Technology

The Spin Torque Oscillator


2007 year's Nobel Prize in physics was awarded Professor Albert Fert and Professor Peter Grünberg for their discovery of Giant Magnetoresistance (GMR). GMR has had profound implications for our everyday life as it enables extreme hard drive storage densities. More recently, MRAM was introduced as the world's 2nd commercial spintronic technology. [i] The 3rd spintronic technology will be Spin Torque Oscillators (STO).

An STO has four specific features that makes it interesting:

  • Small footprint
  • Wide tuning range
  • Fast upstart time
  • Potential for high frequency generation and detection (THz)

All of which can be beneficial in different applications.

More and more components are integrated on chip in contemporary devices. Therefore chip real estate is becoming more and more premium. Today a communication function typically relies on an LC circuit based VCO. The LC circuit need an inductor that will not scale when the transistors become smaller (the transistors follow Moore's law). This means that the inductor becomes a greater part of the system cost as technology becomes cheaper The fully integrated, broadband and tunable STO replaces the inductor. The STO has the benefit of following Moore's law, and it has a minimal footprint. It can also be integrated on any semiconductor material. Further more the wide tuning range enables a few STO to cover the full frequency range of all communication standards, allowing for less complex systems. The cost of the system is proportional to the chip area it occupies. The STO is a factor 10^7 smaller than a traditional inductor. This will allow for a great reduction of production costs. The cost reduction translates to many markets, such as mobile phones, WLAN routers, WLAN cards, and other high volume applications depending on a high degree of integration for cost savings.

The STO can also be utilized for higher frequency applications such as new communication protocols at 60GHz or radar applications. For instance, radar jammers used in aircraft today are cumbersome and take up a lot of valuable space in the fighter jet. This is due to the use of several oscillators to cover the full incoming radar spectra. The footprint of the jammer is thus due to the poor tuneability of contemporary oscillators. The STO has extreme broadband capability and can thus replace the cumbersome existing solution providing a very small footprint in the carrier airplane.

Spin Torque Oscillators (STO) are sub-micron sized, broadband, high-quality GHz oscillators with extremely wide tunability range. [ii], [iii], [iv], [v], [vi], [vii] Instead of a standard LC based resonance element, the STO employs a ferromagnetic resonance in a thin magnetic layer. The resonance is excited by the current and the frequency is tuned by the magnitude of the current. The STO production technology is essentially identical to MRAM but uses much fewer masks and processing steps, and is hence cheaper. It is easily integrated into any semiconductor process where it consumes a negligible chip area. STOs are truly broadband oscillators operating from 1 GHz [viii] to 40 GHz [ix] (and beyond; only instrumentation limited so far). A single STO device can provide a few microwatts RF power with a phase noise of -65dBc/Hz at 1 MHz (world's first STO phase-noise measurement done in our group). STO phase-locking will improve these numbers orders of magnitudes, as will feed-back and amplification circuits. In total, using dividers and multipliers and different STO types, the technology can be applied to radio and radar applications from 100 MHz to 100 GHz.

Our recent unpublished results also include STO means to control the phase shift in radar antenna arrays, solutions for zero-magnetic-field STO operation, and the world's first Cadence simulation of STO circuits. All of wich are an important step towards future applications in everyday life. Details of these inventions are currently being patented.



[i] J. Åkerman, "Applied physics: Toward a universal memory," Science 308, pp. 508 (2005).

[ii] J. A. Katine, et al, Phys. Rev. Lett. 84, pp.3149 (2000).

[iii] M. Tsoi, et al, Phys. Rev. Lett. 93, pp. 036602 (2004).

[iv] S. I. Kiselev et al, Nature 425, pp. 380 (2003).

[v] W. H. Rippard et al Phys. Rev. Lett. 92, pp. 027201 (2004).

[vi] M. Covington et al  Phys. Rev. B 69, pp. 184406 (2004).

[vii] I. N. Krivorotov et al,  Science 307, pp. 228 (2005).

[viii] V. S. Pribiag et al, Nat. Phys. 3, 498 (2007).

[ix] F. B. Mancoff et al, Nature 437, 393 (2005)