Parallel Printing: Mass-Production Printer

Got any crazy ideas?
Post Reply
Posts: 2
Joined: Fri Aug 04, 2017 6:39 am

Parallel Printing: Mass-Production Printer

Post by JacobGold » Fri Aug 04, 2017 8:34 am

Hi All,

I am working on building an open source small to medium batch production printer. The goal is to create a grid of hotends that all move together to create multiple copies of the same object. I am looking to fill the space between injection molding and huge print farms. At maximum capacity there will be six 150mm square print areas with six hotends arranged in a 3 across 2 deep pattern. Since the bed will essentially be one continuous surface, hotends can be moved out of the way to create different print area arrangements.

The premise is that this printer can be treated as if it has one hotend, hopefully both in maintenance and actual use. This sets a standard of ease of use and low maintenance.

This project is particularly difficult because it goes beyond the capabilities of the most advanced 3D printer board on the market (Smoothieboard v1 5x). I will need to add Arduino powered PID control for six hotends, to control both analog thermistor sensor and mosfets for the heater cartridge. Additionally, I have four lead screw stepper motors for my z control. I am working on a manual system to level the nozzles, but I need a way to level the bed with the "effector grid". Since I am using a motion system like the Ultimaker's, there are four outer rails that all the hotends are connected to. They form a plane that I reference as the "effector grid". I need to create a breakout that can control all four steppers both independently for leveling and trimming, but can also run them all at the same time during the 3D printing process as a unified z-axis. Smoothieware will treat this system as if it’s one stepper and one sensor. The breakout system will take the stepper motor data from the smoothieboard and a gcode command line for an internal auto leveling process (leveling the bed with the effector grid). I am planning to use either some sort of endstop or induction sensor as the z limit switch sensors.

Most of this stuff are just creative implementation of pretty standard Arduino systems, however I have very little experience with electronics and Arduino systems. I am first and foremost a mechanical person. Additionally, I need to figure out how to get the smootieboard to interface in the firmware and how to get those actions to happened using gcode.

One other thing I am looking at mostly for cost to performance benefits, is using lgus RJM-01-12 bushings instead of LMUU-12 linear ball bearings. Since the motion these bearings will see is both rotational and linear, I am wondering if there is any degradation of performance or accuracy with the Igus products. (Note. the igus bearing calculator says under these heavier than normal 3D printer load conditions, the bearings would still last forever.) My question really comes down to the motion dynamics and frictional forces put on the system. However, the frictional torque should still be minor as it is on a self-lubricating surface, only at a radius of 6mm, and the steppers should have no problem overcoming the low static frictional force before entering the lower dynamic friction.

Hopefully, it all works out.

I look forward to your thoughts and comments!


General Specs:
• Motion System: 6 Stepper Motors
o (XY) [2x 98Nm Steppers]. Ultimaker Style but with more cross arms. The perk of this layout is that those arms can be connected or disconnected from the belt to create different hotend alignments.
o (Z) [4 lead screw stepper motors].
• Extruder System: [Bowden, 1x 98Nm Stepper]. Since I have a max of six hotends that all should be printing the same part out of the same material, they should all be able to be run off the same stepper motor. I am building a custom extruder using dual gears and a stepper coupled to a 5mm shaft. I will line up six extruder assemblies here. The goal is to be able to handle flexibles.
• Electronics/Control: Smootieboard v1 5x with Arduino-powered breakouts (z-axis control w/ bed leveling, PID loops/auto-tuning and power control for 6 hotends).

User avatar
Posts: 232
Joined: Wed May 24, 2017 1:19 pm
Location: Hessen/Rhön - Germany

Re: Parallel Printing: Mass-Production Printer

Post by Moorviper » Fri Aug 04, 2017 4:13 pm

have fun

you will run into trouble because of different temperatures on the heatbed ;-)
and how often will you print 6x the same Part

and how will you level the 6 toolheads ?

Posts: 2
Joined: Fri Aug 04, 2017 6:39 am

Re: Parallel Printing: Mass-Production Printer

Post by JacobGold » Sat Aug 05, 2017 2:49 am

Moorviper wrote:
Fri Aug 04, 2017 4:13 pm
you will run into trouble because of different temperatures on the heatbed ;-)
and how often will you print 6x the same Part

and how will you level the 6 toolheads ?
I don't think I will run into problems on the bed because I am using a large aluminum plate with a silicon heater. These tend to be very good at distributing heat evenly across a surface. Additionally, this is all going to be in a heated enclosure.

