Lux Sensor RayBans II

Managed to find some RayBans for the Lux Sensor.

The specs in the photo are anti-dazzle driving glasses, which have tiny lenses on both outer corners. I managed to prise them out and put both on the photoelectric eye (you can see them on the photo), which was sufficient to reduce the reading from 800 odd at 08:26 down to 275, thereby leaving some room for increased brightness at midday.

The Chinese script on parts of the Lux Sensor control is intensely annoying, but I've found out what most of it means by trial and error.

Christ! I just realised that the Chinese could control my whole house..... Fiendish.

By 9am the reading was up to 423, at 10:40 it was 637 and continued to climb to 787 at 11:12, when it started to drop off again. Now most of this increase was due to the location of the sensor and had to be compensated for.

The following chart is the measurement over time, showing ups and downs, despite it being increasingly sunny as the day went on. The highs are when the sun was coming into the room from the east and west, with the dip between the two highs being when the sun was round the south of the house.

However, the principle of degrading the reading such that it remained within the upper end of the sensor's scale was proven.

Now comes the logic of the programming;

• I don't want the ASHP kicking in until the sun has sun has done its job via the solar thermal panels, so merely instructing the Lux Sensor to initiate the ASHP above a certain Lux reading could kick it off too early.
• Therefore a programme has to be written to start something after 11:30, that something being the initiation of  a programme that starts the ASHP only above a reading of 230 Lux.
• Similarly, there has to be a concomitant programme to switch the ASHP off if the Lux drops below 230. or else it could go cloudy with the ASHP still chuntering away, but with low solar power to run it, thereby defeating the objective.
• Then everything needs to stop at a certain time - roughly 13:00.

The programmes look like this:

The green routines are running all the time in the background, waiting for their allotted time to kick into action. Upon reaching their time, they pick up the inoperative routines and run them. There's a final programme that shuts them all down again, which isn't shown in the screenshot.

Working out the logic can be quite tricky, as you can easily start thinking like a human, rather than a machine that blindly follows instructions.

I may make one final adjustment and place the sensor in a 5 sided black box, with the open side facing the north-facing window, thereby eliminating the double high midmorning and mid afternoon.