Buildlog Title: The Chimera Project - Laser + 3D Printing
Member Since: 2009-11-22
Tuesday, September 14th 2010 - 2:46 AM
I decided to try a nice long run to see how well the extruder is working. I saw a lot of these bottle openers at the Maker Faire.
This took about 45 minutes to run. The extruder ran flawlessly. My laser etched bed held it down firmly as you can see by the nice raft. I forgot to insert the penny before the picture was taken, but trust me, I am enjoying a beer right now. .
I have a really cheesy bed that I am doing my printing on and the parts don't stick real well. It is basically a scrap of Acrylic that I scuffed up a little. Just for fun I decided to run a cross hatch with the laser. I ran it about 0.50 in out of focus to get a wider sloppier line. I missed selecting 2 lines in the CAM program so there a couple gaps, but it should not matter. Eventually I will get a heated bed, so this is just temporary. I am going to use some magnets to hold the piece in place.
It was fun not having to even remove the part to go to laser mode. .
Is there a "Hello World" object that people print? That sounded like a lot of time, so I just printed a B. I need a real printing surface. This started to move after a while so I held it in place for 10 minutes. .
Extruder is mounted on the carriage. I just duct taped the 3D printer carriage and laser carriage together for the X motion. It worked fine.
The motor has never stalled or even slowed down with either method. It just slips on the ABS. I am using the MK5 drive wheel. I tweaked the design of the idler wheel and the split Acrylic pieces right around the wheel and that seamed to help. I also ran the temp up to about 240C. It is not a very robust design. It would be nice if it just worked...no adjustments needed.
I might upgrade to the whole MK5 system soon. I hate to pay for a whole system, because I have a lot of the parts already. I asked on Thingiverse about the material thickness, but no responses yet. I would like to hold out for the stepper motor too. The Bowden cable needs a high speed feed method.
I finally got my extruder working enough to print a few test prints. I need to make a decent bed to print on. I am just printing on a piece of Acrylic now. I am going to laser a cross hatch pattern into it to give it something to bite into. A heated bed would help too.
I had some trouble with the Z axis misbehaving. It would not move consistently. It appeared to be missing steps. I currently have very high steps/mm. I lowered the max speed and it does a lot better.
I can't get the Bowden cable system to push enough to get decent flow. Here is my setup. The cable is a smooth fit with just enough clearance, but it still adds a couple pounds of force. Note my little tensioner screws the push on the pinch wheel. That allowed me to fine tune the pressure, but still not enough. I was all a temporary setup of course. .
So I decided I need to crawl before I try to run, so I made a smallish mount for the motor right on the carriage. I will give it a try tomorrow. .
Back off the idler for quick feed...yes probably. That is a good idea.
I am not sure I meant the "compression" is totally in the filament. If the filament were incompressible the PTFE jacket might start to stretch to store some "pressure". There might also be some energy stored in the geometry of the loop. A classic Bowden cable is the hand brake on a bike. When you pull hard the jacket moves and and returns. That move back is a release of some energy. There is no visible effect like this, but it would happen slowly on my setup.
Videos: I would like to see just a free feed of material into the air so I get a feel for that. Then a print speed with that setting. I am working with a totally new machines.xml that is probably screwed up too.
I wish I had taken some pictures and videos. I just ran out of time last night. I'll post something soon.
I got the whole thing assembled and installed last night. The extruded heater and thermistor behaved perfectly getting up to temp rather quickly and holding there with a nice flat line. I fed some ABS into my motor and it started pushing it through the long 30" PTFE tube. It took about 5 minutes.
The head started squirting plastic, but the motor was slipping a lot and I did not feel I was getting much material out. I made a quick set of set screws to add a micro adjust method of pushing the idler wheel. That helped a lot and while I was not real happy with the plastic volume coming out, I decided to give a print a try. The sequence of everything looked good, but the head was moving way faster than the plastic was coming out. I ran out of time and stopped there. I need to figure out how to adjust for all that.
The Bowden cable appears to add some extra resistance to the motor. The bowden cable stores a lot of compression in the filament, so it continues to squirt a lot after stopping. I might see if I can add some extra reverse time to the stop sequence to relieve this. I have no real feel for what is right and what is not. I need to do some research and watch some videos.
So I debugged the cable problem pretty easily. I was then able to control the motion via the ReplicatorG control panel. +/- was right on all axes. I know you can easily change it, but how lucky is that.
I am new to RepRap, so I played around with it a little to get familiar with it. The speeds and resolution were all wrong so I needed to create a new driver for my machine. You do this by modifying machines.xml. It took me a while to figure out the one in the program directory is not the one you change on Rev 0018. It is in C:\Documents and Settings\bdring\.replicatorg. I copied the basic CupCake machine and only adjusted the axis parameters. See my settings below.
