BRITE (BRIght-star Target Explorer) Constellation / BRITE Canada
The BRITE constellation is a truly international mission. The original mission concept for BRITE was developed by the Canadian astronomer, Slavek M. Rucinski of the Department of Astronomy at the University of Toronto, as a nanosatellite follow-on to the highly successful MOST (Microvariability and Oscillations of Stars) microsatellite stellar photometry mission. Since BRITE was conceived, the single-satellite mission concept has grown to a six satellite constellation with science teams, engineering teams and funding sources in Canada, Austria and Poland. 1)
Science objective: The primary mission objective of BRITE Constellation is to provide milli-magnitude (0.1% error) differential photometry of bright stars. It so happens that, in Earth’s sky at least, the most apparently bright stars are also among the most intrinsically bright stars. These stars, and in particular the AGB (Asymptotic Giant Branch ) and OB stars (OB stars are hot, massive stars of spectral types O or B which form in loosely organized groups called OB associations), most affect the ecology of the universe by creating and distributing all of the heavy elements that are necessary for life as we know it.
Despite their prominence in the sky, bright stars have not been studied to the same extent as fainter stars. As such, there remain several questions about the life cycles of these stars that BRITE Constellation hopes to help answer using photometry/asteroseismology. The ultimate goal for BRITE is to be able to provide photometric data on all 286 stars brighter than visual magnitude +3.5.
Mission concept: It is well known that massive stars experience periodic, semi-periodic and irregular variations in intensity due to factors such as change in density, magnetic field, surface temperature and internal seismic phenomena. The periods associated with these variations can range from minutes to months. BRITE Constellation fills an observational niche by providing scientists with almost continuous precise photometric time-series measurements with very long baselines (up to six months). By identifying the modes astronomers are able to extract information regarding the internal structure and density profile of these stars.
To achieve its goals, each BRITE satellite will take stellar photometric measurements of a target star field at least 15 minutes per orbit, every orbit, for up to six months at a time. To do this, each satellite in the constellation is equipped with a wide field of view (FOV) optical instrument (24º x 19º). The wide FOV ensures almost 100% coverage of the sky for differential photometry of the target stars.
• The BRITE Constellation began in earnest in 2005 when a single BRITE satellite (UniBRITE) was funded by the University of Vienna, Austria. The collaborative UniBRITE mission with UTIAS/SFL (University of Toronto, Institute for Aerospace Studies/Space Flight Laboratory) was also to be built and integrated by SFL.
• Funding for a second Austrian BRITE satellite (BRITE-Austria, aka TUGSat-1) was secured in January 2006, this time funded by the Austrian Space Agency, named FFG/ALR (Forschungsförderungsgesellschaft/Agentur für Luft- und Raumfahrt), Vienna, Austria. Unlike UniBRITE, BRITE-Austria would be assembled and tested in Austria by engineers at the Technical University of Graz (TUG) using a kit of parts and with mentorship from SFL. While each Austrian BRITE satellite would carry a payload tuned to a different optical band the satellites were otherwise almost identical.
• The third and fourth BRITE satellites were funded several years later, in 2010, by Poland. Like BRITE Austria, the Polish satellites are being assembled and tested at the Space Research Center in Warsaw from a kit of parts delivered by SFL. The last pieces of that kit were delivered to Warsaw in June 2011. The first Polish BRITE is a copy of BRITE-Austria in that it carries the Blue version of the BRITE instrument. The second Polish satellite may carry an ultraviolet (UV) version of the BRITE instrument, pending a favorable feasibility study.
• BRITE Canada represents the third entry into the BRITE constellation. After several delays, the fifth and sixth BRITE satellites were finally funded by the Canadian Space Agency (CSA) in January of 2011, ensuring full Canadian participation in what was originally a Canadian idea. The Canadian BRITE satellites, currently under construction at UTIAS/SFL, will be almost identical copies of the two Austrian BRITE satellites.
All spacecraft in the constellation use the GNB (Generic Nanosatellite Bus) platform, referred to as CanX-3, were developed at UTIAS/SFL.
The reader is referred to the spacecraft description of BRITE Austria / UniBRITE.
