Nothing fancy just a picture dump and a few notes:
Basically the idea is to build a 3D printer styled on those wonderful machines by Zcorp. They are heavily patented so don't think you can ever make money by copying this but I'm sure we can have fun.
The printer I'm using is a HP1360, almost defunct but also almost identical to many others still available.
Step 1: Get the HP acting like a flatbed printer:
First attempts: Using the original feed motor and hoping that all the pre-print dancing did not run it off the end of the rails. It did! Drive is via a spring loaded cable looped around the paper roller on the left. Works well enough and tends to pop off in times of trouble! The box is from the local DIY store, it just looked like a printer.
That plan was less than successful so I had another idea, use a stepper motor to control the gantry and sync it to the encoder mounted on the paper feed roller, then just put the paper feed motor assembly in a little box and forget about it.
Here is is turned into a standalone assembly.
Motor on the right, encoder on the left.
Simple angle and hot-glue encoder mount. Not very pretty but its early days.
The encoder is now mounted on the end of a bearing cartridge from a hard drive arm. This is a push fit in a scrap of delrin.
A quick bend and I have two mounting lugs in suitable places.
To process the encoder pulses I am using a Parrallax propeller microcontroller, these are seriously cool and make development a breeze, they have 8 processors in the one chip and each can have its own job. Now they do proto-boards which are cheaper to buy than the components on them pretty much. Did I mention you can get TV out with just three additional resistors or with a few more VGA?
Success:
First print:
Sorry Zcorp, not funny I know :) Need a new cartridge on the printer and more height for the paper. The ink really jets out and has amazing range just just tend to get stray dots if you are too far away.
And here is a video:
That was so much fun I got carried away printing things:
Balsa and "wall foam". The new pigment inks work on most things.
Blue foam
More balsa and foam.
Fiddler's green sell awesome plans, this one could be made into a RC model directly if it wasn't glued to cardboard, need a vacbed for this stuff!
I know you could do this in photoshop but it just feels cooler
And that is cardboard!
Possibly loosing the plot at this point. Prints to PCB (you can etch it apparently if you get the board cleaning just right and then bake the board to harden the pigment ink). Printing through mesh was pointless but fun.
Then back to the plot. Lots of exciting time spent videoing the printer mechanism of a working printer to decide exactly how to start the synchronization, it turns out there is a reversal of feed direction after the paper is detected and I used this as well as the high speed of the motor just before. Now I have repeatable printing and another video:
Back to 3D:
I've covered the bed in alumnium foil tape to protect it from ink and make it more cleanable. The two piston like boxes have been added a long with a raised area and a hole for excess powder to fall through (to have a Tupperware box beneath it).
Its built a bit like a table but the legs do not reach the bottom of the box so it sits on the rim still. The lower plate will hold two stepper motors.
The source of my pistons, they could be scratch built but I'm trying to do this project in double quick time. I'd also prefer they were rectangular but it should still work fine. The foot plate is unscrewed and a delrin nut inserted into the hollow "con-rod", a threaded rod inserts in to this and is turned by a stepper. A simple aluminum channel and slider (to be finished) is used to stop it rotating.
The pistons, both had a hole in now blocked by tape and hot glue. The round holes are for the print heads to spray through on the cleaning cycles (again into Tupperware lined with absorbent mat).
Back on the box but still with gantry removed, looks almost printer like.
POWDER:
Leveling is done by a counter-roller, this spins fast and in the OPPOSITE direction to gantry when the powder is spread from the supply to the build drums. My first experiments are here:
Anodized ali tube and an electric drill. The powder is plaster just dumped from the packet. The mdf plate was raised up on paper on each pass, this shows about three pieces of papers worth of lift, possibly a little too much.
The roller acs a bit like a toothless cutter pushing the material along while not disturbing the underlying powder.
Not bad for unsieved and random plaster as well as manual feed control.
So here are the threaded rods that lift the pistons. There is a delrin nut in each of the con-rods and the lead screw is glued into a rigid aluminium adapter attached to a pair of (too big) stepper motors. To stop the pistons rotating a delrin tab attached to the con-rod runs in a bit of aluminium channel. I'm wondering if scratch building would not have been easier after all.
Messing around with some plaster, its not the right sort and I have not even sieved it which does make a difference, it tends to create little blocks of plaster that stick to the roller. The 10mm diameter alumium roller is running in needle roller bearings because that is what I had and they are low profile, it is being driven by a stupid little hobby motor with a 16:1 gear ratio and a little belt drive (currently under tensioned). Seems to work fairly well considering. You can see how it has pushed the powder from the round supply bin accross the build area and into the pit of dispair.
What's next?:
Get better plaster and sieve it, experiment with powder until I have it working just right, then replace print gantry and try printing, even raw ink should harden the plaster enough to allow later impregnation with epoxy etc.