TMP235: Follower needed after TMP235

Hi Tomas,

I recommend keeping the follower after the sensor.

This is because the TMP235 has very limited current sinking capability. For the output voltage range of the TMP235 (~ 0-2V), I calculated the current out of the sensor to range from -20uA to +20uA. This is too high for the part to sink (although it has no problem sourcing it). Adding the buffer will cause the opamp to drive the load instead of the sensor. You can double check that the sink and source current capability for the opamp you're using is over 20uA. 

I simulated the circuit with and without the buffer. With the buffer, the current sunk by the sensor is more than small enough, less than 1uA.

Note that in the simulation below the TMP235 is estimated as a voltage source, Vsense.

Without buffer, I_sense goes to -20uA: 

With buffer, I_sense stays above -1uA:

Let me know if there are any further questions or concerns!

Thank you,

Hypatia Magyar

Hi Tomas,

We do have a couple temp sensor parts that will sink enough current. The LMT87 will work with your 5V rail. However, it has a negative mV/C sensor gain, so you would have to rework the other part of your circuit. You might end up with the amplifier anyway to reverse the negative gain.

Best,

Hypatia 

Hey Tomas,

I have an update. There was some part design confusion on our end, and I was mistaken in my earlier messages. The TMP235 is actually quite capable of sinking 20uA of current. I confirmed this in the lab as well. In light of this information, I don't see a need for your first buffer.

I'm sorry about the confusion!

Thank you,

Hypatia