Figure 1: Illustration of the CanX-3/BRITE spacecraft (image credit: UTIAS/SFL)
Figure 2: Exploded view of the CanX-3 nanosatellite showing the structural elements (image credit: UTIAS/SFL)
Table 1: Summary of the CanX-3 / BRITE-Austria / TUGSat-1 / BRITE-Canada spacecraft bus specifications
Launch: The two Canadian BRITE nanosatellites are each expected to launch in 2013.
• The two Austrian BRITE satellites are scheduled for launch Q1 2012 on the PSLV (Polar Satellite Launch Vehicle) of ISRO.
• The first Polish BRITE nanosatellite, BRITE-PL-1, is scheduled for launch in the fall of 2012.
• The second Polish nanosatellite and the two Canadian BRITEs are each expected to launch in 2013.
The accelerated development schedule of the Polish and Canadian satellites with respect to the Austrian ones (which, of course, bore all design and development effort) helps ensure that most or all of the BRITE satellites are likely to be functioning in orbit at the same time, albeit slightly staggered. This arrangement provides a perfect opportunity to both extend the observation baseline in time (from the time the first satellite comes online until the last one dies) while also permitting simultaneous observation of the same targets in multiple passbands (Ref. 1).
The objective is to examine the apparently brightest stars in the sky for variability using the technique of precise differential photometry in time scales of hours and more. The constellation of four nanosatellites is divided into two pairs, with each member of a pair having a different optical filter. The requirements call for observation of a region of interest by each nanosatellite in the constellation for up to 100 days or longer.
BRITE instrument (photometer):
The mechanical design of the BRITE instrument (i.e., science payload) is fairly basic, which helps to ensure that maximum integration flexibility can be provided. The telescope is composed of three modules, the header electronics tray, the optical cell and the baffle. The baffle includes the aperture stop as well as the filter. The optical cell houses five lenses and the spacers that position the lenses with respect to each other.
A detailed description of the BRITE instrument design considerations regarding the optics design and focal plane is provided in Ref. 1).
The electronics tray contains the CCD header board, which includes the CCD and thermal control electronics. The associated computing and CCD driver electronics are contained separately on the IOBC (Instrument On-Board Computer), which is stacked with the other OBCs on the satellite bus to reduce payload size and limit heat dissipation within the instrument itself.
Table 2: Considerations for the selection of the imaging technology
The science payload of each nanosatellite consists of a five-lens telescope with an aperture of 30 mm and the interline transfer progressive scan CCD detector KAI 11002-M from Kodak, along with a baffle to reduce stray light. The optical elements are housed inside the optical cell and are held in place by spacers. The photometer has a resolution of 26.52 arcsec/pixel and a field-of-view of 24º. The mechanical design for the blue and for the red instrument is nearly identical; only the dimensions of the lenses are different.
The Kodak KAI-11002 CCD is an 11M pixel interline, buried channel CCD with 4008 x 2672 effective pixel dimensions. Each pixel is 9 µm x 9 µm in size and BRITE uses the option fitted with micro lenses. At only 37.25 mm (H) by 25.7 mm (V) in size, this imager offers an impressive 66 dB dynamic range with built-in electronic shutter and anti-blooming protection.
Table 3: Characteristics of the Kodak KAI 11002-M CCD detector
Figure 3: Schematic view of the BRITE instrument (image credit: UTIAS/SFL)
Optical design: The science team required from the PSF (Point Spread Function) to overcome undersampling issues and achieve milli-magnitude stellar photometry.
The optical designs each have an aperture of 30 mm with a focal length of 70 mm (focal ratio of 2.33). The theoretical spot diagrams for the two designs at 0º, 5º and 10º from boresight are shown in Figure 4.
Figure 4: BRITE blue and red spot diagrams (image credit: UTIAS/SFL)
Figure 5: BRITE blue and red optical designs (image credit: UTIAS/SFL)
The effective wavelength range of the instrument is limited in the red by the sensitivity of the detector and in the blue by the transmission properties of the glass used for the lenses. The filters were designed such that for a star of 10,000 K (average temperature for all BRITE target stars) both filters would generate the same amount of signal on the detector. The blue filter covers a wavelength range of 390-460 nm and the red filter 550-700 nm; both are assumed to have a maximum transmission of 95%.
