First Result of the SMA Holography Experirnent

Transcription

1 Firs Resul of he SMA Holography Experirnen Xiaolei Zhang Peer Brako Dan Oberlander Nimesh Pae1 Tirupai K. Sridharan A4nony A. Sark December 11, 1996 Submillimeer Array Memorandum, No. 102

2 Absrac This memo summarizes he resul of he firs holography run: which obained a 6-l by 64 beam map of he SK4 anenna 1. Afer Fourier ransform and large scale error fiing, he aperure phase map revealed a dish surface wih rms error abou 130 micron. Subsequen panel fiing and he removal of fied panel errors resul in a residual surface rms error abou 14 micron if including he quadrupod shadow region, and 5-10 micron if no. The Firs Beam Map The firs beam map of he SK4 anenna 1 was aken on December , using he newly-insalled holography daa acquisiion sysem (Zhang e al. 1993), which consiss of wo room emperaure phase-locked receivers, and a phase-locked beacon ransmier locaed on a ower 250 meers away. The gain sabiliy of he holography daa acquisiion sysem is beer han 1%: and he shor-erm phase sabiliy is abou a fracion of a degree. A vecor-volmeer backend is used for deecion. and he posdeecion signal o noise raio is abou 100 db a he boresigh direcion. Figure 1 shows he image of he firs measured beam paern (ampliude only. Phase paern was aken simulaneously bu no shown here). Since a he ime of his measuremen: he boresigh coordinaes were no accuraely deermined (he program we used o measure his paern is acually mean o assis he deerminaion of he boresigh direcion), he beam is seen o be slighly off-cenered. The overall paern, however, is consisen wih wha is expeced of an anenna focused owards 250 meer disance. The signaure of he SMA quadrupod srucure, a paern roaed 7.3 degrees clockwise from verical and horizonal symmery, can also be discerned. Due o he lack of an accurae boresigh direcion, gain and phase calibraion ad he boresigh is no made during he mapping. Our experience so far wih he daa acquisiion sysem, however, shows ha he emporal drif of ampliude and phase levels is minimal. This 64 by 64 map ook abou one hour and 10 minues o acquire. I is expeced ha in he near fuure, we will be able o opimize he daa acquisiion sofware so ha a 12s by 125 map can be acquired wihin an hour. 2

3 Figure 1: Measured beam paern of he SMA anenna 1. 3

4 Analysis The measured ampliude and phase paerns were hen inpu o he analysis package HOLIS (Zhang 1996) for processing..4fer he preprocessing sage o move he phase reference plane o ha in he middle of he dish, and afer a subsequen Fourier ransform, we obain he ampliude and phase disribuions of he elecric field on he anenna aperure: as shown in Figures 2 and 3. Noe ha since we have a large poining offse o he boresigh direcion, he linear phase gradien on he aperure is very large. The phase in fac has wrapped around 3~ a couple imes, and we unwrapped he phase o arrive a he phase disribuion shown in Figure 3. In Figure 2, he ampliude suppression a he locaion of he quadrupod is clearly seen. A blanking of he region inside he subreflecor and ouside he dish, however. was arificially made. Alhough a linear phase gradien dominaes he phase map of Figure 3, some large ring-like ripples are seen on he phase map, indicaing he presence of he secondary mirror diffracion conribuion (Zhang 1995) o he phase paern. Afer aking ou six erms of large-scale errors (consan erm! Y and y linear gradien, defocus: diffracion, as well as he disance uncerainly o he ransmier), we obain a residual map of he small-scale phase errors. caused mainly by he misalignmen of he panels from a perfec paraboloid. Convering he phase error map o a surface error map (Zhang 1995), we hen obain he disribuion shown in Figure 4. From Figure 4, we see ha he maximum surface error is abou 500 micron, and he calculaed surface rms error is abou 130 micron. The locaion of he quadrupod, he separaion beween he differen rings of panels, as well as a few very-badly misplaced panels can also be discerned in Figure 4. The relaively large amoun of surface error near he area where he quadrupod legs connec o he ouer dish is likely o be due o ha, during he assembly phase, he dish was lifed up a hese locaions and hus= has deformed accordingly. The sligh asymmery beween he widhs of he panelsized srucures in he lef and righ hand side of he dish, is due o ha he cener of he dish is locaed on cell (33,33) in his 64 by 64 map, so here is expeced o be a maximum one-cell difference in he size of srucures in he lef and righ. This arifac will be reduced in a 125 by 1% map. 4

5 I,,.,,,,,,, /, / 09 OS OP OE (qax!d) d /., /, d 09 OS OP OE Figure 2: Ampliude paern on he anenna aperur e.

6 Elb I I I1I/ L ISIII/,///,,,,,/,,,I,,,,, 4 \ \ 1 _ / \ \ ---I l --x*$fz<,; 1 I h I III, I 1~11,,,,,,,, I,,,,,.,,,,,I 09 0s OP OE (qax!d) d F r igure 3: gui y. 09 OE OP OE Phase disribuion on he anenna aperure, afer unwrapping a %;r ambi- 6

7 Surface-Error-in Whie/Black: / , _,,

8 Fiing he Individual Panel Movemen The phase error map of Figure 4 is subsequenly used by he panel-fiing rouines in HOLIS o calculae he screw moions needed on each panel o reduce he surface error. The residual error map afer panel fiing is shown in Figure 5. Here we see he dominance of he quadrupod paern as well as is shadow in he ouer dish. The ripples in he ouer ring is expeced o be he aliasing effec produced by he quadrupod shadow. This laer effec is expeced o be reduced once a 125 by 12s map is made. The residual surface rms error afer he panel fiing is abou 14 micron. If we ignore he quadrupod shadow region, he residual rms is beween 5-10 micron. Summary The firs SMA holography es run successfully measured he beam paern on anenna 1. The aperure phase disribuion obained from his beam paern clearly indicaes he four rings of SMA panels, as well as he signaure of individual panel misalignmens. The nex phase of holography effor will concenrae on improving he speed of he daa acquisiion, verifying he repeaabiliy of he measuremen, undersanding he possible source and correcion of sysemaic errors, as well as improving he alignmen of he various componens in he opical pah. Acual panel adjusmens will begin when a high-resoluion (128 by 125), repeaable aperure phase map is made, in he near fuure. Acknowledgemen We hank he SMA saff a he Haysack consrucion sie, a he Cambridge sie, as well as a he submillimeer wave receiver lab for helpful assisance. Marin Levine;. John Tes and Cosmo Papa were involved in he firs sage of he hardware preparaion. Colin Masson and Bob Wilson provided useful echnical advice. We hank Jim Moran, Eric Silverberg and Bill Bruckman for managemen assisance in making he firs holography run happen in a imely fashion.

9 L,. /1/,,,//,,,,,,,,,,,,,/,,,, 1 09 OS OP OE (qax!d) d 09 OS OP OE OZ 01 Figure 5: Residual error disribuion on he aperure afer panel surface error is 14 micron. 9 fiing. The I-ms

10 References Zhang, X. 1996, HOLIS User s Manual, in preparaion. Zhang, X., Levine, M., Brako, P., Tes, J., Papa, C. and Masson, C. Planned Panel Alignmen Procedure for he SMA Anennas Using he h&crowave Holography Technique: S&IA echnical memo, No. 85 Zhang, X. 1995, YZerain Opics Consideraions for he Holography Experimen, SMA Techical memo, No