Analog Angle Blog

When sensors are easy, but sensing is hard

If there’s any area of electronics signal chain that has remained stubbornly analog, it’s sensors. Yes, you can get a sensor output and digitize it very close to the sensor itself, but the sensor itself is almost always analog, as is the physical parameter being sensed. That’s the reality of the physics unless you get down into atomic and subatomic particles that exist in discrete quantum states of one type or another.

Still, in the broadest sense, there’s no doubt that today’s electronics have made it much easier to capture the analog output from the sensor, amplify it, digitize it, correct and compensate it, and then analyze it. But the sensor element or arrangement itself is still analog, and often, trying to capture information is easier said than done.

It’s one thing to have a basic sensor for a physical variable such temperature or pressure, but it’s often another much-more difficult challenge to install that sensor such that it can actually make accurate and robust measurements.

That’s a challenge often faced in vehicles in the weightless environment of orbit or space travel. It’s obviously critical to know with accuracy the amount of liquid fuel left, but the sloshing and poor behavior of the fuel, including “breakup” in multiple unconnected sub-volumes in that setting make this very difficult.

Traditional methods such as “weighing” the fuel’s mass or measuring its pressure are not as accurate as desired; one alternative is to use internal contacts as the fuel is used, but this adds mass and brings new technical issues. A widely used alternative is basic “bookkeeping,” which tracks how much fuel was used in a given burn and then subtracts this from previous value. It’s reasonably accurate when the tank is full, but accuracy declines as the fuel is used, and there are problems with cumulative error buildup of those numbers.

Now, researchers led by NASA technology transfer manager Manohar Deshpande and the National Institute of Standards and Technology (NIST) have developed an experimental system that uses a sophisticated 3D-imaging technique called electrical capacitance volume tomography (ECVT). In this approach, electrodes emit electric fields and measure the target capacitance (see figure below).

The interior of the prototype fuel tank is lined with flexible electrodes that act as plates of capacitors. The measured capacitance values are determined by the mass of fluid in the tank and its locations; shown here is first-level test that is done using a suspended balloon filled with a heat-transfer fluid (standard HT-90) in place of real—and potentially dangerous—rocket fuel. By combining the measurements of the electrode-pair readings, followed by advanced 2D and 3D algorithms and analyses, the gauge can estimate the location and volume of the balloon. Source: NIST

“It’s not a simple single-capacitor arrangement; instead, the tank is lined with an array of sensor electrodes and the capacitance is measured across multiple sensor pairings,” said Nick Dagalakis, a NIST mechanical engineer on the project team, “We measure the difference in transmission for every possible sensor pair, and by combining all these measurements, we know where there is and isn’t fuel and create a 3D image.”

The team used “soft lithography” to create the capacitive-sensor array that lines the tank, which is similar to creating a PCB, except they print the ink pattern on a flexible plastic backing such as Kapton and then etch away the exposed, unwanted copper. The matrix of capacitance data is used to produce a set of 2D images that map the location of fluid throughout the length of the tank and these, in turn, result in a 3D rendition of the fuel in the tank with a volume and thus mass that can be calculated.

It’s interesting and perhaps somewhat ironic that this highly sophisticated approach begins with a set of sensor readings of the very basic parameter of capacitance before it builds on it via multiple sensors and readings plus advanced data analysis. The well-written NIST news article “NIST Designs a Prototype Fuel Gauge for Orbit” provides an overview of the project and also discusses, in brief, how engineers tested it. There’s also a detailed technical paper “Flexible Assemblies of Electrocapacitive Volume Tomographic Sensors for Gauging Fuel of Spacecraft” published in the Journal of Spacecraft and Rockets, but it’s behind a subscriber paywall. However, a search for images using the title of the technical paper will bring up some medium-resolution images from that paper.

Have you ever been in a situation where the basic physical parameter of interest seemed easy enough to sense, but the transition from “sensor” to “sensing” was a challenge? Did you solve the problems with a sophisticated, radical technique or a more-basic “brute force” keep-pushing approach? Was the advanced solution just too complicated, or more of a headache than the problem you were trying to solve?

Related Content

0 comments on “When sensors are easy, but sensing is hard

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.