editor's blog
Subscribe Now

Small single-package IMU

Bosch-Sensortec recently announced a new integrated IMU, the BMI055.

Which, amongst other things, brings up the question: exactly what is an IMU? While researching this for a gyroscope article couple of years ago, I found that the term (which stands for “inertial measurement unit”) was used to refer generically to a class of sensors that use some type of inertia as a way of sensing motion. That inertia might be linear (using an accelerometer) or rotational (using a gyroscope).

The definition Bosch-Sensortec used differed from that, and as I look around now, I see other usage that is similar: an IMU is a combination of sensors – in particular, accelerometers and gyroscopes – for detecting motion. (Some so-called IMUs also include other sensors like magnetometers and possibly even a pressure sensor/barometer, for a so-called 10 degrees of freedom – 6 of which degenerate to 3). Whether this represents a change or simply varying definitions is unclear to me (it’s hard to recreate the internet of a couple years ago). Nonetheless, the term is, to some extent, overloaded; the combo definition seems to predominate now.

While we’re on definitions, you might think of a magnetometer, when used in a navigation application, as a compass (or eCompass) by analogy to an old-school needle compass, which is simply a magnetometer. But that’s not how the MEMS version is defined: a MEMS compass is the combination of an accelerometer and a magnetometer.

To be clear, Bosch-Sensortec announced what they claim to be the smallest combination accelerometer/gyroscope available. It is a multi-die integration (both with respect to the MEMS sensors and the accompanying ASICs); the size advantage comes from housing them in the same package.

As to whether those dice might ever merge, they said that it might happen, but that it’s more likely that the ASICs and MEMS chips will independently merge first, possibly followed by full MEMS/CMOS integration.

They’ve added a power-saving feature through this integration: the accelerometer can wake up the gyroscope. Gyros are notoriously power-hungry; you have to keep the proof mass moving (unlike an accelerometer). So the BMI055 allows the gyro to be turned off. Which isn’t a first, but they’ve sped up the wake-up time from a more typical 30 ms to 10 ms. This is intended to allow the gyro to be woken by the accelerometer without it taking so long that the gyro misses an event. The effect is to cut power in half.

The combined unit comes with free fusion software. There have been two ways of approaching fusion software: using either “tight” or “loose” coupling. Loose coupling means that the data from each sensor is independently processed to some degree before being presented for munging with the output of other sensors. Tight coupling performs the fusion with the raw data from the sensors.

Loose coupling is easier to do (and less reliant on the low-level data format of a sensor), but it’s less accurate. Tight coupling provides a more accurate result, but is more complex and needs to work at the lowest data level (which ties it more closely to the specific sensor).

Bosch-Sensortec uses both: where loose coupling provides sufficient accuracy, they use it, reverting to tight coupling when necessary. Where they make that cut is something they’re keeping to themselves.

You can find more information in their release

Leave a Reply

featured blogs
Nov 22, 2024
We're providing every session and keynote from Works With 2024 on-demand. It's the only place wireless IoT developers can access hands-on training for free....
Nov 22, 2024
I just saw a video on YouTube'”it's a few very funny minutes from a show by an engineer who transitioned into being a comedian...

featured video

Introducing FPGAi – Innovations Unlocked by AI-enabled FPGAs

Sponsored by Intel

Altera Innovators Day presentation by Ilya Ganusov showing the advantages of FPGAs for implementing AI-based Systems. See additional videos on AI and other Altera Innovators Day in Altera’s YouTube channel playlists.

Learn more about FPGAs for Artificial Intelligence here

featured paper

Quantized Neural Networks for FPGA Inference

Sponsored by Intel

Implementing a low precision network in FPGA hardware for efficient inferencing provides numerous advantages when it comes to meeting demanding specifications. The increased flexibility allows optimization of throughput, overall power consumption, resource usage, device size, TOPs/watt, and deterministic latency. These are important benefits where scaling and efficiency are inherent requirements of the application.

Click to read more

featured chalk talk

Outgassing: The Hidden Danger in Harsh Environments
In this episode of Chalk Talk, Amelia Dalton and Scott Miller from Cinch Connectivity chat about the what, where, and how of outgassing in space applications. They explore a variety of issues that can be caused by outgassing in these applications and how you can mitigate outgassing in space applications with Cinch Connectivity interconnect solutions. 
May 7, 2024
39,296 views