A progress review meeting was held at VSL on 14 – 15 February 2023. At the end of Day 2, a teleconference was held with the Technical Advisory Group to keep them informed of progress to date.
The relevant key technical achievements of the project to date with respect to the project objectives are:
1. Traceable Measurements of Active Devices
TU Delft has provided ISS mono-layer (fused silica) and multi-layer substrates to partners, and EM simulations of the corresponding structures have been completed. The transfer standards for power calibration are identified, with diodes as suitable devices for RF power calibration, as successfully demonstrated in the WR5 band. Working with the project’s Technical Advisory Group (TAG) members, the partners have investigated European commercial semiconductor foundries; a list of European GaN semiconductor manufacturers has been produced; and the chosen active GaN devices have been procured by the project.
Test structures and calibration standards have been designed and fabricated, and the devices have been characterized under linear and non-linear operations with 50 ohm and non-50 ohm load terminations. Thermal device characterization using a thermo-reflectance system and Raman measurement methods have been completed.
A behavioral model is developed for vector modulators used in active impedance tuners, allowing accurate uncertainty propagation for uncertainty estimation in load-pull measurements. Furthermore, measurement software is being developed for automated load-pull measurements at VSL.
An investigative study has been conducted into the sensitivity of the transfer function between IQ mixers’ RF-to-LO ports for the DC IQ driving signals. The investigation led to the development of an IQ mixer behavioral model, allowing accurate calculation of the IQ signal voltages for the desired impedance value.
A 25-um-pitch quartz-glass capillary probe has been designed, and the manufacturing process is finalized. The first batch of probes is undergoing validation tests and will be compared to other commercially available probe technologies for a feasibility study of the newly proposed probe type.
2. Signal Integrity (SI) and Power Integrity (PI) of FPGA Chips and PCBs
Ten different on-wafer coupled test structures have been designed using thin-film microstrip line (TFMSL) technology on SU-8 resin as a thin-film substrate. The test structures have been theoretically validated in terms of their S parameter response using 3D EM simulations, which have been compared to circuit simulations and are being manufactured. The test structures have been designed to focus on reflection, crosstalk, and mode conversion effects. Using circuit simulation, S parameter (SI) analysis has been carried out in order to study the effect of losses and line mismatch.
From the results of this, a model has been developed for deriving the power distribution network (PDN) impedance, which can be used for future studies of the resonances of the decoupling circuit for PI evaluation.
3D EM simulation has also been carried out to investigate the interaction between on-wafer probes and components when probing differential transmission lines. The simulation was performed on a generic probe model, and the real probe model has been developed to check for any parasitic probe effect on the accuracy of the measurements.
Suitable reference PCBs that can be used to evaluate the traceability of de-embedding procedures have been designed. The test fixtures, calibration kits, and DUTs have been implemented for both single-ended and balanced microstrip line configurations.
Several FPGA evaluation boards have been purchased, and an FPGA-based VNA has been implemented and used to measure S-parameters of PDN transmission tracks. The PCB has then been modified to allow for a reference VNA to carry out the same measurements and compare the results.
3. Environmental Testing of Electronic Circuits
A test printed circuit board (PCB) on FR4 type material (a composite material composed of woven fiberglass cloth with an epoxy resin binder that is flame resistant) has been designed. The test PCB consists of four metal layers and is designed to be fitted with end-launch connectors for electrical testing at frequencies up to about 10 GHz. A first version of the test PCB has been fabricated, and some preliminary tests have been performed on it. Improvements have been made to the design of the test PCB, and a second version has subsequently been fabricated. Environmental age testing for PCBs has been analyzed from several standard documents to inform the definition of the test procedure.
Environmental chambers have been adapted for use with RF test equipment for PCB environmental in-operando testing. Suitable standards documents to inform the definition of the test procedure have been considered for PCB environmental in-operando testing.
Three types of commercial on-wafer planar circuits are being considered as candidate on-wafer test structures. Some commercial on-wafer planar circuits have been procured for testing. Further work has been done on establishing the test procedures for the environmental testing of on-wafer circuits.
4. Measurements Method for PIM in Communication Systems
The choice of industrial RF connectors and frequency bands has been agreed upon. The chosen industrial RF connectors were 4.3-10 and 7/16, and the frequency bands are around 2 GHz (B3 Band).
For the traceability of RF power measurement capabilities for 4.3-10 and 7/16, these have been developed at CMI and PTB. Using these new capabilities, preliminary measurements of high-power signals have been performed, and new CMC entries were obtained for high-power measurements for CMI up to 55 dBm (~300 W).
In order to establish traceable S-parameter measurements, work on industrial 4.3-10 and 7/16 connectors has begun, and a calibration kit has been purchased. A newly designed 4.3-10 to N-type adapter set is built and characterized. A preliminary uncertainty budget of PIM measurements for 4.3-10 connectors and a scalar measurement system has been established. As expected, the total measurement uncertainty strongly depends on the nominal PIM signal value: a lower PIM is more difficult to measure precisely.
The above-mentioned technical achievements have been published in seven accepted peer-reviewed papers. One of them has been published in the IEEE Instrumentation and Measurement Magazine and the remaining six papers have been published in the proceedings of the following conferences: 2022 IEEE International Symposium on Measurements & Networking (M&N), Conference on Precision Electromagnetic Measurements 2022, and Mulcopim 2022.
