When it comes to measuring pressure sensors, the word "accuracy" becomes quite arbitrary. Accuracy isn't defined by any one set standard, and it becomes more of a qualitative measure than a quantitative one. Several factors contribute to the overall measurement of accuracy. However, these can be inconsistent across manufacturers, so it's important to understand these factors and how they suit your needs before you make the decision about the ideal pressure sensor.
Nonlinearity is usually the main characteristic considered. It describes the curve on a graph that the sensor would produce in comparison to the ideal line (which is, of course, straight). However, even with nonlinearity graphs, it's not always possible to carry out a direct comparison. A graph could be created according to limit-point adjustment or by using the best-fit straight line method. Therefore, all data isn't consistent and can't be fairly compared. The cycle of hysteresis is also seen within pressure sensors and can contribute to their (in)accuracy.
Another uncontrollable factor is nonrepeatability. This is defined as the difference between the largest and smallest measurements, taken several times under fixed conditions, and it is present in every pressure sensor. It's how you measure these results that's important. So long as you test different level sensors under controlled conditions, you can make a fair comparison and decide which is best for you. No sensor will be completely perfect, and errors are to be expected. You just have to choose a level sensor that has the correct degree of accuracy for your needs, based on any available data.
— Nick Murden is a keen writer who helps businesses put thoughts into words. He is currently writing for a number of engineering firms around the UK, including TC Fluid Control.
If marketing and other people would stop using these terms as if they were synonyms, there would be much less confusion. In that I do take issue with the fact that accuracy is defineded as the highes and lowest measurement from a specific point. This is precission. Precission measures the scatter in the data about a known point. Accuracy isspecifically the average of the given points and how close that value is to reality. All the points may have very high scatter, but if the average is correct, then you can call the measurement accurate, despite all the values that make up that measurement to be far off. You can have a device that is not very accurate over a sigle measurement, but is accurate over multiple measurements. For a device to be accurate over a single measurement, it also needs to be precise. Precission is the measure of the error band on the total measurement. The larger the error band, the less precise the sensor is. You can have a sensor, though that is precise, but not accurate. In each of these cases, there are ways of attempting to improve the accuracy. Some of them involve averaging the data, some of them just require a simple calibration value, some require a calibration table along with averaging. Many highly precise sensors will require calibration methods to take out effects from thermal offsets as well as manufacturing differences that cause part to part differences.
You are right, Aeroengineer1. Marketing sometimes has a lot to do with the lack of quality information on a pressure sensor. Likely, it's just a lack of knowledge about the differences you point out. Also, manufacturers don't like to advertise the weaker aspects of their product. I think this second reason happens more often than anyone admits.
Some larger companies have whole departments dedicated to determining all of these key aspects of a pressure sensor by in-house experimental test because the vendor information is incomplete or even incorrect. Some applications require that level of understanding because the application is that critical. For most, having this kind of rigor isn't practical, so we expect the company to provide their known limits of their device. Although some vendors will work with you on getting a better handle of the accuracy and precision, because this information may still be lacking for your particular application, the only fallback is the classical approach, lab or field testing.
This is a good post to remind people that something as simple as a pressure sensor requires some thought the accuracy or precision of that instrument. Knowing that environmental conditions (temperature being one) can shift the measurement is also one more thing to stay keen about. Instrumentation is an important consideration in design.
It is so true that accuracy, precision, and resolution are all quite different. That is why I never attach numbers to the words accuracy or precision when I am writing. The very definitive terms are "uncertainty" and "resolution", with the term "precision" used as a descriptor for items that are more accurate than average. Claiming a maximum of 1% measurement uncertainty is much more definitive, but it still needs to be qualifies as to 1% of what? Both percentage of reading and percentage of range are commonly used. Resolution is much easier to define and is actually independant of both accuracy and uncertainty. A five digit meter can be quite inaccurate while still providing excellent resolution.
The other term that is missing from the discussion is repeatability, which is vital to prodcing useful statistics based on measurements. A source of pain is that it seems that repeatability appears to decrease as uncertainty is decreased and resolution rises.
I frequently have this discussion with customers that have trouble understanding resolution and accuracy. I tell them it's the difference of having a ruler with a single, precise mark on it that is NIST traceable, or a ruler with 1000 marks on it that vary +/- 10 marks.
Interesting coimments all. I would like to add my 2 cents. The actual term I learned was measurement uncertainty (MU) usually expressed as a % of actual reading or full scale of the instrument (not the calibrated range) the determination of which is dictated by the linear plot of the instrument's error data on a Gaussian or normal distribution. The MU comprises of two components. The Precision or Random error and the Bias or Systematic error.
1/The precision error is that whick is inherent to the instrument's circuit, parts and design (pressure elements, manufacturer's calibration facility etc).
2/ The systematic error is the error caused by the installation of same instrument and resultant from things like environment, installation details, ancilliary equipment etc.
The two types of errors (and they are always present) are combined using RSS (root sum square) method the same way the collected data used to calculate the errors. to express the MU. Naturally the more data acquired the more statistically correct (higher confidence level) will be your results, The resolution and repeatablility are totally different and while extrememly important (a highly repeatable instrument can have terrible accuracy but the converse is not acceptable) and a highly accurate (MU) instrument must have very good resolution to adequately reflect the results.
As almost anyone will tell you MU (accuracy/inaccuracy) is one of the MOST misunderstood and misapplied terms in instrumentation and automation!!
Interesting piece which sheds a lot of light on a subject that most manufacturers have to deal with every day. Of the two characteristics to be considered when looking for the most accurate pressure sensor, I have always believed that the repeatability (or non-repeatability, as you put it) of the sensor results should be the most significant factor. The sensor that gives the least differences between independent measurements reflects the smallest variations and is therefore, in my opinion, the most accurate.
The two factors outlined above (linearity and repeatability) can be very useful when choosing other types of sensors besides just the pressure sensors. In fact, they apply to most, if not all, of the sensors within every effective flow control system.
Iterative design — the cycle of prototyping, testing, analyzing, and refining a product — existed long before additive manufacturing, but it has never been as efficient and approachable as it is today with 3D printing.
People usually think of a time constant as the time it takes a first order system to change 63% of the way to the steady state value in response to a step change in the input -- it’s basically a measure of the responsiveness of the system. This is true, but in reality, time constants are often not constant. They can change just like system gains change as the environment or the geometry of the system changes.
At its core, sound is a relatively simple natural phenomenon caused by pressure pulsations or vibrations propagating through various mediums in the world around us. Studies have shown that the complete absence of sound can drive a person insane, causing them to experience hallucinations. Likewise, loud and overwhelming sound can have the same effect. This especially holds true in manufacturing and plant environments where loud noises are the norm.
The tech industry is no stranger to crowdsourcing funding for new projects, and the team at element14 are no strangers to crowdsourcing ideas for new projects through its design competitions. But what about crowdsourcing new components?
It has been common wisdom of late that anything you needed to manufacture could be made more cost-effectively on foreign shores. Following World War II, the label “Made in Japan” was as ubiquitous as is the “Made in China” version today and often had very similar -- not always positive -- connotations. Along the way, Korea, Indonesia, Malaysia, and other Pacific-rim nations have each had their turn at being the preferred low-cost alternative to manufacturing here in the US.
Focus on Fundamentals consists of 45-minute on-line classes that cover a host of technologies. You learn without leaving the comfort of your desk. All classes are taught by subject-matter experts and all are archived. So if you can't attend live, attend at your convenience.