8.1 Conclusion
The thesis work provides an insight to a variety of research topics, including Space Physics Instrumentation, hardware, electronics, sensor systems, DSP and Radio Astronomy. Research methodologies from these areas were applied to design the Priamos and Dimagoras systems.
Chapter 3 describes the hardware implementation a dual-channel cross-correlator, typically found in correlated Radio Astronomy systems. The hardware correlator is seriatim incorporated into Priamos and Dimagoras. The designs of an 8-phase auto- switching dual-channel correlator and a free- input dual-channel correlator are presented, based on the implementation of the cross-correlation function for non- identical incoming signals.
The dual-channel system requires 95,000 equivalent NAND2 gates and the frequency of operation is 81 MHz. The circuit is downloaded to the 6,000,000 gates XC2V6000- 5-FF1152-based ADM XRC-II Alpha-Data board, equipped with the XRM-IO146 module. The bi-directional communication between the FPGA and host is via the 32- bits/33 MHz PCI interface. Data acquisition and visualisation is performed using C/C++ and Matlab software, respectively.
Chapter 4 describes the feasibility study, RF Receiver Unit and peripheral ha rdware design for Priamos. The RF Receiver Unit consists of five functional areas: RF input and calibration, PAs, filters, digital amplifier and ADC circuits. The peripheral hardware consists of the programmable master oscillator and GPS receiver. The bi- directional communication between the GPS receiver and host is achieved via the USB 2.0 and 10/100 Mbps Fast Ethernet FPGA interfaces.
Chapter 5 describes the hardware design of the Priamos DSP Engine Unit. The DSP Engine supports over 300 SDR commands in hardware. Eight mega- functions are implemented into the XC3S1500-4FG456 device to programme or reconfigure the FPGA, RF Receiver Unit, GPS receiver, master oscillator, 16 Mbps SRAM and glueless USB 2.0 and 10/100 Mbps Fast Ethernet Interfaces.
Programmable features include: dual- DDC processing, auto-correlation, sampling frequency up to 250 MHz, integration time, UTC RTC timekeeping, UTC data timestamping, long UTC timestamped data storage up to 144 h etc. The DSP Engine Unit is independent of the antenna type being used and provides a fast prototyping platform for other Space Physics Instrumentation projects.
Chapter 6 describes the feasibility study and system analysis for Dimagoras. A novel applied design methodology for engineering low-power macroscale optimised fluxgate sensors for measurements of the complex solar wind- magnetospheric- ionospheric system is presented. The optimised sensor saturates for an excitation current of +/- 60 mA, four times less the initial 250 mA specifications. Power consumption is reduced by a factor of 16. The sensor’s sensitivity is 151 uV/nT to amply cover the Earth’s magnetic field variation. The tri-axial sensor is built by assembling three single-axis sensors. The sensor determines the Dimagoras datapath processing specifications.
Chapter 7 describes the hardware design of the remaining Dimagoras system, consisting of the following programmable or reconfigurable interfaces: PAs, digital amplifiers, ADCs, DAC, power inverter stage, master oscillator, GPS receiver, 16 Mbps SRAM, FPGA, UART ports, USB 2.0 and 10/100 Mbps Fast Ethernet.
The system samples the sensors’ outputs and cross-correlates them with the reference waveform to detect each channel’s peak-power. The cross-correlated results are interpolated using a cus tom algorithm, updating the DAC once per excitation cycle. The Earth’s field is nullified and the field variation data, due to Space Physics events, are captured and UTC timestamped. The bandwidth is reduced to 10 Hz. A real-time numerical algorithm transforms power measurements to nT.
8.2 Further Work
The Priamos system design is completed. The novel applied sensor design methodology for engineering low-power macroscale optimised fluxgate sensors is verified. The hardware design and simulation of the remaining Dimagoras system is completed. A few remarks regarding Dimagoras are in the following sections.
8.2.1 Remodelling the Sensors using Metglas 2714A
A new material was recently found, which could improve the magnetometer’s specifications. The Metglas 2714A [224] magnetisation curve is in Fig. 8.1. The magnetic flux density saturation value is 0.57 T, compared to supermalloy’s 0.8 T. The basic Metglas 2714A fluxgate sensor would saturate at 159 mA, compared to supermalloy’s 250 mA.
Figure 8.1 Metglas 2714A Magnetisation B (H) Curve (T(A/m)).
Applying the novel macroscale optimisation technique of Chapter 6, the excitation current can be reduced to an estimated value of 40 mA, compared to supermalloy’s 60 mA. The core loss is 1 W/kg at 5 KHz. The power consumption is reduced by a factor of 39, compared to the basic supermalloy sensor.
8.2.2 Theoretical Calculation of the Earth’s Magnetic Field
The system measures the aggregate disturbed magnetic flux density, due to the different ionospheric and magnetospheric events, plus any seriatim generated continental or sea fields, accounting for 2 - 3 % of the Earth’s total magnetic field [222]. Another 2 – 3 % is due to the fields of the different materials at the Earth’s crust. The remaining 90+ % is due to the Earth’s core. The Earth’s core field is included in the digital World Magnetic Model (WMM). The 2006 model and predictions until 2010 are published. WMM takes into account crustal magnetic fields, represented by a constant.
The contribution of magnetometers is not included in WMM. A new research study could use WMM and magnetometer data for short term predictions. The software development would couple the operation of Dimagoras. Since WMM takes into account the altitude variation, the predictive model would also apply to airborne magnetometers.
8.2.3 VLSI Considerations for Priamos and Dimagoras
The two projects could reach the VLSI stage within the next years. A mixed-signal approach is suited for the two projects [225]-[226]. The performance and frequency of operation are increased, while the power consumption and size are reduced. These factors are related to the slow speed of the FPGA’s multiplexed switching matrices. The critical path of an implementation is longer than expected, due to the datapath routing through the different FPGA’s embedded units.
Due to the low- levels of the received power an external antenna is still required for Priamos. Digital magnetometers have been implemented into ASICs for spaceborn applications [106]-[110]. Single- or dual-axis MEMS sensors can be embedded into ASICs. Tri-axial magnetometers utilise an external macroscale sensor. Due to the rapid development of the different processing technologies, it is worthwhile investigating the possibility of incorporating a tri-axial sensor into an ASIC. Investigation is required to determine the lowest limit of miniaturisation that the optimised sensor can be imposed to.