This project (as many projects do) started as a result of another project in which I needed a way to 3D print LOTS of things for the foreseeable future. Each one of things takes 2-3 hours to print so it wasn't really like I could babysit my printer all day. I wanted a little factory
For a while I was banking on PrintrBot releasing their PrintrBelt, which Brook Drumm had teased on a number of occasions. This printer used the typical angled printer head and belt system of an "infinite build volume" printer. The main difference from other infinite build volume printers on the market is that I suspect it would have been of the high quality and the reasonable cost of PrintrBot Products. Little did I know that it would never see the light of day.
Now of course there are some other people that are working on exactly what I wanted i.e. a printer that could just keep spitting out little bobbles until it ran out of filament with minimal human interaction. Most notable that included this project on Hackaday.io. Swaleh Owais did serve as a bit of inspiration for this project but I thought that there had to be a better way. His custom made infinite build volume printer was pretty cool, but I already had a perfectly good Printrbot Simple Metal. Why not just modify the one that I have?
Why don't we just replace the whole y-axis?
Initially I thought the best approach was going to be to replace the bed of the Printrbot Simple Metal entirely. That would mean that all the y-axis movements would be handled entirely by this hypothetical belt. For those of you who are not familiar with the Printrbot Simple Metal, please see the expertly illustrated rendering below for a general idea of what I was thinking.
This seemed like a not so bad idea if you didn't think about it so hard. Of course there were somethings that I had to figure out like how was I going to tension the bed? How was I going to keep this thing as 3D printable as possible without using much after market hardware? What was I going to eat for dinner? Sooo many questions. So the best thing think that I could think of doing was jumping in feet first into some CAD just to get the creative juices flowing and see what I could make. Things were a bit chaotic to say the least.
I initially started out thinking that I was going to need some sort of external truss structure type thing on the outside of the print to support the rollers and hopefully give me an idea about how to get an adjust about tension system. Below you can see a picture of one of the earliest attempts to make my dream of a PrintrFactory come true.
Yah. I am honestly not really sure what I was thinking here? I think that the idea was that the scoopy bit in the upper right hand corner would somehow lock y-axis bed assembly and then the two little do-hickeys on the left would be where the roller would slot. Based on the tone of the last few sentences you can probably guess that this didn't get too far. Despite feeling pretty good about this idea initially, I actually did some Googling to see if anyone had attempted this whole belt-replacement nonsense and sure enough someone had done it before......
...and I would love to show you how they went about doing it but for the life of me I cannot find the video. Seriously I think that the internet just gobbled it up. The gist of it was that the madman had his Printrbot Simple (the older laser cut model) clamped down to their desk and they had set up two big rollers that I believe had a belt made of brown (waxed?) paper stretched between them, with one side being driven by the original stepper motor of the y-axis. From what I could tell their solution wasn't viable due to instability and belt splipage. Oh also it needed to be clamped down to a freakin' desk and was like 3 feet across. So either way this poor fellow did the dirty work for me and led me to abandon my initial approach.
If we can't replace the y-axis bed, why not just make it more betterer?
So my next approach seemed to work a little bit better. If figured if we can't replace the y-axis with a belt, then let's just make the belt move along with the y-axis bed that is already there. Once again, although I wish this was original thought, I did have some inspiration from the guys over at 3Dator. They introduced me to the concept they call the "Belt Bed." Watch the quick video below to see what I mean.
Pretty slick right? Sounded a little crunchy but I have to say it was exactly what was after. I want to point out once again that I wasn't really after and infinite build volume printer, more of just a printer that could print infinitely many, fixed volume things. Make sense?
After this video one of my primary question was what the belt was made out of. Obviously it had the familiar golden hue of kapton so we knew that much at least. However, according to the comment section, the belt was "specially made" for the 3Dator people. Uh....okay? I guess we can come back to that (spoiler alert: I just used kapton tape and printer paper!).
With this new concept of a belt bed in mind I got back to modeling and after a few more sane iterations and virtual measurements, I came up with this cute little guy:
This post is getting a bit long winded and really is only an introduction into my journey so I will cut to the chase. I printed out 4 of these (with the 4th having a special mount for the motor(not pictured)) and screwed them down to the corners of the y-carriage. The heated bed of the Printrbot would then slot in and be screwed down in that little cut out that you can hopefully see. For the bearings I used just normal 608 inline skate/fidget spinner bearings because my hope was to keep this as limited in aftermarket hardware as possible. I cannibalized some from my little brother's fidget spinners and an old roller skate I found in my garage.
I technically modified a servo for continous rotation, and by that I mean it literally cut out the limiting nubby thing on the gears and soldered the power cables right to the motor terminals. That way I had a nice little gearbox. I then soldered a little motor driver circuit that consisted no more than of a transistor, a SMD linear regulator (that had no idea what it was infor), and some male header to connect everything together. As a nice little touch I also added an "override" button that would spin the motor without any sort of need for software input.
With the mounting hardware and the electronics figured out, all that was left was the software side of things. I just needed a way to toggle one of the spare GPIO pins on the PrintrBoard. I figured that there might already be some sort of gcode command for that and sure enough look at that!!
But of course after fiddling around for a few hours and checking out a whole host of different pin naming schemes, I could not for the life of me figure out why the command was not working. So eventually I just broke down and opened up the Printrbot Marlin code and added my own g-code command to do a digitalWrite to the GPIO pin that I found was the only one to actually toggle when it was supposed to. I thought that was going to be the easy part! However once things had been compiled and flashed and all buttoned up, that is when the magic really happened! Below you can see one of the first automated print ejections of the PrintrFactry Mk1 !!
So there we have it for now. If requested I will go into more detail about things however I think that is a pretty good summation of things so far. However, there was still one major issue with this design bed tensioning. Everything may look fine and dandy in the video but let me tell you that thing was wobbling around so much as it printed. So in the Mk2 design I came up with a pretty clever little design that I think you will enjoy.
and yes I am writing all of this in the past tense. I have made some pretty good progress so far and I figured that I would try to play my part on the internet and try to document what I have done this far. Let me know what you think of all of this!