In our digital world and with the rapid development of the Internet of Things and portable electronics, the amount of data transmitted wirelessly is constantly increasing. This imposes continuously increasing data rates. Communication technologies therefore require the development of high-performance measurement tools capable of rapid spectral analysis in order to develop the basic elements of high-speed communication between connected objects.
The key element of a spectrum analyzer is a local oscillator whose frequency can be swept linearly within a time T
via a control signal. In order to analyze a signal whose spectral content can change rapidly over time, it is necessary that the sweep time T, to sweep the frequency of this oscillator in a certain range, is as short as possible. Typically voltage-controlled oscillators (VCOs) are employed which have limited sweep rates (with T >1 microsecond) due to the macroscopic dimensions of the components used. IRIG researchers are working on Spin-Torque Nano-Oscillators or STNOs (
Figure) whose frequency can be swept on time scales below 100 ns, thus much faster than VCOs. It is due to their nanometric dimensions and their non-linear properties of the magnetization dynamics, that one can drastically reduce this sweep time.
Figure: Schematic diagram of a Spin-Torque Nano-Oscillator (STNO). The injection of a current, polarized by the fixed layer (blue), puts in self-oscillation the magnetization of the free layer (yellow). STNOs are stacks of ferromagnetic layers a few nanometers thick and a few tens of nanometers in diameter. This is the first time that researchers have demonstrated experimentally (see inset for more information) that the use of such STNOs in spectrum analyzers can push the usual limits of scanning speeds. Researchers have also shown that it is possible to resolve multiple frequency components simultaneously or to track rapid changes in the frequency of a signal. This makes ultra-fast spectrum analyzers based on STNOs a very promising technology that opens up a new field of applications.
To learn more: A DC current IDC induces self-sustained magnetization oscillations in one layer of the STNO, and generates a microwave voltage signal at the STNO terminals.
Due to the non-linear properties of the magnetization, the frequency of the STNO fSTNO can be swept over a certain range using a sawtooth signal VSW of period T. The modulated microwave signal is then mixed with the external signal Vin of unknown frequency. The mixed signal passes through a matched filter, inducing a narrow Vspec peak whose position in time to 0<to<T indicates the frequency of the signal Vin. The two graphs show an example where the STNO frequency is swept between 8.8 and 9.4 GHz and the frequency of the signal Vin varies in a sawtooth shape. Upper graph: Vspec peaks for 14 consecutive scan periods at T=50 ns. Bottom graph: representation, in the form of a frequency-time spectrogram, of the temporal evolution of the frequency for the 14 consecutive periods, showing that the spectrum analyzer based on a STNO can resolve and detect, on a 50 ns time-scale, the frequency variation of the external signal Vin.
Collaboration: Oakland University, Rochester, USA; INL, Braga, Portugal.
Funding: ERC Magical.