Expected Characteristics of Multibeam Solar Radio Telescope with a Focal Focal Plane Microbolometer Array at 100 GHz

 

V.Khaikin*, (1), A.Luukanen**(2),(3)

 

*(1)The Special Astrophysical Observatory of RAS

195251, St.Petersburg,

Russia, Polytechnicheskaya 21, of.107

E-mail:vkh@brown.nord.nw.ru

(2)**Metorex Oy, P.O. Box 85, FIN-02631, Espoo, Finland

(3)Present Address: VTT Information Technology, Microsensing,

P.O. Box 1207, FIN-02044 VTT, Finland

E-mail:arttu.luukanen.vtt.fi

 

Abstract

 

Achievable  parameters (NEP,NETD, dynamic range) of  a focal plane array with room temperature microbolometers  are calculated in the range of 100-150 GHz. Expected characteristics of 8 m MSRT with a focal plane array at 100 GHz are given. Possible 2D array architecture is presented. For 8 m MSRT 15x15 element focal plane array is to be optimal at 100 GHz as well as 23x23 element focal plane array is near optimum at 150 GHz.

 

Keywords: Microbolometer, focal plane array, multibeam radio telescope.

 

 

1.INTRODUCTION

 

Multibeam receiver system for Solar MM-wave burst observations was initially developed in 1989 for Itapetinga 13.7 m radio telescope to study angular positions of Solar bursts [1]. The multipixel Sun image has been a challenge

in MM band but it can be very  useful for Solar flare monitoring.

 

The current status of Multibeam Solar Radio Telescope (MSRT) of Tuorla Oobservatory is given in [21].  2 m precise MSRT dish is suitable for Solar flare monitoring at 100 GHz where up to 6x6  pixels may be obtained  in Sun disk area.  For MSRT  with bigger dish (up to up to 8 m) more than  20x20 pixel Sun images may be  available  at 100-150 GHz. New electroformed Nickel panel technology developed by Media Lario Co. [32] gives us a chance to adopt  light but very precise sandwich type panels for bigger MSRT dish with Cassegrain optics suitable for thea multibeam moderadio telescope.

 

Modern MMIC technology  was  applied for MSRT receiver array at 10 mm30 . GHz.  Characteristics  of  array  receiver  modules  are given in [43]. Two MMIC array receiver modules at 26-30 GHz10 mm of circular polarization(L/R) are presently under test at MSRT.  2x3 or 3x3 element focal receiver arrays may be then assembled withof these modules. MMIC technology may be also also used to build array receiver  modules at 3 mm  but ifwhen we need dozens or hundreds 36 pixels for Solar flare monitoring with 8 m MSRT at 3 mm the or more pixels for Solar flare monitoring with MSRT anantenna coupled microbolometer array array technology [64]  developed by Metorex Oy, Finland seems to be  more promising for  a focal plane array.

 

 

 

2.MICROBOLOMETER ARRAY

 

 

2.1.Noise and sensitivity of  a room temperature microbolometer

 

Our calculations show that electrical noise equivalent power (NEP) 10 pW/rtHz 10^-12 W/sqrt(Hz) is achievable the nearest yearstoday at 100-150 GHz for an antenna coupled microbolometer in operated at room temperature. Noise equivalent temperature difference (NETD) is given by:

 

                           

 

Where ε is the coupling efficiency, ∆F-bandwidth, τ-integration time. For a realistic  ε ~0.3, ∆F=45 GHz, τ=1 sec NETD=27 K for Ti-air bridge  microbolometer. The use of new composite materials for microbolometer air-bridge with a lower 1/f noise gives us a chance to reach 10 K/sec in the nearest years. Estimated temperature sensitivity is less than the traditionally used to investigate quiet Sun or active Solar regions in MM band but it is quite permissible for Solar flare monitoring with 8 m MSRT where antenna temperature can easily exceed 500-1000 K above quiet Sun level  in the Solar flare case. Wide dynamic range of a multibeam  receiver system is more important parameter in our case.

 

2.2.Dynamic range of a microbolometer array

 

The dynamic range of a microbolometer in the  array  is determined mainly by the resistance uniformity of the array and the read-out electronics and to less degree by the microbolometers themselves. We can estimate the dynamic range of the microbolometer by noting that with 1 s integration time the minimal detectable power is 10 pW, while the bias dissipation is about 0.2 mW. Saturation will take place when optical power ~0.5 Pbias~100 mW, and thus dynamic range = 100E-6/1E-11=1E7=70 dB. If we take  the resistance variations from bolometer to bolometer to be  ~10 % (pessimistic), this corresponds at 1 mA bias current to a signal voltage variation of about 20 mV. The bolometer noise level is about 1 nV/sqrt(Hz), so the 1 s dynamic range requirement for the read-out circuit is 20 mV/1e-9=2E7. Low noise preamplifier, such as the Burr-brown INA103 has an input noise of 1 nV/sqrt(Hz) at 1 kHz and maximum output is 10 V. The signal to the preamplifier will be chopped at 1 kHz, so the noise bandwidth is 1 kHz. Thus, the limit for 1 kHz  dynamic range of the preamplifier is ~3E8. The required second gain stage is capable of this dynamic range too.  In the end, the A/D conversion will finally  yield  the system dynamic range. The digitization  noise  from  the  least  significant   bit (lsb) is 1/2/sqrt(3)=0.3 lsb rms. Thus a 16 –bit  A/D converter for

10 %  resistance nonuniformity allows us to realize noise NETD up to 10 K per 1 sec in 30% band.

