LMT87: 3-wire to 2-wire: LMT87

I don't recommend this circuit, but it could work. Here are some things to consider:

1. Load Regulation. The circuit was drawn for LM34, which has 1mA spec. the LMT87 has only 50uA spec. You must size the resistors such that this 50uA current will not be exceeded.

2. Line Regulation. This circuit will cause the sensor's power supply voltage to vary over the operating temperature range. Line regulation is an output error caused by this supply change.

3. Operating voltage range and voltage overhead. The LM34 can be powered by up to 30V. Modern equivalents like LMT87 can only tolerate 5.5V. Increasing the supply voltage would provide more headroom for the output to vary without causing the supply voltage to dip below the minimum required, but there is less room to increase the supply voltage. The LM34 has no voltage offset; it outputs 0V at 0°F and outputs negative voltages at negative temperatures. Since modern electronics have moved away from large voltages and negative power supplies, modern sensors like LMT87 provide a voltage offset to center the signal in the available range: it's output bottoms out at 500mV at max temperature (due to negative slope.) This means that all of your output voltages on a modern sensor will be higher, and will further eat away at your voltage headroom that is constrained by the low supply voltage.

4. Quiescent Current. This circuit will cause the current that the sensor consumes to contribute error to the output. This is because both the signal current, which is developed through the output resistor, will be summed with the quiescent current of the sensor. The LM34 consumes around 100uA and the load regulation limits us to around 1mA output. This is about 10:1 signal to error. The LMT87 consumes about 5-8uA and load regulation limits us to 50uA output. This is worse than 10:1. 

thanks,

ren