The goal was to get rid of this Pi that has also a second use case to communicate with my BLE body scale. The solution to this will be in Part 2.
SDR - SoftwareDefinedRadio
I could have used the nanoCul on my Thinclient and relocate everything on that one, but the nanoCUL seems a bit flimsy. The Better option I found was an old DVB-T/DAB/FM USB Stick that has a Realtek RTL2832U Chipset. These sticks are commonly used as an SDR Stick for amateur radio operators. It can be used with Airspy SDRSharp Software to look around for interesting signals on the Waterfall plot.
RTL_433
Decoding the signals is quite easy with a Project called rtl_433 that, despite its name, is also capable to decode different frequencies besides from 433MHz. For testing, I first plugged it into my Mac Mini and used Homebrew to “brew install rtl_433”. After fiddling around with the command line settings, I settled on “rtl_433 -f 868.3M -R 117” which will start listening on 868.3MHz for Protocol 117 that is looking for ESA1000 / ESA2000 Energy Monitor signals. And soon I got my first result from my Powermeter:
Final Setup with Docker
I very much love Docker and fortunately there is also a Docker Container for rtl_433. This Container will run on my Thinclient that handles every hosted service on my Network. With lsusb I found out that the RTL2838 DVB-T Stick is Plugged in BUS001 as Device 005 with the ID 0bda:2838. The ThinClient is completely setup and maintained by my Ansible Playbook, where I also added the rtl_433 container with the settings for the MQTT Broker.
- name: Rtl_433
community.docker.docker_container:
name: rtl_433
state: started
image: hertzg/rtl_433
pull: true
restart_policy: unless-stopped
command: -f 868.3M -R 117 -F "mqtt://MQTTIP:1883,retain=0,devices=rtl_433[/id]"
devices:
- "/dev/bus/usb/001/005"
Right after the Container was set up, I got the first MQTT messages.
Unfortunately, in the dataset is no information about how much actual W were used in the last period. This needs to be calculated and was calculated before by fhem. It is quite easy in theory, but it took me a while to implement it in OpenHab. To calculate the Power we need to divide the used kWh between the last two messages, divide it by the time difference between the messages. This Value is then multiplied by 3600s to get kW. [kWhDiff/timeDiff*3600]
Getting this to work in Openhab took a while, but after a bit of googling and some trial and error, I got it working:
rule "Calculate Watt"
when
Item Strom_raw_total changed
then
//Strom_lastTime.postUpdate(Strom_raw_total.lastUpdate.toInstant().toEpochMilli()) //since the new item was empty from the beginning i manually initialised it once!
val time_interval = (Strom_raw_total.lastUpdate.toInstant().toEpochMilli() - Strom_lastTime.state as DecimalType) / 1000 // convert time interval to seconds
Strom_Timediff.postUpdate(time_interval)
Strom_lastTime.postUpdate(Strom_raw_total.lastUpdate.toInstant().toEpochMilli()) //save last timestamp
val watt_value = (Strom_raw_total.state as DecimalType - Strom_raw_total.previousState(true).state as DecimalType) / time_interval * 3600 // convert kWh/s to kW
Strom_actual.postUpdate(watt_value)
end
So every time when the total kWh changes, it calculates the time difference and then the actual kW. Saving the last timestamp gave me some errors, with the newly created item being initialized with NULL. To fix that, I had to run the rule manually with The first line of code active to initialize it.
Now everything is running as it was before with the fhem Setup.
]]>
To find the OSM Plugin in the UI, there is a small arrow which hides the menu. In the “OSM” Tab choose terrain[1] in the Dropdown and then hit the select[2] button which will open a Browser window to select the area of the map that your track is in. Choose it bigger than your actual track. Then paste[3] it into the UI and tick the box “ignore existing georeferencing”[4]. After that, you can hit import[5]. Depending on the size, you might not be able to see it, because it’s bigger than the workspace. We will shrink it later.
Now choose GPX[6] in the dropdown and navigate to your GPX track[7]. Tick the box “project GPX-Track on terrain”[8] but do not “Ignore existing georeferencing”! Now, you can also hit import[9].
To make it smaller, choose the terrain, and start to reduce the size with the scale in the “Item” menu. A good starting point is 0.005, but depending on the size it may be much smaller until it’s almost fitting your printer. If you want, you can exaggerate the height to make it more extreme. It is important to scale X and Y direction the same to keep everything in place. Note your scaling factor somewhere and apply it also to the track.
Add a cube to act as the cutout (Add → Mesh → Cube) and then scale and move it to fit the exact area you want to print. To make it easier to move around, adjust the “Viewport Display”[1] to only show the wireframe of the Cube. When the position and the size of the cube fits, click on the wrench[2] of the Terrain, “Add Boolean modifier”[3]. Use the intersect tool and as object, use the Cube[4]. This should present you a cutout of the terrain with a solid bottom. If everything is fine, hit Apply[5]. Now we have already a terrain model which would be printable.
To create the track, choose the track curve and in the Wrench menu it automatically has the shrinkwrapmodifier activated[1]. This will put the track onto the surface of the terrain. As a standard it has a distance of 0.5 m which we will set to 0[2] then hit apply[3]. The next step is making it solid.
In the “Object Data Properties”[4] we have to set a couple of things.
In the 3D print Menu[1] we can now export it as STL[2] to make really solid in Meshmixer.
Load your exported track.stl into Meshmixer. It should already look quite okay, but to be sure we will use the Make Solid Tool in the edit Menu[3]. I Basically turned “Solid Accuracy” and “Cull Edges Threshold” to max and “Mesh density” to sth. in the middle. Hit Update, accept[5] and then export the new solid STL.
Back in Blender, reimport the solidified Meshmixer STL. It should be in the exact same place as before.
Now choose the terrain model[1] and add a Boolean modifier[2]. Choose difference and then as Object your reimported Solid track. If you Hide the Solid STL in the view, you should see a nice carve in the surface of the terrain model. If everything looks good, hit apply. You could now export the Terrain as STL, and it would be ready to print.
To make it easier to remember the Hike, I added some informational text on the side. To do this, Add → Text. Press Tab to edit the Text, and Tab again to finish editing the text. To make it solid is similar to the track. Go to “Object Data Properties” and in the “geometry” part define the thickness with the Extrude Value[1]. Use the “Item” menu to scale it down and move it to the right position. Make sure it penetrates the model just a bit.
Now you have two options. You can either just print it as an “imprint” on the outer body, or you could also export the letters as STL to print it in color.
Imprint:
Convert your text to mesh. Scale it to just make a slight dent into the outer surface. Choose the terrain and add a boolean modifier Difference with the object text. Then hit apply. You can now export the Terrain as STL and print it together with the colored track
Colored Text:
Convert your text to mesh. Scale it to be a bit thicker inside and being almost flush with the outer surface. Choose the terrain and add a boolean modifier Difference with the object text. Then hit apply. Export the Text and the terrain as STL[3]. You might need to make the text manifold[2] before exporting (3D-Print menu → Cleanup → Make Manifold).
Violet: HR, Yellow: Targetpower, Green: Speed