No-Silicon digital clock

Yesterday I finished putting all the tube sockets and I placed some of the tubes, just to see if everything fits nicely. Sure it does!

I have also a beefy transformer for the filaments of the double triodes and rectifier tubes for the power supply. The dekatrons and nixie tubes don't have a filament but the double triodes do!

I also have selected a transformer for the HV.

Yesterday, I drilled out all the holes to the intended size, ranging from 3mm to 44mm. I made one mistake and drilled 35mm where it should have been 32mm. That will not be a large problem. I have polished the base plate to a medium shine. I have started to fit the tube sockets.

Today I manufactured the base plate. This process was quite hellish. I started with a design in FreeCad, designed al Aluminium plate 340x180x3 mm, placed all the tubes and their mounting holes (one large hole, plus two screw holes each) until it looked good. Some tubes will just sit in a ceramic tube socket. The nixie tubes are much shorted and will be raised on a separate base plate. The GS12C's have no tube socket, just solder lugs and will be mounted on a lowered plate. 

I had to really refresh my FreCad knowledge, because it is not such an easy programme to work with. I could have done the same work in KiCad, that would have been much easier.

I then wanted to generate a gcode file with the locations of pilot holes with the intention to drill small holes (1.5mm) on my small cnc machine in all the right places. Then later to enlarge these holes by hand to the desired size. I asked chatgpt how to do this, and it seemed to go quite well. FreeCad designed a path along which the drill was supposed to move from one hole to the next. All went well until the last step: generating the output file. This generated nothing bu the error message: Running the Python command 'Path_Post' failed: traceback[...] '_TempObject' object has no attribute 'InList' and no file is generated.

I put a question on social media and got several advices, one of which to upgrade to the latest FreeCad version. That helped, I had to jump through a lot of hoops but finally got a gcode file that was suitable for my cnc machine.

Then I realized my design was to large for the cnc machine, The job had to be split in half. I ended up with an excel sheet with all the x, y-coordinates of the holes and generating two gcode files from there. I sorted the y coordinates and did some excel wizardry to create a zig-zag cnc motion pattern.

Then started the struggle with the cnc machine. My advice if you want to buy a cnc machine: don't buy the <€150 ones. The usb connection keeps resetting (which means you have to do everything again). Someone on the net said it might be the power supply, so I tried with four different power supplies (ended with my max 50V (set to 24V), 5A bench supply. But it kept crashing. And several lead screws came loose and had to be redone during the process. Sigh.

After the best part of Saturday struggling, I had my base plate with 89 holes 1.5mm. On the drill press I enlarged all holes to 3mm and deburred them. See the photo for the result. Next job is to drill the big holes for the tubes with a step drill.