Pressure Transducers
In industry, we deal extensively with pressure. We do in our day-to-day lives as well. For decades, many buildings had a barometer, a pressure sensor, to provide information about weather patterns. Small changes in atmospheric pressure can have a massive effect on whether the sun shines, or if we are covered in rain. But how do we actually measure it?
To understand how we measure pressure, we must first think about what it is. Pressure is force spread evenly over an area. When we say a fluid is ‘at a pressure’, what we really mean is that the internal energy of the fluid is such that it exerts a force of that pressure on its surroundings. It does so in all directions evenly.
So, to measure that, we just need to work out how to measure that force, right? But how do we do that? We’ve looked at a number of transducers over the past months; none of them can directly convert a force to a signal. But we did look at one that converts acceleration to a signal. Through F=ma, we know that force and acceleration are directly proportional to each other. You might remember from that article on accelerometers that they are essentially all built around the deflection of a mass on a spring. So, could we use the same principle again? Can we use deflection to measure pressure?
One of the big challenges we run into in this case is that pressure is everywhere. Let’s say we tried to measure the deflection of a plate in response to pressure; we encounter the issue that pressure pushes on both sides of the plate. If we want the plate to deflect, we need to produce a difference in pressure across both sides of the plate. We can swap the plate for a rubber diaphragm or other flexible membrane, or otherwise seal a chamber such that the pressure on one side of the plate is different from the other. We could have one side be the pressure we want to measure and one side be atmospheric, or we could have one side be a fixed pressure, maybe 1 bar (about the same as atmospheric pressure), and the other be open to the atmosphere if we want to measure variances in atmospheric pressure.
So now we just have to measure the deformation of the membrane. If we just need to read the pressure off, we can link the membrane, or another moving component, mechanically to some sort of indicator. This can be a classic rotating dial or, far more simply, a rod that gets pushed out against a spring, a common design used to check tyre pressure; these designs are decades if not centuries old. Another method, instead of a solid object that deforms, is to fill a tube with a liquid, often mercury as it doesn’t evaporate. Pressure pushes on a pool of the liquid and pushes it up a low-pressure tube in response. The height the liquid makes it up the tube can be read off as the pressure. In fact, this has given rise to one of the units of pressure, millimetres of mercury.
The astute among you will have noticed that if we are using a sealed chamber as a ‘reference’ pressure, when the moving object, be it a fluid, a diaphragm, a piston in a cylinder, etc., moves, the size of that chamber changes and the pressure in that chamber will also change. This changes the ratio between the pressures. However, if the deformation or movement of the object is linear in response to a pressure difference, then we can still calibrate a linear scale as the change in pressure due to volume change is also linear. If the response becomes non-linear, we can compensate by using a very large chamber such that the change in pressure of the reference due to movement of the object is essentially nothing. This issue does not arise for pressure transducers that measure relative to open atmosphere, which is part of why they are so common.
When we want to get an electric or electronic signal, things become slightly more complicated. Some will remember that one of the transducers we looked at early in the series was the proximity probe, and that was one of the ways of converting a displacement to an electrical signal, a signal that we can then convert to digital using a digital to analogue converter. Indeed, there are some pressure transducers that use a capacitive approach, measuring deflection of a diaphragm by measuring changes in capacitance.
Another very common way to do it is using another transducer we’ve looked at: strain gauges. By having strain gauges attached to a diaphragm, you can measure its deflection as induced strain in the gauges. The standard issues with temperature response arise, but there are many compensatory techniques that can be used by the designer of the transducer to compensate for this when designing the circuitry that produces a signal.
The vast majority of pressure transducers available today produce either an analogue signal, be that volts or amps, typically used in industry, or a digital signal from MEMS sensors, often used in consumer electronics. However, with the increasing quality of MEMS sensors and the increasing digitisation of industrial systems, this line is blurring. It is very rare to directly measure the strain gauge resistance change in the same way that we would use a strain gauge typically. This allows a lot of signal conditioning to be designed into the transducer, such that many of the transducers available today are highly accurate and high-quality devices with errors in the region of 1% for comparatively low-cost sensors.
Flintmore have worked with pressure transducers on many occasions, both industrial grade and MEMS sensors. They can be a bit odd to work with, depending on if they measure relative to ambient, to 1 standard atmosphere, what percentage of their range is actually being used, and various other factors. However, generally, we have found that they belong in the same class as proximity probes; they just work, they do their job.
One of their oddities is the lack of a way to sanity check them. A known pressure source is quite rare to find, other than the atmosphere, which is typically so close to a zero for a pressure transducer that it’s not useful. In this situation, we can attach a mechanical gauge, which has a heap of zero drift and calibration issues, to give us an easy check with the given caveats. Another approach is to get a certified transducer that has a set calibration and certificate of use. It has been tested and verified as accurate to a given range. Often, when buying an industrial transducer new, it will come with a certificate. Certificates that have expired can usually be renewed for a fee with the sensor being rechecked.
Do you have a job requiring the measurement of pressure? Flintmore have experience in the field and the expertise needed to get the job done! Get in contact today!
Further Reading
The wikipedia article on pressure measurement is quite detailed, and, largely, factually accurate. It gives a great overview of the subject.
https://en.wikipedia.org/wiki/Pressure_measurement
Omega engineering do fantastic articles on transducers, if sometimes a little brief. They have two great articles on pressure transducers that are a great alternative to ours.
https://www.omega.co.uk/prodinfo/pressure-transducers.html
https://www.omega.com/en-us/resources/pressure-transducers-how-it-works
MEMS sensors are a subject in and of themselves, this article from polytec gives a great insight to the crossover of MEMS technology and pressure transducers:
