Hmmm... for a 3D printer I can see you'll need an extruder, ~$100-200 depending on route taken. 4th stepper controller, $15 (for the extruder drive), possibly a heated bed for $35 (required for ABS with that size of build surface) and a few miscenallous parts.
I'd personally extend the Z axis, which would obviously add to the costs... but most of the existing parts can be reused.)
You'd easily be able to get away with ~$200 - $250, I'd think as long as your existing controller can control the 4th channel.
Thanks for the input. Actually the magic number would be five, because I have two steppers on the Y axis. As far as the Z axis goes, I have something in particular that I would like to model on a regular basis, and the table should be just about right for this purpose. However, I am sure there will come a time when I will wish I had adjusted the Z axis height.
$250 sounds reasonable.
Anyhow, thanks once again for the input. I will have to dig around in this thread for a bit just to get more info.
coming from some one who has studied the riprap 3D printer for a year and still goes "huh?"
lets make a prop based 3D printer!
suggestion inkjet filled with an active agent sprayed on a powder.
plaster of paris+ colored water= pretty but delicate
baking soda + thin 2 part epoxy split into color ink tanks = strong but exspensive sort of.
use an old parallel printer have the PROP print a "page" as a slice in your powder spread more powder and lower platform print another page and so on. i have had this ideal but lack the money to try it currently. plus i lack knowledge of epoxies.:nerd:
Actually the magic number would be five, because I have two steppers on the Y axis.
Bruce, those two steppers can still be controlled via the same controller, so you still only need four. In fact, you can use two steppers on the Z axis as well, like the Prusa Mendel does for a nice easy heavy lift scenerio.
Bruce, those two steppers can still be controlled via the same controller
Alright Bill, you have my attention, because I have never attempted that and it would save me a controller. Considering that I will be using GeckoDrive's G251X, how can this be accomplished, and is it possible to run both motors at full current capacity.
Additionally, there was a fairly large Steele family in Indiana several years ago with the siblings being named, Jean, JoAnn, Charlie, Tom, John, and one other pretty little thing, but I forget her name. Are you any relation?
The Two Z-axis motors on that Prusa are simply directly connected to the stepper driver.
One source indicated serial connection, another suggested parallel was best. I tried both, I think the ended up about the same. (Right now the wires are covered, so I can't see how I left it. But, it works just fine with
NEMA 17 motors http://www.pololu.com/catalog/product/1200/faqs
and
pololu 4988 stepper drivers http://www.pololu.com/catalog/product/1182
Sal estimates we might have something in April.
We are looking at a prop based drop in replacement for Sanguino electronics on RepRap, with a G-code interpreter implemented in propforth.
Actually, I'm in Indiana too, but I don't know any of those names. I'm from the West Harrison area.
As for the motors... the controller is the limitting factor. I'd imagine that you could run both motors at 'their' capacity as long as the controller could drive both within its limits. That GeckDrive looks like it can handle 3.5 amps... which is enough considering that I drive my large laser with half that for two motors!
I got it to slice, its really fast, but it starts printing at 1.0 mm Z (above the bed). I only have an STL file for this part, and I don't have access to the original application that created the part.
KISSlicer automatically drops (or raises) the lowest point in the model to sit right at Z=0. If you select a grid or pillar raft then the object will sit on that, naturally, and the thickness of the bottom layer of the raft is a function of the "Bed Roughness" setting. The only other thing I can think of that would affect the Z values in the resulting G-code is the "Offset...Z" box. If the Z values in the G-code look correct, but the head is physically 1mm up off the bed. then we're looking at a different class of issues. Would you mind zipping up the output G-code file and emailing it to me? (Best email address for this is info at kisslicer dot com.)
Like others before me, I have followed the thread beginning to end/now, and thought I would throw in my 2 cents. A) Hoosier, take that for whatever you think it is worth. My Prusa has run with Marlin firmware. It is currently on Sprinter due to some printing issues. The toolchain is OpenSCAD/slic3r/pronterface. C) I have been curious, is Cluso going to have headers for endstop and thermistor connections?
The Marlin fw has some useful capabilities built in that allow for an LCD, encoder wheel, and an SD card to be all you need to make the printer print aka standalone.
I have also seen a guy build a healthy(steel square tube 2"x3"x1/8") CNC machine and try to use it to print with. It was a failure and he will gladly tell you as much. There was too much mass to move quick or even well enough to print with. I have seen other mention it. I really don't imagine being able to use a 3d printer as a CNC mill. Structural integrity is once again the culprit. I saw someone comment 3d printers aren't fast. My friend built a Prusa with linear bearings and can easily print at 150mm/s. He has printed at better than 200mm/s.
