I saw somewhere that had a CNC machine cutting foam with a 1/32" bit. They used vacuum bed to keep the foam flat - you'll probably need to accommodate for the ability to do that with this. My Dollar Tree NEVER has perfectly flat (aside from the waves) foam. You might also want to look at spring loading the drawing implement as well so the waves don't play into any issues. The spring would not need to be all that strong either, you don't want it to press into the foam when it hits a high spot.
I'm going to follow this, because I've been considering building a full CNC to use for ALL purposes, but it would be neat to have a relatively inexpensive machine that I could set up inside for foam plan drawing/cutting. Honestly just being able to PLOT the plans on the foam would save me TONS of time.
I've been following that thread for a bit now. The Mostly Printed CNC is brilliant. I'd love to end up building one that I can swap out a Dremel and 3d Printer heads. I was looking at what MDF and angle Aluminum I can source locally at my Lowes.
I have a car I need to get back in working order but once it is out of the garage I'm jumping into prototype fabbing.
Well the car is dead. The price of a new transmission outweighs the value of the car. SOOOO! That at least frees up some time.
I had written off trying to use a true linear rail system for this because what turned up in my searches was way to expensive. However, I came across the NYCNC video below that had a full set of links to all the parts used and they are surprisingly affordable. I've started running some numbers for the total cost of parts. (Feel free to peek at the list(s) here)
Being able to switch to manufactured rails, bearings would allow the final machine to be much more rigid and accurate as well as greatly simplify it's construction. I'm still figuring out what the rail version would look like but here's a WIP idea of it.
The deeper I get into this the more I like the idea of going the next step and making it a Needle Cutter.
This video is about as close as I've come to a single source for what I'm looking to do. Everthing from the Gcode shield to the timing belts.
New design with 'real' rails and bearings. Based on John's details from NYC CNC (Who, as of 2014, is back in Zanesville Ohio, go Buckeyes) John has a lot of great videos on his channel and you can really feel the love of the work this craftsman has for what he does. If you watch this part of his video he's built a plotter in much the same way I was originally thinking. I'm just going to be scaling it up to have a 20x30 workable area so I can draw directly on a full sheet of dollar store (or FT Water resistant) foam. As you can see from his test plot there were a few bits of jitter or play in the lines it drew. I've used his idea of having two of the rail bearings on each side of the gantry to hopefully limit or eliminate that play in my design.
Back to the final(?) design. I've reworked it to use the parts outlined in one of his projects. This will make it so that I'll have to order more parts from online sources but the prices seem low enough that it would be worth it and it should be repeatable for anyone who wants to build their own. Switching to these rails and bearings should also make the final build much more rigid than what I was originally going to get using the Lowes angles.
I was originally thinking I'd only use one stepper for the Y (long) axis but the deeper I got I felt that even with good bearings, the far side would be dragging so I've got a stepper for each side. In theory, I don't think this will add much complexity as both would be sent the same signals from the CNC shield.
I just pulled in some bearing models from the SketchUp warehouse so I'm not sure if these would be exactly the size I'd need, I just wanted to get the whole thing out of my head and into something that looked like it should work. I'll go back and get the size of the bearings that I'm actually purchasing and adjust these plans before I start cutting materials.
I've also found a reasonably priced source for bearing supports. Going with manufactured supports and rails will allow me to be sure they are positioned as accurately as possible and will greatly simplify construction. However, the clearance of the timing belt is pretty tight if I'm going to use it mounted under the edge of the gantry as I'm doing in this design. Again, I'll need to make sure my bearings are clear.
Things not shown in the model that I'll have to figure out:
Wire/cord management.
timing belt pullies at each end of the belt run
Pen lift/Z axis control. Still thinking simple servo should work.
12v regulated power supply
I'm cheap and that's why I'm a bit of a measure seven times, then know exactly what parts you need, then cut type of guy. My parts list estimate puts this thing at $180~ or so. I now feel confident enough with the design that I'll start buying the parts and get to building.
Here are the steppers I got for $12 each and that I've been playing with using the Arduino and motor shield.
I'm not an engineer, so if anyone is and you see some red flags in how I'm going about this design, please offer any feedback you can.
Watching this one to see how it works for you. Looks like a great project.
