BEAR 1 & 2 Temperature Sensors

A LM234 adjustable current source IC is used as a 1 uA / degree Kelvin temperature sensor. The 1 uA/K calibration was done by hand by selecting two 1/10 watt 5% resistors that in series result in a current of 1 uA/K from the LM234 (that by calculation should be 227 ohms) and 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 assembly.
The shielded probe was encapsulated in a piece of heat-shrink tubing with an inner wall that melts and seals the 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 −90C to +100C 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 1C 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 0C 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