• This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn more.

Cutting foam sheets... with a needle!

dkj4linux

Well-known member
hey how do i connect my cutter to my cutter. what program do you use to upload files and control the printer with.
Eric,

I responded to your last request for similar information back in post #2393. I know you are new here but you really need to provide more information in order to receive the help you're looking for. Answer questions as well as ask.

I'll give you a "nugget"... the needle-cutter really doesn't need to be connected to the computer at all; i.e. it can free-run. That may not be ultimately the way you'll want it but, for now, it can be that simple.

Please respond to us. You're very welcome here and we'd love to help but you've got to talk to us. Please join in the conversation and we'll help you get going.

-- David
 

dkj4linux

Well-known member
WRT FoamRipper, when I last left off, I was starting to wander off into an area I have less interest in, and practically no use for, given that I'm no longer really cutting, building, decorating, and flying planes. So, once again, I've chosen to concentrate on the machine itself... specifically, the FoamRipper's Y-carriage and needle-cutter. The current carriage was originally designed for the Phlatprinter-style machine I built a couple of years ago but, for FoamRipper, there are clearance issues with the fixed guide rods that extend down uncomfortably close to the workpiece surface. So, I revisited some of the discussion Tommy and I had, back when I had expressed interest in a more compact carriage assembly. A new carriage and lift assembly is starting to take shape and initial tests are looking pretty good.

Y-carriage.png


I printed the motor mount, a temporary end plate, and the bearing block. I also printed a series of test bearings to find the ID that gives a nice sliding fit without too much slop... and printed them in groups of four. The coupling nut and linear bearings will be press-fit into the bearing block. All the parts laid out to start basic assembly...

20190327_171310.jpg


20190327_171603.jpg


20190328_083620.jpg


20190328_090219.jpg


20190328_172706.jpg


Using my bench vice, I pressed the coupling nut and the linear bearings into the bearing block. There are a total of 8 linear bearings -- four each pressed into the top and bottom of the block -- to provide good support for the guide rods, which are simply 6", smooth-shanked, 5/16" hex bolts from the big box store. I then started assembling the various bearings, printed parts, and fasteners into the major sub-assemblies.

20190328_172927.jpg


20190328_173948.jpg


20190328_174019.jpg


20190328_183124.jpg


I "borrowed" the motor and leadscrew assembly from a previous carriage design and then built up the sub-assemblies on the Y-rails. I've tried to keep it as modular as posssible and easy to build "in place"... without lifting or removing the gantry from the machine

20190328_184845.jpg


20190328_211331.jpg


20190328_211345.jpg


and then tested for basic operation...


I still need to capture the bolts heads on the motor mount. of course, to insure the guide rods move with the motor and mount when lowering toward the work piece. I could always turn or cut away the heads and thread the end to look like the other end but I'm trying to avoid that to keep all the hardware as standard as possible... to make it easy to duplicate. I also need to start thinking about the lower end, where the needle-cutter and/or other tooling will mount. I also need to devise a carriage arm... to attach the Y-axis drive belt to the carriage.

But, all in all, the operation is smooth and I'm pleased with how it's looking so far.

Later. -- David
 

dkj4linux

Well-known member
Thanks, Tommy. I wish you could, too. Adulting calls.... too late for me. I'm reverting to childhood... only harder-headed and more opinionated. ;)

Finally got to the point I needed to clear the rails of the old carriage assembly. Here's the reason for using a two-part roller assembly... I can't take the old roller off without removing a rail

20190401_111503.jpg


20190401_111629.jpg


Rails cleared, new parts ready... also printed a couple of cutter bodies, straight and sideways (slightly more compact)...

20190401_093540.jpg


20190401_092253.jpg


Assembled needle cutter with slip-on flywheel and wooden stack preguides...

20190401_123209.jpg


and mounted on new carriage...