The photometer instrument has a mass of ≤ 0.9 kg and a power consumption of ≤ 3.5 W. The instrument uses a custom set of electronics to operate the imager. The electronics include four A/D converters (14 bit) to convert the analog pixel values, and 32 MB of memory to temporarily hold a full frame image. The imager and memory timing and signals are being controlled using a CPLD (Complex Programmable Logic Device).
Figure 6: Photo of the BRITE telescope (image credit: UTIAS/SFL)
CCD driver design challenges:
The main challenges introduced by the shift of imaging technology were the increase in power and the complexity of the electronics required to operate the CCD. At the same time, the instrument design still had to meet the primary mission requirements including:
1) Perform differential photometry with error less than 0.1% per 15-minute observation.
2) Set exposure time between 0.1 and 100 s with an accuracy of 0.01%.
3) Achieve an integrated signal to noise ratio (S/N) of 3000 for a cumulative exposure taken over 1000 s on a V= +3.5 star.
This set of requirements can be translated into goals of providing very stable analog signal levels that operate the CCD, stable thermal control, high fidelity clocking and very low noise.
A two-board photometric CCD driver design:
The performance of CCDs is extremely sensitive to temperature. Generally speaking, lower imager temperature brings higher SNR. The original STL-11000M camera assembly contains a two stage thermal electric cooling unit with liquid assist that can bring the operating temperature down to 50ºC below ambient. Unfortunately active thermal cooling is not achievable on BRITE due to volume and power constraints for the BRITE mission. These constraints drove the design towards a two-board system that would minimize the heat dissipated alongside the CCD, in the payload itself.
The first board, known as the CCD header board, contains the CCD and heater control electronics. The second board, known as the instrument on-board computer (IOBC) contains all on-board computing, power regulation and CCD driver electronics. The IOBC board is an amalgamation of a complete set of standard GNB on-board computer (OBC) and the entire driver electronics required to operate the CCD imager. Only the driver design aspects will be discussed.
The two-board configuration not only mitigates thermal issues, it also isolates the imager from the noise generated by supporting electronics located on the IOBC. The tradeoff made here is the increased distance that the data, in analog form, must travel in order to arrive at the ADC (Analog-to-Digital Converter). In order to safe guard against signal depreciation, an amplifier unit was placed on the CCD header board and a coaxial cable is used for signal transfer. These two implementations help preserve the high SNR achievable by the CCD.
CCD thermal control:
Perhaps one of the largest challenges in CCD-based astronomical missions is thermal control of the imager itself. CCD detectors are susceptible to two main detrimental effects: noise, and dark current. Both can significantly reduce the detector’s ability to image faint stars, and great measures are taken to reduce these effects by lowering the CCD temperature, sometimes to as low as -110ºC.
Due to the small size, extremely limited power and full sky coverage requirement of the BRITE satellite, neither active nor passive (radiative) cooling schemes were feasible. Fortunately, since BRITE will image only the very brightest objects in the sky and because the chosen CCD has relatively low noise, even at room temperature, reducing the absolute temperature of the detector is a secondary concern. Of higher importance is the thermal stability of the CCD over the course of an observation window. Since an observation consists of a number of co-added exposures, it is essential that all exposures be taken with the detector at the same temperature. The BRITE requirements state that temperature of the CCD must be stable to ±2.5ºC throughout an observation, with a goal of ±0.5ºC.
As controlled cooling of the BRITE instrument is not practical, thermal stability is achieved by way of heating, which is much simpler to implement. Located towards the center of the spacecraft, the thermal fluctuations of the instrument are significantly lower than those of components located closer to the surfaces. Trim heaters are used to bring the CCD temperature to just above the highest temperature it normally experiences during the orbital cycle. Figure 7 shows a theoretical uncontrolled CCD temperature fluctuation throughout the orbit, as well as the temperature stabilized with heaters.
Figure 7: Illustration of CCD thermal control strategy (image credit: UTIAS/SFL)
Figure 8 shows the CCD mounted to the header board with four resistive heaters placed near the corners of the CCD. Four temperature sensors are located underneath the CCD and are coupled to it with a thermal gasket. A microcontroller is located on the opposite side of the header board. The microcontroller collects temperatures from the four temperature sensors, controls power to the heaters, and communicates with the IOBC.