To answer your second question, assuming that the bed and effector grid are level to each other, I am just going to let the hotends do the work for me. By this I mean I'm just going to bring the bed up and then loosen all the toolheads so they all sit contacting the bed. Then I tighten them all up and drop the bed. Since I plan on using inductive sensors, I can have the bed go to specific positions close to the effector grid. (This is true because the bed homes to the top).

Posts: 1
Joined: Sat Aug 05, 2017 12:23 pm

Re: Parallel Printing: Mass-Production Printer

Post by A10 » Sat Aug 05, 2017 12:48 pm

A cartesian Prusa setup with a stretched X axis with multiple hot ends linked at a fixed distance from each other would be another configuration worth considering. It would be easier to set the Z adjustment than with a matrix of hot ends, to monitor build progress and also allow for longer Y build space.

Posts: 1
Joined: Thu Aug 03, 2017 3:07 pm

Re: Parallel Printing: Mass-Production Printer

Post by dualite2 » Thu Aug 10, 2017 3:26 pm

When i read this i begin to think why not move only the build plate => static grid of extruder.

For the extruder part, a big steeper with gears for product the 6 movement that move the 6 filaments

For the part with the four stepper for each edge of the build plate i probably go for manually setting the alignment and just use one big stepper.

With this approach you can rest on standard electronic card and firmware (Smootieboard or other...).

If you not really happy with the big stepper approach, you probably need a card that can copy one input to X output stepper driver :
  • like 1 to 6 for the extruder
  • like 1 to 4 for the bed
However this approach also need manuel alignment for bed / hotend.

I hope i'm clear because i'm not really good with english.

Posts: 3
Joined: Sat Jan 20, 2018 11:31 pm

Re: Parallel Printing: Mass-Production Printer

Post by JeffDM » Wed Jan 24, 2018 2:09 pm

A Duet board might help you get what you want without having to develop custom controllers.

I suspect you might want to break out each extruder with its own motor. The extruder torque shaft might distort too much to be useful down the line.

Posts: 1
Joined: Thu Feb 22, 2018 8:09 am

Re: Parallel Printing: Mass-Production Printer

Post by RUWO » Thu Feb 22, 2018 3:46 pm

I am more then happy to help with the electronics part of the project. I can design schematics, write the code and design the pcb, if you want to market the product. The only thing i can’t do is actually build it for you, unless you live near to me in belgium(which i doubt). I am also capable of cad moddeling, would that be requiered. As for the leveling, i would keep the effector grid stationary and just level the aluminium plate. For such a large size bed, mesh bed leveling is almost a must. I know its almost impossible to do that the classic way on a multi head machine, but since the bed will be driven by four independant motors, i could modify the firmware to take inputs from different inductive sensors on the hotends, and change the bed level accordingly by turning the steppers different amounts. As for driving those four steppers at a time, you could use an approach simmilar to prusas multi extrusion. I would design a custom circuit which would operate in different modes: moving all the steppers the same amount, or just opperating one specific steppper individually. If you are interested in letting me help on this project, i would love to hear from you. Do you have discord/skype? That would speed up conversation a lot.

User avatar
Posts: 4
Joined: Sat Jun 24, 2017 12:30 pm

Re: Parallel Printing: Mass-Production Printer

Post by DevadathP.R » Fri Apr 27, 2018 4:11 pm

dualite2 wrote:
Thu Aug 10, 2017 3:26 pm
why not move only the build plate => static grid of extruder
Actually, bed moving in x and y axis is highly space inefficient . In a printer with hotend moving in x and y axis [eg: Ulitmaker, corexy] the size of the printer will only be a little more than the size of the bed. In a prusa i3 style printer, the size of the printer will be slightly more than 2 times the length of the bed in y axis, effectively doubling the footprint of the printer compared to a corexy with same build volume. Now in a printer with bed moving in X and y axis, the length of printer will be more than double the length of the bed in x and y axis, effectively the printer will be having a footprint more than 4 times the size if the bed. That's why i went for corexy for my printer build and i would only consider bed moving in x and y if the hotend happens to be a giant pellet extruder.

moreover, having 6 hotends, you probably have a big printer with a big bed made of aluminium which means it will be heavy, it is not a good idea to have a heavy moving mass anyway. So hotend moving in x and y axis would be your best bet. ;)

Post Reply

Who is online

Users browsing this forum: No registered users and 1 guest