Looking for temperature sensor with +/- 0.5 degree accuracy at 25 degrees.

[StephenL]

I am looking for temperature sensor with +/- 0.5 degree accuracy at 25 degrees. The DS18B20 would do the job well, but does not work over the range of 1.8V to 3.6V. Other type such as the DS75LV, and LM94022 are not accurate enough.

Digital output is preferred, but can work with analogue output. The sensor is to be built into a probe with a 2 meter connecting lead.

Thank you for any suggestions.

 

[Hero999]

The low voltage of 1.8V is going to be the problem.

What about using a precision thermistor?

 

[StephenL]

A thermistor is a good idea.

The uP I am using is the SiLabs C8051F930. The A/D full scale is 1.68V. This would require an amplifier or else the self heating of the thermistor would be high. Also I would need a precision current source.

This job is quite price sensitive. Just hoping for the equivalent of a DS18B20 that runs at 1.8V.

 

[Hero999]

You don't need a current source: use the power supply as the A/D reference and connect the the thermistor as part of a potential divider. Self heating won't be much of a problem as long as you keep the power dissipation low.

I doubt you'll find an IC, high accuracy, low voltage and low cost are not compatible.

 

[StephenL]

We have gone with the LM94022 which give very low power consumption (6uA).  The accuracy is only +/- 1.5 degrees C at 25 degrees so it will need calibration.

Thank you for your suggestions and avoidance of a constant current supply.

You are right that low voltage ==> higher prices.

 

[Hero999]

Good you came up with a solution.

Thanks for returning to tell us about how you did it, most people don't bother.

 

[rakeshkoku]

good post keep it up