solar 3-D printing with sand !!!!! after seeing the video now I want to build something like it I have the lens from a 50 inch rear projection TV.
large enough to print a 3-D House out of sand. ... it would take some time. And yes when you're house is finished it would be completely filled with sand. But in theory you could print the foundation with the plumbing pipes, plumbing fixtures(tubs,sinks,etc), counter tops, cabinets, furniture even down to the plates and cups.
I saw a robot that was just a concrete form on an arm, and it would "print" a house out of concrete. If you take the rig from the video, and put is on rails, you could move the whole mechanism and print all the walls. Better still, mount it on caterpillar treads and just drive all over the work site.
Imagine a rig big enough to operate on Mars. It could run for years building structures to have waiting for the colonists arrival. "Ferb, I know what we're going to do today!"
The only real drawback is, being made of sintered sand, after you move in you could no longer throw stones.
I just visited Pumping Station One last night, a Makespace/Hackerspace in Chicago. They were showing and working with an assortment of CNC in various stages of assembly. They have a laser cutter, a couple RapRap Mendel, a makerbot cupcake and a Thing-O-matic.
Reading about these and other machines only goes so far, actually seeing them work in person really made a difference.
please writ ein detail what are these differences!
When I started this thread almost 6 months ago, I was interested in the idea of a 3D printer, but did not know anything specific. I had read that one would interpret g-code and control stepper motors and an extruder head. Sounds simple enough, eh? So I looked at the input from the contributors to the thread, and gather more information. I found out interesting things about what exists and how they are used. As I collected more information, I discovered there was ever more critical information that I did not have. Most of the input from successful projects seemed to come from folks that had a lot of professional experience with professional equipment. This gave the impression that my meager knowledge and resources put this project out of reach.
But until I actually saw a couple machines running, I didn't really get feel for how the "information" relates to "real life". The Cupcake "sings" when it runs, you can make the stepper motors play a song in three part harmony. The thin plywood acts as a resonant surface or soundboard. That tells me exactly how much vibration and therefore what kind of positional error it will experience. Now it makes sense how a larger model would vibrate. Now I have a feel for what can be done with different size motors, and how much juice they need. I saw the machine print an example part, a gear with helical threads that is "very difficult to do with subtractive processes". Since I was able to hold the finished part in my hand and examine it, I know exactly what the machine produces.
Seeing the difference between an older models and a newer designs is kind of stunning how much personal 3D printer tech has advanced in a short time. There were also a couple design teachers from a local university, talking about DIYLILCNC. This is a do-it-yourself-CNC open source project. These guys are part of an effort that is moving towards standardization and open source hardware. Since I have no immediate need for a specific part to be printed right now, it seems to continue to be prudent and watch and wait, as today's unit is noticeably improved over yesterday's.
Maybe I see things different than other folks, maybe I'm just lazy. Seeing the machines operate and talking to the builders in person made all the information clear and put it in perspective.
thank you very much for your description. I really enjoyed reading it. It is the same with me. Before I was working at a company that developed CNC-engraving machines I thought oh what a high-tech
thing CNC is. Know that I have developed software for those machines and after lathing and milling some parts on conventional machines this technology is much closer to me and I can see what can be done
with different equipment. It is the same with hobbyist: you drive up the learning curve with your experience and improvements.
Reading your post points me again how important it is for students to work practically. Now a days I work as a teacher with different kind of students. The education system in germany is quite different from the USA.
But if I try to compare middle-school, high-school, and people learning a job like carpenter, baker etc. (I'm teaching the basic theoretic things related to lathing, milling etc. but I don't know the exact english words for jobs like industrial mechanic, building heating systems, steel stairs etc)
A part of my studentshaven't even graduated middle-school and get "a second chance"
it is important to win self-confidence through doing projects like the "egg-bot" of which they first think of it "I never could do that!" And then make the experience "Wow! great I can do it!"
Following the links from the 13 peice, $100 dolloar machine I ended up here. http://mtm.cba.mit.edu/
Among other things there is a machine made of cast cement. This would seem fairly cheap to make, easier to work than steel stock, and a bit more rigid than plywood. There are also examples of compensating for backlash in software.