 

2.3.Microbolometer array architecture

 

conditions  that allows us to realize in 30% band noise equivalent temperature difference (NETD) up to 10 K per 1 sec. A lithographic log-spiral with one circular or dual circular (L/R) polarization may be applied as a microbolometer antenna in a widebandthis case. The Nb metal (Nb or Ti) air-bridge microbolometer is placed in the insert connecting spiral arms.

Mentioned temperature sensitivity is less than it is traditionally needed to investigate quiet Sun or active Solar regions but it is quite permissible for Solar flare monitoring when MSRT antenna temperature can easily exceed 500-1000 K above quiet Sun level  at 3 mm in Solar flare case. Necessary dynamic range more than 50 dB is reachable in a microbolometer array.

To reduce 1/f noise an optical modulation of the entire array at ~1 kHz is needed. A rotating chopper with an archimedes’ spiral shape is required to circumvent the large 1/f noise of the bolometers below 1 kHz-100 Hz. The optical coupling elements, i.e. the substrate lenses may be necessary to avoid 30 % signal loss at the Si-Air interface.   

 

Possible 2D architecture and the design of microbolometer focal plane  array for MSRT are given in Fig.1-2.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 


Fig.1. Possible 2D architecture of microbolometer focal plane  array for MSRT

2D topology of 30x30 element microbolometer array  at 100 GHz for MSRT  is proposed. Expected MSRT characteristics with such a focal array are considered. Results of  multibeam simulation at 100 GHz for 8 m primary dish in the optimal Cassegrain case  are presented.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 


Fig.2. Possible design of the focal plane  microbolometer array for MSRT

 

3. CHARACTERISTICS OF MSRT WITH THE FOCAL PLANE ARRAY AT 100 GHZ

 

3.1.Simulation

MSRT is able to work with 60% antenna efficiency at 100 GHz and single pixel microbolometer tests may be done today at 100 GHz or even at 150 GHz. If  Media Lario Co. panels are applied for bigger radio telescope MSRT  dish may be  completely smooth up to 300 GHz. So off-axis aberrations and gain losses for peripheral  array elements are the most important. If gain loss for peripheral  array elements does not exceed  -1.5 dB  flux sensitivity of  8 m MSRT  with a microbolometer focal plane array should  not be worse  than 1 s.f.u./per beam/per sec.

 

For double reflector simulation we used GRASP8-SE software packages in PO+GTD approach  and Gaussian feed approximation. In simulation process  position of the secondary  mirror  was  optimized  to  reduce  the sidelobe  level  for  on/off  axis  feed. Let us take for simulation the diameter of  the main and secondary mirrors 8 m and 0.8 m, f1/D=0.4, f3/f2=2 for  hyperbola foci that gives less aberrations with reasonable illumination angle.  Fig.3. shows radiation patterns  for on/off axis feed of circular polarization (L/R) at 100 GHz with  λ/2 step.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 


Fig3. Radiation patterns of 8 m MSRT  for on/off axis feed of  circular polarization (L/R)  at 100 GHz with  step  λ/2.

 

 

 

 

 

 

 

 

 

 

 

 


Fig.4.Gain, HPBW, first sidelobe level (FSL) and beam deviation (BD) as function of feed removal from  the focus.

 

 

Calculated gain, HPBW, first sidelobe level (FSL) and beam deviation (BD) as function of removal from  the focus are presented in Fig.4 for a feed of linear polarization. Isolines of radiation patterns (Peak-3dB and Peak-30 dB) for on/off axis array feed are given in Fig.5.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 


Fig.5. Isolines of radiation patterns at 100 GHz, peak-3dB(left) and peak-30 dB(right) for on(top)and off (bottom) axis array feed

 

3.2.Optimization of the focal plane array

 

To reach maximum number of independent pixels and minimum aberration level spacing of  array  element must be close to λ/2 in air. The use of a dielectric substrate with higher ε  can reduce size of antenna element in sqrt(ε) but it also adds mismatch and insert loss. The dipole, patch  or spiral type antennas have the smallest size and is almost free from spillover effects when they are well isolated. However an attempt to achieve a closely packed multi-element system will reduce the sensitivity of each pixel due to "crosstalk" or mutual coupling[5]. Antenna efficiency  of each antenna element may be also reduced in an array. To avoid these effects antenna elements  and spacing are to provide low enough level of mutual coupling (not less than -20 dB) and practically spacing is seldom less than 0.7-08 λ. It was shown in [5] that mutual coupling of neighboring patch elements of an array may be less than -20 dB for spacing 0.8 λ  and ε=3.  To improve isolation different polarization of adjacent antenna elements is preferable (fig1.)[4]. A log-spiral and a spiral slot may be chosen  for a wideband and a narrowband focal plane array of circular polarization. 

