Cutting foam sheets... with a needle!
- Thread starter dkj4linux
- Start date
Awesome. Thanks for stopping by to share your idea. I have one question already. How do you deal with the timing of the needle? I've used a sewing machine to cut the foam board manually and it is very sensitive to needle breaks if you try to push the foam faster than the feed of the machine.
Timing with the needle is dealt with by having control over the feed rate and picking a needle RPM that gives at least 10 punctures per mm of feedrate. When you're making that many punctures needle drag isn't really an issue. You can push it and start to see the needle start to deflect and drag as the puncture per mm rate starts to drop. But you have to go pretty far before it becomes an issue.
The needles made from piano wire are also considerably more flexible than sewing machine needles. So rather than just break they bend and deflect.
But overall with the feedrate and needle RPM matched to give that 10 punctures per mm prevent most needle drag and breakage issues.
The needles made from piano wire are also considerably more flexible than sewing machine needles. So rather than just break they bend and deflect.
But overall with the feedrate and needle RPM matched to give that 10 punctures per mm prevent most needle drag and breakage issues.
dkj4linux
Elite member
An update on my MPCNC rebuild...
Finally got off my lazy butt and prepared a little first MPCNC milling project. I tried to duplicate a small, milled, box enclosure for a pot I installed on one of my earliest BuildYourCNC-inspired machines. I took measurements, prepared a DXF using Onshape, brought it into Estlcam, and generated the gcode file… using Ryan’s intermediate milling tutorial as my guide (https://www.v1engineering.com/estlcam-2-5d-routing-intermediate/). Then, of course, I found the rattiest, most twisted, piece of pine I could find… just to make this more of a challenge. Oh… and it was a full 3/4″ thick, which required pressing the stop collar off my 1/8″, single-flute, endmill. Finally, after all that, a part was produced… useless but successful
Lots of “new” stuff [to me] with this machine… the latest firmware, full-graphic LCD, T8 leadscrew, series stepper wiring, standard tool mount, etc. Whoa… overload!
Old pattern part… I left off the notch…
Surfaced 1mm off top of material first… shows the twist pretty well…
Bottom side… almost complete through cut…
New part with old pattern parts…
I’m pleased. Hopefully, having now worked through the procedure, I’ll not be so lazy/reluctant about doing more of this stuff in the future…
— David
Finally got off my lazy butt and prepared a little first MPCNC milling project. I tried to duplicate a small, milled, box enclosure for a pot I installed on one of my earliest BuildYourCNC-inspired machines. I took measurements, prepared a DXF using Onshape, brought it into Estlcam, and generated the gcode file… using Ryan’s intermediate milling tutorial as my guide (https://www.v1engineering.com/estlcam-2-5d-routing-intermediate/). Then, of course, I found the rattiest, most twisted, piece of pine I could find… just to make this more of a challenge. Oh… and it was a full 3/4″ thick, which required pressing the stop collar off my 1/8″, single-flute, endmill. Finally, after all that, a part was produced… useless but successful
Lots of “new” stuff [to me] with this machine… the latest firmware, full-graphic LCD, T8 leadscrew, series stepper wiring, standard tool mount, etc. Whoa… overload!
Old pattern part… I left off the notch…
Surfaced 1mm off top of material first… shows the twist pretty well…
Bottom side… almost complete through cut…
New part with old pattern parts…
I’m pleased. Hopefully, having now worked through the procedure, I’ll not be so lazy/reluctant about doing more of this stuff in the future…
— David
Last edited:
So I have been having much difficulty with my needle cutter since I went with a 3D printed holder. I am still working through it but I needed to cut out a plane for a friend. I put on the DW660 and a .035 bit (labeled a dental bit) worked great for score and 50% cuts, but it would get a melted foam "nub" on it and through cuts were not possible. Not enough thread length to pull the foam up. I knew I needed multiple passes, but with the "nub" that would not work. I then tried a .0625 bit and that worked well. Lots of paper dust but it did all of the cuts. The Kerf is a little wide though. Anyone know what bits the spindle guys are using when they cut foam. I like how fast I can cut this way vs. the needle cutter.
dkj4linux
Elite member
Rob,
Sorry to hear you're having so much difficulty with your needle cutter. What kind of difficulty? Pictures showing some detail of your cutter would help immensely.