This brings me back to the fanciful and unrealistic goals of 5x improvement in accuracy and 10x reduction in cost expressed earlier in this thread. Does this sound like these goal might now be achievable?
I'm no expert about building CNC machines. I guess with cement the parts have to be bigger and then will have more weight which will reduce speed-perfomance.
About 10x cost reduction: It depends on what is the reference cost. As long as you don't have big forces that bend the machine while milling plywoord might be good enough or maybe using a
little more special wood called multiplex. Backlash-compensating in softwre is doable. To increase precision reasonable you would need some kind of measuring equipment that is connected to the software.
This will cost too. But I think an anti-backlash nut will be easier. 5x more precision referenced to what? If the reference is 0,1mm so 5x more is 0,02mm this might be achievable.
If the reference is already 0,001mm and you want to go to 0,0002mm you go into the area where even standard ballbearings are not precise enough and you need a room with a constant temperature.
Another approach is to share a machine in a group. So cost reduces through the sharing. Or the parts for the things you want to build are adjustable to make them fit to each other.
Things like this might be an aproach too http://www.lignoshop.de/kugelvollauszug-seitlichen-montage.html?&campaign=froogle
Some of them are cheap and quite stiff with a low play.
If anyone wants to start with prop code, I am doing the hardware pcbs for the prop control. It has taken me a long time with other projects on the go too.
However, most of my circuit design is completed and the pcb has the parts placed but not routed. My concept is for a prop controller with SD (microSD), display (option for LCD/TV/VGA) and keyboard (PS2 mode - most USB keyboards still contain PS2 mode so you don't necessarily require a minidin6 version) and drives a pluggable pcb containing pluggable driver boards (initially using the pololu drivers). My intention is to also make a compatable pcb that uses an AVR so that users who prefer AVR can still do so, and then switch to the prop to get the extra features (SD, video, keyboard).
I have done some basic design modifications to simplify the mechanics of the reprap mendel http://forums.reprap.org/read.php?4,57039 . The Prusa mendel contains a lot of simplifications, but I still think there is more simplification possible. However, the electronics are similar, excepting for the motor sizes and hence driver pcbs. That is why the drivers need to be separated from the main electronics.
I have started on coding but there are several other propeller-heads that are working on it.
What is your opinion about trying to establish some basic standards on how the code should work.
I started an approach with using G-Code. I don't know much about reprap, Makerbot etc.
Am I right that they are using STL-files? Do I understand right that your control will be standalone without PC?
So the interface to PC is the SD-card containing the information what to build?
Are you thinking about a GUI? or do you think a textbased "G"UI so to say a TUI will be enough?
I guess a GUI will immediatly deserve one prop for graphics and a second prop of interpreting G-code or STL-coordinates
rep-rap is a layerbased additional process (I mean adding layer for layer on the created object) . Milling is a substractional process. Should the software can do both?
Maybe discussing all these details needs too much time. So let's set a timelimit How about disussing this for a week or two and then start coding even if not everything
is solved and some branches will exist. Like in the hardware too reprap, makerbot, DIY-CNC etc.
STL is a format that specifies the coordinates of lots of triangles that make up the surface of the solid, the order in which the vertices are listed specifies the external surface of each triangle. To 3D print you need to slice the STL file and then produce tool paths, the tool paths will be g-code or a similar variant. So the PC software for maker bot and similar will take in STL and produce the tool paths. Don't try and touch STL files with the propeller, they are big and processing them is inefficient and horrible.
Stefan: What I was/am aiming to do first is to just replace the arduino code with a prop. It should be reasonably simple because it is only a matter of converting the code (probably into Catalina C), separating the drivers out from interrupts.
Next, would be to display on a screen the temperature, etc as it prints. Then next would be to get the files held on the microSD and launch from the keyboard.
The prop is ideal for this sort of program and the code is open sourced. So, actually, I wasn't planning anything too complex from what has been done.
@Graham: thank you for clearing up about STL-files. Does this mean that - to the reprap/makerbot-control goes g-code?
