I would agree with the other David on this. If you do not have many drawings, it would be less expensive to just have a print shop print them for you.

 

Good idea. I will see if I can get the info together in the near future to add separate build for this. I did redo the pen gantry to use a pancake nema 17 which will give it a normal Z-axis rather than the servo motor. I have not had time to test it & configure yet but works seems to work well when manually moving the parts around & it also has a Max Z limit switch. Here is what it looks like. Essentially the same design as servo motor but put a pancake & limit switch on there. I also added option to screw another mount to where the limit switch hits so I could use an optical limit switch also.

 

My experience with servos for use with this pen plotter has been that great. I probably didn't do enough research into the use of servos with pen plotters. Is Inkscape & a GRBL type controller required to get the proper gcode to use the S-commands? The servos I am using get quite hot & I had one break on me. I am using a MG90S servo with metal gears. Seems like the Nema 17 would give more flexibility if GRGL is required.

 

Thanks for the info. It is good to have that info in this thread. I had not looked at using OpenBuildsCAM for this, mostly because I wanted it to be portable & be able to run either WiFi or with Raspberry Pi attached. I see there is a build of OpenBuildsCAM for Linux. Is there by chance a version that runs on a raspberry pi?

I did have the servo configured to raise the pen up & gravity with the small springs do the pen down, so not as much stress on servo on the pen down. It will be the same way for the Nema 17. The Nema 17 will give me more flexibility in how to create & send the gcode to plotter. I was using Klipper with Octoprint as it gave me a way to create my own gcode macro for pen up & pen down. If I kept that solution, I was going to write some post processing program to easily swap the necessary gcodes. I was using Estlcam V11 to create the gcode. I did recently try FluidNC on a MKS DLC32 board which has the ability to use a servo as a Z axis, but I couldn't get it to work for me. I also didn't spend a lot of time on it. I did get the firmware uploaded to the controller ok. If I played with this more & asked some questions in the discord, it would probably work. I also played around with v1pi (jeffeb3.github.io) which has the ability to run either Octoprint or CNC.js on a raspberry pi & also be able to turn the pi into its own hotspot so it doesn't need a WiFi network.

A better-quality servo motor would probably help, but that looks like about the same cost or more as a pancake Nema 17.

 

The BlackBox does seem like a really nice controller. I think it would be overkill for me in this situation though.

 

I looked at doing it that way, but after playing with that idea a while, decided to use a similar setup as I was with the servo motor. It will also make it easier for anyone to use either the motor or servo with similar hardware setup.

 

I read your message a little closer & see that A0 size paper is 1189mm x 841mm (46.811" x 33.11"). That is a pretty big sheet size. Biggest sheet size we used when I worked at A&E firm years ago was 42"x30". If the accuracy of this machine is good enough for you, this plotter with 1000mm length V-Slot 2040s gives me about 820mm width & most any length along wheel axis. Since you probably have a border width 12mm or so, this should be wide enough. Alternatively, you could use 1500mm length V-Slot 2040s which should give you about 1320mm width along the carriage axis. Since the inaccuracy will happen along the wheel axis, this would probably be a better configuration for your situation except the machine will take up more space because of the extra width. You will also need about 1 meter longer belt because of the extra 500mm length V-slots. Think that would be around 3.5 meters of belt. I think the belt length & V-Slot length would be the only change from the BOM I published on thingiverse. That reminds me that I do still need to update the files area for this build log to reflect the latest I have on thingiverse.

Does that give you enough information to answer your question?

 

That looks like a nice simple design that would be a lot easier to build. You should not have any accuracy problems with that once you have it configured. What size & length linear rails are you using?

 

I decided to add a drag chain to this rolling plotter. I spent more time than I care to admit getting this the way I wanted. My 1st big mistake was making the links so they could rotate close to 180 degrees. I had printed quite a few of the links before I realized the links across the top would hit the bottom links if they rotated that much. I also made some V-Slot Channel guides. I was initially going to use a 2020 for the guide rail, but the 2040 gave a wider plane for the chain to ride on. I also added some channel guides to make sure the chain stayed on path. I printed the chain links in sets of 8 for a total of 48 links. I will probably also add some press fit end caps to the top V-Slot 2040.

There is not a lot of weight on that top 2040, so only needed a bracket on each end to tie it to the bottom 2040 & only connected on the one side of both 2040, since I might limit the travel of the carriage.

I remixed this from another design but doesn’t look much like that design now. Here is my remix.
Drag Chain for my Rolling Plotter by GeoDave - Thingiverse