Herbert Zirath

A D-band transition from silicon–germanium (SiGe) monolithic microwave integrated circuit to waveguide (MMIC-WG) is proposed in this letter. The transition includes a slot etched in the chip’s ground plane and a two-stage step in a metal waveguide. The slot works as a launcher for contactless electromagnetic field coupling, fed by a shorted microstrip line, while the step is an impedance transformer. As the assembly requires splitting the waveguide along the H-plane, glide-symmetric holes in a multilayer waveguide (MLW) are introduced as an electromagnetic bandgap (EBG) structure to suppress leakage, different from metal pins used in an H-split rectangular waveguide (RWG). As a proof of concept, a back-to-back (B2B) transition is fabricated based on MLW technology with an additional H-split RWG design as a reference. Measurement results show the proposed transitions have 2-dB insertion loss, in good agreement with the simulation. In addition, packaged receivers are also presented to validate the feasibility of the proposed transitions for sub-THz system integration.

This work proposes a methodology entirely based on processing S-parameters to determine the gate-to-source voltage-dependent mobility of charge carriers in the graphene field-effect transistor channel, without requiring any information about the material properties. Furthermore, regressions from experimental data of transistor arrays with different geometries are not required, thus avoiding uncertainties related to process variations and device-to-device measurement conditions. Hence, one key advantage of this method is its applicability to analyze the performance of different devices under the same operating conditions, or the performance of a single device under varying conditions. As part of the methodology, the effects of the parasitic series resistances associated with the source and drain access paths are considered, thereby overcoming the well-known disadvantage of direct-current methods, where the corresponding de-embedding is cumbersome. This method has been used to determine the carrier mobility in graphene within a gate-to-source voltage range, starting from the Dirac voltage and extending to the region where electron conduction dominates. A small-signal model with the extracted parameter values shows an excellent agreement with the experimental S-parameters up to 20 GHz for the dynamic response of different devices, including two devices that have not been used during parameter extraction. Throughout the development and application of the proposal, a mobility model accounting for the degradation caused by the transverse electric field has been considered.

A novel stand-alone generic analog predistortion (APD) MMIC is fabricated, implemented, characterized, and experimentally demonstrated at E-band (72.78 GHz). The APD has a wide-range tunability of both the fundamental and the 3rd-order coefficients, mapping directly in predistortion functionality to a 3rd-order memoryless polynomial digital predistortion (DPD) solution, with the output bandwidth of 62–84 GHz. Thus, the proposed APD topology takes the form of a stand-alone and generic module, suitable for the predistortion of different E-band power amplifiers (PAs). System-level characterization is performed using 16-QAM 250-MHz wideband modulated signals, following the ETSI spectral mask standard. Performance is measured through the mask-breaking output power (MBOP), error-vector magnitude, and normalized mean-square error (NMSE). Improvements of up to 3 dB in MBOP are observed. In addition to a demonstration of the tunability of the APD, the system performance comparison between implementations with a stand-alone APD, a stand-alone DPD, and a cascade of DPD and APD is provided. It is experimentally shown that with the advantage of an APD, similar predistortion performance to the counterpart DPD can be achieved.

This article presents a D-band cascode balanced up conversion mixer in a 130-nm SiGe process. The proposed mixer employs a double mixing technique in its cascode configuration, where the local oscillator (LO) and intermediate frequency (IF) signals are applied at the base of the common-emitter (CE) configured transistor, and the common-base (CB) configured transistor remixes the $f_{\mathrm {LO}}$ and $f_{\mathrm {IF}}$ signals at the collector of the CE transistor, generating additional RF signal power and thereby enhancing the overall mixing process. This up conversion mixer achieves a saturation output power of 5.9 dBm with a dc power consumption of only 69 mW. The measured output referred 1-dB compression point ( $\mathrm{OP}_{\text {1 dB}}$ ) is 4.8 dBm. The measured peak conversion gain (CG) of the mixer is 9.8 dB at 139 GHz, with 3-dB bandwidth exceeding 30 GHz for different LO/IF powers. The LO-RF isolation is below −40 dB across the 120–170-GHz range. The chip size is $620\times 360~\mu $ m². This mixer exhibits superior performance concerning output power and linearity as well as core chip area.

Graphene field-effect transistors (GFETs) exhibit negligible transconductance under two scenarios: for any gate-to-source voltage when the drain-to-source voltage is set to zero and for an arbitrary drain-to-source voltage provided that the gate-to-source voltage equals the Dirac voltage. Hence, extracting the channel and the parasitic series resistances from S -parameters under these conditions enables analyzing their dependence on the gate and drain biases. This is fundamental to assess the portion of the output resistance that is controlled by the gate. Besides, the drain bias dependence of the drain and source resistances is also evidenced. Within the proposal, resistive components accounting for the lossy nature of the gate capacitance are incorporated into the model, which exhibits a broadband correlation with experimental data. This avoids the series resistances to be considered as frequency dependent in the model.

A novel co-designed transition from InP monolithic microwave integrated circuits to silicon micromachined waveguides is presented. The transition couples a microstrip line to a substrate waveguide sitting on top of a vertical waveguide. The silicon part of the transition consists of a top and a bottom chip, fabricated in a very low-loss silicon micromachined waveguide technology using silicon on insulator wafers. The transition has been designed, fabricated and characterized for 220-330 GHz in a back-to-back configuration. Measured insertion loss is 3-6 dB at 250-300 GHz, and return loss is in excess of 5 dB.

Due to government and other agency regulations, microwave bands, which range in frequency from 6 to 42 GHz, often have a small bandwidth of a few hundred megahertz, which lowers the data rate for communications. On the other hand, high data rate communications can be realized in millimeter-wave (mm-Wave) frequencies (30–300 GHz). Tens of gigabits per second of bandwidth are available in mm-Wave bands, such as the D-band, from 110 to 170 GHz, and H-band, from 170 to 260 GHz [1] .