Articles by Thiemo Nordenholz

  1. Testing the HA-B04 in-wall switch PCB

    I have built a prototype of the 6-way switch circuit board (photos below), and done an initial, very basic test: I can communicate with the microcontroller when I power the thing with 12V through its bus connector, and can use the on-board 16MHz crystal oscillator module (as opposed to the ...

  2. First in-wall switch board revision is a FAIL

    The six-button switch for in-wall mounting that I wrote about in March (no update for quite some time!) has been designed, the board layed out and manufactured.

    Today I wanted to populate the first board to have it handy while implementing the device firmware, however, I noticed that the main ...

  3. Planning an in-wall switch assembly

    I have started to work on a switch assembly to be mounted in standard in-wall mounting boxes, so that the finished thing will blend in with my other light switches and wall outlets, but have a bit more control capabilities.

    The switch buttons will be white (as are my other ...

  4. Fresh boards

    As noted in the last logbook entry, there are some new CAN I/O module boards which I have started populating. Hot-air soldering worked well for the output controller module's SMT parts, and the dispenser needle was actually still usable after two days in a sealed plastic bag. However ...

  5. Fume extractor for the E-Lab

    Finally I found the time to build a solder fume extractor for the "production" place in the lab:

    Fume extractor in window opening over work table

    The window insert is made out of plywood and lined with pipe insulation to give a tight fit against the walls around it. Mounted to it are two handles for easy insertion ...

  6. Screw mounts

    Experimental findings regarding screw mounts, comparing CAD data to actual PLA-printed parts. Prints were done with 0.1984mm layer height (i.e. "medium" quality).


    • 2.5mm is a tight fit
    • 2.8mm is more comfortable, but can be pulled out with certain force


    • 3.0mm is a ...
  7. End-to-end test successful

    Today I have debugged the problems constructing CAN datagrams described in a previous post.
    I found some flaws in my handling of the various parameters handed to the CAN message object constructor. As soon as I had fixed these, the end-to-end testing was successful!

    I press a switch, and the ...

Page 1 / 3 »