20190401_123558.jpg


Applied velcro to flat backs of guide rod keepers to mount ESC and servo tester... and connected and tensioned drive belts using
zip-ties

20190401_132017.jpg


20190401_132010.jpg


Ready to test...

20190401_131403.jpg


20190401_132037.jpg



Need to lube a few areas and...

-- David
 

dkj4linux

Well-known member
Needle cutter preliminaries...

The simplest needle cutter system -- and the one I use exclusively -- is stand-alone. It uses nothing more than a ~1000kv brushless motor, connected to an ESC, which, in turn, connects to a 12-volt power brick, and the 3-wire servo lead connects to a servo tester.

20190402_104059.jpg


20190402_103207.jpg


20190402_103221.jpg


20190402_103241.jpg


This allows the servo tester's potentiometer to be the speed control. Using software to control the needle cutter system really isn't necessary, as once you've found the cutter rpm and feed rate that results in the best cuts for the material you are working with, you really have little reason to change any of the settings during operation.

The most important thing, initially, about the needle cutter is to set the cutter speed... 6000-8000 rpm is a good starting point. An inexpensive (~$15 USD), contact-less, tachometer is a tremendous help; i.e. I strongly recommend you get one. I'll almost guarantee that you'll be running far too high rpm's without it if you try to use the printed scale on the servo tester to judge how fast you are running. Though the printed scale is linear and works fairly well with servos, it can't be trusted to set in the cutter rpm's because ESC's are incredibly non-linear... at 12-volts, with a 1000kv motor, you'll get 10k - 12k rpm at full scale setting... but you'll pass 6000-8000 rpm in the bottom quarter/third of the scale.

Too high cutter rpm's will result in -- at best -- poor cuts... but it quite probably will result in broken/'flying needles first. It will also undoubtedly result in shorter cutter life as this simple needle cutter is, of necessity, an inherently "pretty sloppy" system and wear/tear/friction-heating can, and will, be problematic. So, shoot for 6000-8000 rpm and go from there... this typically results in decent cuts in DTFB with feed rates in the 600-1000 mm/min range.

So, consider yourself warned... ;):eek:;):eek:;):eek::p

More to come...

-- David
 
Last edited:
David,

The new system looks really good!! Great Engineering!!
You inspired me many months ago to build my own MPCNC with a version of your needle cutter attachment. Its been up and running with no issues for several months now. I am running my MPCNC with mirco-steppers and a BOB with Mach3.

I really like this new design of yours, but I have been also using my MPCNC on wood projects as well (with a router). I am sure this new design could be adapted for wood projects. I wish this was around before I built my MPCNC 'cause I would have built this one. I am running out of room in my garage or I would give this new one a try!!

Keep on doing this stuff, it's awesome!

Chalie

MPCNC pic is from when I was building the MPCNC., Needle cutter attachment and one of my cuts.

MPCNC_1.jpg MPCNC_2.jpg MPCNC_3.jpg
 

dkj4linux

Well-known member
David,

The new system looks really good!! Great Engineering!!
You inspired me many months ago to build my own MPCNC with a version of your needle cutter attachment. Its been up and running with no issues for several months now. I am running my MPCNC with mirco-steppers and a BOB with Mach3.

I really like this new design of yours, but I have been also using my MPCNC on wood projects as well (with a router). I am sure this new design could be adapted for wood projects. I wish this was around before I built my MPCNC 'cause I would have built this one. I am running out of room in my garage or I would give this new one a try!!

Keep on doing this stuff, it's awesome!

Chalie

MPCNC pic is from when I was building the MPCNC., Needle cutter attachment and one of my cuts.

View attachment 127899 View attachment 127900 View attachment 127901
Thank you, Chalie, for the kind words and encouragement. I struggle sometimes with self-doubt and wonder if I'm not just being vain in posting my lengthy tomes to this thread... and feel it would probably just die if I didn't. There are a few regulars who post here and they truly are the ones who keep me going... but I do still wonder sometimes if this has become more just a "personal diary" or "engineering notebook" and I'm really the only one who benefits. So thank you for speaking up and sharing how it has helped you. That means a lot to me.