I am not sure how 2 steppers on the same axis will work for you. If you think your system is not rigid enough that one belt will drive the Y-axis across that width, how about another idea? Eliminate stepper B and instead run a long shaft all the way across and add a second pulley at the far end to drive your other timing belt. That way, you have one motor driving 2 belts timed exactly together. Not sure if you would have to use a motor with more torque for this.
But...the quality on their bearings and rods isn't the greatest. For something smaller like the Prusa i3 printer they're sufficient...but for something bigger like a CNC capable of cutting 20"x30" foamboard I'd personally want something better. The bearings they sell tend to be rough and noisy and the chrome rods aren't the smoothest (and sometimes not the straightest.) Also those rods are really tough to cut and don't really come in sizes that look optimal for a 20"x30" working area.
I think your idea of using angle iron and bearings has more potential to give higher quality at lower cost while being more repeatable and easy to source.
If you do go with the rods I'd suggest using 3 bearings on the X carriage like most Prusa printer designs. 2 on the driven side and 1 on the drag side. I'm not a ME but have some friends who are and they assure me that gives the best constraint of motion...though there is apparently an argument in favor of using 4 bearings if they're low quality. Using 2 as you've setup isn't likely to give good constraint.
I'm still working on reprinting all the parts for my MPCNC in PETG before wiring it up - but so far the motion from it juts pushing it around feels great. It's gone a long way towards convincing me that cheap 508 bearings and conduit can make a good large and cheap motion platform.
Can't wait to see what you end up with though, either way you go I'm excited
I am not sure how 2 steppers on the same axis will work for you. If you think your system is not rigid enough that one belt will drive the Y-axis across that width, how about another idea? Eliminate stepper B and instead run a long shaft all the way across and add a second pulley at the far end to drive your other timing belt. That way, you have one motor driving 2 belts timed exactly together. Not sure if you would have to use a motor with more torque for this.
I think you're right. I'd initially gone with the two steppers solely based on what I'd seen others do with real CNC machines. I think the connector rod idea would be cleaner and easier to build. I also realized that in this design I have the steppers facing each other so that would have complicated things. I'd have to work out how to mirror the signals to one of them or rearrange it so they are both facing the same way.
My gut says I should be able to get away with still using these small steppers since they only have to be strong enough to overcome the drag of A) the bearings and B) the drag of the pen point across the paper.
I'm still trying to decide what route I'm going to go for the motion, Rails or Rods. Based on what PMjhitesma shared I may try the rails on version one. I've been hunting all over town for a local place to get simple roller bearings but the only place that has had anything is the Tractor Supply Company but theirs aren't actually flat. The bearings have a little lip on one side. I live just outside of Columbus Ohio, a decent sized city so I'm sure someplace has what I'm looking for but I'm just not finding what I need. Yet.
However, the TSC has much better looking/higher quality angle rails than Lowes or Home Depot so if I do go rails for v1.0 I know where to get them.
Running multiple steppers on an axis isn't very hard. The stepper drivers can easily handle the current and all you have to do is swap the wiring on one of the coils of one of the steppers. The guy behind the MPCNC has a wiring kit for it that's basically plug and play: http://www.vicious1.com/product/wiring-kit/
It's actually kind of simpler than the jackshaft idea and less moving weight and mechanical things to go wrong. It does require a 2nd stepper for the axis...but can work with cheaper lower torque steppers.
I'm still trying to decide what route I'm going to go for the motion, Rails or Rods. Based on what PMjhitesma shared I may try the rails on version one. I've been hunting all over town for a local place to get simple roller bearings but the only place that has had anything is the Tractor Supply Company but theirs aren't actually flat. The bearings have a little lip on one side. I live just outside of Columbus Ohio, a decent sized city so I'm sure someplace has what I'm looking for but I'm just not finding what I need. Yet.
Might want to try a skateboard shop for the bearings. It's not the cheapest way to get them, but they're the same size used in skateboard wheels so they're easy to find
You can get them even cheaper ordering from China...but slower shipping and could be lower quality...
I got my first 8 bearings at a local skateboard shop (needed 2 of them for an upgrade on my 3D printer and didn't want to wait for mailorder) and it was something like $20 for 8 Probably could have found them cheaper from a specialty shop but it was Sunday, not many places were open and I didn't want to wait another day. I was surprised the skateboard shop didn't have any cheaper generic bearings honestly.