<machine> <name>Buildlog.Net G540 Laser</name> <geometry type="cartesian"> <!-- different pulleys on X and Y axii --> <axis id="x" length="300" maxfeedrate="1800" scale="49.2126"/> <axis id="y" length="300" maxfeedrate="1800" scale="49.2126"/> <axis id="z" length="300" maxfeedrate="500" scale="5669.2913"/> </geometry> <tools> <tool name="Pinch Wheel Extruder v1.1" type="extruder" material="abs" motor="true" floodcoolant="false" mistcoolant="false" fan="true" valve="false" collet="false" heater="true"/> </tools> <clamps></clamps> <driver name="sanguino3g"> <!-- optional, defaults to first serial port found. <portname>COM1</portname> --> <!-- required: we need 8 bit and 38400 baud. --> <rate>38400</rate> <parity>8</parity> <!-- optional, defaults to 1. <databits>1</databits> --> <!-- optional, defaults to N. <stopbits>N</stopbits> --> <debuglevel>0</debuglevel> </driver> <warmup> </warmup> <cooldown> (Turn off steppers after a build.) M18 </cooldown> <machine>
Since my extruder is not setup, I did not try to run a real STL file. I did create G-Code for one so I could study it. I made a simple "draw a box" program and it worked. There is no acceleration parameter, so you need to keep the speeds pretty low to avoid jerking.
I then wanted to have some fun so I wrote a very basic post processor for Vectric Aspire. I drew a quick Hello World and generated a g-code file from it. .
rr_post.jpg (10.25 KiB) Viewed 6282 times
I imported it into ReplicatorG and gave it a whirl. It is drawing HELLO WORLD. It would actually do it with the laser, but the speed is a little low and it would not turn of the laser between strokes because the laser enable is not hooked up to a RepRap. .
Next step is setup up the extruder and mounting the printer head on the track.
I temporarily mount all the reprap parts on a piece of MDF. This will keep them from moving around while I test. I used my Sparkfun "Premium Jumper Wires" to attach them to a 25Pin breakout board.
I got a nice 600W ATX power supply on sale at Fry's.
Unfortunately the RepRap motherboard is not moving the motors yet. I get what appears to be good Step and Direction signals to the G540 when viewed on my scope, but no movement yet. Does anybody have any ideas? I am not worried about things like steps/mm yet, I just want to see some movement.
I have the extruder motor section complete. I mounted it on the inside wall on a short piece of extrusion. I think I am going to move it to the outside. I don't think there is a good way to spool the plastic here. If I mount it to the outside I can build a little reel there too. There will be less to reconfigure when switching to laser mode. .
I think I am going to go the easy route and make a very simple MaketBot adapter PCB. This will allow either a parallel port or a MakerBot motherboard to control the steppers. I guess you could even hook this up so a parallel port controlled standard MakerBot stepper drivers. I choose the most common stepper board parallel port pinout.
I should be able to get about 25 of these for around $100 from Gold Phoenix. .
I cut the parts and test loosely assembled the extruder. This uses a stock MakerBot nozzle, barrel and insulator. The rest is custom. I threaded the lower metal plate. I thought this might reduce stress on the insulator. The tubing threads into a 1/4-20 coupling nut. .
I got the tubing from McMaster already. McMaster is amazing. I ordered it at about 1:30pm yesterday and it was delivered the same day! i am about 15 miles from the warehouse.
I thought a great way to attach this would be to thread the tubing into each end. It did an experiment on both the HDPE and PTFE, which are both 1/8" ID 1/4" O.D. I think it turned out great. That is a 1/4-20 thread. The trick is holding onto the slippery tubing. I used a 1/4" lathe collet. I put a 1/8" brass rod inside first to keep it straight and keep it from collapsing. The thread hold strong (about 40lb pull) on a standard thickness nut. The thread is very slippery so it will easily loosen. I think I might use a long coupling nut to get more thread engagement.
This is the PTFE tubing. It has a tighter ID and is more slippery, so that will be my first test. .
I decided to run with the suggestion rEd86 proposed. This moves the motor away from the moving head and feeds the material Bowden Cable style. This significantly reduces the size of moving head. I show it next to the laser lens carriage, but it could simply quick mount to the carriage.
I decided to go with mostly stock MaketBot parts to start with. I am sure many people can suggest improvements, but I need to get something going so I can get up to speed on how these things work.
I ordered the parts from Makerbot and have some tubing coming from McMaster. i will try both HDPE and PTFE. I want to get a feel for the flexibility and the friction. .
I cut the parts for one z carriage. There a few things I might change, but it should work for testing. The extrusion parts are on order and should be here mid next week. I want to get some Weld-On 3 Acrylic Cement to permanently fuse these parts together. I am not sure if I want to use 3 or 4 v groove wheels, so I allowed for both options and made it symmetrical. .
I also started modeling the MakerBot Plastruder. Their design purposely elevates the extruder head. I actually need to get it lower because the Z surface is lower on my design. I am going to order a unit from them soon. .
Chimera: An Animal descended from two genetically different animals.
After seeing all the quality 3D printed parts at Maker Faire I really wanted to build my own printer. While I was trying to justify the $600-$900 it would probably cost, I realized that my laser could probably be converted to a dual purpose machine. It already had all the motion control needed. That would save a lot of time and money. Also all the others building my laser could benefit.
Currently the Z Axis is not ideal for a printer because it is not constrained from XY motion while moving. It wiggles a bit when moving. I decided to add another v wheel track for the Z. I have also heard that some people are having a little wiggling when engraving at high speed. This should help.
I am using the widest Misumi 25mm extrusion available 80mm. I think that should work. I do not know if I need this on both ends, but I have material on order for two. I decided to try the smaller v-rail this time. I donâ€™t know why I went with the next size up for the laser. I think I had some already. I got the material and it should work fine. This rail could save about $50 on a build.