And I'm so thrilled to hear that your MPCNC and needle cutter is working for you so nicely. It's also interesting that you are using Mach3... are you using the parallel port with the BOB? It's been great fun getting back into LinuxCNC/EMC2 (where I actually started my CNC journey) and figuring out how to interface all these neat CNC machines through the parallel port. I don't know for sure but I think the "Inexpensive LinuxCNC interface for MPCNC" circuitry I recently developed would work for Mach3 (using parallel port) as well.

Anyway, thanks again for sharing.

-- David
 

dkj4linux

Well-known member
More needle cutter considerations...

An optional -- but recommended -- step is the addition of a cotton-wad oiler/stabilizer to help with cutting down on friction-heating in the needle and guide.

Since I printed the cutter body and use wooden guide blocks above and below the plastic platform, I have a ready-made space to add in the oiler. It's easier to put the needle and guides in place before packing in the cotton...

20190402_165426.jpg


Start filling in the area around the needle with little twists of cotton, until the cavity is full...

20190402_165734.jpg


Then add drops of sewing machine oil... here some 50+ year-old stuff inherited from parents/grandparents...

20190402_165928.jpg


20190402_170324.jpg


And, finally, button it all up and it should be ready to go...

20190402_205300.jpg


Periodically -- every few sheets of foam? -- apply a couple of drops of oil to the needle shaft, above the guides, and let gravity and capillary action pull the oil into the cotton below. It doesn't take all that much and you certainly want to stop short of over-filling and allowing the oil to migrate all the way down the needle to stain the foam.

Moving forward slowly... a couple of errands to run. Later...

-- David
 

dkj4linux

Well-known member
David,

What are you using to attaching the needle to the flywheel? Is it better than the grooved bearing?

Drew
Drew,

I'm now working on and testing a slip-on flywheel... sized for the bell of the Emax CF2822 motor I've been using all along. It is a fairly tight slip fit over the motor bell and encloses a press-fitted 8mmx4mmx3mm bearing in one of three cavities in the face of the flywheel.

slip-on_flywheel.png


20190402_103905.jpg


20190402_103836.jpg


The eccentric bearing has an M3 machine screw through it, captured by a regular nut, the loops of the needle, and a Nyloc nut to keep it in place. Directly across the face is a captive nut and a M3 machine screw and an assortment of nuts/washers that serve as counter-weights to bring the whole thing into reasonable balance.

The motor's shaft is trimmed to length so it doesn't interfere with the rotation of the needle and the flywheel is slipped fully over the bell... making sure the needle attachment is directly above the main needle guide...

20190402_104059.jpg


20190402_205300.jpg


This approach is probably "iffy" for some users because it requires the use of a pretty "dialed in" 3d printer and the flywheel needs to be sized to a specific motor. I use Onshape to do the CAD and tend to work in "tenths" (of a millimeter), and occasionally "hundredths", to get the press-fits for various components. But this flywheel is very quick to print and results in a shorter overall cantilever length, motor mount to needle plane, and allows for a more compact needle cutter. The Onshape model is public and can be accessed here, if you are interested.

If you are making your own flywheel and can completely enclose the eccentric bearing... do it. Then needle connection to a threaded free-wheeling post is considerably easier than attaching to the smooth/grooved outer race of the bearing (though there are many, many cutters out there today using this very approach).

-- David
 
Last edited:
I posted this in another thread but I figure Ill post it here for archive purposes -

I've connected it to my ramps board. I ran a few extra cables to the servo pins, and extra wires to the heated bed power output. I use these to power either the laser, or the esc, and use the servo pins to power the esc and the laser controller. I power on and set the motor speed via gcode. You have to go into the firmware and enable the servo controls.