I considered just going to a goodwill or other thriftstore and looking for some old rollerblades/roller skates/skateboard to see if I could scavange some cheaper...but not much is open around here on Sundays
Might want to try a skateboard shop for the bearings.
...
I considered just going to a goodwill or other thriftstore and looking for some old rollerblades/roller skates/skateboard to see if I could scavange some cheaper...but not much is open around here on Sundays
Running multiple steppers on an axis isn't very hard. The stepper drivers can easily handle the current and all you have to do is swap the wiring on one of the coils of one of the steppers. The guy behind the MPCNC has a wiring kit for it that's basically plug and play: http://www.vicious1.com/product/wiring-kit/
It's actually kind of simpler than the jackshaft idea and less moving weight and mechanical things to go wrong. It does require a 2nd stepper for the axis...but can work with cheaper lower torque steppers.
I have been thinking about a CNC similar to this, and use it to cut 1200x800mm sheets of depron.
My idea was to use V-Slot Rails and run 2 Y axis rails and then a single X axis rail which would be bolted onto 2 'Carts' that run on the Y Rail. The great thing with V-Rail is that its all open source so you can mill/ 3d print the parts.
I have been thinking about a CNC similar to this and use it to cut 1200x800mm sheets of depron.
My idea was to use V-Slot Rails and run 2 Y axis rails and then a single X axis rail which would be bolted onto 2 'Carts' that run on the Y Rail. The great thing with V-Rail is that it's all open source so you can mill/ 3d print the parts.
Hmmm... so now there is a third design option that seems like it would be within the price range. Which one to try first. Leaning to the original, 'open v' design as I think I'll call it since the bearings are not 'captured' but just sit on the aluminum angle rails.
But I'm going to see what the cost would be for a V-Slot based table.
Please do keep posting your progress and research - I've build a little (4x6x8") 3D filament printer and a little (4x4) laser cutter, and am seriously interested in taking on a CNC carving machine that would be able to handle simple/thin MDF wood templates as well as foam cutting, and I think you are on a roll with this design path! Thanks to everyone in this thread for sharing their info!
The openbuilds stuff is nice. I have an upgrade for my printer waiting for me to install it that switches it from chrome rods to 20x60 extrusion and one of their plates. Feels so much smoother and quieter I can't wait to install it...just don't want to have to reset my z probe since the printer is printing so well...but seem to have knocked that out of alignment tonight. So guess I now have an excuse to finally install it
Some have speculated that the openbuilds stuff isn't as high precision since it uses small low rated bearings. But just playing with it on my desk it feels a lot nicer than the low end linear bearings/rods from Folger/sidewinder.
On the other hand...the openbuilds stuff isn't exactly cheap. The rail is cheap (and you can get it on ebay even cheaper) but the gantry packs are $30 each. So you'd be looking at $90 just for those. Of course all you'd really need is those and the rail for motion parts.
Decision made: the first prototype will be Open rails and skate bearings. I found a skate shop that should have 5 bearings for $9.99. That along with being able to get decently straight angles from TSC and the MDF from Lowes means I may be able to start building this weekend. I think I'm going to try the thrift shops for a cheap tower computer with a power supply I can scavenge.
And this little guy arrived today. I'm no expert but I'm impressed with what I'm seeing for the price I paid. No bent pins, no bad traces from what I can see. The soldering looks good and the silk screening is crisp.
Based on what I've found in all the image searches I think I've got the drivers installed right side up. I'm going to try and track down some spec sheets before I put any power to this thing.
Remember to adjust your current on those drivers before hooking up your steppers. They usually come set way too high. You set the current by measuring voltage between the center of the adjustment screw and ground ( don't try to probe ground on the driver itself, too easy to slip, just probe ground where you feed 12v to the board.) There's a bit of math to calculate the amperage based on the resistors that are installed on your particular drivers. There are details about it here: http://reprap.org/wiki/Pololu_stepper_driver_board
Picked up some skate bearings at the local skate shop. I would have gotten two matched sets but the Mini-Logo were on sale. Sixteen bearings for $15.00 after tax. Not bad.
Also got GRBL 0.9 loaded into the Arduino UNO R3
Toying with trying out kevlar fishing line instead of timing belt. The PrintrBot Simple (Wood) seemed to have good results with it. It would be easier to source locally than timing belts, noched pulley, etc.. If it didn't work it wouldn't be a huge investment lost.
