable distance7. By moving the secondary and tertiary
mirrors with respect to the spatially-fixed pupil and primary mirror, the telescope can be zoomed
over a relatively wide range of focal lengths while maintaining good Image quality. Due to the limited
number of optical elements, these configurations are all of the uncompensated type, I.e., the Image
moves with zoom.

All three mirrors share a common optical axis although their mechanical axes are displaced from one
another. This forms a plane-symmetric optical system. Since rotational symmetry is absent, even-
order ray aberrations appear along with the odd-order aberrations common to rotationally-symmetric
systems. In general, it seems that the center of the field-of-view should be located approxImately 0.7
FOV units below the optical axis. In other words, the nominal line of sight Is not parallel to the
optical axis of the telescope8. The advantage of the off-center FOY over the centered FOV for this

SPIE Vol. 3129. 0277-786X/971$10.OO 97

Downloaded from SPIE Digital Library on 24 Oct 2009 to 76.29.229.28. Terms of Use: http://spiedl.org/terms
 system is that residual spherical aberration and other aberration residuals aboutihe optical axis are
used to achieve balance with the astigmatic and coiuatlc aberrations present farther offaxis. This
approach allows realizadon of good resolution over relatively large FOVs. Images formed by systems
having highly eccentric pupils have unusual characteristics9. Since this type ofzoom telescope
utilizes an offcentered FOV, the active field-of-view or Image area lies entirely below the optical axis.
This Implies thatthe angular FOV has only positive slope angles with respectto the pupil axis. As a
consequence ofthe zoom telescope being a plane-symmefric configuration, the Image Is slightly
anamorphic Le., the azimuth and elevation focal lengths are notidenticaL This anamorphic error is a
type of distortion ofihe Image. DistorIiou° patterns In the image look a bit unusual until one
recos that the Image is not centered on the optical axis and that distortion is influenced by all of
the Seldel aberrations (unlike centered-pupil systems with rotational symmetry).

Figure 1 Illustrates the Interesting variety ofimage distortion patterns one may observe when
considering an off.centered or eccentric pupil optical system. The leftmost column shows the
patterns typical ofdassical third-order distortion. The following three columns showthe patterns for
image distortion when only third-order coma, astigmatism, and Petzval are presen1 respectively. The
rows are for positive, negative and zero values ofthe respective aberrations. The anamorphic
distortion due to coma and astigmatism is rather evident Spherical aberration causes a constant
error contribution to the overall distortion. This error manifests itselfas a shift ofthe entire Image
without changing its shape Spherical aberration In centered optical systems appears as a rotaflonally
symmetric image error. For an off-centered pupil optical system, spherical aberration causes image
errors that are rather uncommon to most designers and look something like a mixture of astigmatism
and coma. Figure 2 illustrates such Images for an eccentric pupil optical system having an effective
focal length (efi) of 10 Inches and operating at F/5 (eli divided by pupil diameter). The center of the
pupil Is offset by 2, 3, and 4 Inches, and the effect of defocus Is also shown.

For an off-centered pupil optical system, coma aberration causes Image errors that are Illustrated in
Flg 3. The plane.symmetrlc nature ofihe system Is clearly observed in the comatic Images. For
reference, the comatic Image for the same optical system having the pupil centered Is shown In the
lower right corner ofthe figure Another Interesting characteristic ofthls type ofthree-mlrror zoom
optical system Is thatthe Image area moves parallel to 11w optical axis and also orthogonal to It since
the effective focal length Is changing while the center ofthe FOV Is at a constant angular OffsettO the
optical axis. Due to the nature ofthe optical design, the Image area can be maintained as a flat
surface. Furthermore, such all-reflective zoom systems having an unobscured pupil and an offset
field-of-view have certain attractive features In terms oflight energy throughput and compactness.
The angular offset ofthe FOV.is determined by the bias angle. The three mirrors comprising the
optical system share a common optical axis. The mirrors are conic surfaces with higher order
aspheric coefficients. In a zoom configuration, the secondary mirror, tertiary mirror, and the Image
plane are movable during zoom; however, the remote pupil remains spatially stationary with the
primary mirror.

