HDC2010: Abnormal value of the humidity after the power supply injection

Hi,supports,

As I was not able to attach a wave pattern of I2C, I send it.

Regards,

Toshi

Toshi - 

The waiting time after application of input voltage, before beginning any I2C transactions is 3mSec. Please see this FAQ regarding firmware handling of the HDC2010. 

https://e2e.ti.com/support/sensors/f/1023/t/878721

In here I show the timings after power up and the resulting logic captures. 

Dear Josh

I already confirmed for the waiting time.
3.3V is turned on before 900 msec to send a command.
I confirmed the smallest limit time between commands.

330usec and 920usec were limit time in this table.
I think that the total of these time is equivalent to 1270usec.
I have enough waiting time in comparison with the limit time.

I am thinking that this trouble may occur in power-on reset process and asking these questions.
Is there the precaution about the setup of the power supply?
Please re-examine Qestion.1-3 based on my comments.

Regards,
Toshi

Toshi - 

At 14 bit resolution output setting, you need to wait at least 1300uSec (i use 1500uSec) waiting time before asking for data from the part. You can also read register 0x04 for 0x80 before reading registers 0x00 through 0x03 to make sure it is ready and you can also use the INT/DATA READY pin for redundancy. 

Also, I do not use the reset of the part in normal operation - I think it is a waste of clock cycles, you should consider to remove it. 

Please pay close attention to the FAQ. It was made specifically for this kind of question you are asking. 

Dear Josh

I understood that you did not use reset frequently in normal operation.
And, I understood that I should confirm FAQ carefully.

However, please tell me a little more about Q1 and Q2.
I change the contents of the question.

1. When a noise is carried on 3.3V at the time of power supply, does a trouble occur for the power-on reset sequence of HDC2010? In addition, is there the method to know it when a trouble occurs?

2.Does the IC work normally if I hold the voltage at 1.8V during 100 msec at the time of power supply and change voltage to 3.3V?

For my future analysis, I want to hear your opinion by all means.
Regards,
Toshi

Toshi - 

1. You would have to share what sort of noise is present on the line. We recommend putting in bypass capacitor on the Vin pin of the part and located close to it, which is common practice for any IC, to help remove any noise or ringing on the line. The value of this bypass capacitor can be changed by characterization of the speed and type of the noise or ringing present. It really depends on the specifics of a given situation. I don't use the reset as said, because I have not seen a reason to use it and it wastes embedded codespace and clock cycles, and this is just my view - I prefer compact and efficient code because I know the value of it and if it was needed (as you may see other competitive parts require it) I would of course share that with you.    
2. From power up at any valid input voltage (from 1.62VDC to 3.6VDC) , the IC needs at least 3mSec before I2C communications are started, so 100mSec is more than enough time. I am not sure why you would switch the voltage afterwards, but the device will continue to operate. 

Dear Josh

I think that it has a hint of the analysis to sometimes occur only at the time of power supply injection.
I confirm again. May I think that the trouble that data turn into at a power-on reset process does not occur?
(e.g.) Read error of calibration data

Regards,
Toshi