At least one accomplished needle cutter user, kkmalu, still prefers using a rotary bit (1/32" IIRC) when cutting non-papered materials but has said that he uses the needle cutter when cutting DTFB... presumably for the same reasons you cite.
The Rascal CNC guys use a rotary bit as well and have a thread going on this forum... if you haven't checked in with them already, you probably should.
Please keep us posted on what you decide and any progress you make. I'd love to help you get your needle cutter going again but will need some detail regarding the issues you're seeing.
-- David
Sorry to hear you're having so much difficulty with your needle cutter. What kind of difficulty? Pictures showing some detail of your cutter would help immensely.
At least one accomplished needle cutter user, kkmalu, still prefers using a rotary bit (1/32" IIRC) when cutting non-papered materials but has said that he uses the needle cutter when cutting DTFB... presumably for the same reasons you cite.
The Rascal CNC guys use a rotary bit as well and have a thread going on this forum... if you haven't checked in with them already, you probably should.
Please keep us posted on what you decide and any progress you make. I'd love to help you get your needle cutter going again but will need some detail regarding the issues you're seeing.
-- David
dkj4linux
Elite member
Alright! What should be the last update to my MPCNC milling adventure…
The initial project successful — but pretty useless except for the learning aspect — I decided to commit to a real project… Ryan’s LCD enclosure endplates. Using Onshape, I imported the “left.DXF” file (from his "intermediate" tutorial) into a sketch and mirrored it to model both ends. Exported the dxf and used Estlcam to create the gcode. Found a 10mm thick, laminated flooring, cutoff from my scrap pile, and set it up on the MPCNC. The laminated hardboard flooring milled beautifully and I couldn’t be more pleased with the result…
-- David
The initial project successful — but pretty useless except for the learning aspect — I decided to commit to a real project… Ryan’s LCD enclosure endplates. Using Onshape, I imported the “left.DXF” file (from his "intermediate" tutorial) into a sketch and mirrored it to model both ends. Exported the dxf and used Estlcam to create the gcode. Found a 10mm thick, laminated flooring, cutoff from my scrap pile, and set it up on the MPCNC. The laminated hardboard flooring milled beautifully and I couldn’t be more pleased with the result…
-- David
David,
The first problem was heat related. I did not use an insulator between guide and PLA and it melted off. I tried experiment with the flywheel using a shouldered servo screw vs bearing. Similar to the crankshaft theory in operation but with a flywheel. The servo screw got so hot that it melted the PLA flywheel (might not have had enough play in the needle). I then went to the typical crankshaft (that is what I have been doing before I needed to change mount) and then my motor came off of the PLA mount. I noticed the motor mount was very warm so I am guessing it started to melt the PLA and pulled the screws out. I have a few plans in place, just need to find the time to implement them.
1) Printed a flywheel in PETG and will try the servo screw theory again (or try to find a small shouldered bolt).
2) Print a flat adapter to the MPCNC so I can use a "mouse trap" mount for the crankshaft style.
3) Try to find a small bearing and go with the typical setup
I ran the crankshaft style cutter through 100's of sheets with the wood mount that I had on the original Z axis mount. When I needed a new Z assembly I went with the latest design which did not accommodate my original mount. I am sure it will get sorted out, just need to find some time.
Thanks,
Rob
The first problem was heat related. I did not use an insulator between guide and PLA and it melted off. I tried experiment with the flywheel using a shouldered servo screw vs bearing. Similar to the crankshaft theory in operation but with a flywheel. The servo screw got so hot that it melted the PLA flywheel (might not have had enough play in the needle). I then went to the typical crankshaft (that is what I have been doing before I needed to change mount) and then my motor came off of the PLA mount. I noticed the motor mount was very warm so I am guessing it started to melt the PLA and pulled the screws out. I have a few plans in place, just need to find the time to implement them.
