44
c
HAPTER
2:
Cooling System Design and Temperature Control
Model 372 AC Resistance Bridge and Temperature Controller
make sure the package is “strain free” or makes some accommodation to reduce
stress on the resistive element during thermal cycling. Many sensors are also offered
without packaging for applications that require smaller size or lower mass than pack-
aged sensors provide. These must be installed with the greatest care and all manufac-
ture recommendations should be followed.
2.10.1.6 Environment
All cryogenic temperature sensors have to contend with high vacuum and drastic
temperature changes but some experiments create an even more hostile environ-
ment by adding high magnetic fields, ionizing radiation and more. These additional
requirements make sensor selection more difficult because they can limit the useful
range of many sensors. The Lake Shore Temperature Measurement and Control cata-
log includes magnetic field dependence for most temperature sensors. Other specific
data on environmental factors is included when available. Lake Shore Cernox® RTDs
are an excellent choice for many of these critical applications.
2.10.2 Sensor
Calibration
The Model 372 can convert the resistance of a sensor to temperature, using either lin-
ear or cubic spline interpolation methods, if a temperature response curve is loaded
into the instrument. Most curves are generated from an individual sensor calibration
because few sensors follow a standard curve at low temperatures. The instrument
can use curve options, called CalCurves, available with Lake Shore calibrated sensors
or curves generated by the user. The Lake Shore Temperature Measurement and Con-
trol catalog gives complete specifications for calibrated sensors and gives the calibra-
tion range available for each sensor.
The Model 372 uses either linear or cubic spline interpolation to convert sensor resis-
tance to temperature. Linear interpolation tends to be simpler to implement and
more predictable. However, this advantage is a tradeoff to the loss of accuracy along
non-linear regions of the temperature curve.
Cubic spline interpolation offers a smoother curve fit, due to the continuous nature of
the first and second derivatives, and tends to be more accurate than linear interpola-
tion methods. In addition, it does not suffer from oscillations and instabilities on cer-
tain regions of the curve that higher order polynomials tend to exhibit.
Cubic spline curves can only be loaded and modified over the remote computer interface.
If you wish to calibrate your sensors in house, Lake Shore recommends using the lin-
ear interpolation method. If you wish to use a cubic spline curve with your existing
Lake Shore sensor, please contact Lake Shore.
When calibrating sensors in house, remember to account for errors caused by linear
interpolation between curve breakpoints. The Model 372 allows curves with up to
200 breakpoints. If linear interpolation errors are larger than desired in a specific
temperature range, more breakpoints can be added to reduce the error. NTC resistor
breakpoints are entered as Log
)
vs. K which acts to linearize temperature response
data.