Figure 8: CCD header board containing the CCD heaters and support electronics (image credit: UTIAS/SFL)
As the heaters are resistive, the heat produced is basically equal to the power dissipated. This suggests that the power output of the heaters can be controlled by either varying the voltage across it, or the current through it. However, analog circuitry to accomplish such control would add complexity and increase the part count. Instead, it was decided to use a PWM (Pulse-Width-Modulation) scheme, in which the heater is either fully on or fully off, with a variable duty cycle. The duty cycle would then correspond to the average power produced by the heaters.
A PID (Proportional-Integral-Derivative) controller was selected to control the PWM duty cycle of the heaters. This scheme was selected due to its simplicity and ubiquity in digital control applications. A PID controller is also very intuitive to tune, and does not require substantial computational resources to execute.
Bench-top tests conducted at room temperature and atmospheric pressure have shown that the controller can achieve a control accuracy of ±0.1ºC, which is 25 times better than the required control accuracy. Figure 9 shows these results. The green line represents the average CCD header temperature. The time interval from 0 to 4200 s shows an uncontrolled state, with temperature fluctuations due to changes in ambient temperature. Once turned on at 4200 s, the controller quickly achieves the target of 28ºC and successfully maintains it for the remainder of the test. The blue line represents the duty cycle of the PWM signal applied to the heaters. Due to the naturally slow response time of the system, it was found that a mostly proportional controller can achieve good results. Comparable performance was achieved while performing thermal vacuum testing on the integrated spacecraft, whereby the spacecraft was subjected to the thermal oscillations expected in orbit.
Figure 9: CCD thermal control – bench-top performance (image credit: UTIAS/SFL)
Mission operations concept and software:
The main objective of each satellite in the BRITE constellation is to observe the brightest stars in the sky and measure variability in their brightness over time. As target fields can be anywhere in the sky and there will be times during which the target fields will be obstructed by the Earth, the sun, or he moon, a minimum 15 minute interval of view of the target is expected for each orbit. This window will henceforth be referred to as the observation window. During an observation window the satellite will point at its target and perform as many observations as possible. An observation consists of a number of individual exposures which are co-added together to produce a single data point on the brightness curve.
To enable differential photometry, target fields will be selected such that more than one star of interest falls into the FOV (Field of View) of the instrument. The region around a star of interest is called a raster, and up to 16 rasters could be imaged during any given observation. To accommodate the satellite’s limited downlink capacity, most of the data processing will occur onboard the satellite, with only minimal information being stored and forwarded to the ground. Full or even partial images will be downloaded only for commissioning and debugging purposes. All data sent to the ground will be compressed on-board first.
To accommodate the observation plan described above, an operational concept was developed and a high-level operations cycle defined for each of the satellites in the BRITE constellation. This concept was then flowed down to define the required ground and flight software. The top-level operational cycle is shown in Figure 10. 2)
Figure 10: Schematic view of the mission operations cycle (image credit: UTIAS/SFL)
The operational cycle begins with the science consortium, which consists of 12 lead scientists and investigators for the mission. The science consortium collects observation requests from the astronomy community, selects targets and times for observations and releases this information through the observation plan. The observation plan is entered into the BRITE Target ground application, which takes into account the time of year, positions of stars in the sky, and the orbit of the satellite to determine feasible observation times as well as the required spacecraft attitude. The BRITE Target produces two files: an Observation Setup File that is uploaded directly onto the satellite’s instrument computer, as well as an Observation Schedule File that is passed to the BRITE Schedule ground application.
The Observation Schedule File contains the times at which the target will be visible to the satellite, and the attitude to which the instrument must point. BRITE Schedule converts this information into a series of timed commands in a format understandable by the spacecraft. It also adds other commands that are necessary to initiate an observation. BRITE Schedule outputs these commands to a TimeTag Script file, which is then uploaded to the housekeeping computer on the spacecraft through the TimeTag ground application.
The Observation Setup File contains instructions for the instrument computer that specify the number and duration of exposures to be taken for each observation, the location of the targets in the imager’s field of view, and the type of processing that must be done on the raw data.