@Cluso: you suggested to port the arduino-code (which is C?) to propeller-C as Catalina C. I'm familiar with Pascal/Delphi and SPIN.
I have done some very basic things in C with an embedded Artists ARM7-development-board.
But I would prefer to code in SPIN. I guess you know C very well too. What would be the pro's annd con's for Catalina C or SPIN?
I know almost nothing about Catalina C. Do you think that the control-software could be made modular so that
f.e. the part that creates three axis step and direction-signals could be written in catalina C
but could work the same way as an PASM-driver? meaning add PASM-code to a SPIN-file in the DAT-section?
Can you prvide a link to the arduino reprap sourcecode. I would like to look inside to see how complex this code is
and how many lines of code it is
Stefan: No I am not a c programmer although I have done a few c programs years ago. But it does not matter as long as something is running. Code can be changed. I just thought Catalina because it is LMM pasm, it would be so much faster than spin and easier to get running.
Driving the axes, etc, would be done most likely in pasm, and in separate cogs. IIRC someone started on a Gcode interpreter. The maths would be done by C or whatever, and the move control done by pasm cogs.
The reprap source is linked to on the wiki reprap pages. Probably best to try Gen6 software - there is a cute little name for a verson of Gen6 but I cannot recall atm.
I just ordered an ultimaker RepRap machine (www.ultimaker.com) and am planning to replace the electronics with a propeller board. As a result I started translating the firmware into SPIN.
As I will not have the hardware until Xmas (it is a looong wait I started with a very generic Gcode parser, which apart from line numbers and checksum it does not do any interpretation of the Gcode
what kind of propeller-hardware do you already have?
If you have at least a propeller EEPROM, chrystal and prop-plug you can test everything but the sd-card through a serial terminal program
like parallax serial terminal br@y's terminal or what ever
the terminalsoftware sends a single line of g-code to test the rest on a real propeller.
There is a special terminalsoftware emulating tv-output called PropTerminal. Downloadable from www.insonix.ch
I have an ybox2 that I want to start using for this (ethernet enabled 3D printer anyone?). I have not thought any further then that, but I will most likely need another prop for adding motor control and keep the ybox2 for the user interface part.
Sofar I have used spinsim for the development of this and as a result spinsim's conio (API of FDS) and fileio (API of FRSW). So it should run without much problem on real HW. But I first need to get the ybox2 running again as I messed it up when I tried to use a 6MHz crystal to play with the USB stack.
I've update the G-code parser (Stefan used your code as inspiration) and have attached the new version.
Started working now on support for the different G/M-codes, but I'm still fiddling around with how to best support coordinates. Also been looking at a lot of different firmwares for reprap based 3D printers and feel like this one has a nice way of working with the coordinates and has been shown to be pretty fast: https://github.com/kliment/Sprinter/
So I've started work with that one as base. Nice thing about this is also that it has implemented some M-codes to deal with SD cards.
I've ordered some Pololu stepper motor drivers and stepper motors so I can play around with driving these. Who knows before I get the ultimaker I might have build one from scratch ;-)
I am just waiting for my pcbs to arrive back from production. I have a A4982/A4984 equivalent Pololu pcb and a daughterboard with 12 prop pins and 12V to plug into a prop (e.g. my BaseBoard1). I haven't sent out the FET pcbs yet (for the extruder and platform heaters). I will post pics as soon as the pcbs arrive.
I have a reprap machine here driven by ramps. I'd love to see a competing prop board. My mind is bristling with ideas but I can't help wonder how cool it would be if it used the demo board that can be bought off the shelf from Radio Shack.
Are you planning on publishing your code + schematics on github so both communities can follow and contribute?
At the moment I'm not thinking about github, mainly because I can not really test the SW as I don't have an actual 3D printer yet. In terms of HW I was thinking of simply replacing the arduino with a propeller using the same driver hardware as used by the Ultimaker (as that is what I've ordered but I will not get it before Xmas, as I have it delivered to my parents place). I do have the code in a GIT repo so it should be easy to move this to github when the time comes.
Cluso99 seems to be much closer to getting a 3D printer up and running.
