PARAMETER DERIVATION

1. If the pressure uncertainty of ± 12.5 hPa is not acceptable, calibrate the Barograph by means of parameters k and r. To calculate the calibration parameter k, first the current local Barometric pressure (reference pressure) must be known. This info can be obtained from a Torricelli mercury barometer or from a metereological site on Internet. On Internet normally barometric pressure is given at sea level: to convert it to your altitude H, use the following expression:

Slp = Sea Level Pressure found on Internet, hPa

H = Altitude in metres.

G = 7000 in metres (universal reference parameter)

Shp = Pressure at altitude H

Then: Shp = Slp * EXP (-H/G) where the expression EXP means e^(-H/G)

EXAMPLE: Slp = 1015,2 hPa; H = 230 metres; Shp = 1015,2 * EXP(-230/7000) = 982,4 hPa.

2. To find calibration parameter k, we proceed as follows.

Using [1] we compute the sensor output voltage if the sensor were calibrated:

Vo = Vcc*((0,01067*P)-0,32667) [1]

Then we measure the actual sensor output Vo’: the actual sensor calibration parameter is k = Vo/Vo’

EXAMPLE: suppose our reference pressure is: Shp = 982,4 hPa. Remember that P in [1] is in kPa, then Shp must be divided by 10 and becomes 98,24 kPa. Then:

Vo = 5*((0,01067*98,24)-0,32667) = 3,6078 V = 3607,75 mV this is the output voltage the sensor would have if it were accurately on the spot.

Now we turn on the Baro circuit and we read, not 3607,75 mV, but Vo’ = 3,6403V = 3640,3 mV

Using the inverse function [2] we can find the pressure as read by the sensor:

P = ((Vo/Vcc)+0,32667)/(0,01067) [2]

Po’ = ((3,6403/5)+0,32667)/0,01067)*10 = 988,50 hPa

Thus the static error of this particular sensor is: 988,5 – 982,4 = + 6,1 hPa

Therefore constant k = Vo/Vo’ = 3607,75/3640,3 = 0,9911. This correction brings our Barograph on the spot.

3. To find parameter r, used to find parameter h, as shown in Part 3, we measure sensor output at different Vcc voltages: 4,8 ; 5,0 ; 5,2 V and we record the sensor output voltages. Then we calculate the average deviation and finally we divide the average deviation by the supply voltage variation, to obtain the sensor output voltage deviation per unit of supply voltage variation: this is parameter r. Using one of the two Excel examples:

Derivation of r for Sensor KP-235 #1:

For Vcc = 4.800 mV ------------Vo = 3.571,4 mV ---------ΔVo- = -149,10 mV

For Vcc = 5.000 mV ------------Vo = 3.720,5 mV ---------ΔVo = 0.00 mV

For Vcc = 5.200 mV ------------Vo = 3.854,7 mV ---------ΔVo+ = +134,20 mV

ABSOLUTE AVERAGE +/-mV = ((ΔVo+) - (ΔVo-))/2 = (134,20 +149,10)/2 = 142 mV

Measured ratiometric error factor: r = 142/200 = 0,7083

While the procedure above provides the actual Ratiometric Error, it must be noted that the formula for Ratiometric Error percentage given in § 2.5.1, page 13 of the KP-235 data sheet, is not correct.

At the moment the Barograph is giving excellent performance and is undergoing a one week test (168 hours). Test results will be posted in PART 5.

## THE DrDAQ NEW MINIATURE BAROGRAPH - PART 4

### Re: THE DrDAQ NEW MINIATURE BAROGRAPH - PART 4

Hi Glovisol. Please watch out for the "DrDAQ" section in the next Pico Test & Measurement newsletter – I'll give a special mention your projects.

Jeff