2. TECBNICAL OBJECTIVES

The basic requirements for the unobscured all-reflective zoom optical system are a 2 to I zoom range
to operate continuously from an eli of 600 mm to 1200 mm, a fixed entrance pupil diameter of 150
mm; and a FOV which varies from 4.4 degrees to 2.2 degrees11. The resultant f-number varies from
Ff4 to F/S. At the 600 mm eli position, the system matches the operating parameters of an existing
fixed-focus collimator having a configuration similar to that described in 161. The thermal pixel array
to be used with this optical system has a pixel pitch of 175 p.m. The optical performance should be
near diffraction-limited over the zoom range and across the FOV when used in the 3-5 p.m or 8-12 p
mspectral regions. Anamorphic error and distortion should be less than 3-5 percent The physical
envelope enclosing the optical system and Its mechanical structure Is desired to be less than 610 mm

98

Downloaded from SPIE Digital Library on 24 Oct 2009 to 76.29.229.28. Terms of Use: http://spiedl.org/terms
by 305 mm. The distance from the remote pupil to the closest surface ofthe optical system or the
Image plane should be at least 350 mm.

3. EVOLUTION OF OPTICAL DESIGN

One ofthe most Important and often perplexing questions an optical designer has to ask himself In
underlaking a new design activity is "where do you start?" In most Instances, the most desirable
starting pointis an existing design that comes close to meeting some ofthe requirements and which
holds the promise of meeting all ofthe technical objectives.

One such starting pointis the existing fixed-focus unobscured all-reflective collimator which has a
600 mm effective focal length and operates at Ff4 over a 4.4 degrees (square) FOV. This is a very
compact system with diffraction-limited resolution over the entire FOV. However, using this system
as a starting point In an attempt to modify the system Into a zoom collimator proved unsuccessful
prhnarlly because the Image quality gotworse the further the mirrors deviated during zoom from the
Initial fixed position.

The next attmptto find a suitable slarting polntwas based upon an examination ofexistlng designs
ofrelevaut uuobscured reflective zoom systems such as those discussed In [3J. Figures 4-7 depIct
four candidate configurations where the mirrors are shown In several positions coveting the zoom
range. The tertiary mirror and Image plane locations for the wide and narrow FOV zoom positions
are denoted by "Wide" and "narrow." All ofthese designs have a flat Image surface. Table 1
summarizes the characteristics ofthese four designs. All ofthe entrance pupil diameters were scaled
to 100 mm to assist In comparing these designs. The angular resolutions listed are based upon the
RMS spot size where the first number corresponds to the wide FOV position and the second number
corresponds to the narrow FOV position.

Design FOV F/# Zoom Resolution Bias Overall

Range Range Ratio (p&radians) Angle Length
(deg.) (deg.) (mm)
1 3-1.5 3-6 2:1 59/30 -5 1092

2 2-0.5 6.5-2.5 4:1 325/83 -5 559

3 2-0.5 4-16 4:1 85/40 -5 534

4 2-0.5 4-16 4:1 45/12 -2 711

Table 1. CharacterIstics of DesIgns 1-4

It should be observed that the definition of effective focal length for this type of optical system Is not
adequately expressed by the standard Gaussian definItion. For this situation, the eli is given by efi =
AX/A0 and efi = AY/AOy where V Is In the plane of symmetry, X Is In the orthogonal direction, and
0 Is the field angle.

It is evident from an examination of Table 1 that none of these designs meets all of the requirements.
Specifically, none of them covers the 4.4 degrees square FOV. Although Design I comes closest to the
required FOV, it is much too long for serious consideration. Design 2 has the worst resolution and
has the largest F/#. Design 4 is also excessively long. Design 3 was selected as the starting point
because it appeared to be the most likely to meet all of the requirements, even though it covers only a
2 degree FOV. In the Initial stages of the design, the portion of the zoom range from 450 to 900 mm
focal length operating over a 4 degree FOV was selected for optimization, still using the InItial 100

99

Downloaded from SPIE Digital Library on 24 Oct 2009 to 76.29.229.28. Terms of Use: http://spiedl.org/terms
 mill pupil diameter. When reasonable Imagequality was obtained for this setof parameters, this
preliminary design was scaled up to 600 to 1200 mm focal length range. The scaling increased the
pupil diameter to 133.3 mm. At this point the pupil diameter was increased to the required 150 mm.
This change degraded the Image quality which resulted In the addition ofhigher-order aspheric
coefficients on the primary and tertiary mirrors In order to achieve acceptable performance levels.
The reason for using higher-order aspheric coefficients to correct Image quality Is thatwhen a mirror
segmeutis used slgnlflcanlly off-axis, low-order surface deformation coefficients begin to have little
effect upon surface figure. In a like manner, high-order aspheric coefficients have essentially no
effect on surface figure In the region near the optical axis. Consequently, high-order aspheric
coefficients can he used to controithe surface deformation when the mirror segmeutis used
significantly off-axis. A consideration In the design process was to determine the optimum bias angle
This parameter was Included In the design variables, butwas constrained to produce systems that
conformed to the physical envelope requirement

