The challenges of implementing MEMS in edge devices.
The world on the edge of the IoT provides a rich microcosm to explore. While much of the attention paid to IoT is on big data applications in the cloud or the world of aggregating massive data provided by the real-world edge devices in the wild though cellphones and gateways, the edge devices themselves present a cornucopia of design challenges and exciting applications. Sensors and actuators make up the edge of the IoT where the digital Internet crosses over to interact with the analog real world.
One way to describe the IoT is to use a three-tier model. “Life on the Edge of IoT” is where sensors and actuators interact with the real world – across farms, cities, industries, oceans, and as wearable devices that people interact with on a daily basis. Data from the edge devices can be aggregated by sensor hubs, gateways, or smart devices like smartphones. The aggregated data is then sent across the internet to Cloud servers where applications interact with the data and data analysis takes place.
Let’s take a look at a simple IoT edge device:
This sample device contains a MEMS sensor that interfaces to the Internet. The sensor signal is sent to an analog signal processing device. The output connects to an A/D converter to digitize the signal. That signal is sent to a digital logic block that contains a microcontroller or a microprocessor. The sensor telemetry is sent and control signals are received by a radio module that uses a standard protocol such as WiFi, Bluetooth, or a custom protocol. The radio transmits data to the Cloud via a gateway or to a smartphone. If the sensor is an accelerometer, we have an edge device that can transmit velocity and orientation data. Because the device contains a CPU, that data can be analyzed before transmission.
There are many uses for a device like this. An example that you might not have thought of is a system for monitoring whether or not prized horses are moving around and what position they are in. Apparently, if a horse stops moving or lays down for long periods of time, that can indicate a health issue that must be addressed immediately.
A casual glance at our sample IoT edge device points out a real design challenge: the design brings together the worlds of digital, analog and mixed signal (AMS), RF, and MEMS design. Each of these individual design domains brings challenges. Combining them together, often on the same IC, with stringent power, area, and cost requirements can seem impossible. IoT edge design requires that all four domains are designed and work together, especially if they are going on the same die. Even if the components are targeting separate dies that will be bonded together, they still need to work together during the layout and verification process. The design team needs to capture a digital, AMS, RF, and MEMS design, layout the chip, and perform both component and top-level simulation.
One of the most challenging aspects of IoT design is implementing the MEMS device. This challenge mainly centers on model accuracy and the fact that, unlike the IC design ecosystem, the MEMS ecosystem can be very dependent on the foundry that you work with. Of course, another big challenge is integrating the MEMS device with digital and AMS electronics.
Successful design of highly-integrated IoT systems requires simulating MEMS components together with the peripheral circuitry to verify functionality at the system level. For example, self-sustained oscillators contain both mechanical and electrical components. Even though the mechanical oscillators can be designed independently from the IC using traditional mechanical analysis tools, closed-loop simulations with the control and feedback circuits are still necessary to test the design at the system level and this can reveal design defects or instabilities that are hard to find in isolated device-level simulations.
To simulate environmental effects on the MEMS device, you could create a 3D model of the device using a 3D analysis tool and then analyze its dynamic response using a finite element analysis (FEA) tool. But, to fabricate the MEMS you need a 2D layout mask and deriving a 2D mask from a 3D model is error-prone and difficult to validate. A better approach is to follow a mask-forward flow that results in more confidence that the actuator will not only work correctly but that it can be successfully fabricated:
This mask-forward MEMS design flow starts by creating the 2D mask layout in the Tanner layout tool
L-Edit. Then, use the SoftMEMS 3D Solid Modeler (integrated within L-Edit) to automatically generate the 3D model from those masks and a set of specified fabrication steps. Perform 3D analysis using your favorite finite element analysis (FEA) tool and then iterate if you find any issues. Make the appropriate changes to the 2D mask layout and then repeat the flow. Using this mask-forward design flow, you can converge on a MEMS device that you are confident can be fabricated correctly because you creating the 3D model directly from the masks that will eventually be used for fabrication, rather than trying to work backwards from the 3D model.
FEA tools use conventional numerical analysis methods for simulations in the mechanical, electrostatic, magnetic, and thermal domains. But, these tools can have unacceptable run times and sometimes cannot converge on a solution. To alleviate these limitations, Reduced Order Modeling (ROM) bridges the gap between traditional FEA tools and electrical circuit simulators. Designers can use the generated reduced-order models in combination with traditional finite element models in a mixed mode to speed up FEA simulations. Or, they can be converted into analog hardware description languages, such as Verilog-A, using the SoftMEMS Compact Model Builder and then exported into electrical circuit simulators as black boxes for system-level co-simulation.
If you are visiting Austin for DAC during the week of June 6th, be sure to stop by the Tanner booth (number 1828, across from the DAC Pavilion) to see the IoT edge design flow in action. In the meantime, learn more about the challenges of IoT edge design with this whitepaper.