Having a look at the image on the website and in git I have discovered that the sensors on my board do not seem to be correct. The problem presented itself when I was testing each sensor and noticed that the LED’s were not coming on in the correct sequence when cycling a test read.
In a clockwise manner from the top the sensors seem to be ordered accorging to the picture
Top - AS72651 and directly opposite the 405nm UV LED.
Right - AS72652- and directly opposite the 5700k White LED.
Left - AS72653 and opposite that the 875nm IR LED.
In my case the LED’s are in the right place but it appears the sensors are
Top - AS72653 and directly opposite the 405nm UV LED
Right - AS72652- and directly opposite the 5700k White LED.
Left - AS72651 and opposite that the 875nm IR LED
I have confirmed this by covering over each sensor taking a reading, then covering it over and taking another reading. That tells me what sensor is responding and what position its in. It does not tell me if that sensor is a UV or IR sensor. For this I put th unit in the dark and only turned on the LED then checked the sensor and the one with the highest level or non zeros would be the sensor for that spectrum.
Why I was chasing my tail on this is that I am working on a unified library that handles the Triad and the independent sensors and makes it easier to set up an array behind a MUX.
While polling the sensors and the LED’s I notices the wrong LED was showing for the sensor that was being read. UV was for the IR and IR was on for the UV hence why I dug a little deeper.
Any thoughts on this or was I just chasing my tail?
The sensors are individually marked, so please post a focused, closeup photo of your board, clearly showing the package markings.
The sensors on my board are mounted in agreement with the stock photo posted above, but it is possible that at some point, a pick and place machine was incorrectly configured.
Thanks for posting. As @jremington suggests, please do post a close-up photo of your board. Maybe we got the sensor tape reels mixed up? It’s important we find out.
As you can see, there are no markings.
I also checked the un opend board and that seems to be the same, no markings on the chips. I also tried the macro lense at varying angles to see if the markings were etched and not showing due to the light reflection but still no go.
The 875nm IR LED will probably give the strongest response on the AS72651 - it’s wavelength is closer to band W than band K. But the AS72652 will get plenty of signal too. This may be confusing you?
AS72651 drives the white LED.
AS72652 drives the IR LED.
AS72653 drives the UV LED.
You should be able to tell if the AS72653 is in the right position by turning on the connected UV LED. Point the light at some white paper. If you see blue / purple fluorescence, you know the UV LED is emitting. Beware that the white LED also emits a load of UV, in addition to the visible “white” hump at 575nm. The IR LED will appear a faint dull red to your eyes; beware that it is emitting a load of light which your eyes are not very sensitive too. Don’t stare at it closely for too long. Your cell phone camera probably has good IR sensitivity; point that at the IR LED to see if it is on.
I hope this helps. Please pick your way through this. If you’re still convinced the sensors have been swapped around, please let us know.
