For this test I decided to use a much simpler object than my arm in order to test whether or not the complexity of my arm was causing the problems. While the paper tube solved one problem, it was so rigid that any force that was put on it would push it out of the way making the rest of the tattoo fail. I have decided that now I am going to try to re-scan my arm with 140 pts rather than 40, and see if that makes a difference.
I have had some problems with the gears moving out of place when I calibrate and also when I apply the tattoo. Earlier in the project I placed a long screw at the end of the timing pulley to keep the whole end piece pressed up against the back servo motor, but there were still problems. With the new piece I still need the screw there to keep the timing pulley in place, but the new piece keeps it from moving around too much.
Not only will this allow me to have higher quality results when calibrating, but it will also let me form my hand into a fist allowing more of my arm to be scanned and tattooed.
Once I get the marker off my arm I will be ready to test my full sized tribal tattoo (not that I would ever want a real tattoo like this). The design has 4x more points than either of the first tests so hopefully I can get a high resolution tattoo.
While it is just a straight line, it confirms that everything works as it should. There is a perfect amount of pen pressure which draws the line to be about 1 cm in width, enough so that any subtle movement in the machine wont register in the tattoo. Towards the end of the video the gears get caught up and stop spinning but that’s because I got excited and held onto the end piece too tightly.
Before I can call this project a success, I will need to do one last complicated tattoo, but other than that everything works exactly the way I want it to. I would do the complex tattoo now, but I am going to wait until I wash the green stripe off my arm so I can get the best results.
This video shows my first test of the drawing portion of the machine. The red squiggly line is the path that I designed for the machine to follow, and the green circle is the position of the pen along the path. This is not supposed to be a drawing anything in particular, just a good test for the machine to run. While I was very happy with the results of the test, there are quite a few factors that I haven’t been able to fully test yet.
The first and possibly the most problematic piece I have yet to test is how well the machine draws. While I can control the motors sufficiently, I still don’t know if I can get enough pressure from the pen servo to leave a mark on your arm, let alone a perfectly sized, consistently drawn line.
Another problem, this one clearer to see from the video, is that I physically can’t use it on my arm. To control the machine I need to move the support arch piece back and forth with one hand and control the position slider with the other, which means that I have no arm to put into the machine. I will either need to automate the slider, or get someone else to stick their arm into the machine for testing.
This half-minute video shows the full 40 point calibration process at 16x speed (the process takes 8 minutes when not sped up). The video also shows a fairly clear view of the computer monitor on the right which depicts how the 3d model is made simultaneously with the calibration.
The far left side of the video shows a missing third servo motor that I had to remove to make the machine work. I redesigned the end support arch to help the rods move, but unfortunately it wasn’t enough. Now that this is my third time attempting and failing this particular piece, the plan is to just abandon the third servo motor unless I can come up with some sort of a solution in the next few days.
I am currently uploading the video which shows the full 40 points being taken from my arm, however it doesn’t show a comparison between the two arm calibrations. With less rod wiggle and more points I have a much more arm-like result than I did earlier.
I still don’t have a specific distance that I allow the pen press into my arm, but I will determine that based on how well the pen draws on my skin at this level.
After playing around with different ways of securing the final servo to the old arch, I decided I should just bite the bullet and print out a new one. The new support piece is over 3x as thick as the previous design, and has a lot more support at the base of it. There is very little movement to it and all of the pieces fit into it much better.
I was not planning on putting the final motor in for a few more days, but I had to take an unscheduled break from programming and testing which freed up some of my time. Unfortunately the battery that powers the stepper motor has died and the charger wont arrive until later this week, so I won’t be able to work on the final part of the coding until that shows up.
I correctly assumed that the final servo would be a quick job with less than an hour of printing and assembly. And while everything did fit together and it was only 40 minutes of printing for the two parts, the piece didn’t work.
The video shows the idea I was going for, and in theory it works, but for now it just isn’t strong enough to support the weight. I actually think that the servo is strong enough, it is just the top of the structural arch that wants to bend with the new weight. I haven’t had too much time to think about how to fix this, but whatever I need to do, hopefully it doesn’t include printing a new arch piece.