After nearly a year of tweaking I’ve replaced the printed extrusions with aluminum. It worked well enough but flexed during operation and broke pieces occasionally. Now it can print nearly 900mm tall and I decided to bolt it to the wall for stability. Still dialing things in, the prints are getting better every day.
One thing that is nice is when I find another use for my printer. My Father was over one day and I knew he had modeled his house in SketchUp when doing some work on it. I decided to try and print out this floorplan
Once you have a floor plan or a model in a 3d modelling program, all you need to do is find the proper tool to export into a format the printer can use. In my case I had a plugin for SketchUp that allowed me to export the model as an STL file.
As I put this machine together from a kit and have been customizing since I have often wondered how fast it could really print. There are many articles about the limits of printing, I have always liked experience through trial and error.
When you are designing a part you need to keep this in mind. PLA shrinks as it cools, very slightly, but enough that it makes things not fit. My new printer is using 8mm rods and linear bearings for the carriage. To mount the rods I had made 8mm holes in the model where they needed to be. The rods do not fit, it is close but not great. To resolve this currently I am just going to buy an 8mm drill bit and enlarge the holes slightly.
If you are trying to print a large model on a small printer, you will come to a point that you need to attache the pieces to each other. Plastic epoxy and super glue are usually the first two that you would go to for this. Unfortunately, both of these options leave much to be desired when it comes to bonding strength. In all cases, even with special care to prep and sanding, the bond broke at the joint.
Because the end result is going into a new printer, it had to be stronger than what could be achieved with epoxy or glue. I have found the strongest way to bond PLA is with a solvent glue. This dissolves part of the print, which then hardens when the solvent evaporates fully. The resulting bond is so strong that the layers separate before the joint breaks. This is as strong as you can get, and hopefully the result will be strong enough to withstand the forces during printing. The solvent that I used is Weld-On #4.
One of the most tedious tasks owning a 3D printer is leveling the bed. The nozzle must be level with the bed surface at all points, within about 0.1mm. It can be done by hand, and must be done every time you remove the nozzle or bump the bed too hard. If you get it wrong, your print will not stick.
To fix this, I added a capacitive proximity sensor to my printer. It is adjustable and once it is set up the printer will measure the distance of the bed every time. There is still some calibration when removing the nozzle but it is a much simpler process and only needs done when the nozzle is changed. One problem I ran into was the sensor ran on 6V – 36V, but the microcontroller for the printer runs on 5V. Sending the sensor less than 6V didn’t work, but sending a signal to the microcontroller over 5V would break it. I ended up making a connector that had a voltage divider built in. The sensor then ran of the 12V that the heater, stepper motors, and hot end run off of, but sent the signal back through the voltage divider to the microcontroller. The end result is just under 5V at the microcontroller and a working bed leveling.
Here is the printer leveling the bed using the new sensor.
Mistakes or errors happen often when printing. How you recover from them is just as important. Sometimes you can let it just keep on going and get a partial success, especially when you are making many duplicates of something. For my new printer I am trying to make it out of mostly printed parts. This includes the beams that make up the shape of the printer.
Often tall slender objects will fail because they lose contact with the build plate. The leverage the nozzle has against the holding surface only increases as the parts grow taller. In this picture one of the beams had come loose from the surface about halfway through. It took it 6 hours to get this far, with another 3 to go. In this case it was better to just let it finish, having 5 successful beams instead of 6.
After learning so much about 3d printer, I am excited to continue. My current printer can make some pretty nice things, but it is limited in the size that it can make. There are many things that I would like to print that it just isn’t quite big enough. After doing some research of printer designs, I have decided to build me a CoreXY style printer. The first step was to design the XY section. There will be two stepper motors with two belts and pulleys to control the XY motion. I am building it out of 2020 extruded shapes but instead of using aluminum I am trying to do it with all printed pieces. This may not work, but it is worth a try.
Throughout the learning process more often than not you will end up with a blob of plastic. Maybe the settings weren’t right for the model, or it was too difficult for the printer to print. In this case, it was a model I had modified for the corner of a new printer and it had an error in it. The printer didn’t care that it was impossible to print, it tried anyways. The result was the mess you see here.