HDC2080: Accuracy issue with mechanical

Dear Sumit - 

if you could convey the details of the issue you are having, then we can assist you. 

Hi Josh, 

My bad if i did not convey all msg. But i think i already shared all data. Issue is in accuracy with mechanical and those accuracy is more in negative temperature side like below 0C.

And i shared PCB layout image for your reference to see if that much no copper area near HDC2080 is sufficient?

Let me know what all inputs do you need other than above explained things. 

Dear Sumit - 

what exactly is the offset you are seeing here?

Dear Josh,

we don't have fix offset actually and that is one of the reason not to get it fixed by firmware by simply adding same offset in firmware.

Below is the observation:

1. On Negative temperature specially when i measured then offset is too much high.

For example, I have done testing of two devices with everything same configuration, Mechanical , firmware and tested with NABL certified chamber and master probe.

One sensor has offset of 1.68C while other has 0.58C, both devices kept even for long time (Approx two hours) to get it settled.

2. Attached excel sheet of two different temperature one with chamber temperature of -30C and other at 0C.

3. In first sheet we have temperature sensor with three sides vent (checked with two vents also as per guidelines) but no luck.

4. Then done one more experiment and in that i have kept one thermal pad on HDC2080 Ics and then one metal part which is exposed outside with mechanical. In that accuracy is coming better than other device configuration but humidity reading is not possible because here Ic is covered by metal and thermal pad.

Texas HDC2080 Discussionxlsx.xlsx

Sumit - 

unfortunately - it is still not clear what your issue is. it seems as though you are describing a temp difference, but i cannot tell where your reference probe is vs the sensor. if you are looking for correlation, you need to get reference sensor as close as possible to the device under test. So your issue may be in the way you are measuring it - perhaps some closer and clearer images of what you are doing here might help alot. 

Also, at cold temperatures, the temp sensor on the HDC2xx devices as shown on figure 3 on page 6 of the datasheet will deviate typically as shown. You should not attempt measuring %RH at the cold temps, as this really makes little sense in most applications. 

Lastly - on HDC2080, you need to subtract 40.62 at the end of the equation when calculating temp. (vs. just subtracting 40)

Dear Josh,

I am not doubting on measurement error because if that is the case then why there is better performance in expose mechanical pad. It should have same variations? 

I followed same measurement in all devices.

Secondly, I am wondering that you still did not get an issue after my all explanation. What is the meaning of temperature accuracy from your end ?

If HDC2080 is saying that it has accuracy of 0.2C to 0.5C then it should meet as compare to NABL/NIST certified probe. i am not doubting on Ic and seeking your help to see if i am following layout and mechanical as per guidelines.

Lastly, why are you suggesting to subtract 40.62 instead of 40?

as per datasheet it should be 40 only. Please suggest.

Hi Josh

I am referring same datasheet for temperature calculation and it is suggested to subtract 40 instead of 40.62, Please refer attached image for your reference.

From where you got the subtraction digit of 40.62?

Hi Josh,

which document you are referring for -40.62? Please share that, i wanted to go through that once.

Hi Josh

Please share snapshot got the same.

 I am not getting for your suggested unit for subtraction.

Hi Josh,

That is understood. I was asking about that whether same is mentioned in the document.

I just read carefully and found that looks like you are telling , this will be updated in datasheet. Is it ?

This 40.62 we have to use for whole temperature range?

One more query i have, In section 7.5, temperature accuracy is mentioned for 5C to +60C but in Figure 3 it is mentioned for complete range.

Why there are two different sentence? What is the point to not mentioned accuracy in Section 7.5 for negative range also?

Hi Josh,

Thanks for sharing information and i really appreciate this. 

Just last doubt i have, this formula will be applicable for negative range only or both negative and positive range?

I am getting good result in positive temperature range and if we will use this latest formula then we will have worst result.

Please suggest,

Hi josh

So are you saying that the modified formula(i.e. the one which subtracts 40.62) is applicable for all temperatures(positive as well as negative temperatures)?