TMP112: Conversion time and reliability of first data after powering up

Hi Saito-San,

Thank you for your post, please see below for my answers. 

Takahiro Saito said:

Looking at Figure 14, it says "Delay is set by CR1 and CR0". I believe "Delay" means "conversion time".
According to electrical characteristics, conversion time is fixed value.

"Delay" in this case refers to the "Conversion Modes" section of the electrical characteristics table. It reflects how long the TMP112 will wait between conversions. For example, if you set CR1 = 1 and CR0 = 1, the TMP112 will trigger one temperature conversion every 8 seconds.

Takahiro Saito said:

1. Is there any relation with conversion rate and conversion time?

They are independent of each other for this device. The conversion time is how long it takes the TMP112 to measure temperature, and the conversion rate is how often the TMP112 measures temperature. 

Takahiro Saito said:

2. Could you make sure whether conversion time is changed by CR1 and CR0 or not?

Conversion time will not be changed by these bits, it will always be 35ms (max).

Takahiro Saito said:

3. Could you confirm if the reliability or accuracy is the same between first data and subsequent data?

Yes, you should get good data as long as you wait at least one conversion time (26ms typical, 35ms Max) after power up before reading the register the first time. If you try to read before this you will get 0C instead of the actual temperature. I would recommend always waiting 35ms after startup to be safe.

Takahiro Saito said:

4. Is there any power up condition that degrades the reliability of first data?

This behavior has not been observed in the TMP112 before, and temperature sensors do not typically have many restrictions on powerup sequence. The only time I could see you encountering an issue is if the device power supply dipped outside of the recommended operation conditions during this first conversion. 

Takahiro Saito said:

5. To make sure the first data reliability, could you check how much time the internal oscillator needs to be stable after powering up

It is possible that I am misunderstanding you question, but as I mentioned above, 35ms after power up is the maximum limit for the TMP112's first conversion (first reliable temperature data). Please let me know if there is any additional information needed here. 

Best Regards,
Brandon Fisher