In today’s smart devices, the RF Front-End (Also called RFFE) represents all the circuitry between High frequency data received from Antennas to near-zero frequency baseband signals.
As the technology is now moving to 5G, those RFFE circuits will need to handle much more data at a very high speed while handling always higher frequencies.
RFFE generally includes many components such as power amplifiers, filters, clocks, switches and low noise amplifiers.
Now the question which is interesting to ask is the following:
Why do RF Filters play such a big role in the performance of the New generation of Mobile Devices RF Front-End Circuits?
Today’s mobile devices receive signals from multiple sources and all those signals have different frequencies. A normal smartphone now has 4-5 Antennas to receive signals from 3G, 4G, GPS, Bluetooth, Wifi…
In order to be able to receive all of those signals and treat them efficiently, the RF Front End needs to separate very well each frequency band those signals belongs to. For that, RF Filters which are also pass-band filters are used!
The RF filters play a key role as they are the main components allowing the accurate and precise selection of a very narrow range of frequencies.
For 5G to work, the system architecture of those RFEE circuits will have to be re-engineered to handle more than 100 RF filters in the extremely limited space allowed by today’s smartphones. In current 4G smartphones, a maximum of 50 RF filters are required, so 5G will likely require 2-3 or even 4 times more RF Filters in the same amount of space!
Engineers will have to keep cost low, while making this new architecture power efficient and easy to manufacture in large quantities.
What are the current most important challenges to design RF Filters?
In order to reduce the size of current RF Filters while improving their performance, some innovation in the design process will have to be considered. One of the problems of the current design process for example is that engineers still use 1D mathematical models to represent the filter characteristics.
Those 1D models fail to provide accurately many needed design parameters, such as:
- The source of parasitic spurious modes due to lateral vibration modes
- The real effect of mounting conditions and topology on device performance
- Energy trapping effect based on electrodes and stack layer configuration
- Effect of temperature on the vibration of the fundamental and spurious modes
- Activity dips occurring due to the cross-over of spurious modes and fundamental drive mode as a function of drive level power
Engineers and Filter Designers will have to adopt new methodologies of 2D and 3D simulation which will allow them to test more designs before even starting physical prototyping. Up to not long ago, 2D and 3D simulation of filter models was unthinkable, because those models are just too large and computational intensive!
This prevents designers to create digital Design of Experiments (DoE’s) as they are forced to analyze only a few design iterations.
Thanks to OnScale and the ability to run an unlimited amount of large models on the cloud, this “simulation barrier” is now removed.
What are the Key Performance parameters for a winning RF Filter Design?
First of all, we want a low insertion loss of the desired signal in the passband, which means that we want to decrease as much as possible the power of the input signal lost after going through the RF Filter. Then, we want also to attenuate the undesired interference in the stopband as much as possible. One of the most important key performance parameters for an RF filter is its quality factor Q which determines the filter insertion loss and filter roll-off sharpness. The second key performance parameter is the coupling factor, which determines the sustainable filter bandwidth.Both the Q and coupling factor are directly dependent on the resonator design, which makes resonator design the key to achieve the required filter performance!
Thus, the trick to achieve good filters is to connect high-Q resonators of various impedances having the desired bandwidth.
Why is the Q factor so important?
While there are several ways to calculate the Q factor for filters and resonator devices, a simple way is to divide the center frequency at peak response and to divide it by the difference of the upper and the lower frequencies at -3dB power level.
From this curve, we understand graphically that increasing this Q factor will lead to sharper filter response with less insertion loss and a better ability to filter very specific frequency bands.
Why OnScale for RF Filters?
Solving filters in 3D provides a wealth of insights that cannot be obtained otherwise, such as:
- KPIs: Insertion loss, Q, resonant frequency, etc.
- Mode Analysis: Dispersion, elimination of spurs, etc.
- Device-level Effects: Coupling to bulk silicon, etc.
- System-level Effects: Impedance matching, packaging, thermal, electrode loading, etc.
Want to know more about how to simulate all that?
Check our Simulation Guides for RF Filters
You can also check our Simulation-Driven Optimization of 5G RF MEMS Filters White Paper