 

Let us consider 15x15 element focal plane array at 100 GHz  for MSRT with 8 m dish and f1/D=0.4. We shall take  -3 dB level of beam overlapping though it is possible to consider  a lower or even higher level. Reasonable element spacing is 0.7 λ in air that  gives  beam deviation 2.15’ per array  step while on-axis   HPBW=2.08’(Fig4).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 


Fig.6. MSRT beams for 15x15 element focal plane  array

90 degree quarter of the focal plane array  is 4.9 λ x 4.9 λ  and correspondent HPBW is within the range 2.08’-2.35’. Gain loss (-1.5 dB) and FSL (-17 dB) for peripheral array elements (Fig.4) are acceptable. Simulated radio telescope beams with such an array   are given in Fig.6  for x, y and diagonal directions. Beams starting  from  numbers (Nx=0,Ny=-4) and (Nx=4, Ny=0)  overlap at less than -3 dB but within -3dB+0.5 dB. Thus spacing  0.7 λ is near optimum for Cassegrain system with f1/D=0.4 at 100 GHz. For longer focus system aberrations  will  grow

 

because of less beam deviation and then less element number will be available. For shorter focus system less spacing is needed with unavoidable growth of mutual coupling. For 150 GHz f1/D=0.6 seems to be near optimum as we have both less  beam deviation and HPBW value. 15x15 element array with the 0.7 λ step  has the optimal field of view for the Sun  32’x32’. 23x23 element array at 150 GHz  with the 0.7 λ step  has the same field of view for f1/D=0.6 but a lens correcting coma-like aberrations may be needed in front of a focal plane array to provide gain loss at given or less level. Final choice of  the central frequency of  a focal plane array will also  depend on  atmospheric opacity at the radio telescope site,  that requires season measurements of atmospheric optical depth in both transparency windows (100 GHz and 150 GHz).

 

 

4.CONCLUSION

 

The  room  temperature microbolometer may be used with a middle size radio telescope for  Solar  flare monitoring at 100-150 GHz. Achievable parameters  of a microbolometer at 100-150 GHz  and characteristics of  8 m MSRT with a microbolometer  focal plane array  at 100 GHz are given. The focal plane 15x15 element array  at 100 GHz with 0.7 λ step is to be optimal for  f1/D=0.4 and 23x23 element array at 150 GHz is near optimum for f1/D=0.6.

 

5.REFERENCESeferences

 

1.Herrmann R, Magun A., Costa J.E.R., Correia E., Kaufmann P., 1992, Solar Phys. 142,157 NASA ADS.

21.V.Khaikin,  S.Yakovlev, A.Kazarinov,  I.Efimov, A.Volkov,  E.Valtaoja. Multi-beam Solar Radio Telescope of Tuorla Observatory : test operations. In Proceed. of  URSI/IEEE XXVII Convention on Radio Science, pp.84-87, Espoo, Finland, Oct.2002.

32.G.Valsechi, J.Eder, G.Grisoni, C.G.M. van’t Klooster, L.Fanchi. High precision electroformed Nickel panel technology for sub-millimeter radio telescope antennas. In Proceed. of 25-th ESA Antenna Workshop on Satellite Antenna Technologies, pp.307-313, Noordwijk, The Netherlands., Sept.2002.

43.V.Khaikin, V.Dobrov, M. Parnes , V.Volkov ,  A.Golovkov, Yu.Rybakov. Multi-channel array receiver mudules for a radio telescope at 26-30 GHz. In Proceed. of  URSI/IEEE XXVII Convention on Radio Science, pp.176-178, Espoo, Finland, Oct.2002.

5.V.Khaikin,E.Majorova.Yu.Parijskij, M.Parnes, R.Shifman, V.Dobrov, V.Volkov and S.Uman. 7x8 Element MMIC Array at 26-30 GHz for Radio Astronomy Applications. In Proceed. of International Conference "Perspective on Radio Astronomy: "Technologies for Large Antenna Arrays", The Netherlands, April 1999, pp.171-182.

64.A.Luukanen, V.-P.Viitanen. Terahertz imaging system based on antenna-coupled microbolometers. Proc.SPIE vol.3378, p.34-44, Passive Millimeter-Wave Imaging Technology II, Roger M.Smith; Ed., 1998.