Measuring pressures have many different applications which require proper encapsulating of the sensors, but at the core there is a limited amount of techniques.

To my understanding, the two leading technologies in high accuracy pressure sensing are:

  1. Measuring resonant frequencies in e.g. quartz crystals
  2. Piezo-electric elements

And there are two leading technologies in pressure transfer from ambient pressure to the pressure sensing element (force collectors)

  1. Diaphragm
  2. Bourdon tube

I found a easy to understand reference to precision and accuracy of pressure sensors Pressure sensor fundamentals: Interpreting accuracy and error. Here they divide Accuracy into the following categories

  1. Nonlinearity
  2. Hysteresis
  3. Nonrepeatability
  4. Zero point error
  5. Span error
  6. Temperature impacts
  7. Long term drift

Most high accuracy sensors promise an accuracy of 0.01% of full scale, but how does this distribute among the different types of errors?

Are there any sensors promising significantly better accuracy?

Is there some technologies that are better than others on e.g. hysteresis or temperature impacts?

Has there been scientific comparisons between different sensors to measure these errors which constitute the total accuracy?

What is the best way of going about to find the sensor best suited for my purpose?

  • 1
    $\begingroup$ While I would love to see a thorough answer to this, I think this question needs to be narrowed down a bit. $\endgroup$ – Rick Sep 16 '15 at 18:36
  • $\begingroup$ I could narrow it down to fit my application, e.g. by specifying a pressure range. I am working on deep subsea pressure measurements. But then others might find the question less interesting. In subsea measurements waves poses such a high level of noise that most don't need accuracy better than 0.01% therefore I was thinking other areas might have more information on high precision pressure measurements. Do you have suggestions on how to narrow it down? $\endgroup$ – Martin Vatshelle Sep 17 '15 at 13:51
  • $\begingroup$ With an accuracy of 0.01%, surely the problem will not be the sensor, but its implementation $\endgroup$ – CL22 May 4 '16 at 9:49

The reason sensor error is distributed among all those variables is because different applications have drastically different requirements. Let;s take out the errors you can't do anything about: the non-repeatable errors and hysteresis. They aren't linear or stationary, and as such, it's hard to compensate for them.

There are many sources of non-linear noise, one of which is temperature effects. Those, if you so choose, you could map out the non-linearity in detail and calibrate the sensor to remove much of them.

But ultimately, it comes down you what it is you want to detect. Sensor datasheets will provide a noise density. The larger the measurement bandwidth, the more noise the system will let in. So match the sensing bandwidth to the signal you want to detect. Understand the temperature effect and calibrate/compensate if necessary. That's as general an answer as your original question.

  • $\begingroup$ MEMS silicon membrane capacitive sensors have virtually no mechanical hysteresis because silicon microstructures exhibit near perfect elasticity, as long as they are not pushed pass their elastic limit. The nice feature of these devices is that they enable one to make a "smart sensor" because the control electronics are incorporated right in the die with the sensing diaphragm. $\endgroup$ – William Hird May 3 '16 at 2:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.