The Pressure Calibration page provides an overview of how we calibrate our strain gauge pressure sensors and the equations we use to do so.  This page does not discuss the calibration process or approach for quartz pressure sensors

The page also includes information on pressure ranges, accuracies, and precisions.  For information related to the pressure sensors themselves - sizing, headers, materials, design, etc. - please click on the Pressure page under Technical / Parameters. 

Applied Microsystems Ltd utilizes three equations for calibrating strain gauge pressure sensors.  The choice of equation is instrument model specific.

There are two equations used for temperature compensated pressure calibrations.  The first compensated equation was used on a small number of early SV Plus V2 instruments.  It uses the form:

P (dBar) = A+B*Nc+C*Nc2+D*Nc3+(E+F*Nc+G*Nc2+H*Nc3)*Np+ I +J*Nc+K*Nc2+L*Nc3)*Np2

Where: A through M are calibration coefficients, Nc is the thermal compensation number (Nc=Npt-M), Npt is the raw pressure sensor thermal compensation measurement, and Np is the raw pressure sensor measurement.

The second equation is used on a much wider variety of instruments, including all Micros and current SV Plus V2 and CTD Plus V2 instruments.  The pressure calibration equation for these instruments is:

P (dBar) = A+B*Npt+C*Npt2+D*Npt3+(E+F*Npt+G*Npt2+H*Npt3)*Np+(I+J*Npt+K*Npt2+L*Npt3)*Np2

Where: A through L are calibration coefficients, Npt is the raw pressure sensor thermal compensation measurement, and Np is the raw pressure sensor measurement.

For older, non-temperature compensated instruments like the CTD Plus and SV Plus instruments the calibration equation used is:

P (dBar) = A+B*Np+C*Np2+D*Np3

Where A, B, C and D are calibration coefficients, and Np is the raw pressure sensor measurement.

Older non-temperature compensated instruments are calibrated at room temperature using a dead weight tester.  Ten to twelve pressure points are taken spanning the full scale range of the pressure sensor.  The raw data from the pressure sensor measurement is then curve fit to the actual pressure applied by the dead weight tester.

Newer instruments such as the Micro series  and V2 series have thermally compensated pressure sensors.  The calibration procedure for these instruments is significantly more involved.  Each pressure sensor is calibrated in a temperature controlled bath.  10 to 12 pressure points are taken at each of 5 temepratures distributed throught the temperature range of the sensor. The 50 to 60 calibration points collected have three values associate with each point.  The actual pressure applied by the dead weight tester, the pressure sensor temperature compensation measurement (Npt), and the pressure measurement (Np).  All this data is utilized in a three dimensional curve fit to generate the calibration coefficients.

The pressure sensor temperature compensation measurement is performed at the Wheatstone bridge of the pressure sensor.  This helps to avoid thermal lag, which results in data spiking in high speed applications.

Applied Microsystems offers calibrated pressure sensors in the following ranges, measured in dbar.  NB: Dbar and metres are roughly similar:

• 0 to 50 dbar
• 0 to 100 dbar
• 0 to 200 dbar
• 0 to 500 dbar
• 0 to 1000 dbar
• 0 to 2000 dbar
• 0 to 4000 dbar
• 0 to 5000 dbar
• 0 to 6000 dbar

For an explanation of the relationship between calibrated pressure range and absolute accuracy, please see below.

The calibrated range of a pressure sensor is the normal operating range of the sensor where the pressure readings will fall within the sensor's stated accuracy limits.   Applied Microsystems allows user to select from amongst 12 different pressure ranges. 

Pressure sensors can be over-ranged by 50% without causing damage to the sensor.  In such a case, the sample points that lie beyond the calibrated pressure range of the instrument may not fall within the accuracy specifications of the instrument.  However, the sensor will not suffer irreparable damage.

Burst pressure is that pressure at which irreparable damage is done to the sensor.  In such a case, the instrument will flood and the sensor - and possibly the electronics - must be replaced.  Burst pressure is normally 75%  more than the stated pressure range on the sensor.

Pressure Sensor Ratings

	Increment Over Calibrated Range	Sample 200dbar Pressure Range
Calibrated Range	Range as Stated	200 dbar
Allowable Over Range	Range as Stated plus 50%	300 dbar (200bar plus 50%
Burst Pressure	Range as Stated plus 75%	350 dbar (200 bar plus 75%)

Given Applied Microsystems' detailed calibration methodology, instruments with strain gauge sensors are frequently shipped with an RMS error that is better than our specified field accuracy of 0.05%FS. It is not unusual for the RMS error on our pressure calibrations to be 0.028%FS or 0.029%FS. 

Pressure Specifications

	Accuracy	Precision	Resolution	Response
Strain Gauge	+/- 0.05%FS	+/- 0.03%FS	+/- 0.005%FS	10 milliseconds
Quartz	+/- 0.01%FS		+/- 0.000001%FS

Absolute error is different from relative error and is a function of the pressure range that is chosen and the relative accuracy of the sensor.  For example:

• a pressure sensor with a range of 2000 dbar and a relative accuracy of 0.05%FS will have an absolute error of 1 dbar (ie. 0.05 * 0.01 * 2000 dbar)
• a pressure sensor with a range of 4000 dbar and a relative accuracy of 0.05%FS will have an absolute error of 2 dbar (ie. 0.05 * 0.01 * 4000 dbar)

Many vendors of oceanographic instrumentation refer to accuracy and precision interchangeably.  They are not interchangeable.  In effect, accuracy refers to how well a sensor performs against a known third party standard.  For example, a temperature sensor may be +/- 0.001 C, as compared to a Black Stack themistor module.  Precision refers to the repeatability of the readings of a given sensor.   A sensor is precise when it repeatedly provides the same reading, regardless of how accurate that reading is.

A good analogy is a dart board.  The thrower of darts is accurate when he or she is able to reach the target, the bulls-eye.  He or she is precise if, having thrown three darts, all three land in the same location, irrespective of whether or not that location is the bulls-eye.

For high accuracy pressure measurements, temperature compensation is absolutely required.  Without such compensation, the user may experience errors of up to 0.25%FS in their pressure reading as temperatures vary.  Equally important, temperature compensation at the bridge and not at the thermistor - the instrument's own temperature sensor - will help to avoid thermal lag, which results in data spiking in high speed applications.

LIke many other oceanographic sensors, there is no magic answer.  In general, the greater the number of pressure cycles, the more likely that a pressure offset will begin to occur.  The likelihood of such an offset increases with the prolonged exposure to the upper pressure range of a sensor's calibrated range, for example in long term in-situ applicatons. 

A deformation in the diaphgram of the pressure sensor is likely to impact the calibration of that sensor.  Such a deformation can occur if the sensor is poked with sharp objects.  A pressure sensor should always be returned for recalibration if it suffers such a poke or prod.