A LM234 adjustable current source I.C. is used as a 1 uA / degree Kelvin temperature sensor. The 1 uA / °K calibration is done by hand selecting two 1/10 watt 5% resistors that in series result in a current of 1 uA / °K from the LM234 (by calculation should be = 227 ohms). This is more accurate than using the closest value single 1% resistor.

The probe is connected to the remaining circuitry with a miniature 2 conductor shielded cable.
This photo shows the interconnecting cable shield partially wrapped over the probe assembly that has been enclosed within a piece of heat-shrink tubing.

A piece of shielding braid, obtained from a scrape of small coax, is positioned and ready to be slipped over the probe.
Here, the shielding braid has been slipped over the probe assembly.
The shielded probe is encapsulated within a piece of heat-shrink tubing that has an inside wall that melts and seals the completed assembly.
The 1 uA / degree Kelvin current from each sensor is converted to 19.6 mV / deg K. This voltage is level shifted −2.5 volts to keep the resulting voltage, over the −90°C to +100°C measurement range, within the 0 to +5 volt input range of the MIM.

Making each °K = 19.6 mV (the MIM's A/D 1 bit resolution) means each 1°C change of measured temperature will result in a change of the telemetry temperature A/D value also by 1. Thus, rather than having to use the formula (( A/D Value x .0196) - 2.89) / .0196 to calculate the measured temperature, one can simply remember that 0°C is equal to the telemetry value of 147 and subtract 147 from the current telemetry value to find the temperature.

Update: It is hard to shield against RFI when working within inches of a 5 watt transmitter. Probe shielding did help some, but it is not the total solution. On the next BEAR mission I am going to try using the Dallas digital temperature sensors to overcome RFI problems.
To BEAR Home Page