DAPL Operating System | Processing Command
Module SENSORM :: THERMOPOLY
Calibrated temperature conversions for thermocouples.
Syntax
THERMOPOLY( VIN, CJT, [NORDER,] VCOEFFS, TEMPC )
Parameters
- VIN
- Input thermocouple potential measurements
- WORD PIPE | FLOAT PIPE
- CJT
- Input cold junction temperature measurements
- FLOAT PIPE
- NORDER
- Order of the conversion polynomial function
- WORD CONSTANT
- VCOEFFS
- Coefficients of the conversion polynomial
- FLOAT VECTOR
- TEMPC
- Output results, temperature Centigrade
- FLOAT PIPE
Description
This command is an alternative to the THERMO
command provided by the DAPL system for systems that need
calibrated temperature measurements for maximum accuracy.
Input values are received from pipe VIN.
The polynomial characteristic defined by vector VCOEFFS
is applied to each input value, producing the corresponding
thermocouple temperature difference. Thermocouples do not
return an absolute temperature, so independent measurements
of the "cold junction temperature" must be provided in pipe
CJT, one cold junction temperature for each
thermocouple measurement. The measurements are combined and
the resulting absolute temperatures in degrees C are placed
into the TEMPC pipe.
For determining the temperature, the usual way to interpret
the potential across a thermocouple to apply published conversion
curves as determined by standards organizations. Manufacturers do
their best to produce devices match these standard curves, but the
match is never perfect. Individual devices are very repeatable, however. If
operated consistently over a bounded range, without severe thermal
stresses, it is possible to apply adjusted curves and get a total
measurement accuracy within about 1 degrees C. The THERMOPOLY
command supports adjustable conversion curves; in contrast,
the THERMO supports only predetermined standard
curves.
The nonlinear conversion mapping is defined by a polynomial. The
polynomial coefficients are defined in the VCOEFFS
vector. The coefficients start with the zero-order term, followed by
the first-order term, followed by the second-order term, etc. up to
the order specified by the NORDER parameter. If you
omit the NORDER parameter, the command will count
terms and assume the corresponding maximum order. A third to fourth
order polynomial is typically sufficient to span an interval of a
few hundred degrees C to full accuracy. A second order polynomial is
sufficient over a lesser range, around 100 degrees C or so. A
first order polynomial is the equivalent of offset and gain
corrections using the SCALE command.
The measured thermocouple potential is very small, so it must be
amplified to obtain a signal level large enough to digitize. Small
offset or gain errors are amplified along with the signal and can
have a significant effect. The polynomial of the THERMOPOLY
command can fully account for these effects at the same time
that it corrects for nonlinearity. It can also cover any required
scaling changes related to converting from A/D digital readings to
physical units. Consequently, the THERMOPOLY command
does not need a SCALE command to condition the signal
prior to conversions.
Examples
VECTOR VCF FLOAT = ( -0.01897, 25.41881, -0.42456, 0.04365 )
...
THERMOPOLY( PKIN, PCJC, VCF, PTOUT )
Use a fast third-order approximation for a type K thermocouple
to measure water and steam temperatures (in the range 0 to 200
degrees C). Take junction measurements captured (with a gain of
1000) from pipe PKIN. The third-order polynomial
function defined by coefficient vector VCF is applied
to the input data to determine temperature differences. Read the
corresponding ambient temperature values from the PCJC pipe and
combine with the temperature difference to produce the final
output temperature measurements in pipe PTOUT.
See also:
SCALE, GENPOLY, THERMO