Hi Jason,

Thank you for your post. I am unable to see your image. Can you try to reupload the image using the insert/edit media box dialog in the E2E post editor?

Jason Tu said:

   Our product had been using in Japan where the humidity may over 80%RH. And we found that from the spec, "Long exposure outside the recommended operating conditions may temporarily offset the RH output. The recommended humidity operating range is 20 to 80% RH (non-condensing) over 0 to 60°C." And refer to other thread, we got that baking can help to it.

   After 36hrs baking at 100℃, these hdc2010 sensor all recovers, but not completely. The shift decrease from +5% to 2.5%. 

1. How to deal with the remaining 2.5% shift? Is a calibration OK for it?

What temperature and humidity are you measuring the offset at? Also, are the device and reference enclosed in a controlled humidity chamber during this test?

Jason Tu said:

2. Do you have any suggestions if the operation environment  can exceed 80%RH? 

Typically exceeding the recommended operating conditions is a bad idea. For high-humidity and reasonable temperatures, the primary concern is condensation. When at high-humidity, rapid changes in temperature can promote the forming of condensation, that can saturate the sensing element and cause damage if not corrected. Your idea of running the heater at high humidity is the correct solution.

Jason Tu said:

3. We want enable the integrated heater to bake the sensor. But when should it be used? Our plan is when the sensor detect the envirment humidity exceeds 80% for a period of time(T1), then enable the heater for a period of time(T2). Do you have some suggestion about what the T1 and T2 should be?  

If you are regularly seeing humidity above 80%, i would not recommend using the heater at that point. It might be better to use the heater when you see a higher threshold, such as 90%. This higher-than-normal value would indicate the sensor was accumulating moisture. If you see 100% you would definitely want to turn the heater on.

I wold focus on selecting a good value for that threshold rather than an exact T1. it is unlikely that you could find a value of T1 that would work for all cases if you often see humidity from 80-90%. For T2 you may have to determine that value experimentally. It will depend on how long it takes to remove the accumulated moisture. 

Best Regards,
Brandon Fisher

Hi Brandon 

          "What temperature and humidity are you measuring the offset at? Also, are the device and reference enclosed in a controlled humidity chamber during this test?"

We did the test on a table of  a lab where the temperature and humidity is controlled. The temperature is 25~26℃, and the humidity varies slowly from 40%RH to 60%RH. The 2.5%RH shift is the average value. Our product target accracy is ±5%RH. So we want to calibrate the remaining +2.5%RH. Is it fessiable?

HI Brandon

    Can you share your suggestion for the remaining shift of about 2.5%RH after 36hrs baking at 100℃? Currently we have no way to calibrate unit by unit, because all of them had been ship to our custiomer for some time. It seems that continuing to increase the baking time does not have as obvious effect as before.

Hi Jason,

I wouldn't typically expect many parts to be 2.5% off from an ideal reference, and baking will only work to remove any excess moisture that accumulated on the part over time. What is the guaranteed accuracy of your reference? 

Also, how old are the parts you are assessing and have the boards you are using been washed at any point? 

Best Regards,
Brandon Fisher

Hi Brandon

Our meters are Rotronic HygroPalm HP21(accuaracy: ±1%RH) and Testo 175H1(accuaracy:±2%RH).
These parts D/C is 20171210. And We had check with manufacture that these boards had never been washed. And we also checked the Chemical material in the list of datasheet, but none of them are used in the production process of these boards.