15. Filling curve as an exemple

And now let's try to draw something like this interesting curve know as "Moore curve".
The underlying algo allows to fill a square in the methodical way.

The main purpose remains to test the ability to move the pen on the paper with the plotter:

And it works pretty well.

Actually I tried it with different different size of the square.

I also tested on a millimeter paper how precise my drawing was: its already nice yet there's room for improvement.

14. Z: moving the pen up & down

An additional stepper motor is required to move the pen up and down. For that purpose, a small 28BYJ48 should be enough:

Imax = 200 mA

To learn how A4988 driver works, here is a good explanation:



As per this explanation and documentation, VRef x 2 = Imax

Therefore VRef should be set to 0.100 V. Done.

Now the mechanics and electronics are ready:

And we can proceed with the first test, with a ball pen:

13. X, Y and a pen

It starts to look like what it is meant to be: a 2D plotter.

I just need to add a pen, which is achieved with nice linear shafts on the vertical axis. Yet no command to lift the pen at this time: it draws a continuous line for now.

It allows to make a simple test to check how precisely it closes a rectangle or write over the same lines several times. Here it is in motion:


Precision-wise, I'm not so happy: around 2 mm while I was targeting around 0.5 mm. I still have a bit of a way.

12. X and Y

One more axis.

I left my 2D plotter away for some time. Back with the 2nd axis: Y

For this Y-axis, I use 2 stepper motors, hence 2 servos, one for Y and the other one for A, configured to duplicate Y. I added two jumpers as shown here:

Initially, both stepper motors were turning in the same direction, which considering their position on the board is wrong. I needed one of the to turn in the opposite direction: easy, it's just a matter of inverting its connection on the CNC shield. Note the wires: black, green, red, blue on the left-inside (A) and opposite on the right in-side (Y).

I also created a wood box to hold electronic parts together.

Closed, the PC sits on the top and I have an emergency stop on the left in-side that works !

And now the result in motion:

Note just for the record: few useful G-code commands

• ! = Hold
• ~ = Start cycle
• ?  = Current Position/Status
• $$ = Current Settings

11. Not yet lost in translation

Got there. It's not perfect but it's a good progress.

I managed to built the gantry over which some kind of a cart can move, controlled by one axis of the CNC shield + Arduino.

Here it is in action:

Looks ok so far.

Yet I was surprised that the belt vibrates so much. Also the stepper motor is quite nosy, especially as the whole aluminum structure resonates.

Apparently the quality of piloting the stepper motor depends on the drivers I am using: I bought another set of drivers and I will check one of these days if it makes any difference.

10. Well it's complicated... :-)

Well I feel more comfortable with software, a little less with electronics. I am not really into mechanics, so I need to learn everything from scratch.

How to properly cut to get everything squared and aligned. How to drill a hole that fits my design, with the right position, angle...

All sorts of little details that seem so obvious until you do it.

I bought quite a few tools for that, such as a vertical drilling machine, a tap and die set,... and parts.

I also purchased wheels and belt for the transmission and drilled a hole for its axis: this is exactly when I discovered that it was not that easy... :-)

09. Designing the mechanics

Now stepping into mechanical design: after reviewing plenty videos and websites looking for the kind of cheap mechanical parts I can buy, I took my pencil, an eraser, my old wood board and found back some yellowish Canson paper I had bought last century.

Here is how my design look so far.

Front view

View from above

Two versions of a device to adjust belt tension