I've seen slic3r, but did not try it yet. I just got Kisslicer to run properly, and it is very much faster that pronterface/SFACT/Skeinforge.
My question is about slic3r's "easy configuration and little or no calibration". Don't we need to measure the output filament width, heights, and adjust the flow rate, etc?
I have a bunch of problems with the bottom solid layers being overfilled, and the top layers not filling in solid. I think this is due to improper calibration, but I don't know the process for Kisslicer. The process for Pronterface/SFACT/Skeinforge is pretty thorough, but I only got as far as I did (due to inexperience).
Do you use think filament stock 3mm, or thin filament stock 1.75, a wide nozzle 0.3 mm or think nozzle (what ever the small one is?) , PLA or ABS?
My printer uses 3mm, and, squirts it through a .35mm nozzle. Most of the time I am printing PLA. My printer has been used to make some ABS parts, but, it was difficult from what I hear. The problem as I understand it was getting the heat required while not heating up the plastic before the nozzle (I have no fan).
So far, I've only printed in ABS... and just now got my first roll of PLA in. I have no problem printing anything I want with it. I use a heated build platform and set the extruder to a fix temperature (based on the color of ABS I'm using.)
I'll be firing up the PLA tomorrow... I understand it smells like waffles cooking when you use it, so I can't wait to experience it myself.
The problem as I understand it was getting the heat required while not heating up the plastic before the nozzle (I have no fan).
I used a 40mm fan from either an old pentium 1 or a vga cooler. I just wired it to a spare header on the power supply, along with some Ikea LED book shelf strips, and they run whenever power is on.
Can I throw my hat into the ring here? Just finished my Linear Prusa build. I'm using Marlin firmware and Pronterface as recommended by the local 3d printing community group. I have some PLA and ABS filament, and a manufacturer just around the road to make up some 'specials'
I'm using a ramps 1.4 driver board. Ultimately I'd like to replace the Mega2560 with a prop based board, so we can have something like the Pronterface GUI rendered by the prop. The easiest approach I have found so far would be to adapt Ariba's PropinoPP (one of the Smorgasboard boards) - it could involve as little as half a dozen jumper wires to make this work.
Just taken a step further....
Ordered Prusa V2 plastic with LM8UU bearings.
Looking to use GT2 6x2mm 36t pulleys and belts
Looking at the J-Head Mk V-B Hot End
Just finished my Linear Prusa build. I'm using Marlin firmware and Pronterface as recommended by the local 3d printing community group. I have some PLA and ABS filament, and a manufacturer just around the road to make up some 'specials'
I'm using a ramps 1.4 driver board. Ultimately I'd like to replace the Mega2560 with a prop based board, so we can have something like the Pronterface GUI rendered by the prop. The easiest approach I have found so far would be to adapt Ariba's PropinoPP (one of the Smorgasboard boards) - it could involve as little as half a dozen jumper wires to make this work.
So where is everything at as of now?
My Prusa has http://reprap.org/wiki/Sanguinololu Sanginololu (1.3a) hardware. I'm still using Sprinter and Pronterface. The "stock" parts are working well, so we have an establish baseline. Also, we're waiting on Sal's printrbot, which was promised some months ago but has not shipped. The intent is a drop in replacement pop solution for the electronics and firmware. There are many alternative being explored on different threads.
My project, propforth, is looking to create the driver firmware in forth. The stepper motor control has already been demonstrated. In order to facilitate testing, we are completing the the test automation for the propforth 5.3 kernel. (The test automation is a huge effort and a huge step forward, and has been taking most of the effort since release 4.6 almost a year ago.)
All the software pieces are there, we just have to bring them together. We think the propforth firmware will allow better positional accuracy, better interface to the PC, allowing control using a PC side Go application, and network accessible control and monitors of the printer application. But, we have to do one step at a time, so we can't promise when anything will be ready.
I have a pcb (I sent one to tubular too) that is just a 1.75"sp pcb carrier for 4x Pololu Stepper drivers (I have a stepper driver pcb too based on the A4982 & A4984 TSSOP packages). This board plugs into my BaseBlade1 pcbs (14 pin connector 5V, Gnd, P0..P11).
I have yet to build the Mosfet pcb to drive the heated bed and extruder. Again I expect it to be 1.75"sq and plug into my BaseBlade1 pcb on the other side, making 3x 1.75"sq pcbs.
I expect to do an AtMega644P BaseBlade1 equivalent so that I can substitute the Prop and AtMega to make code transition to the prop easier.
Thanks for the update, Prof. I hadn't really followed the Printrbot stuff but will go and have a good look around. Your propforth driver sounds interesting, good to hear it will support Peter's drivers as these sound very impressive.
Cluso got your boards late last week, thanks! Might need some help assembling them (what parts where etc). Having a 644P base sounds like a great idea too. The small mosfet PCB will need lots of copper - i put the power meter on my Prusa and is drawing about 210 watts during heatup, though the TO220 fets themselves don't get too warm.
I think my next step will be to get the Prop into an "observation" and HMI role - perhaps by mounting the PropinoPP board in the middle of the Mega<>RAMPS sandwich, or Cluso's board beside it. From there it should be easy to drop off the Mega and issue commands from the Prop.
Cluso got your boards late last week, thanks! Might need some help assembling them (what parts where etc). Having a 644P base sounds like a great idea too. The small mosfet PCB will need lots of copper - i put the power meter on my Prusa and is drawing about 210 watts during heatup, though the TO220 fets themselves don't get too warm.
WIth the low RDSon these days being <10mohm, even at 20A the disipation is only 200mW. Of course providing the Mosfet is driven hard on.
I just checked the pricing of the ATMega324/644/1284. From all the "know-it-alls" on this forum, I was expecting a really cheap chip. It's not much better than the prop!!!
tubular: Just let me know when you are ready to assemble them.
BTW I just ordered the GT2 kits. My parts will be yellow and black for the hot parts. see www.create3d.com.au
I am looking at some much cheaper 35oz Nema17 motors for the Z axis and extruder. The X & Y will be 77oz Nema17.
I have a few current thoughts. My BaseBlade1 has provision for placing the prop on my 1"sq daughterboard. I could put an ATMega644 on a similar 1"sq daughterboard. Then it would be just swapping out the Prop vs ATMega daughterboard.
In preparation for the arrival of my Prusa V2 plastics, I am just about to send off a prototype board 1"x2" containing 2 MOSFETs for driving the Extruder and Heated Bed. The connector to the prop is 1x4 (5V,GND,Pa,Pb). Input to the pcb is a PC 4pin molex (the same as on a HardDrive) requiring 12V,GND,GND,5V. The heated beds normally use around 9A @ 12V. Connectors to the heaters are 1x2 5mm screw terminal blocks.
Cost to me is $5 ea inc postage. And it will cost me $3 to post unregistered airmail to anyone interested. Anyone interested in a pcb please PM me.
I also can get more Stepper Driver motherboards (see my website for more details) done at the same time. This pcb will change when I add the above mosfet board and the prop chip to a final pcb soon. I can also get more BaseBlade1 pcbs done too (propeller board), but you could use any prop proto type pcb. You will require 16-18 prop I/O pins for the basic reprap controls (8x X,Y,Z&E STEP & DIR plus 2x optional ENA & RST, 2xHeater, 2xTemp +1 for the prop ADC sigma delta, 3x X/Y/Z endstops).
So this is just an offer to those currently building a RepRap and who want to experiment with prop code. I will be doing a final pcb with all the requirements shortly.
(tubular: I will get a mosfet pcb to go with the other pcbs I sent you)
What instructions are you using? The references I used were a little lacking in places particularly in truing the chassis, calibrating the extruder, and getting a consistent flow.
Well despite not paying enough attention to calibration, leveling etc (it's not in its final location yet), I managed to print something resembling a Teal & Pintail Duck Whistle last night.
I'm using black PLA for now because my heatbed keeps tripping the 11A polyswitch. The heatbed seems to be close to 1 ohm, so 12A flowing. I think 14A polyswitches are available
Thanks for your notes Prof Braino. It's exciting to have come this far. Now for the addictive phase...
ps Cluso are you going with 1.75 or 3mm filament? I am using 1.75, can send you samples in ABS (5 colours) or PLA (black) if you're going 1.75. Sadly no glow in the dark filament just yet
Comments
I'd personally extend the Z axis, which would obviously add to the costs... but most of the existing parts can be reused.)
You'd easily be able to get away with ~$200 - $250, I'd think as long as your existing controller can control the 4th channel.
Bill
Thanks for the input. Actually the magic number would be five, because I have two steppers on the Y axis. As far as the Z axis goes, I have something in particular that I would like to model on a regular basis, and the table should be just about right for this purpose. However, I am sure there will come a time when I will wish I had adjusted the Z axis height.
$250 sounds reasonable.
Anyhow, thanks once again for the input. I will have to dig around in this thread for a bit just to get more info.
Bruce
In case you didn't buy you parts already, here are my notes.
http://reprap.org/wiki/PrusaBuildNotesBraino#Prof_Braino.27s_Prusa_Build_Log
You might shave a couple bucks here or there.
Is that the unit with the UHDP bearings?
Yea that is the unit that you saw, however the bearings are UHMWPE. Thanks for the link, I will look it over when I have more time.
Bruce
lets make a prop based 3D printer!
suggestion inkjet filled with an active agent sprayed on a powder.
plaster of paris+ colored water= pretty but delicate
baking soda + thin 2 part epoxy split into color ink tanks = strong but exspensive sort of.
use an old parallel printer have the PROP print a "page" as a slice in your powder spread more powder and lower platform print another page and so on. i have had this ideal but lack the money to try it currently. plus i lack knowledge of epoxies.:nerd:
Bruce, those two steppers can still be controlled via the same controller, so you still only need four. In fact, you can use two steppers on the Z axis as well, like the Prusa Mendel does for a nice easy heavy lift scenerio.
Bill
Alright Bill, you have my attention, because I have never attempted that and it would save me a controller. Considering that I will be using GeckoDrive's G251X, how can this be accomplished, and is it possible to run both motors at full current capacity.
Additionally, there was a fairly large Steele family in Indiana several years ago with the siblings being named, Jean, JoAnn, Charlie, Tom, John, and one other pretty little thing, but I forget her name. Are you any relation?
Bruce
One source indicated serial connection, another suggested parallel was best. I tried both, I think the ended up about the same. (Right now the wires are covered, so I can't see how I left it. But, it works just fine with
NEMA 17 motors http://www.pololu.com/catalog/product/1200/faqs
and
pololu 4988 stepper drivers http://www.pololu.com/catalog/product/1182
We are looking at a prop based drop in replacement for Sanguino electronics on RepRap, with a G-code interpreter implemented in propforth.
As for the motors... the controller is the limitting factor. I'd imagine that you could run both motors at 'their' capacity as long as the controller could drive both within its limits. That GeckDrive looks like it can handle 3.5 amps... which is enough considering that I drive my large laser with half that for two motors!
Bill
I can't wait!!!
Bill
A fellow Hoosier
http://www.kisslicer.com/
I got it to slice, its really fast, but it starts printing at 1.0 mm Z (above the bed). I only have an STL file for this part, and I don't have access to the original application that created the part.
Any suggestions?
If the former then download a trial of rhino 3D, import it, move it and then export it again. I think you get about 25 saves in the trial.
Graham
KISSlicer automatically drops (or raises) the lowest point in the model to sit right at Z=0. If you select a grid or pillar raft then the object will sit on that, naturally, and the thickness of the bottom layer of the raft is a function of the "Bed Roughness" setting. The only other thing I can think of that would affect the Z values in the resulting G-code is the "Offset...Z" box. If the Z values in the G-code look correct, but the head is physically 1mm up off the bed. then we're looking at a different class of issues. Would you mind zipping up the output G-code file and emailing it to me? (Best email address for this is info at kisslicer dot com.)
thanks,
Jonathan
The Marlin fw has some useful capabilities built in that allow for an LCD, encoder wheel, and an SD card to be all you need to make the printer print aka standalone.
I have also seen a guy build a healthy(steel square tube 2"x3"x1/8") CNC machine and try to use it to print with. It was a failure and he will gladly tell you as much. There was too much mass to move quick or even well enough to print with. I have seen other mention it. I really don't imagine being able to use a 3d printer as a CNC mill. Structural integrity is once again the culprit. I saw someone comment 3d printers aren't fast. My friend built a Prusa with linear bearings and can easily print at 150mm/s. He has printed at better than 200mm/s.
My question is about slic3r's "easy configuration and little or no calibration". Don't we need to measure the output filament width, heights, and adjust the flow rate, etc?
I have a bunch of problems with the bottom solid layers being overfilled, and the top layers not filling in solid. I think this is due to improper calibration, but I don't know the process for Kisslicer. The process for Pronterface/SFACT/Skeinforge is pretty thorough, but I only got as far as I did (due to inexperience).
Do you use think filament stock 3mm, or thin filament stock 1.75, a wide nozzle 0.3 mm or think nozzle (what ever the small one is?) , PLA or ABS?
I'll be firing up the PLA tomorrow... I understand it smells like waffles cooking when you use it, so I can't wait to experience it myself.
Bill
I used a 40mm fan from either an old pentium 1 or a vga cooler. I just wired it to a spare header on the power supply, along with some Ikea LED book shelf strips, and they run whenever power is on.
I'm using a ramps 1.4 driver board. Ultimately I'd like to replace the Mega2560 with a prop based board, so we can have something like the Pronterface GUI rendered by the prop. The easiest approach I have found so far would be to adapt Ariba's PropinoPP (one of the Smorgasboard boards) - it could involve as little as half a dozen jumper wires to make this work.
So where is everything at as of now?
Ordered Prusa V2 plastic with LM8UU bearings.
Looking to use GT2 6x2mm 36t pulleys and belts
Looking at the J-Head Mk V-B Hot End
My Prusa has http://reprap.org/wiki/Sanguinololu Sanginololu (1.3a) hardware. I'm still using Sprinter and Pronterface. The "stock" parts are working well, so we have an establish baseline. Also, we're waiting on Sal's printrbot, which was promised some months ago but has not shipped. The intent is a drop in replacement pop solution for the electronics and firmware. There are many alternative being explored on different threads.
My project, propforth, is looking to create the driver firmware in forth. The stepper motor control has already been demonstrated. In order to facilitate testing, we are completing the the test automation for the propforth 5.3 kernel. (The test automation is a huge effort and a huge step forward, and has been taking most of the effort since release 4.6 almost a year ago.)
We are looking at supporting both the "standard" A4988 motor drivers, and Peter Jackaki's L6470 http://forums.parallax.com/showthread.php?135225-L6470-Microstepper-%28128-step%29-sample-Forth-code/. Supporting other drivers or configurations should not be a problem, once testing is solved.
All the software pieces are there, we just have to bring them together. We think the propforth firmware will allow better positional accuracy, better interface to the PC, allowing control using a PC side Go application, and network accessible control and monitors of the printer application. But, we have to do one step at a time, so we can't promise when anything will be ready.
I have yet to build the Mosfet pcb to drive the heated bed and extruder. Again I expect it to be 1.75"sq and plug into my BaseBlade1 pcb on the other side, making 3x 1.75"sq pcbs.
I expect to do an AtMega644P BaseBlade1 equivalent so that I can substitute the Prop and AtMega to make code transition to the prop easier.
Cluso got your boards late last week, thanks! Might need some help assembling them (what parts where etc). Having a 644P base sounds like a great idea too. The small mosfet PCB will need lots of copper - i put the power meter on my Prusa and is drawing about 210 watts during heatup, though the TO220 fets themselves don't get too warm.
I think my next step will be to get the Prop into an "observation" and HMI role - perhaps by mounting the PropinoPP board in the middle of the Mega<>RAMPS sandwich, or Cluso's board beside it. From there it should be easy to drop off the Mega and issue commands from the Prop.
I just checked the pricing of the ATMega324/644/1284. From all the "know-it-alls" on this forum, I was expecting a really cheap chip. It's not much better than the prop!!!
tubular: Just let me know when you are ready to assemble them.
BTW I just ordered the GT2 kits. My parts will be yellow and black for the hot parts. see www.create3d.com.au
I am looking at some much cheaper 35oz Nema17 motors for the Z axis and extruder. The X & Y will be 77oz Nema17.
I have a few current thoughts. My BaseBlade1 has provision for placing the prop on my 1"sq daughterboard. I could put an ATMega644 on a similar 1"sq daughterboard. Then it would be just swapping out the Prop vs ATMega daughterboard.
Cost to me is $5 ea inc postage. And it will cost me $3 to post unregistered airmail to anyone interested. Anyone interested in a pcb please PM me.
I also can get more Stepper Driver motherboards (see my website for more details) done at the same time. This pcb will change when I add the above mosfet board and the prop chip to a final pcb soon. I can also get more BaseBlade1 pcbs done too (propeller board), but you could use any prop proto type pcb. You will require 16-18 prop I/O pins for the basic reprap controls (8x X,Y,Z&E STEP & DIR plus 2x optional ENA & RST, 2xHeater, 2xTemp +1 for the prop ADC sigma delta, 3x X/Y/Z endstops).
So this is just an offer to those currently building a RepRap and who want to experiment with prop code. I will be doing a final pcb with all the requirements shortly.
(tubular: I will get a mosfet pcb to go with the other pcbs I sent you)
Now to get the parts assembled...
my notes: http://reprap.org/wiki/PrusaBuildNotesBraino
I'm mentoring a kid building a Wallace, he might try to make improvements to the instructions, any input might help.
I'm using black PLA for now because my heatbed keeps tripping the 11A polyswitch. The heatbed seems to be close to 1 ohm, so 12A flowing. I think 14A polyswitches are available
Thanks for your notes Prof Braino. It's exciting to have come this far. Now for the addictive phase...
ps Cluso are you going with 1.75 or 3mm filament? I am using 1.75, can send you samples in ABS (5 colours) or PLA (black) if you're going 1.75. Sadly no glow in the dark filament just yet