Once the basic system requirements were achieved, the anamorphic error and distortion became the
next Image defects to be addressed. Anamorphic error can be defined as the difference In image
length In the Y direction as compared with the Image length In the X direction: In other words, the
difference In efl Ofir Distortion Is defined In this case as the difference In Image length In the +Y
direction as compared with the -Y direction. These values were approximately 10% at this stage In
the design process, which was unacceptably high. Operands were then Included In the merit function
to mfnimlie these values during subsequent optimization12. Anamorphic error varying from 3.3%
to 4.4% over the zoom range was achieved without adversely affecting hnage quality. The
corresponding distortion varied from 3.5% to 18%.

Up to this point In the design process, performance was monitored primarily through the analysis of
rim ray plots and the calculation ofRMS spot size. Improvement at specific zoom posItions and field
points was accomplished through the use ofRMS wave error operands. Next MTF operands at one
wavelength were used during the final optimization and MTF was then used to assess image quality.
Although three zoom positions were used for design, It was determined that In later stages of the
design activIty five zoom positions were needed In order to avoid peculiar behavior between design
zoom positions.

4. BASELINE DESIGN CONFIGURATION

The final design configuration Is shown In Figs. 8-10 at the short, intermediate, and long focal length
positions. The pupil decenter Is 156.4 mm, and the bias angle Is -3.3 degrees. The overall length, as
determined by the maximum separation ofthe secondary and tertiary mirrors, Is 642 mm while the
minimum distance between the Image plane and the tertiary mirror Is 573 nun. The distance from
the remote exit pupil to the front of the housing which encloses the complete system Is 345 mm. A
noteworthy feature of this zoom system is the short travel required of the secondary, tertiary, and
Image plane In order to achieve the 2 to 1 magnIfication range. The movement of the image plane
over the zoom range was about 57 nun along the optical axis and approximately 33 mm In the
orthogonal direction. Higher-order aspherlc coefficients up to 14th-order were utilized on all three
mirrors In order to obtaIn the desired optical performance. The surface deformation on each of the
mirrors Is well behaved.

5. OPTICAL PERFORMANCE

The MTF was evaluated at the Nyquist sampling frequency of the thermal pixel array which Is 2.86
cycles/mm. Using this sampling frequency makes It possible to perform comparative evaluations at
all zoom positions. The results of the evaluation are contained In Table 2. The values In the table are
ratios given by the actual diffraction MTF divided by the diffraction-limited MTF at the spatial
frequency of 2.86 cycles/mm. The actual MTF Is an average of the tangential and sagittal with each

100

Downloaded from SPIE Digital Library on 24 Oct 2009 to 76.29.229.28. Terms of Use: http://spiedl.org/terms
2mArq irube '3q2xeM wooZ suopjsod3JE WOJJ jioqs 03 200JJE30J ipuei Plel! suopjsodair w suuej
eq
s juep. qq
Jo aAJ4EIaJ ppg sai2ur '() pur aai2epjo Luuojun aao asp aqjuapeij a&jo
jnoq2noaspasp suooz aBusu

suooz
50'LP l'o 0'0 i-'o Vl o'i l-'1
0Z9 PS Z8 8 cs f8 BL
689 08 88 06 SL PS
9Z6 SL 08 Z6 98 Z8 99
LOll fB tB 98 88 LB 89
PZZI t8 18 6L 98 Z8 P9

aqrj jjj auewaojaed espjo auasrq ulajs1Cs

'Cpuanbas,jasp iruuasit texid Craar rio sopepp £rurpe,roi sq asp ersu eurid M enbai asp
uosnpiq Ejo opup aq prduq rjo opupi no asp ppdo auruuojiad srt pau!uuapep 04 eq