1) Printed a flywheel in PETG and will try the servo screw theory again (or try to find a small shouldered bolt).
2) Print a flat adapter to the MPCNC so I can use a "mouse trap" mount for the crankshaft style.
3) Try to find a small bearing and go with the typical setup
I ran the crankshaft style cutter through 100's of sheets with the wood mount that I had on the original Z axis mount. When I needed a new Z assembly I went with the latest design which did not accommodate my original mount. I am sure it will get sorted out, just need to find some time.
Thanks,
Rob
dkj4linux
Elite member
Rob,
With the needle cutter, any metal subject to friction heating needs to be insulated from the plastic... especially the needle guide. I'm not sure what "latest design" you tried... Jason's design (https://www.thingiverse.com/thing:2450613) is probably best and most popular, I think. His uses the 4 bearing pre-guide setup that will *reduce* the friction in the main guide... but the wooden plate is still recommended to thread the Mig-welding tip or inflation needle into. It also uses a shorter motor if I recall. My latest version hasn't really been used by too many folks to my knowledge but my recently-passed flying/fishing buddy was using one pretty regularly that I'd made for him. I dismantled his needle cutter machine to build the MPCNC milling setup I'm currently playing with but here are some closeups of the needle cutter (less the needle) I removed from that machine...
I'm actually shocked you were as successful as you were with the crank-pin needle drive. I found it quickly wore out -- usually cutting through the crank-pin after cutting a few sheets -- and offset any advantage of minimal balancing requirements. A properly balanced flywheel with eccentric bearing runs far smoother and lasts much longer In my experience... I think Jason said his needle lasted for over a year and cut many sheets of foam once he switched from the crank-pin version to flywheel version.
I wouldn't bother with any PLA/PETG flywheel with a crank-pin bolt/screw... the friction-heat will eventually soften and destroy the flywheel. If you want to try it, I'd at least use a metal flywheel. Better... replace the crank-pin with a small bearing as is usually done and it should work like a champ.
The motor should never be stressed and get hot in this application. If it got hot enough to pull loose from the needle cutter body, something was drastically wrong... most probably bad bearings. It could be that using the crank-pin drive for as long as you did, the vibration finally destroyed the bearings in the motor.
Please let me know how I can help further. -- David
Brett_N
Well-known member
Well, you guys FORCED me to take the plunge and start printing off bits and pieces. I'm 100 layers into the 1st pair of corner brackets - they're a little ugly, but they'll work
Oh, and hello from another fellow Ex-TI'er - although I was in Houston in the C2000 group!
I've decided on a 30x48 inch size designed to be somewhat portable / storable (hang on the garage wall)
So far I've got:
All of the print files (for the 1 inch diameter, figure it's a little "stronger")
Ordered a 100 pack of bearings from Amazon - $20
Ordered 10 meters of belts and all of the pulley ends - another $20
Already have a spare RAMPS board which should work fine with some tuning
(I do motor control for a living, so I've got dozens of spare NEMA 17 steppers around)
Nuts bolts and stuff I'll grab from ACE hardware
Quick question though - I started playing with Inkscape and figured out how to layer and generate G-Code based on the template back on page 73 - do you load this directly to the RAMPS board? I have a 3D printer but I use Repetier Host to talk to it....but it won't accept any of the G-Code from Inkscape. Do I even bother with a slicer?
I'll start taking pictures!
Oh, and hello from another fellow Ex-TI'er - although I was in Houston in the C2000 group!
