@tomtom1234 I've been meaning to do that, will take me a few days.
@Simbungu would you mind sharing your Stevenson Screen design?
Thank you for the hard work, I'm really enjoying the station! I'm using MQTT: base station -> mqtt server -> weewx -> https://www.pwsweather.com/station/pws/SD4S92127
@Simbungu Thanks for your donation, I appreciate it. :-)
That makes sense. I also have a M3 washer in between the AnemoFlyersMount & bearing in AnemoReedMount. I think the washer may be touching the seal on the (2RS) bearing.
AnemoFlyersMount.stl is what I reference with “upper part”. Actually, this last discussion should go to part 5.
The upper part is sitting on the bearing. This is the point where friction is biggest. Initially, I had two washers between the upper part and the bearing. This results in a small distance between moving plastic parts and replaces plastic on bearing friction by metal on metal friction. In the latest version of the upper part, there is a deepening / recess for a nut like this: https://www.frantos.com/de/matrix/index/sCategory/3250 The round head is pointing down and actually sitting on the inner part of the bearing in the lower part. It stands out a bit so there is a small distance between the plastic parts too. Both the version with the washers and the new one start turning at around 0.3m/s. But true, it needs to be fine tuned. In case the M3 axis is slightly bent (which can happen), this will add a high resistance for starting too. I’ll check if I have a good picture of this area.
I'm having problems with the anemometer reading very low. Using an accurate handheld anemo, the wind has to be around 2.5m/s before the printed anemo starts turning - it just has too much mass and/or resistance. I've experimented with larger 634 bearings and a M4 shaft, which seems to help but it is still takes a lot of wind to make it move. Any suggestions? I think I may try larger cups and thinner walls to reduce mass.
Yes, this solar tracking will not be required for regions like yours. For us, at 54°N, winter time means that the sun is coming up 20° above horizon only on the shortest day. During this time, and with gray clouds for weeks, I had to change the battery pack 4 times this winter. Not sure a tracker would have changed that a lot, a bigger panel is probably better.
I notice some people use a pyranometer to measure solar radiation, not sure I'll go that far. I've been monitoring the Lux values I get during different weather conditions. I find it useful to track daylight hours, the readings are accurate enough inside the screen to use.
BTW I don't think I have to build the solar tracker, we have so much sun in Southern California, the 2x18650 station batteries are recharged in around 2-3 hours!
To measure sun hours (which is defined by a certain Lux threshold measured exactly pointing to sun position), either a diffusor or some sun position tracker is required. I have experimented with a mathematical compensation calculated from sun position and sensor characteristics - but this has not been too successful. Most use a diffusor because it is the simplest solution.
I used the VEML7700 which measures luminance in lux. I may add or replace this in future with a VEML6075 UVA UVB and UV Index Sensor.
@Simbungu Please share which ambient light sensor you have used.
I modified this excellent design slightly by using alternate components in the wind sensor and adding an ambient light sensor. I also added a color LCD on the base station to display current readings. I designed my own Stevenson Screen to house the weather station and sensors and several brackets to mount all the components.
The software is comprehensive, well designed and written. I added code changes for the wind vane sensor, LCD, ambient light sensor and an MQTT client, all of which were fairly simple to integrate.
This is by no means a trivial project, and will test all your DIY skills, but I found it extremely rewarding!