9 LDNVIflOLSISA'IVNV

v Liruiuillaid a3uriaIo3 51$AC0Esrq uaeq pauuojiadioj asp eaiq iouiui uiooz ppdouiess
a3ur.Ieosepçqsues aia pezLpur r aq 'jaoqs 'pjw pur uoi tao.i qj$ua suojsod aq &iprds
ueeat3aqaqp Lnj.1a3 JoLqut pur eruq auiqd sr&t peso sr r aorsnoduioa 30d5 azs poE LLF'1
auriuiopad aia pasn su aqja3uruuopad rpapp a3urIaIojdnpnq sr pruipsa Cq surw jo or
UO4fl33 airnbs) ;ooa aqjo wns aq4po (saiv.nbs se paujuuapaprq 3znivuuoJIid sr&
jsoiu apsu.s04 a3urIøo4 dnpjnq aqjrpw uosod pur sra suas jraq jaoqs roj qua
iiopjsod o asraiep.qp p.riups. pods snprt woajanzrlaIoJ'dnpjnq soq sauraalo3 air
suo 3q) aia punoj 03 aq q jsom :aAJ)snas upids uaajaq Lwiupd pu 'Lnpuo3as A aapuxop no
q3 '1CJIlU!Jdpu p oqi IXE 00 qj auqid snsa qjo iwaaio s!spuB aa pazituwns

L S1%TOISf1r1DMOD

uv paansqounjpapajj wooz iirndo urnsAsaojuouzn uj paijuj wnapds qajq

sap&oad Auuojun uopiqos&i-qjq Lsnonupuo noqnoiip
jo
aq tuooz a2uta pu JaAo aip
siq uaq pusp sapzsuowapaqjjqsuajjo uzjn uu pinsqoundnd pue Lw
piati uu
o zjuqxui jndqnojq iq pEduiojuawauuur aill juiudopapJo
jasgo asaAjO-ppg
ad4ojoidot asuouipaip jjpjpud sjqjjo pdo usSs SJ P_ppuL
sJ4rahIotJa'IMoNNDv
aqj io pai,iodauj sjq juauinop s
p3uuojid ipun aq diqs.iosuods qjjo 'q3JI3sa
JSJ pu UOflpA UOSJh4j ABaqJO'pUBlUtUOOUO 0E330 3UllIGAJflS
iaju s
jo qas
jUUIdOpA3a
juawa jo ffoo-3-c6-1oo99I[pJ3uo3 psj ja uwi2oajsj papunj £q aLit
'unuoiji4j jq 'unqiuqy aaoz •itc

LOt

Downloaded from SPIE Digital Library on 24 Oct 2009 to 76.29.229.28. Terms of Use: http://spiedl.org/terms
 zoom posItion (1) (2) (3)

nominal RMS spot radIus (mm) .0523 .0493 .0643

(averaged over FOV)

estimated RMS spot radIus .0530 .0521 .0625

(using RSS summation)

estimated change In radIus .0007 .0028 .0012

Tolerances utilized for all three zoom positions are:

spacing bet. prim. & sec. .0254 mm

spacing bet sec. & tert. .1000 mm

X decenter, prImary .0254 mm

V decenter, prImary .0254 mm

X decenter, secondary .0254 mm

V decenter, secondary .0254 mm

X decenter, tertiary .1000 mm

V decenter, tertiary .1000 mm

tilt aboutX, all mlrrors(TETX) .0100 deg.

tilt about Y, all mlrrors(TETY) .0100 deg.

TETX measures tilt about the X axis, whIch changes the tilt about theY direction.
TETY measures tilt about theY axis, which changes the tilt about the X direction.

Table 3. Results of the tolerance analysis at (1) short, (2), mId and (3) long focal length positions.

102

Downloaded from SPIE Digital Library on 24 Oct 2009 to 76.29.229.28. Terms of Use: http://spiedl.org/terms
 REFERENCES AND FOOTNOTES

1.Walter E. Woehi, "An all-reflectivezoom optical system for the Infrared," Opt Eng. 20(3), 450-459
(May/June 1981).
2. Reynold S. Kebo, "All-reflective zoom optical system," U.S. Patent No. 5,144,476, (Issued I
September 1992).
3. R. Barry Johnson, "Unobscured reflective zoom systems," Proc. SPIE, Vol 2539,218-225 (July
1995).
4. R. Barry Johnson, "Reflective Zoom Optical Systems," SPIE Short Course SC2I, SPIE Annual
Meeting, San Diego (July 1995).
5. Allen Mann, Chair/Ed., Zoom Lenses, Session 4, "Reflective Zoom Systems," Proc. SPIE, Vol 2539
(July 1995).
6. R. Barry Johnson, "Wide Field-of View Three-Mirror Telescopes Having a Common Optical Axis,"
Opt Eng. 27(12), 1046-1050 (Dec. 1988).
7.The pupil can be either an entrance or exit pupil depending upon If the telescope Is being used as a
sensor or prQjectOr. When used as a projector, the pupil size Is often determined by the physical size
of the pupil of the unit under test
8. This Is also known as an off-center field or off-set field.
9. Dletrich Korsch, Reflective Optics, Academic Press, San DIego (1991).
10. DIstortion Is an aberration of the principal ray and affects the spatial (angular) position of a point
Image without Impacting Its optical quality.
11. R. Barry Johnson and Allen Mann, "Evolution of a Compact, Wide Field-of-view, Unobscured,
All-reflective Zoom Optical System, Proc SPIE 3061,47 (AprIl 1997).
12. The ZEMAX Optical Design Program was used for the design and analysis of this optical system.