I would personally avoid the Thing-O-Matic and makerbot in general. They are nice machines but they are WAY overpriced, like double. You're paying for the Makerbot branding.
I own a reprap prusa and have access to a makerbot at the hackerspace.
My prusa prints faster, has a larger build area and produces identical or better quality prints.
A Makerbot is $2,500 assembled, $1,300 in kit form and is one design.
A Prusa can be scaled with cost and quality from $350 if you self source, $800 for an all-in-one-kit (higher quality components) to $1,400 tested and assembled.
Actually, Red, the MakerBot is just another design derative of the RepRap... with that said, it's an open source design. My second CupCake was built without any parts from MakerBot at all just using the design files from Thingiverse. It only cost me about $400 for everything, including the heated build platform.
I've also modified it so that it is a much "cleaner" design with all the electronics and pullys that were on the outside now nicely hidden in the bottom. The next modification I'm making is to "invert" the x/y and z axis so that it has a flying print head (like the ultimaker) with a stationary (except z axis) platform for the printed part. This will give me a build volume of 6"x7"x9" inside the box. (I've also upgraded to the stepstruder for finer control of the output... and can now easily print out in .1mm layers.) Btw... with the flying head, I should be able to achieve the same speeds as the Ultimaker since the only moving part now is just the hot end.
I have a prototype of my "Cluso's Micro-Mendel" that should hopefully be even simpler that the Prusa. Only time will tell. However, I found that I could extract the roundbar from old printers and used laser cartridges. The threaded rod was easy to buy. I found some bearings and parts from old rollerblades.
Electronics
I anticipate the electronics (Propeller based of course ) should cost <$100. Here is how I propose this...
Stepper Motor Motherboard
Small 45x45mm t/hole pcb has power connections for 4 stepper motors. There are 4 sockets for 4x Stepper Driver smt pcbs (0.6x0.8" Pololu compatible pcbs with Allegro A4982/3/4/8 stepper chip). This permits easy/cheap replacement for a blown driver chip. I have a Pololu compatible pcb. The interface to the is via 10 I/O pins (14 pin header) (Step & Dir for each stepper, plus common Reset & Enable).
Processor Board
Propeller based pcb preferred, although for software transition purposes, an Arduino can be used. I have a few small 45x45mm Propeller smt pcbs that can be used with the 14 pin header. The differences in my pcbs is the optional I/O such as microSD, TV/VGA, Keyboard. The Prop Protoboards can be used quite simply, as can many of the other Prop pcbs with expansion ports.
Heater/MOSFET Board
A small 45x45mm smt pcb with 2 MOSFETS, each to control heating output 12V-24V @ <15A. These 2 channels control the Extruder heater and optional Heated Bed. My current design (wip) uses an ATTiny84 to do the PWM control and sensor (thermistor) for each channel. It can be (re)programmed by a 4pin header to the propeller. The normal connection to the Propeller (or Arduino) is via 2 pins which can be serial or I2C. It is intended to be a set the temp and forget control. Because the MOSFETs are P channel and are small smt versions that should not require a heatsink (very low Rdson), this pcb is standalone. It is designed to be cheap in an event of failure. Power connections are via screw terminals utilising user-made cables with the readily available (and cheap) automotive style crimp connections.
Stepper Motors
Nema17 motors are preferred. I have not found a good cheap source for these yet. My mechanical design hopefully will only require 4 motors, being 1 each for X, Y & Z and 1 for the Extruder. The Prusa design simplified the Z axis at the expense of using 2 stepper motors. IMHO this increased the cost for a reduction in complexity. I believe I have a simpler design that removes the extra Z axis motor. Only time will tell.
Extruder
This is one area that is still under development as no solution seems to be adequate. Wade's Extruder seems to be the best, along with the lever modification and an alternate threaded drive bolt modification. I think this is going to be a buy-in design, at least for starters.
Heated Bed
This is another area that is still being developed. My initial design is small and so I may get way without the heated bed, at least initially.
Mechanical Design
My design is on the RepRap forum here. http://forums.reprap.org/read.php?4,57039
Note this thread is quite old and has evolved. I have a prototype cludge built using scrap timber for the joints to prove the concept. Unfortunately, lack of time over the past year has meant while I followed the improvements of the RepRap, I have been unable to pursue my interest as I would have liked.