At the time specified by the TimeTag commands, the satellite’s housekeeping computer turns on the instrument computer, performs initialization routines, and begins the observation. Observations continue until the time at which the target is no longer visible, is polluted by stray light or fine pointing lock is lost, at which point the instrument computer is shut down.
During every pass over the ground station observation results are downloaded through a File Downloader application. The observation results are then returned to the science consortium for processing, concluding the observation cycle.
Spacecraft-level operations sequence:
Each BRITE satellite is designed around SFL’s GNB (Generic Nanosatellite Bus). Each BRITE satellite contains three onboard computers: the HKC (Housekeeping Computer), the ADCC (Attitude Determination and Control Computer), and the IOBC (Instrument On-Board Computer). Additionally, a dedicated microcontroller is responsible for the thermal control of the CCD. In nominal operations, the HKC receives all ground commands and distributes them to the appropriate destinations on the spacecraft. Likewise, all outgoing traffic is forwarded to the ground through the HKC. Since the CCD heater controller is connected only to the IOBC, all commands to it must be forwarded by the IOBC. Figure 11 shows a simplified interconnect diagram for the onboard computers.
Figure 11: C&DH (Command & Data Handling) interconnect diagram (image credit: UTIAS/SFL)
The HKC houses the TimeTag command interpreter. Each entry in the TimeTag queue contains a command as well as the time at which it should be executed. When the command time elapses, the HKC forwards the command to the appropriate computer for processing. TimeTag provides a necessary means of automation as observations will most often occur outside of contact periods with the ground.
During a ground pass, a TimeTag Command File and an Observation Setup File are uploaded to the HKC and IOBC respectively. At a predetermined time before the observation is scheduled to begin, the HKC begins configuring all of the satellite subsystems. First, the attitude determination and control system is enabled and sensors are given a warm-up period. If necessary, a slew is performed to point the instrument at the target. Following the slew, commands are issued to the power distribution module directing it to turn on the instrument, IOBC, and CCD thermal control subsystem. The CCD heater controller is enabled and allowed to settle to the desired temperature. With the satellite pointing in the right direction and the CCD thermally stable, the instrument is configured and the observation cycle initiated.
Observations continue autonomously until the HKC issues a time-tagged command to halt observations. While in an exposure, the IOBC continuously polls telemetry from the ADCC and CCD Heater Controller to ensure that the spacecraft attitude and CCD temperatures are within an acceptable range for imaging. Once observations are finalized, to conserve power the power distribution module is commanded to power down all instrument electronics and the satellite returns to its pre-observation state. This cycle is repeated once per orbit until new targets are selected and new observation scripts are uploaded.
Instrument computer operations sequence:
Once an observation start command is received by the IOBC, it begins an autonomous cycle of observations with the parameters defined in the observation setup file. An observation window spans the time between an observation start command and an observation stop command. It may consist of multiple observations, and each observation may be composed of multiple co-added exposures. The number of exposures, the locations and sizes of rasters within the FOV, and the type of processing to be performed on them are all defined within the observation setup file. Figure 12 shows the state transitions occurring on the instrument computer during an observation window.
Figure 12: IOBC level observation sequence (image credit: UTIAS/SFL)
1) Jake C.T. Cheng, Jakob Lifshits, C. Cordell Grant, Mihail Barbu, Robert E. Zee, “The BRITE Constellation Space Telescope Design and Test of a Wide Field, High Resolution, Low Noise Optical Telescope for a Nanosatellite Constellation,” Proceedings of the 25th Annual AIAA/USU Conference on Small Satellites, Logan, UT, USA, Aug. 8-11, 2011, paper, SSC11-I-1
2) Jacob Lifshits, Robert Tee, “Embedded Systems Development for SFL Satellites,” MASc Thesis, Graduate Department of Aerospace Science and Engineering, University of Toronto, Toronto, Canada, 2010, URL: https://tspace.library.utoronto.ca/bitstream/1807/25764/1/Lifshits_Jakob_201011_MASc_thesis.pdf
The information compiled and edited in this article was provided by Herbert J. Kramer from his documentation of: ”Observation of the Earth and Its Environment: Survey of Missions and Sensors” (Springer Verlag) as well as many other sources after the publication of the 4th edition in 2002. - Comments and corrections to this article are always welcome for further updates.
The BRITE Constellation: