avatar

EYE on NPI: XENSIV™ BGT60LTR11AIP and Radar Shield2Go

Adafruit Industries
Adafruit Industries
Episode • Feb 23, 2023 • 11m
This week's EYE ON NPI is sixty-giga-cool with 60 giga-Hertz of Doppler radar sensing. It's the Infineon XENSIV™ BGT60LTR11AIP Radar and Shield2Go Dev board (https://www.digikey.com/en/product-highlight/i/infineon/xensiv-bgt60ltr11aip-radar-shield2go), featuring Infineon's super-low-power 60 GHz Doppler radar sensor with antenna-in-package (AIP). This device is an ultra-tiny 60 GHz radar-on-a-chip sensor that even includes the normally-large antenna's seen in most Doppler radars baked right into the top!
RADAR is literally "Radio Detecting And Ranging" - a way of bouncing 'high frequency' electromagnetic waves off of objects to detect the reflection. We're familiar with radar (https://en.wikipedia.org/wiki/Radar) in a few contexts: originally developed for military use, it worked great for detecting metal things in the sky (airplanes) as they would approach a city to possibly bomb it. We still use it a lot in military contexts: detecting planes, bombs, boats and submarines. It's also super handy in domestic contexts like air traffic control, if you can detect planes you can verify their positions and make sure each one comes in or out of the airport without collision. Radar is also useful for weather pattern detection (although this sensor isn't good for that purpose, just a fun fact for you to know).
Another common purpose for radar you're probably familiar with is the dreaded speeding radar gun (https://en.wikipedia.org/wiki/Radar_speed_gun) a small-enough-to-be-handheld radar device that can be pointed at a car to determine its speed without requiring a difficult-to-calculate "time between two points" measurement. These reliable devices are apparently now are replaced with LIDAR based ones!
Most folks who have used IR sensors, Time of Flight, or sonar sensors can understand how to detect an object by counting the delay between sending a signal vs when it's received and dividing by the wavelength. How can you determine velocity - normally you'd need to detect location twice, then determine the time between the two locations whereas radar can do it one measurement. This is thanks to the 'Doppler Effect' (https://en.wikipedia.org/wiki/Doppler_effect) the same scientific principle that causes ambulance sirens to be higher pitch as they approach and lower pitch as they drive away from you due to the 'bunching up' of the waves as the wave source moves. Likewise with Radar, the frequency will shift slightly depending on the speed of the object, which can be detected by calculating the minute variation in frequency.
What's particularly neat about this sensor is that, compared to other radar modules that have a trace PCB antenna, the antenna is embedded right into the chip itself. Compare that with the BGT24 modules that come on a PCB with an 'Arduino shield size' antenna (https://www.digikey.com/short/dn48cfhr) The BGT60LTR11AIP comes as a BGA component, with a lot of pins! Thankfully, Ultra Librarian has a package footprint for this part that Digi-Key helpfully provides, so you can import and get started with integration quickly (https://app.ultralibrarian.com/details/2071C153-54B1-11EC-9033-0A34D6323D74/Infineon/BGT60LTR11AIPE6327XUMA2?ref=digikey)
There's also a few different development boards available, including a tiny module 'shield' SHIELDBGT60LTR11AIPTOBO1 (https://www.digikey.com/en/products/detail/infineon-technologies/SHIELDBGT60LTR11AIPTOBO1/16343839) that requires only 4 pins: power, ground, output direction and output motion detect (https://www.infineon.com/cms/en/product/evaluation-boards/demo-bgt60ltr11aip/) There's also a fancy "SHIELD2GO" board (https://www.digikey.com/en/product-highlight/i/infineon/xensiv-bgt60ltr11aip-radar-shield2go) that has all of the user-useful GPIO and SPI interface accessible, and a full demo dev kit that also includes the mini shield. (https://www.digikey.com/en/products/detail/infineon-technologies/DEMOBGT60LTR11AIPTOBO1/13563722)
There are two ways to integrate the sensor: via simpl