I am building a micro version because I believe I can make it simpler and test it. The only part that will not be usable when I build a larger model is the actual mechanics (rods, bars and some plastics). All the motors, electronics, extruder, and some plastics should be re-usable. This is why I am not using cheaper and smaller stepper motors. I do not mind if I have a smaller build area.
Processor Chip
I want to use the Propeller chip because IMHO it is far more suited to the control of a RepRap than the other processor usually used. By making the electronics design modular, I see that my design can be used with the "Arduino" for those that wish it. They can then convert simply to using a "Propeller" based solution quite cheaply without having to throw away most of their electronics. And we prop users can use any pcb that we already own to do the job.
The future......
I am really interested to see what others have done, think etc.
Comments
I saw a robot that was just a concrete form on an arm, and it would "print" a house out of concrete. If you take the rig from the video, and put is on rails, you could move the whole mechanism and print all the walls. Better still, mount it on caterpillar treads and just drive all over the work site.
Imagine a rig big enough to operate on Mars. It could run for years building structures to have waiting for the colonists arrival. "Ferb, I know what we're going to do today!"
The only real drawback is, being made of sintered sand, after you move in you could no longer throw stones.
Reading about these and other machines only goes so far, actually seeing them work in person really made a difference.
please writ ein detail what are these differences!
best regards
Stefan
When I started this thread almost 6 months ago, I was interested in the idea of a 3D printer, but did not know anything specific. I had read that one would interpret g-code and control stepper motors and an extruder head. Sounds simple enough, eh? So I looked at the input from the contributors to the thread, and gather more information. I found out interesting things about what exists and how they are used. As I collected more information, I discovered there was ever more critical information that I did not have. Most of the input from successful projects seemed to come from folks that had a lot of professional experience with professional equipment. This gave the impression that my meager knowledge and resources put this project out of reach.
But until I actually saw a couple machines running, I didn't really get feel for how the "information" relates to "real life". The Cupcake "sings" when it runs, you can make the stepper motors play a song in three part harmony. The thin plywood acts as a resonant surface or soundboard. That tells me exactly how much vibration and therefore what kind of positional error it will experience. Now it makes sense how a larger model would vibrate. Now I have a feel for what can be done with different size motors, and how much juice they need. I saw the machine print an example part, a gear with helical threads that is "very difficult to do with subtractive processes". Since I was able to hold the finished part in my hand and examine it, I know exactly what the machine produces.
Seeing the difference between an older models and a newer designs is kind of stunning how much personal 3D printer tech has advanced in a short time. There were also a couple design teachers from a local university, talking about DIYLILCNC. This is a do-it-yourself-CNC open source project. These guys are part of an effort that is moving towards standardization and open source hardware. Since I have no immediate need for a specific part to be printed right now, it seems to continue to be prudent and watch and wait, as today's unit is noticeably improved over yesterday's.
Maybe I see things different than other folks, maybe I'm just lazy. Seeing the machines operate and talking to the builders in person made all the information clear and put it in perspective.
thank you very much for your description. I really enjoyed reading it. It is the same with me. Before I was working at a company that developed CNC-engraving machines I thought oh what a high-tech
thing CNC is. Know that I have developed software for those machines and after lathing and milling some parts on conventional machines this technology is much closer to me and I can see what can be done
with different equipment. It is the same with hobbyist: you drive up the learning curve with your experience and improvements.
Reading your post points me again how important it is for students to work practically. Now a days I work as a teacher with different kind of students. The education system in germany is quite different from the USA.
But if I try to compare middle-school, high-school, and people learning a job like carpenter, baker etc. (I'm teaching the basic theoretic things related to lathing, milling etc. but I don't know the exact english words for jobs like industrial mechanic, building heating systems, steel stairs etc)
A part of my studentshaven't even graduated middle-school and get "a second chance"
it is important to win self-confidence through doing projects like the "egg-bot" of which they first think of it "I never could do that!" And then make the experience "Wow! great I can do it!"
best regards
Stefan
Among other things there is a machine made of cast cement. This would seem fairly cheap to make, easier to work than steel stock, and a bit more rigid than plywood. There are also examples of compensating for backlash in software.
