DL200 Datasheet, PDF (335/670 Pages) Motorola, Inc – Sensor

Freescale Semiconductor, Inc.
AN935
90
80
70
60
50
40
30
20
10
00 _{10 20 30 40 50 60 70 80 90}
PRESSURE (torr)
Figure 1. MPX12 Linearity Analysis Raw Data
0.5
0.4
0.3 B0 = 0.1045 – 0.00295*(SPAN)
0.2
0.1
0
– 0.1
– 0.2
– 0.3
– 0.4
– 0.5
20 25 30 35 40 45 50 55 60 65 70
SPAN (mV)
Figure 3. MPX10 Linearity Analysis —
Correlation of B0 Vout = B0 + B1 (P) + B2 (P)2
5.0
4.0
3.0
2.0
1.0
0
–1.0
0 10 20 30 40 50 60 70 80 90
SPAN (mV)
Figure 2. MPX10 Series Span versus Linearity
1.1
1.0 _{B1 = 0.2055 + 1.598E – 3*(SPAN)}
0.9
+ 1.723E – 4*(SPAN)2
0.8
0.7
0.6
0.5
0.4
0.3
0.2
20 25 30 35 40 45 50 55 60 65 70
SPAN (mV)
Figure 4. MPX10 Linearity Analysis —
Correlation of B1 Vout = B0 + B1 (P) + B2 (P)2
COMPENSATION FOR NONLINEARITY
The nonlinearity error shown in Figure 1 arises from the
assumption that the output voltage changes with respect to
pressure in the following manner:
Vout = Voff + sens*P
[1]
where Voff = output voltage at zero pressure
differential
sens = sensitivity of the device
P = applied pressure
0.0030
0.0025
0.0020
B2 = –1.293E – 13*(SPAN)5.68
0.0015
0.0010
0.0005
It is obvious that the true output does not follow this simple
straight line equation. Therefore, if an expression could be
determined with additional higher order terms that more
closely described the output behavior, increased accuracies
would be possible. The output expression would then become
Vout = Voff +(B0+B1*P+B2*P2+B3*P3 +. . .)
[2]
where B0, B1, B2, B3, etc. are sensitivity coefficients. In
order to determine the sensitivity coefficients given in equation
[2] for the MPX10 series of pressure transducers, a polynomial
regression analysis was performed on data taken from 139
devices with full scale spans ranging from 30 to 730 mV. It was
found that second order terms are sufficient to give excellent
agreement with experimental data. The calculated regression
coefficients were typically 0.999999+ with the worst case
being 0.99999. However, these sensitivity coefficients
demonstrated a strong correlation with the full scale span of
the device for which they were calculated. The correlation of
B0, B1, and B2 with full scale span is shown in Figures 3
through 5.
20 25 30 35 40 45 50 55 60 65 70
SPAN (mV)
Figure 5. MPX10 Linearity Analysis —
Correlation of B2 Vout = B0 + B1 (P) + B2 (P)2
In order to simplify the determination of these coefficients
for the user, further regression analysis was performed so that
expressions could be given for each coefficient as a function
of full scale span. This would then allow the user to do a single
pressure measurement, a series of calculations, and
analytically arrive at the equation of the line that describes the
output behavior of the transducer. Nonlinearity errors were
then calculated by comparing experimental data with the
values calculated using equation [2] and the sensitivity
coefficients given by the regression analysis. The resulting
errors are shown in Figures 6 through 9 at various pressure
points. While using this technique has been successful in
reducing the errors due to nonlinearity, the considerable
spread and large number of devices that showed errors >1%
indicate this technique was not as successful as desired.
Motorola Sensor Device Data _{For Mor}w_eww_In.m_foot_ro_mro_ala_t.c_ioom_n/s_Oem_nic_To_hnd_isuc_Pto_rr_os _duct,
Go to: www.freescale.com
3–189