I have D6 as my laser signal, D11 as my esc signal and have a ground wire I run to both d6 and d11 grounds, and share depending on which mod I'm using. For the power wires, I had to actually run a relay, as turning off the heated bed just shuts off the ground, and the ground from the speed control pin completes the circuit, so I connect the ground direct to the ground pin on the heated bed power out, and run the positive through a relay, which I control with pin d4.

To power my esc-

M280 P0 S0 - Set the esc to 0
M106 P0 255 - Enable the ground pin on the heated bed
M42 P4 S255 - Enable the relay to provide power to the esc
M280 P0 S60 - Set the esc (60 for me gets me around 7000rpm which works well for me)

To use my laser, which is 0-5v instead of pwm, do the same except use

M42 P6 S255 - Full power to the laser


I have a XT60 connector that powers either the laser or the esc, and two 3 pin servo connectors, with opposite pins connected so that only the specific tool with the correct end fits, either the laser signal or the esc signal.
 

Attachments

Also an update on my cutter attachment, and my hold down table adventures.

Finished the hold down table by first peeling off the plastic on this pink foam from home depot, and gluing it down to my table, then using the cnc to cut out 150 holes at the intersections. I just connect the vacuum to the port on the bottom, and have this switch to control the vacuum power and speed.

Needle cutter uses the hold down foot, and a fly wheel that I balance with a single steel bolt that I slide in and out in the cutout until it stops vibrating, then tighten it down.
 

Attachments

rockyboy

Skill Collector
Mentor
I posted this in another thread but I figure Ill post it here for archive purposes -

I've connected it to my ramps board. I ran a few extra cables to the servo pins, and extra wires to the heated bed power output. I use these to power either the laser, or the esc, and use the servo pins to power the esc and the laser controller. I power on and set the motor speed via gcode. You have to go into the firmware and enable the servo controls.

I have D6 as my laser signal, D11 as my esc signal and have a ground wire I run to both d6 and d11 grounds, and share depending on which mod I'm using. For the power wires, I had to actually run a relay, as turning off the heated bed just shuts off the ground, and the ground from the speed control pin completes the circuit, so I connect the ground direct to the ground pin on the heated bed power out, and run the positive through a relay, which I control with pin d4.

To power my esc-

M280 P0 S0 - Set the esc to 0
M106 P0 255 - Enable the ground pin on the heated bed
M42 P4 S255 - Enable the relay to provide power to the esc
M280 P0 S60 - Set the esc (60 for me gets me around 7000rpm which works well for me)

To use my laser, which is 0-5v instead of pwm, do the same except use

M42 P6 S255 - Full power to the laser


I have a XT60 connector that powers either the laser or the esc, and two 3 pin servo connectors, with opposite pins connected so that only the specific tool with the correct end fits, either the laser signal or the esc signal.
Which ESC are you using? I've tried almost the exact same commands and while I can get my ESC to beep (sometimes) it never successfully completes the initialization sequence to turn on.

I've been keeping a servo tester hooked up to the ESC and manually turn the motor on and off with that, which sucks whenever I'm not paying close enough attention and it parks the head with the needle still running, and then augers itself into the table. :(

Thanks!
 
Emax 30a. You will basically need to calibrate it the same way you would a plane, set it to 100 or so (That worked for me), power the esc, then set it to 0.
 

rockyboy

Skill Collector
Mentor
Emax 30a. You will basically need to calibrate it the same way you would a plane, set it to 100 or so (That worked for me), power the esc, then set it to 0.
Hmm... thinking about it I didn't try doing the calibration through the RAMPS board - just with the servo tester. Maybe that will help!

Thanks!
 

dkj4linux

Well-known member
Thanks, Verris, for sharing your hold-down table and needle cutter. Looks good!

In conjunction with the carriage and needle cutter development, I've also been revisiting/resurrecting my vacuum pad scheme... using a router speed controller to control the shop-vac and reduce the noise substantially. Here are my first cuts made using the new Y-carriage/Z-axis/needle cutter assembly... and with shop-vac running at minimum setting on the speed control.