This brings me back to the fanciful and unrealistic goals of 5x improvement in accuracy and 10x reduction in cost expressed earlier in this thread. Does this sound like these goal might now be achievable?
About 10x cost reduction: It depends on what is the reference cost. As long as you don't have big forces that bend the machine while milling plywoord might be good enough or maybe using a
little more special wood called multiplex. Backlash-compensating in softwre is doable. To increase precision reasonable you would need some kind of measuring equipment that is connected to the software.
This will cost too. But I think an anti-backlash nut will be easier. 5x more precision referenced to what? If the reference is 0,1mm so 5x more is 0,02mm this might be achievable.
If the reference is already 0,001mm and you want to go to 0,0002mm you go into the area where even standard ballbearings are not precise enough and you need a room with a constant temperature.
Another approach is to share a machine in a group. So cost reduces through the sharing. Or the parts for the things you want to build are adjustable to make them fit to each other.
Things like this might be an aproach too http://www.lignoshop.de/kugelvollauszug-seitlichen-montage.html?&campaign=froogle
Some of them are cheap and quite stiff with a low play.
keep the questions coming
best regards
Stefan
However, most of my circuit design is completed and the pcb has the parts placed but not routed. My concept is for a prop controller with SD (microSD), display (option for LCD/TV/VGA) and keyboard (PS2 mode - most USB keyboards still contain PS2 mode so you don't necessarily require a minidin6 version) and drives a pluggable pcb containing pluggable driver boards (initially using the pololu drivers). My intention is to also make a compatable pcb that uses an AVR so that users who prefer AVR can still do so, and then switch to the prop to get the extra features (SD, video, keyboard).
I have done some basic design modifications to simplify the mechanics of the reprap mendel http://forums.reprap.org/read.php?4,57039 . The Prusa mendel contains a lot of simplifications, but I still think there is more simplification possible. However, the electronics are similar, excepting for the motor sizes and hence driver pcbs. That is why the drivers need to be separated from the main electronics.
I have started on coding but there are several other propeller-heads that are working on it.
What is your opinion about trying to establish some basic standards on how the code should work.
I started an approach with using G-Code. I don't know much about reprap, Makerbot etc.
Am I right that they are using STL-files? Do I understand right that your control will be standalone without PC?
So the interface to PC is the SD-card containing the information what to build?
Are you thinking about a GUI? or do you think a textbased "G"UI so to say a TUI will be enough?
I guess a GUI will immediatly deserve one prop for graphics and a second prop of interpreting G-code or STL-coordinates
rep-rap is a layerbased additional process (I mean adding layer for layer on the created object) . Milling is a substractional process. Should the software can do both?
Maybe discussing all these details needs too much time. So let's set a timelimit How about disussing this for a week or two and then start coding even if not everything
is solved and some branches will exist. Like in the hardware too reprap, makerbot, DIY-CNC etc.
keep the questions coming
best regards
Stefan
Graham
Next, would be to display on a screen the temperature, etc as it prints. Then next would be to get the files held on the microSD and launch from the keyboard.
The prop is ideal for this sort of program and the code is open sourced. So, actually, I wasn't planning anything too complex from what has been done.
@Cluso: you suggested to port the arduino-code (which is C?) to propeller-C as Catalina C. I'm familiar with Pascal/Delphi and SPIN.
I have done some very basic things in C with an embedded Artists ARM7-development-board.
But I would prefer to code in SPIN. I guess you know C very well too. What would be the pro's annd con's for Catalina C or SPIN?
I know almost nothing about Catalina C. Do you think that the control-software could be made modular so that
f.e. the part that creates three axis step and direction-signals could be written in catalina C
but could work the same way as an PASM-driver? meaning add PASM-code to a SPIN-file in the DAT-section?
Can you prvide a link to the arduino reprap sourcecode. I would like to look inside to see how complex this code is
and how many lines of code it is
keep the questions coming
best regards
Stefan
Driving the axes, etc, would be done most likely in pasm, and in separate cogs. IIRC someone started on a Gcode interpreter. The maths would be done by C or whatever, and the move control done by pasm cogs.