I've decided on a 30x48 inch size designed to be somewhat portable / storable (hang on the garage wall)
So far I've got:
All of the print files (for the 1 inch diameter, figure it's a little "stronger")
Ordered a 100 pack of bearings from Amazon - $20
Ordered 10 meters of belts and all of the pulley ends - another $20
Already have a spare RAMPS board which should work fine with some tuning
(I do motor control for a living, so I've got dozens of spare NEMA 17 steppers around)
Nuts bolts and stuff I'll grab from ACE hardware
Quick question though - I started playing with Inkscape and figured out how to layer and generate G-Code based on the template back on page 73 - do you load this directly to the RAMPS board? I have a 3D printer but I use Repetier Host to talk to it....but it won't accept any of the G-Code from Inkscape. Do I even bother with a slicer?
I'll start taking pictures!
I am using Inkscape for cleaning up and making my preferred adjustments then saving that file as a DXF. Estlcam is the software I am using to generate the G-Code from that DXF. That code can go right to the ramps board. I use Repetier Host to view the cut file, like a slicer, but the DXF from Estlcam is what gets cut.
There are ways to generate gcode within inkscape using certain plugins...but I've not had great luck with them. Estlcam is much more powerful and easier to use for generating gcode in my experience. But...it's not free and only runs on windows (without jumping through some hoops with wine to get it running on mac or linux.) Fusion360 can be used to generate gcode with a post-processor...but I found it to be "too powerful" and more complex than needed for the stuff I do. Plus Estlcam runs better on my older computers.
Brett_N
Well-known member
Quick question on the MPCNC - my main objective is to be able to cut DTFB which is nominal measurements 20" x 30"
I as originally thinking 24x36 - does that give me enough leeway for a 20x30? it's a little confusing reading the printable dimensions versus frame size. Does it make more sense to go 30x48?
I as originally thinking 24x36 - does that give me enough leeway for a 20x30? it's a little confusing reading the printable dimensions versus frame size. Does it make more sense to go 30x48?
dkj4linux
Elite member
Brett, there's a desired working area spec... and a "footprint" (outer dimension) spec for the MPCNC. Essentially you'd add about 11" to your desired X and Y working values to get a minimal footprint and then add some leeway. So, on builds for DTFB (with no leeway) you'd have 31" and 41" for a minimal footprint... and that is also the minimum lengths of your X and Y rails (conduit), 3 of each. On all my builds I've rounded those minimal values to 36" and 48" and that gives plenty of room to add hold-downs, clamps, pins, weights... whatever you want to use to hold down the DTFB sheet while it's being cut. So I cut 3x at 36" and 3x at 48" pieces of conduit for the X and Y rails. Works for me...
dkj4linux
Elite member
Many of you will be familiar with the popular, humorous sign, "DO NOT FEED THE ENGINEERS!" There was a time…
But having been cast out like an old, worn-out, shoe in 2005, after 30+ years, and now living alone… I’ve had to change my tune. I followed David Walling's video
to create my version of the sign (the first signage with my buddy's rebuilt MPCNC)…
-- David
But having been cast out like an old, worn-out, shoe in 2005, after 30+ years, and now living alone… I’ve had to change my tune. I followed David Walling's video
to create my version of the sign (the first signage with my buddy's rebuilt MPCNC)…
-- David
How balanced does the flywheel need to be? I spun the wheel on a piece of piano wire and at first it always stopped with my "counter weight" down. I swapped one of the large washers with a smaller one and it seemed to stop spinning at random places. I spun up a "empty" fly wheel on the motor and then the one with the needle holder. I do feel a small vibration when the needle fly wheel is turning, but not to the point where it is shaking out of my hand.
It does run a lot smoother than my crankshaft so I think I am good. Just waiting on the new bracket to finish printing and then I am ready to try it out.
It does run a lot smoother than my crankshaft so I think I am good. Just waiting on the new bracket to finish printing and then I am ready to try it out.
dkj4linux
Elite member
If it is running a lot smoother than your crankpin version did, then I'd say go with it... that should serve as a good starting point. You can always add/remove a little counterweight -- add a nut, use a shorter screw, etc. -- and see if it smooths out a bit more. In time you should be able to get it dialed in and running nicely.
Similar threads
