Reply
Posts: 4
Registered: ‎07-14-2016

SI1133 chip UV index range too big

Hi, I am evaluating the SI1133 chip for UV-Index measurement. Based on the document, http://www.silabs.com/Support%20Documents/TechnicalDocs/AN968-Si1133-UV-Index-Sensor-Electrical-and-... We are using the High Performance Compact Assembly. an ADC gain mode of 9, and only one A/D sample per reading, k is equal to 0.0187. But the raw data are range from 90 to 170 which means the UVI are range from 2.2 to 5.2. Under the same condition with the develop board, the raw data are range from 115 to 125 (UVI 3.2 to 3.47). We don't understand why the range is so big compare with the develop board. Please help us to identify the problem. Thank you
Posts: 97
Registered: ‎04-09-2015

Re: SI1133 chip UV index range too big

The default coefficients k&m are provided as a reference. Even though you are using the High-Performance Compact Assembly, the optical design won't be exactly the same as the one on our EVB. It is required to calibrate the UV result on your system.

 

Regards,

Tony

Posts: 4
Registered: ‎07-14-2016

Re: SI1133 chip UV index range too big

Hi Tony; Thank you very much for the quick reply. My real question is why the range of raw data is so big compare with dev board. I understand the system need calibrate the k & m. The UVI value is related with k & m values but the raw data range do not effect by k & m value. The raw data range is too big. Regardless any k & m values, the UVI range is big. Actually we think the UVI range of 2.2 to 5.2 is very close to the mean value of dev board reading (Range 3.2 to 3.47). Can you provide the k & m value to make the raw data range 90 to 170 with smaller UVI range? We think the actually UVI value is close to 3.3. We also did the test that put UV lamp on top of the sensor. The raw data reading is very stable (much smaller range). It matches the reading from the dev board (similar range). However, under the sun light, the data range is huge but dev board doesn't. Maybe this info can help you identify the root cause.
Posts: 97
Registered: ‎04-09-2015

Re: SI1133 chip UV index range too big

Basically, you're saying that the standard deviation of the UV raw data is higher in your system than on our EVB. I think that's caused by the optical design. You can remove the optical design in your system and the diffuser on our EVB to perform the UV measurement and see if the results are close.

 

I cannot provide the k&m for your system. You'll have to calibrate that by yourself: take multiple measurements at different UV index value and fit the result to a second order polynomial equation. That will give you the coefficients to calculate UV index in your system.

 

Regards,

Tony

 

 

Posts: 4
Registered: ‎07-14-2016

Re: SI1133 chip UV index range too big

Thanks Tony. We will try to put our optical design into the dev board to test. What is the reasonable standard deviation of raw data with proper optical design? Is there a method to connect our chip/optical design to dev board through I2S ?
Posts: 97
Registered: ‎04-09-2015

Re: SI1133 chip UV index range too big

You can put the switch to EXT on our EVB and jump wire from your system to the SCL, SDA, VDD, GND, INT pin through test points. The GUI can work with your sensor after that. If you switch to the Si115x Control Panel while running the demo, you can see the reported mean and std dev value. I don't have an exact number off the top of my head.

 

Regards,

Tony