The reprap source is linked to on the wiki reprap pages. Probably best to try Gen6 software - there is a cute little name for a verson of Gen6 but I cannot recall atm.
is being presented next week if you are near Chicago
http://blog.workshop88.com/2011/07/13/the-shapeoko-a-great-looking-personal-cnc-machine/
I will have something to report shorly on pcbs for propeller electronics
As I will not have the hardware until Xmas (it is a looong wait
good start!
what kind of propeller-hardware do you already have?
If you have at least a propeller EEPROM, chrystal and prop-plug you can test everything but the sd-card through a serial terminal program
like parallax serial terminal br@y's terminal or what ever
the terminalsoftware sends a single line of g-code to test the rest on a real propeller.
There is a special terminalsoftware emulating tv-output called PropTerminal. Downloadable from www.insonix.ch
I wrote a g-code parser for the elements of a g-code line
you can download it from thris thread at post #204 http://forums.parallax.com/showthread.php?129448-Prop-based-CNC-3D-printer&p=1010848&viewfull=1#post1010848
please take a look at it if we could merge our codes together.
keep the questions coming
best regards
Stefan
I have an ybox2 that I want to start using for this (ethernet enabled 3D printer anyone?). I have not thought any further then that, but I will most likely need another prop for adding motor control and keep the ybox2 for the user interface part.
Sofar I have used spinsim for the development of this and as a result spinsim's conio (API of FDS) and fileio (API of FRSW). So it should run without much problem on real HW. But I first need to get the ybox2 running again as I messed it up when I tried to use a 6MHz crystal to play with the USB stack.
@Stefan: I'll have a look at your code.
Started working now on support for the different G/M-codes, but I'm still fiddling around with how to best support coordinates. Also been looking at a lot of different firmwares for reprap based 3D printers and feel like this one has a nice way of working with the coordinates and has been shown to be pretty fast: https://github.com/kliment/Sprinter/
So I've started work with that one as base. Nice thing about this is also that it has implemented some M-codes to deal with SD cards.
I've ordered some Pololu stepper motor drivers and stepper motors so I can play around with driving these. Who knows before I get the ultimaker I might have build one from scratch ;-)
I have a reprap machine here driven by ramps. I'd love to see a competing prop board. My mind is bristling with ideas but I can't help wonder how cool it would be if it used the demo board that can be bought off the shelf from Radio Shack.
Are you planning on publishing your code + schematics on github so both communities can follow and contribute?
Cluso99 seems to be much closer to getting a 3D printer up and running.
Bill
I own a reprap prusa and have access to a makerbot at the hackerspace.
My prusa prints faster, has a larger build area and produces identical or better quality prints.
A Makerbot is $2,500 assembled, $1,300 in kit form and is one design.
A Prusa can be scaled with cost and quality from $350 if you self source, $800 for an all-in-one-kit (higher quality components) to $1,400 tested and assembled.
I've also modified it so that it is a much "cleaner" design with all the electronics and pullys that were on the outside now nicely hidden in the bottom. The next modification I'm making is to "invert" the x/y and z axis so that it has a flying print head (like the ultimaker) with a stationary (except z axis) platform for the printed part. This will give me a build volume of 6"x7"x9" inside the box. (I've also upgraded to the stepstruder for finer control of the output... and can now easily print out in .1mm layers.) Btw... with the flying head, I should be able to achieve the same speeds as the Ultimaker since the only moving part now is just the hot end.
Bill
I have electronics pcbs back from mfr but I am away from home so cannot build them this week. I have discussed some info here on Phils MakerBot thread http://forums.parallax.com/showthread.php?133002-MakerBot-Build-Log-Prop-Control-Discussion/page4
and also here
http://forums.parallax.com/showthread.php?134134-Cluso-s-new-Propeller-stackable-and-pluggable-boards-(MultiProps-too)&highlight=pluggable
Electronics
I anticipate the electronics (Propeller based of course
Stepper Motor Motherboard
Small 45x45mm t/hole pcb has power connections for 4 stepper motors. There are 4 sockets for 4x Stepper Driver smt pcbs (0.6x0.8" Pololu compatible pcbs with Allegro A4982/3/4/8 stepper chip). This permits easy/cheap replacement for a blown driver chip. I have a Pololu compatible pcb. The interface to the is via 10 I/O pins (14 pin header) (Step & Dir for each stepper, plus common Reset & Enable).
