HDC1080: Correct Device Behavior?

Thanks Josh. I'm curious, though, whether you've had problems reading temperature and humidity measurements when configured for the acquisition mode where Temperature and Humidity are acquired in sequence, Temperature first. One read transaction returns 4 bytes - two bytes for temperature and two for humidity.

I get the temperature MSB then and ACK. But then I get 0xFF followed by NACKs for the remaining 3 bytes.

Is there some delay I need to implement between bytes?

Craig - 

do you have an LSA or 'scope capture of what you are sending out and RXing? That would probably help. 

This is the best I can do at the moment. When I manually perform a read transaction, works fine:

This is what my Verilog is doing:

Hi Josh,

I was modeling the read transaction according to Figure 14 in the datasheet. One transaction sending the pointer byte first (h81). I'll have to go back and re-review your original post. 

Josh,

I tried a single read transaction. Reading temperature. I trigger first h80/h00. Wait about 32 ms. Then read temperature (h81). After I get temperature MSB (h66), I should get an ACK rather than a NACK. Maybe the temperature LSB is actually hFF. But that seems unlikely.

What am I missing?

Craig - 

Looks like the master is taking control/not taking control (depending on how you look at it) of the SDA line perhaps at the end of the first byte of the response and leaves it high for the duration. Is that the case? this would result in the NACK you see at the end of the first byte and the 0xFF you see on the second byte of what should be the response from the part.