20190405_220729.jpg


20190405_094723.jpg


The vacuum pad consists of three sheets of DTFB... two slotted sheets (glued together) with a loose "holes" sheet that serves as the replaceable spoilboard. The 3-layer stack sits on the work surface, with nothing between, and the vacuum pad not only pulls down the foam workpiece but also pulls itself down against the work surface... even with the shop-vac running at minimum setting on the router speed control.

20190405_221443.jpg


20190405_221431.jpg


Placing a clean sheet of foam on the stack and jogging X and Y to a suitable home position, I turn on the vacuum, and start the cutter... setting in about 8000 rpm. With the cutter running, I lower Z until the needle just starts to "tickle" the surface of the workpiece... it's very obvious from the change in sound when the needle touches the work surface. Set this in as Z-home and start the job...

20190405_215545.jpg


The bottom side shows the depth of cut wasn't quite deep enough here... another millimeter would do it. I had to strip the paper off the backside to pop these parts out but, with that done, every piece came out easily and cleanly...

20190405_221150.jpg


20190405_231545.jpg


20190405_231552.jpg


20190405_231845.jpg


20190405_233255.jpg


Turned the sheet to find a blank area on the other end... cut the obligatory crown (of MPCNC fame) at a slightly deeper DOC. The crown is actually lightly stuck to the spoilboard underneath and stayed in place when I lifted the sheet of foam... possibly the cutter rpm and/or depth of cut could be slightly reduced. These particular cuts were achieved with the cutter running approximately 8000 rpm (unloaded) and a feed rate of 610 mm/minute...


20190405_222349.jpg


Gently working the piece loose... everything loosened, with paper intact....

20190405_222933.jpg


Top and bottom sides...

20190405_223147.jpg


20190405_223200.jpg


20190405_223216.jpg


Without doubt, the cut is improved with the vacuum pad holding the work piece flat and secure. And, with the router speed control set to minimum, the shop vac supplied plenty of vacuum to hold everything in place... with substantially less noise than when plugged into wall power. I was pleasantly surprised... these are on par with some of the nicest cuts I've ever attained with DTFB, paper on. Note the foam is cut very cleanly and only the paper shows any "fuzziness""... un-papered foams should cut very cleanly.

Time to turn in... it's been a good day.

-- David
 
Last edited:

dkj4linux

Well-known member
FoamRipper job setup...

Please turn down the sound... not all the way, however. Also, please ignore the old man in the background... he's a ne'er-do-well I occasionally feed and he supposedly sweeps behind my machines every year or two. There's also a cameo appearance by hairy gorilla, who thinks he's doing something useful by running his hand (and goobers) over the surface of my foam... idiot! ;)


Actually, this all goes much quicker than shown in the video. The procedure is to position the vac stack on the work table, turn on the vacuum, jog X and Y into position and set home, turn on the needle cutter and set speed, lower cutter until it just touches work surface, and set Z home. Double-check everything and start the job. This really doesn't take more than a minute or so.

Here, for demonstration purposes, I turn on the shop-vac at full speed... and then switch to variable speed, running up to full-speed and back to minimum, to demonstrate the contrast in the noise level. You can see the foam stack compress when vacuum is first applied and the work piece is held securely enough for cutting, even at minimum vacuum setting.

The needle cutter is then powered up (you should hear it boot, just like when on an airplane) and, here, I'm using a non-contact tachometer to set the cutter speed to about 8000 rpm (unloaded)... it will drop a few hundred rpms, once cutting is started. This doesn't really need to be done every time... mark (or note) where on the servo tester dial this setting is and then simply go straight to it on future runs. The cutter speed doesn't need to be exact... this is close enough. Spot-check it occasionally if something doesn't seem right.

Jog X and Y into the desired position and then home X and Y. Then lower the cutter, still running, until it touches the foam surface. You can easily hear when the needle touches... now, home Z.

Double-check everything and start the job. Nothing to it... :)

-- David