Processor Board
Propeller based pcb preferred, although for software transition purposes, an Arduino can be used. I have a few small 45x45mm Propeller smt pcbs that can be used with the 14 pin header. The differences in my pcbs is the optional I/O such as microSD, TV/VGA, Keyboard. The Prop Protoboards can be used quite simply, as can many of the other Prop pcbs with expansion ports.
Heater/MOSFET Board
A small 45x45mm smt pcb with 2 MOSFETS, each to control heating output 12V-24V @ <15A. These 2 channels control the Extruder heater and optional Heated Bed. My current design (wip) uses an ATTiny84 to do the PWM control and sensor (thermistor) for each channel. It can be (re)programmed by a 4pin header to the propeller. The normal connection to the Propeller (or Arduino) is via 2 pins which can be serial or I2C. It is intended to be a set the temp and forget control. Because the MOSFETs are P channel and are small smt versions that should not require a heatsink (very low Rdson), this pcb is standalone. It is designed to be cheap in an event of failure. Power connections are via screw terminals utilising user-made cables with the readily available (and cheap) automotive style crimp connections.
Stepper Motors
Nema17 motors are preferred. I have not found a good cheap source for these yet. My mechanical design hopefully will only require 4 motors, being 1 each for X, Y & Z and 1 for the Extruder. The Prusa design simplified the Z axis at the expense of using 2 stepper motors. IMHO this increased the cost for a reduction in complexity. I believe I have a simpler design that removes the extra Z axis motor. Only time will tell.
Extruder
This is one area that is still under development as no solution seems to be adequate. Wade's Extruder seems to be the best, along with the lever modification and an alternate threaded drive bolt modification. I think this is going to be a buy-in design, at least for starters.
Heated Bed
This is another area that is still being developed. My initial design is small and so I may get way without the heated bed, at least initially.
Mechanical Design
My design is on the RepRap forum here.
http://forums.reprap.org/read.php?4,57039
Note this thread is quite old and has evolved. I have a prototype cludge built using scrap timber for the joints to prove the concept. Unfortunately, lack of time over the past year has meant while I followed the improvements of the RepRap, I have been unable to pursue my interest as I would have liked.
I am building a micro version because I believe I can make it simpler and test it. The only part that will not be usable when I build a larger model is the actual mechanics (rods, bars and some plastics). All the motors, electronics, extruder, and some plastics should be re-usable. This is why I am not using cheaper and smaller stepper motors. I do not mind if I have a smaller build area.
Processor Chip
I want to use the Propeller chip because IMHO it is far more suited to the control of a RepRap than the other processor usually used. By making the electronics design modular, I see that my design can be used with the "Arduino" for those that wish it. They can then convert simply to using a "Propeller" based solution quite cheaply without having to throw away most of their electronics. And we prop users can use any pcb that we already own to do the job.
The future......
I am really interested to see what others have done, think etc.
What price would be attractive for you?
http://www.watterott.com/de/Schrittmotor-Unipolar/Bipolar-200-Steps/Rev-42x48mm-40V-1200mA-NEMA-17
this is a german company selling NEMA 17 steppermotors at 15,50 Euro aprox $22.
I know with shipping much more.
http://www.reichelt.de/Schrittmotoren/QSH4218-40-033/index.html?;ACTION=3;LA=444;GROUP=C39;GROUPID=3299;ARTICLE=62653;START=0;SORT=artnr;OFFSET=16;SID=15YYoUvtS4AQ8AAG3eavQ8bb72a2d5a9813a1ea1edd6ed9b9c869
http://www.voelkner.de/products/83676/Schrittmotor-42mm-0.8-Bis-1.5a.html
keep the questions coming
best regards
Stefan