Herzlich Willkommen beim beam-Verlag in Marburg, dem Fachverlag für anspruchsvolle Elektronik-Literatur.


Wir freuen uns, Sie auf unserem ePaper-Kiosk begrüßen zu können.

Aufrufe
vor 7 Jahren

6-2016

  • Text
  • Komponenten
  • Radio
  • Technik
  • Filter
  • Oszillatoren
  • Quarze
  • Emv
  • Bauelemente
  • Filter
  • Signale
  • Messtechnik
  • Wireless
  • Modulation
  • Software
  • Simulation
  • Messung
Fachzeitschrift für Hochfrequenz- und Mikrowellentechnik

RF & Wireless Figure 3:

RF & Wireless Figure 3: Expanded view of full sensor currents will confirm that parasitic coupling exists, but they do little to identify and fix the problem. • Time-domain simulation results generated using the Finite-Difference Time- Domain (FDTD) method allow engineers to see where the coupling occurs and redesign the layout to prevent it. • Current distribution on ground planes. At 25 and 77 GHz, grounding structures are no longer equipotential surfaces. As seen in Figure 2, the ground plane has strong currents – 10 dB below the maximum – on its edges that need to be considered during the design. • Effects of secondary sources like the local oscillator (LO) line. Secondary sources can couple onto other conductors and even generate unintended radiation. Both of these problems can be identified and quantified through simulation. Analysis of RF Board with Sensor A properly designed RF board is an indicator of future success, but there is more work to do before OEM specifications are met. For starters, the RF board needs to be placed in the sensor’s case and covered by a radome. These structures will change the performance of the antenna. An expanded view of the full sensor model is shown in Figure 3. The model contains the radome, RF board, digital board, packaging, data connector, and sensor case which bring the overall dimensions to 106 x 63 x 21 mm. An FDTD simulation includes all of the model’s complexity so no simplifications to the sensor are needed. This gives engineers a more realistic picture of how the sensor will perform if it is built. For example, the data connector is a relatively large structure that is not in the immediate proximity of the radiators so one may remove it from the simulation to reduce RAM requirements. Including it in the simulation, however, increases accuracy because energy couples onto its pins and is then radiated from these dipole-like structures. The radome is one of the most important structures of the sensor because it sits directly in front of the antenna arrays and can have a large impact on the antenna’s radiation patterns. By parameterizing the imported radome model, its design can Figure 4: Far-field gain results for different radome thicknesses Figure 5: Sensor mounted behind fascia be honed to meet the desired performance. The results from a basic parameter sweep of the radome thickness are shown in Figure 4. Parameterizing the geometry and setting up the simulations can be accomplished in minutes, which is significantly less than the time it would take to create and measure five different radomes in a lab. Analysis of Sensor Behind Fascia Ultimately, the installed sensor’s performance will dictate the sensor’s ability to accurately identify targets. Here, engineers are interested in understanding how mounting brackets, paint color, and curves in the fascia will degrade the antennas’ radiation pattern. Models of fascia, obtained from an OEM, can be imported into XF like any other CAD model. Figure 5 shows an example of a fascia model included with the sensor. The corresponding simulation space is 195 x 204 x 74 mm. Application and design engineers benefit from simulation because it allows them to identify the optimal placement of a sensor behind a fascia or troubleshoot problems with an installation. Similar to parameterizing the radome’s thickness, the sensor’s location relative to the fascia can be parameterized. This, coupled with the ability to visualize trapped modes between the radome and fascia, allow engineers to understand which aspects of 68 hf-praxis 6/2016

RF & Wireless the installation are impacting the results. Run Time and Memory Requirements The ability to complete simulations in a timely manner is an important factor in determining the usefulness of a simulator. Combining graphics processing unit (GPU) technology and FDTD allows engineers to perform multiple design iterations much faster than previously possible. The chart in Figure 6 compares the memory requirements and run time for three simulations: RF board only, RF board with full sensor, and sensor with fascia. The microstrip structures in the RF board generated a minimum cell size of 0.037 mm and the grid definition around the RF board was maintained as the problem size grew with the additional geometry. For the benchmark, XF utilized four NVIDIA GPUs with the Kepler architecture. GPUs provide a massively parallelized computing platform with 2,800 cores per card. The FDTD algorithm efficiently utilizes this parallelization and 50x speed improvements over CPUs are commonly achieved. This combination enables full sensor with fascia simulations to complete in under seven hours. Summary Engineers are pushing the limits of sensor technology in order to meet OEM requirements and Figure 6: Run time and memory requirements improve transportation safety. FDTD simulation provides the tools they need to understand an antenna’s performance. At the board level, sources of parasitic coupling or variations in ground potential can be identified and mitigated. This type of analysis carries through to optimizing radome structures and determining the best location for a sensor behind a fascia. Coupled with GPU technology, engineers are able to perform this analysis in hours, thus reducing the overall development time. Antennas Small Flexible FPC Antennas for the 3G Mobile Data Market Armata antenna SRFC011 folded and inserted in OBD II automotive device Armata Antenna P/N SRFC011 Armata, P/N SRFC011 folded Antenova Ltd. is targeting the fast increasing 3G mobile device market with two new flexible printed circuit board antennas for 3G and MIMO. Both antennas operate at GSM850, GSM900, DCS1800, PCS1900 and WCDMA2100 and will provide full coverage of the 3G bands in all world regions. The two new FPC antennas are designed to be mounted inside a small mobile or wearable device, and are supplied with a cable and a connector so that they are very easy to integrate into a design. At just 0.15 mm thick, they can be curved or folded, to fit into tight spaces. They have a ready applied adhesive with a peel-off strip making them easy to fix inside the casing of a product. Antenova is offering these two very similar antennas in two sizes because the performance of the antennas is related to the size of the PCB where they are used. The first antenna, named Avia, part number SRFC025, measures 71 x 12.5 x 0.15 mm, and is suitable for use in products like smart meters, Femto/Pico base stations, tracker devices, remote monitoring, and POS terminals. The second antenna, Armata part number SRFC011, measures just 30 x 28.5 x 0.15 mm and is suitable for smaller applications such as OBD-II units and other in-vehicle devices. Both antennas come in a choice of three cable lengths, with options to customise the cable length and connector type. The Avia and Armata antennas are available to order from Antenova now. For more information, and to see Antenova’s full range of FPC and embedded antennas, please visit www.antenova-m2m.com. ■ Antenova Ltd. www.antenova-m2m.com Avia Antenna P/N SRFC025 hf-praxis 6/2016 69

hf-praxis

PC & Industrie

© beam-Verlag Dipl.-Ing. Reinhard Birchel