The recent very clear and sunny weather has really tested a couple of my sensors.
The LTR559 light level measurement seems to have topped out at 38k lux, which is well below the datasheet max of 64k+, and this was on a sunny window sill indoors.
In addition, my UV sensor output of UV Index, a mean of uvA and uvB, shot up to over 12 when I put it on the sill outside. That is well above the ‘dangerous’ level and not believable.
Has anyone else had these kind of values and corrected them?
I can’t help you with your main question, but want to point out something not every one is aware of.
Be aware that today’s plastics and most glass windows will block UV light, and mess with your sensor readings.
The LTR559 light level measurement seems to have topped out at 38k lux, which is well below the datasheet max of 64k+
So at a glance 38K is low but still within the range of outdoor light levels. Is there a particular reason you think it should be higher? From reading the datasheet if you’re getting readings of 38K then the sensor is set up to take readings right up to 64K, so it seems to be working fine.
In addition, my UV sensor output of UV Index, a mean of uvA and uvB, shot up to over 12 when I put it on the sill outside
Which UV sensor, and which library are you running it with?
EDIT:
This PDF on lights interfering with bats says that British summer sunshine is about 50K lux (page 6), so 38K doesn’t sound bad.
Thanks for the replies.
I understand UV is reduce by normal glass and some plastics, however my sensors are currently ‘naked’, although I am considering putting some PTFE tape over the UV sensor to widen its reception angle. Perhaps it is the continuous cloudless sky, I was just suspicious that the level flattened off at 38k. Maybe 38k lux looks about right as it has hit that for a couple of days now.
Sorry, yes, the UV sensor is a VEML6075 using the Pimoroni standard library.
I’ve honestly no idea why you’d get UV ratings that high. DEFRA monitoring suggested it was nudging the mid 7s. I wondered if it might be a setting on the sensor but the datasheet is pretty crap, it mentions a dynamic range but only says that flipping a bit “allows for increasing the dynamic”, whatever that means.
I wouldn’t go dilberately trying to increase the amount of light that the sensor collects, that will screw with the calculations (it will have been calibrated to calculate the UV index given its normal light collecting properties, if you increase the light collected you’re going to increase the index values it calculates).
I have an LTR559 and a VEML6075 UVA/B Sensor all hooked up and ready to go on my Breadboard rig. I’ll try and take it outside latter today and see what I get for readings. From the looks of our local weather forecast, today will be the last sunny day for a while. Five or six days of rain starting tomorrow.
The VEML6075 sensor is very directional and the PTFE tape is to reduce that, it will also probably reduce the level a bit. It is actually mentioned in the sensors __init__.py
https://github.com/pimoroni/veml6075-python/blob/master/library/veml6075/init.py
def convert_to_index(self, uva, uvb, uv_comp1, uv_comp2):
result = 0
# These values can be adjusted for calibration
uva_calib = 1
uvb_calib = 1
uv_comp1_calib = 1
uv_comp2_calib = 1
'''Coefficients for open air sensor and thin Teflon diffuser up to 0.25mm
For diffuser thickness of 0.4 mm and 0.7 mm other / lower IR coefficients need to be used. These are: uva_b_coef = 1.17 and uvb_d_coef = 1.58.
The visible cancellation coefficients stay the same.'''
uva_a_coef = 2.22
uva_b_coef = 1.33
uvb_c_coef = 2.95
uvb_d_coef = 1.74
uva_response = 0.001461
uvb_response = 0.002591
uva_calc = uva - ((uva_a_coef * uva_calib * uv_comp1) / uv_comp1_calib) - ((uva_b_coef * uva_calib * uv_comp2) / uv_comp2_calib)
uvb_calc = uvb - ((uvb_c_coef * uvb_calib * uv_comp1) / uv_comp1_calib) - ((uvb_d_coef * uvb_calib * uv_comp2) / uv_comp2_calib)
uva_index = uva_calc * (1 / uva_calib) * uva_response
uvb_index = uvb_calc * (1 / uvb_calib) * uvb_response
result = (uva_index + uvb_index) / 2
return uva_index, uvb_index, resultFor my location, Sydney Nova Scotia, Canada, Environment Canada reports the UV index as 8.
My Si1145 which approximates the UV index based on ambient and IR light is saying its 9
The VEML6075 gave me the following when placed in direct sunlight.
UVA: 12978, UVB: 13925, Comp 1: 2387, Comp 2: 971
UVA index: 9.33, UVB Index: 13.457, AVG: 11.3945
For LUX I got 37704 on the LTR559
It’s a bright sunny day with no clouds. 25c temperature
Similar figures here in Cornwall.
I have put a 0.7 multiplier in the code for the UV reading based on… instinct !
I think the figures look a little more sensible?
28c, 53%, 1006mb here in my bedroom at 3PM. Readings taken by a BME680.
I’m expecting the barometer to drop as the forecast is for rain or showers for the next few days.
