Example 22: corr_pres5 - Temperature correction of Pressure

This equation computes a temperature corrected pressure value, with the temperature, Tsensor, derived from a voltage ratio corresponding to temperature of the sensor itself Rt, as follows, corrected by the pressure ratio signal R:

T_{sensor} = \sum\limits_{n= 0}^{3} K_{t_n}(R_T - \sum\limits_{m=0}^{3}Kc_m(A \cdot R + B)^m)^n

Where the Kt terms are the polynomial coefficients for the temperature calculation, the Kc terms are the polynomial coefficients of the pressure correction, while A and B are coefficients of the measurement circuit linear model.

Once the temperature is known, we proceed almost exactly as in example 3 above, So the pressure reading, without correction for the effect of temperature on the sensor, is given by a cubic polynomial:

Praw = c_0 + c_1 \cdot R + c_2 \cdot R^2 + c_3 \cdot R^3

where  c0...c3 are the core coefficients of the cubic polynomial equation, and Praw is the uncorrected pressure output, reported in dbar for RBR instruments.

The equation which accounts for residual temperature sensitivity of the pressure sensor is:

Pcorr = Pcal + \dfrac{(Praw - Pcal) - Kp_1 \cdot (T - Tcal) - Kp_2 \cdot (T - Tcal)^2 - Kp_3 \cdot (T - Tcal)^3}{1 + Kp_4 \cdot (T - Tcal)}

Casting this into the form used by the logger would yield:

Pcorr = x_0 + \dfrac{(Praw - x_0) - x_1 \cdot (T_{sensor} - x_5) - x_2 \cdot (T_{sensor} - x_5)^2 - x_3 \cdot (T_{sensor} - x_5)^3}{1 + x_4 \cdot (T_{sensor} - x_5)}

where

  • Praw is the cubic polynomial in R as in example 3,
  • x0 is the calibration pressure 'Pcal' in dbar,
  • x1, x2, x3, x4  correspond directly to the constants "Kp1" through "Kp4",
  • x5 is the calibration temperature "Tcal" in °C,
  • x6,x7 correspond to the circuit gain parameters A and B  
  • x8,x9,x10,x11 correspond to the constants "Kc0" through "Kc3"
  • x12,x13,x14,x15 correspond to the constants "Kt1", through "Kt3" used to compute the sensor temperature
  • n0  is the index of the sensor temperature ratio channel,
  • Tsensor is the computed sensor temperature in °C, 
  • Pcorr is the corrected output in dbar.

Examples

>> calibration 3 type
<< calibration 3 type = pres28

Confirm the channel type.

>> calibration 3 datetime = 20171201000000, c0 = 0.2346, c1 = 120.9873, c2 = 2. 7356, c3 = 0.7

Set the core coefficients.

>> calibration 3 datetime = 20171201000000, x0 = 9.983, x1 = 0.2003, x2 = 0.2943, x3 = 0.0721, x4 = 0.1049, x5 = 21.29
>> calibration 3 datetime = 20171201000000, x6 = 0,  x7 = 187.06043
>> calibration 3 datetime = 20171201000000, x8 = 14.219203,  x9 = -0.30655386, x10 = 0.0012941403, x11 =0.0000039787190
 
>> calibration 3 datetime = 20171201000000, x12 = -3727.6430,  x13 = 2.8672020, x14 = -0.00075268429, x15 = 0.000000068588312
>> calibration 3 datetime = 20171201000000, n0 = 5

Set the cross-channel correction coefficients.

>> calibration 3
<< calibration 3 type = pres28, datetime = 20171201000000, c0 = 0.2346, c1 = 120.9873, c2 = 2. 7356, c3 = 0.7, x0 = 9.983, x1 = 0.2003, x2 = 0.2943, x3 = 0.0721, x4 = 0.1049, x5 = 21.29, x6 = 0.0, x7 = 187.06043e+000, x8= 14.219203, x9 = -306.55386e-003, x10 = 1.2941403e-003,   x11 = 3.9787190e-006, x12 = -3.7276430e+003, x13 = 2.8672020e+000, x14 = -752.68429e-006, x15 = 68.588312e-009, n0 = 5

Request confirmation of all calibration coefficients.