Open scruzin opened 4 months ago
could be a silly question, but would a photometer pick up the light from the LEDs and potentially distort the data? or is there a particular wavelength that photometers are concerned with that will have nil impact from the LEDs?
I don't think so. The photometer would be aimed up, above the nav light LEDs. However, if there's stray light, we can mount the photometer inside a small upward-facing cone.
Also, the nav light LEDs should only come on at night (although we don't always bother).
How precise are we looking to be? From my brief research it seems like to get a good indication of solar irradiance we need a bit more circuitry then just a single sensor (See here). However, if we just want to measure brightness (across different wavelengths) then we could consider an option like this.
We need to decide on this component before I can finalise the new PCB design. If we are happy that the sensor will just be analog, I can leave an analog pin free for future development. However if we need another interface, we need to consider this before I can finish the design.
I think that Adafruit TSL2591 fits the bill nicely. As an I2C device, we could potentially integrate it with the existing ESP8266, not just the ESP32.
Also, at 19mm x 16mm, it looks like it is small enough to sit on top of the mast light LEDs. We'll just need to design a nice holder for it that integrates with the latter.
I think that Adafruit TSL2591 fits the bill nicely. As an I2C device, we could potentially integrate it with the existing ESP8266, not just the ESP32.
Also, at 19mm x 16mm, it looks like it is small enough to sit on top of the mast light LEDs. We'll just need to design a nice holder for it that integrates with the latter.
I will go ahead and order a couple for testing
I have received the light sensors for testing, and got them running with the ESP32 dev board that I have on hand.
SDA - GPIO21 (ESP32 I2C default) SCL - GPIO22 (ESP32 I2C default)
I have tested with the example program supplied by adafruit here. Which gives the following serial output from my desk:
[ 665001 ms ] IR: 5503 Full: 13122 Visible: 7619 Lux: 240.654953
[ 665863 ms ] IR: 5267 Full: 12641 Visible: 7374 Lux: 234.004257
[ 666725 ms ] IR: 4009 Full: 9834 Visible: 5825 Lux: 187.698395
[ 667587 ms ] IR: 3840 Full: 9391 Visible: 5551 Lux: 178.496414
[ 668449 ms ] IR: 3544 Full: 8749 Visible: 5205 Lux: 168.454254
[ 669311 ms ] IR: 2974 Full: 7434 Visible: 4460 Lux: 145.561340
[ 670173 ms ] IR: 2709 Full: 6812 Visible: 4103 Lux: 134.439636
[ 671035 ms ] IR: 2638 Full: 6647 Visible: 4009 Lux: 131.536224
[ 671897 ms ] IR: 2611 Full: 6583 Visible: 3972 Lux: 130.374771
[ 672759 ms ] IR: 2651 Full: 6678 Visible: 4027 Lux: 132.103928
[ 673621 ms ] IR: 2767 Full: 7069 Visible: 4302 Lux: 142.423523
[ 674483 ms ] IR: 2877 Full: 7221 Visible: 4344 Lux: 142.161224
[ 675345 ms ] IR: 2826 Full: 7108 Visible: 4282 Lux: 140.328156
[ 676207 ms ] IR: 2832 Full: 7122 Visible: 4290 Lux: 140.576111
[ 677069 ms ] IR: 2733 Full: 6887 Visible: 4154 Lux: 136.301865
At present, we have no direct way of measuring solar radiance (light intensity).
While the PV voltage loosely correlates with solar radiance, it cannot be relied on to measure the latter. Further more, there are times when panels are occluded and/or damaged, even when solar radiance is high. Having an independent way to measure solar radiance would enable us to detect such situations.
It is therefore proposed to mount a photometer inside the dome on top of the nav lights, at the very top of mast.