103

Downloaded from SPIE Digital Library on 24 Oct 2009 to 76.29.229.28. Terms of Use: http://spiedl.org/terms
 SPHERICAL ABERRATION
0FF-CENTER ENTRANCE PUPIL
FOCAL LENGTH —10, F/S

•
DISTORTION, PUPR.OFFSEr.4
OFF-CENTERED ENTEANcE PUPIL
ABERRATION
VALUE

POSITIVE
__________
•
- PUPILOFFSET-3

NEGATIVE _______

PUPILOFFSET 2

ZERO
__ 4
DISTORTiON COMA ASTIGMATISM PETZVAL °°° -SS750O 0.00000
DEFOCUS
0.07500 0.15000
0.0500 IwCi-s

Figure 1. The Influence of the classical third- Figure 2. Image of a point source due to
order aberration on the image distortion for an spherical aberration present in an eccentric
off-centered pupil optical system is shown for pupil optical system with an efi =10 Inches and
positive, negative and zero values of the operating at F/S.
respective aberrations. (After f9J, Chapter 6.)

COMA
OFF-CENTER ENTRANCE PUPIL
FOCAL LENGTH = 10, F/S
PUPIL OFFSET =2, 1 OFF-AXIS

UP
LEFT
RIGHT

DOWN

CENTERED PUPIL
(IMAGE MAGNIFIED BY 2)

Figure 3. Image of a point source due to coma present

In an eccentric pupil optical system with an efi
=10 inches, pupil offset of 2 inches, object located
I degree off-axis, and operating at F/S.

104

Downloaded from SPIE Digital Library on 24 Oct 2009 to 76.29.229.28. Terms of Use: http://spiedl.org/terms
 NARROW

iigure 4. LPesigii 1: Loom opticai system navmg a LA zoom range, square iUV range oi I.Y-J
degrees, Ff3 at FOV 3 degrees, and low distortIon (1%) and anamorphic error. Notice that the
tertiary mirror is much larger than the primary mirror.

NARROW WIDE NARROW

Figure 5. Design 2: Zoom optical system having a 4X zoom range, square FOV range of 0.5-2 degrees,
Ff6 at FOV =2 degrees, and low distortion and anamorphic error. Notice that the tertiary and
primary mirrors are about the same size.

105

Downloaded from SPIE Digital Library on 24 Oct 2009 to 76.29.229.28. Terms of Use: http://spiedl.org/terms
 I

NARO;;'rDE
7
NARROW

Figure 6. Design 3: Zoom optical system having a 4X zoom range, square FOY range ofO.5-2 degrees,
Ff3.6 at FOV = 2 degrees, and low distortion and anamorphlc error. The image plane moves hi the
opposite direction with zoom when compared with the prior two designs. Notice lhatthe tertlaiy
mirror Is much larger than the primary mirror.

Figure 7. DesIgn 4: Zoom optical qstemhaving a 4X zoom range, square FOY range o(O5-2
degrees, F/3.9 at FOV = 2 degrees, and low distOrtion and anamorphic error. Notice that the tertiary
mirror Is much larger than the primary mirror.

106

Downloaded from SPIE Digital Library on 24 Oct 2009 to 76.29.229.28. Terms of Use: http://spiedl.org/terms
 D LPVOUT
zoci MTPPI3? S'VT[M. cs4cPT £L
SC.E ø2 aaø CENT1HTPS

Figure 8. FInal design In the wide field-of-view configuration.

ZOON PIRRO SYSTE4. P110 EFL

SCRE: iø. CENTIP1E7ERS

Figure 9. Final design In the mid-range field-of-view configuration.

D LPv0UT
P1IWR SSTE14. LUP EFL
9Cm... ieee cE,4rplETEps ____________________

Figure 10. FInal design in the narrow field-of-view configuration.

107

View publication stats Downloaded from SPIE Digital Library on 24 Oct 2009 to 76.29.229.28. Terms of Use: http://spiedl.org/terms