‘Rolled-up’ RF transformers look to smaller, more efficient wireless IoT
To create the micrometre-sized rolled-up RF transformers they describe in a paper titled “Three-dimensional radio-frequency transformers based on a self-rolled-up membrane platform” published in Nature Electronics, the researchers rely on silicon nitride’s intrinsic mechanical properties and built-in stress upon deposition, which when released from a sacrificial layer, rolls-up naturally to ease the strains.
transformers developed by Illinois professor Xiuling Li’s
team. Image courtesy Wen Huang.
Interestingly, the SiNx-based self-rolled-up membrane (S-RUM) nanotechnology platform described in the paper allows designers to completely parameterize the final 3D configuration of the primary and secondary windings of an RF transformer, all from an original 2D layout lending itself to conventional photolithography manufacturing processes.
structure from the planar layout with parasitic
parameters and ‘coil cells’ labelled.
Through theoretical modelling of the deposited strained bilayer SiNx nanomembrane used as a substrate for 2D conductive patterns, the scientists are able to calculate the precise dimensions of the finished rolled-up membranes and the electrical performance of the resulting air-core RF transformer they want to design.
This includes the inner diameter and the number of turns of the primary and secondary coils according to the turns ratio and inductance requirements of the end design. Since the membrane structure of S-RUM transformers can include multiple layers of dielectric and conductive thin films, each layer with its own built-up stress, self-coiling results from the stress imbalance from the whole stack.
showing the transition of the 2D primary and
secondary strips to 3D coils.
In the end, it is the vertical membrane structure, thickness and stress that determines the inner diameter after the device is rolled up, while the horizontal layout determines the spatial configuration of the coils and their spacing. The authors report excellent manufacturing yields, designing RF transformers of different diameters from 18.7 to 50μm with different turn ratios.
Samples RF transformers were designed with turns ratios (n) from 1.5:1 to 2.5:1, with self-resonant frequencies from 11.5 GHz to over 20GHz. At 7GHz, the paper reports maximum Q factors for the primary and secondary coils of 1.65 and 1.45, respectively. With a footprint as little as 0.003mm2, the devices were stable until annealed at 350°C.
For the samples they fabricated at a turns ratio of 2.2:1, the authors report an index of performance of 235, which they claim represents a 47% improvement over the best on-chip planar counterpart reported so far for the same turns ratio. What’s more, the devices’ coupling efficiency and performance increase as the turns ratio scales up, making the devices easy to scale down while being fully compatible with all planar semiconductor processing, including CMOS and MEMS technologies.
What about the practical packaging of these devices as standalone or embedded RF transformers?
The researchers measured the RF transformers’ stiffness, orders of magnitude larger than that of a suspended MEMS high-Q factor spiral inductor with X-beams (taken for comparison), meaning the structure shows some degree of mechanical flexibility and could easily withstand external forces during packaging or shocks with a large g force. When designed with more turns for higher inductance, the S-RUM on-chip transformers could be made even more robust.
University of Illinois – www.illinois.edu
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :
