Cutting foam sheets... with a needle!

[dkj4linux]

Wish I got to go on some fishing trips...just as long as not too many took the bait and made me actually have to do something more than sit and relax

...

Looking forward to seeing the new designs come together. For those who aren't following the FT Fans group on Facebook....there's some competition in the works https://www.facebook.com/groups/FlitetestFans/permalink/1991279561090581/

Looks like a nice design, but I'm gently trying to suggest they look into a needle cutter instead of a spindle if their primary design goal is cutting DTFB I'm also not a big fan of the "traditional" parallel port driven controller setup...especially driven from windows. Still - it's nice to have options! And I like their plans to start a site dedicated to sharing plans ready to cut...though I'm crossing my fingers that they don't limit it to gcode as I'd rather see things shared in a format that's usable across a variety of machines.

Oh, and David you must have way more patience than me...or a nicer Chromebook than the one I got my wife for Christmas I tried onshape on there...and it was usable...but definitely not as snappy as on my notebook. But to be fair I find the touchpad on both equally annoying for CAD without an external mouse!

The "tug" [on my line] is my drug of choice. There are times when we're pulling them in fast and furious but right now we're in the Texas summer doldrums/heat and you've really got to work at it a bit to catch fish (or even sit and relax). I prefer the faster action...

Thanks for sharing the eclipse viewer experience at your daughter's school. I debated whether I could/should include one of the pics you put on the MPCNC/Vicious1 site in my "Ya know..." post but elected not to... and then you stepped up to the plate and provided a great, and timely, report. Everybody came out looking good... you're a great dad/maker!

I agree with you about some of the odd choices (parallel port, Windows, etc) being made with that CNC machine. They can always strap a needle-cutter on it, of course... but, until then, I suspect much of the initial excitement will wane when they see what it takes to cut DTFB with a laser. Best case, they get a laser setup that rivals the FT speed build kits in cut quality... and that's fine if the under-cutting, etc. is what you want.

My little Acer Chromebook 14 with 4G of ram is not the speediest machine around but it's not bad at all IMO... I've had far worse performance from much-higher powered and more expensive dedicated machines. I have it setup with Crouton/Ubuntu and use a USB mouse with it... agreed, touchpad sucks. I've got a couple of 11" Acers with 4G of ram that do equally well... though of course the screen resolution isn't nearly as good. Other than not having a working Estlcam -- yet -- these little, less than $250, machines do everything I want to do. [RANT] I have virtually no desire to mess with Windows any longer... it rarely cooperates with what *I'm* trying to do and continually throws up nags/restrictions that do nothing but anger and frustrate me. MS/Windows wants to control what I can, and can't, do with *my* machine... I hate that and won't put up with it any more. [/RANT]

[Deep breath, David... deep breath...]

 

[jhitesma]

I agree with you about some of the odd choices (parallel port, Windows, etc) being made with that CNC machine. They can always strap a needle-cutter on it, of course... but, until then, I suspect much of the initial excitement will wane when they see what it takes to cut DTFB with a laser. Best case, they get a laser setup that rivals the FT speed build kits in cut quality... and that's fine if the under-cutting, etc. is what you want.

Don't think they're looking to go the laser route (though a few people brought it up in that discussion) but rather are going for a more traditional spindle. I like having a spindle as an option on my machine...but for cutting foam - the needle is just so much nicer. Less kerf, simpler work hold down requirements, NO MESS, no vacuum needed, quieter (though still annoying to listen to), tighter/sharper inside radiuses (which is great for FT planes - like where the sides and top of a fuselage come together), better for score cuts. Yeah, you can get really small end mills - but they're crazy fragile and get expensive fast. $2 worth of music wire will make a lifetime of needles and honestly I've yet to wear out a needle - I've had to freshen up the edge on mine once or twice but ran the last one for over a year and through almost 2 cases of foam board and it was still doing fine when I replaced it.

Driving a machine from windows instead of a dedicated controller just seems like a poor idea to me. Windows just doesn't have the stability I'd want to see and isn't a real time OS. I'm not sure what software they're using either but I haven't see much in the way of open/free CNC software on windows. Seems like everyone is using Mach3 which starts at $175. The TB6600 they plan on using is basically just a stepper controller - so it's the tethered computer doing all the gcode processing. That made sense in the 80's. But now with the state of microcontrollers it just makes so much more sense to me to offload that from a non RT computer to a dedicated micro.

I like the mechanics so far and am interested to see where it goes. But I like the approach being taken in these threads with the low rider and phlatprinter style mechanisms that are more space efficient and the microcontroller based control...and of course...the needle


Speaking of needles - when I was experimenting with cardboard a week or two ago and kept knocking my needle off the flywheel bearing I decided to try something. Put a single tiny drop of super glue on the needle. Wasn't sure if it would actually bond to the metal of the needle and bearing and was very sparing with it because I didn't want to risk getting any glue in the bearing. Well...that needle is ON there now. Hit one of my T-pins with the needle the other day and noticed it was dragging and not cutting as clean in one direction afterwards - sure enough the tip had developed a bit of a fishhook shape. Went to take it off to hone it - and then remembered the super glue. I ended up having to take the needle and bearing off together to sharpen it. I'm sure I could get it off with a little work and maybe a drop of acetone...but might be worth trying if anyone is having issues with keeping the needle on the bearing and doesn't want to groove the bearing.

My little Acer Chromebook 14 with 4G of ram is not the speediest machine around but it's not bad at all IMO... I've had far worse performance from much-higher powered and more expensive dedicated machines. I have it setup with Crouton/Ubuntu and use a USB mouse with it... agreed, touchpad sucks. I've got a couple of 11" Acers with 4G of ram that do equally well... though of course the screen resolution isn't nearly as good. Other than not having a working Estlcam -- yet -- these little, less than $250, machines do everything I want to do. [RANT] I have virtually no desire to mess with Windows any longer... it rarely cooperates with what *I'm* trying to do and continually throws up nags/restrictions that do nothing but anger and frustrate me. MS/Windows wants to control what I can, and can't, do with *my* machine... I hate that and won't put up with it any more. [/RANT]

Sounds like the same or similar to the one we have. It's an Acer C201. Does well for my wife - light and the battery lasts forever. She's still running chromeOS which is a little limiting but fine for her - and onShape will still run. To be honest I haven't tried Onshape on it in a few months and didn't really know what I was doing last time I did try it on there. I shoudl give it another try now that I'm more familiar with onshape and see what I think again. Though if I did pick one up for myself I'd probably wind up with a more traditional linux on it as well

There are only 2 reaons I still have any windows computers. Photoshop and estlcam. Older versions of PS will run in wine but with my wife having a small photography business we have to keep up with recent versions. And I've yet to get estlcam working in there. Gimp is ok...but it's no photoshop. I've been using PS since v3 and every time I try gimp or something else I get frustrated. Plus we have a nice high end larger format Canon printer and doing color management for the screen and printer under linux is a huge pain. It's a pain in windows as well...but not as big of a pain

I have to admit though I'm loving this macbook - but after 9 months of using it I still hate macos's keyboard mappings and finder. But drop to a terminal and it's wonderful almost normal unix. Reminds me of the SGI Irix I got to use in the early 90's which at the time I refered to as "what apple's os should be".

 

[quimney]

Driving a machine from windows instead of a dedicated controller just seems like a poor idea to me. Windows just doesn't have the stability I'd want to see and isn't a real time OS. I'm not sure what software they're using either but I haven't see much in the way of open/free CNC software on windows. Seems like everyone is using Mach3 which starts at $175. The TB6600 they plan on using is basically just a stepper controller - so it's the tethered computer doing all the gcode processing. That made sense in the 80's. But now with the state of microcontrollers it just makes so much more sense to me to offload that from a non RT computer to a dedicated micro.
".

I totally agree with your reservations concerning running with windows. I use LinuxCNC on mine and love it. I do have a plotter running GRBL on an arduino being fed by GRBL Panel. At least the windows machine just sends gcode and all the critical timing is left to the arduino. It hasn't failed yet but I'm waiting for that one windows update to just take me out mid-file...

Speaking of needles - when I was experimenting with cardboard a week or two ago and kept knocking my needle off the flywheel bearing I decided to try something. Put a single tiny drop of super glue on the needle. Wasn't sure if it would actually bond to the metal of the needle and bearing and was very sparing with it because I didn't want to risk getting any glue in the bearing. Well...that needle is ON there now. Hit one of my T-pins with the needle the other day and noticed it was dragging and not cutting as clean in one direction afterwards - sure enough the tip had developed a bit of a fishhook shape. Went to take it off to hone it - and then remembered the super glue. I ended up having to take the needle and bearing off together to sharpen it. I'm sure I could get it off with a little work and maybe a drop of acetone...but might be worth trying if anyone is having issues with keeping the needle on the bearing and doesn't want to groove the bearing.."

Take a good look at Shurakair's flywheel. Instead of grooving the bearing he embedded the bearing in the flywheel. Then the needle can be bolted on using the center hole in the bearing. It works great and the needle never comes off until you want it to.

 

[jhitesma]

Take a good look at Shurakair's flywheel. Instead of grooving the bearing he embedded the bearing in the flywheel. Then the needle can be bolted on using the center hole in the bearing. It works great and the needle never comes off until you want it to.

How did I miss that? I've looked at his so many times (I still lust after that custom oilite guide!) and never noticed that. Not sure I could pull that off on my current setup, I would probably have to make the flywheel a little thicker and that would push the needle out of alignment...but...might be possible. I may have to play with that idea a little....

 

[dkj4linux]

I totally agree with your reservations concerning running with windows. I use LinuxCNC on mine and love it. I do have a plotter running GRBL on an arduino being fed by GRBL Panel. At least the windows machine just sends gcode and all the critical timing is left to the arduino. It hasn't failed yet but I'm waiting for that one windows update to just take me out mid-file...



Take a good look at Shurakair's flywheel. Instead of grooving the bearing he embedded the bearing in the flywheel. Then the needle can be bolted on using the center hole in the bearing. It works great and the needle never comes off until you want it to.

I, too, used EMC2/LinuxCNC in my pre-MPCNC days... it's what I started with. I was really into Linux advocacy at the time and it was a no-brainer to simply download the LiveCD and install it on an old desktop I had... worked like a champ. OS plus all the CNC goodies rolled into one image... it's a great setup and a really powerful CNC control software.

Shurakair's needle cutter, from day 1, has been one of the nicest design out there. He's fortunate to have friends like you and your machinist buddy to help with the fabrication. That's definitely the way to go if you if you have access to the machines and the talent to run them. In my experience, the needle guide has always been more problematic (once I started 3d printing the cutter body anyway...) than the needle itself... and that's where Shur's custom oil-lite(?) needle guide really impressed me. I really think that's one of the keys to the precise way that cutter operates.

Only recently have I revisited the needle's attachment to the flywheel... some folks seemed to be having trouble with flying/breaking needles when using the flywheel. I'd honestly forgotten about Shur's design -- except for his needle guide -- and Moebeast's little printed needle holder/keeper is the one I'm currently using. Moebeast has cut a lot of foam at the last two FliteFest events and seems very happy with the simplified needle fabrication and operation... and it's been great in the testing I've done. I'll have to revisit Shur's design... it should be pretty simple to come up with a new printed flywheel that embeds the bearing and uses the bore for needle attachment. It makes a lot of sense.

-- David

 

[randyrls]

It hasn't failed yet but I'm waiting for that one windows update to just take me out mid-file...

Go into the Control Panel and disable Windows Update completely!! I usually make a drive image about once a month, then tell Windows Update to go ahead. Easy to reverse if problems.

Take a good look at Shurakair's flywheel. Instead of grooving the bearing he embedded the bearing in the flywheel. Then the needle can be bolted on using the center hole in the bearing. It works great and the needle never comes off until you want it to.

I am so going to use that idea. THANKS Shurakair!!!

 

[jhitesma]

Well, I like the idea of the encapsulated bearing with the needle attached to the screw instead...but Just don't see a way to make it work on my current setup without some fairly major changes:





There just isn't room to clear the screws I use to attach the flywheel to the motor The flywheel screws to fit entirely in their recesses so I could move the bearing out and get more stroke...but I already get more stroke than I need. I could make the flywheel thicker and have the bearing in front of the screws...but then I'd have to lengthen the base of the cutter to keep things aligned...and I kind of worked hard to make this one as short as possible to keep it stiff and minimize vibrations.

So for now I'll stick with what I've got. It's working well for my afterall. And if I do start to have problems with it coming off I'll try mobeasts attachment thing or buying a grooved bearing.

 

[Shurakair]

Take a good look at Shurakair's flywheel. Instead of grooving the bearing he embedded the bearing in the flywheel. Then the needle can be bolted on using the center hole in the bearing. It works great and the needle never comes off until you want it to.

I apologize for my lack of communications regarding the flywheel design. Looking back at what I posted I can see how it was overlooked. I have learned a few CAD skills and hopefully this will help illustrate what is going on:

Blue: Nut & Bolt
Red: Bearing
Silver: Washers
The needle is held around the bolt, between the two washers

My original design had provisions for a set screw in the base of the flywheel to hold it onto the motor. This has proven to be unneeded as the flywheel is a press on fit to the motor shaft and it holds it well enough that it never slips.

Shur

 

[Shurakair]

Well, I like the idea of the encapsulated bearing with the needle attached to the screw instead...but Just don't see a way to make it work on my current setup without some fairly major changes:

jhitesma,
It appears as though you are attaching the flywheel with 4 screws? What about just using 2 screws and put the bearing 90 degrees from those? It seems like that would work and it wouldn't require major changes. Just a thought.

*edit* One concern I have is clearance for the needle's bolt head. I'm not sure how thick your flywheel is but it would have to be thick enough to encase not only the bearing but the bolt head as well.

Shur

 

[jhitesma]

jhitesma,
It appears as though you are attaching the flywheel with 4 screws? What about just using 2 screws and put the bearing 90 degrees from those? It seems like that would work and it wouldn't require major changes. Just a thought.

I considered that...but that flywheel kind of scares me It's small and light enough it shouldn't...but 8krpm+ spinning mass still feels like it deserves some respect Those screws that attach it are just M2 screws so they're pretty tiny. Honestly like yours it's probably ok even without them as it's a tight press fit...but especially cutting something tougher like I was with cardboard I could see it slipping.

*edit* One concern I have is clearance for the needle's bolt head. I'm not sure how thick your flywheel is but it would have to be thick enough to encase not only the bearing but the bolt head as well.

That was another concern I had - but I think that would be ok. My flywheel is 8mm thick, and the bearing is 4mm, so that leaves 4mm for the screw head. I'd have to measure a bolt head to be sure...but I think it would probably fit. I could always make the flywheel thicker - but then I'm back to having to redesign the cutter body to keep things aligned.

If I eventually make another cutter I'll probably go with your style of mount since I really like it. But on this one...I think it's more effort than it's worth to switch over since the only issue I have currently was with cutting cardboard at higher speeds....which I don't regularly do anyway

 

[dkj4linux]

Progress towards a new machine...

...

I like the mechanics so far and am interested to see where it goes. But I like the approach being taken in these threads with the low rider and phlatprinter style mechanisms that are more space efficient and the microcontroller based control...and of course...the needle
...

While playing with my Moebeast-inspired carriage, I mounted it onto my now-defunct camera slider for testing... and was once again "stricken" by visions of a Phlatprinter-style machine for needle-cutter, laser, etc. Pretty happy with the carriage assembly, I started thinking about the moving-material axis.

Here's a demo of a rack-and-pinion axis consisting of dual-motor pinions driving racks attached to either side of window-screen frame which will be fitted to carry the material to be cut... here 12" x 12". As expected, the slaved drive motors allow the frame to track relatively smoothly back and forth, with virtually no racking/skewing...

The herringbone rack and pinion (https://www.thingiverse.com/thing:37787) were printed as is... and then the rack was committed to Onshape so that I could adjust the number of teeth to fit a given space. The frame is made from an inexpensive 48" x 48" window-screen frame kit from Lowes -- with custom printed corners for mounting the racks -- and provides more than enough frame material to make frames to hold DTFB sheets... and larger. I intend to fashion printed clips to fit inside the frame and securely hold the foam sheet, and will need to relocate/rework the pressure bearing to clear those clips, and add a pressure bearing at the top of the frame...



















Sneaking up on it, we are...

-- David

 

[jhitesma]

Wow David, you've been busy

That Y is really quick. (not that we need that kind of speed...but neat to see!) My only concern is the frame still seems like it would take a good bit of space. I guess I need to take a closer look at the phlatprinter (he says for the 20th time) since I just assumed it was more of a pinch roller type of setup like desktop printer. Or like this music box punch I ran across today:

But...the nice thing about the frame is you can get full edge to edge coverage easily. With the pinch roller you're going to loose some of at least one edge where it's not supported enough to feed. Which combined with my office being uncomfortably hot right now is why I haven't tried putting anything together to test my ideas yet

 

[dkj4linux]

Wow David, you've been busy

That Y is really quick. (not that we need that kind of speed...but neat to see!) My only concern is the frame still seems like it would take a good bit of space. I guess I need to take a closer look at the phlatprinter (he says for the 20th time) since I just assumed it was more of a pinch roller type of setup like desktop printer.
...
But...the nice thing about the frame is you can get full edge to edge coverage easily. With the pinch roller you're going to loose some of at least one edge where it's not supported enough to feed. Which combined with my office being uncomfortably hot right now is why I haven't tried putting anything together to test my ideas yet

Phlatprinter actually requires a full-sheet's worth of space in front and back of the machine -- since the sheet material isn't flexible and can't be rolled or allowed to drape/hang like paper, vinyl, etc -- to allow the extremes to be reached but the machine itself has a smaller footprint on the tabletop. Make the frame removable (as is the plan) and the machine takes up far less space than a conventional X-Y machine. In operation, the machine could/should be raised above the tallest stuff surrounding it and/or in range... or possibly a vertical/slanted orientation against a wall might be used?

Yeah, it's fast here but I haven't adjusted the steps/mm yet. For demo/test purposes, I didn't want to cut off the motor connections to splice/reverse in the one motor to the other yet so just used the RAMPS Z-axis dual-connection setup (so I could reverse one motor from the other) and lived with what steps/mm setting was already set up in the firmware that was already loaded. For now I was really more interested in seeing smooth/strong/positive movement of the frame... so gingerly tested to see what RepetierHost thought full-range motion was for the frame and then set up one of the scripts to run it back and forth repeatedly.

In keeping with KISS -- and cheap/inexpensive! -- everything here is stuff we can print or we've already got on hand... or can get from the big-box stores. Conduit, 1/4" threaded rod, aluminum window-screen frame, and the standard hardware, nuts/bolts, etc... MPCNC-style. The printed rack can now be customized to make as many sections as you need to fit the required frame space and/or your print bed (..maybe worth a Thingiverse customizer effort?)

I thought about the pressure roller method that Phlatprinter used... and the grip-the-edges method of the little HP7475 desktop plotter I had on my desk and used way back when... and ruled them out for the [perceived] complexity of the material grip/ungrip mechanism, special rollers/bearings/coatings, outrigger requirements, and loss of active cut area. A lightweight, strong-enough, frame/carrier can easily be fashioned from a $10-$12 48" x 48" window-screen frame kit that seems to address most/all those issues so is the approach I've taken here. I've also been intrigued by the printed rack and pinion setup that seems accurate enough for our purposes, can be made/adjusted to any practical length, and mounts rather nicely on the frame... and IMO makes for a simpler, more positive, material drive mechanism.

I've still got a lot of work to do on the machine and material frames, material clips, above/below-material support, etc. I need to adjust the steps/mm in firmware and wire up the other axis to start pen-plotting and checking the accuracy. I also need to get my printed parts organized better in Onshape and uploaded to Thingiverse as they start to take final form. And when I up-size the design for DTFB, I'm thinking I'll mount the racks on the short side (20") of the material frame to minimize their length and the space required in front and back of the machine while in operation. Etc...

That's my plan anyway... thoughts/questions/concerns? I'd love to hear them!

-- David

 

[moebeast]

Wow David, you've been busy

That Y is really quick. (not that we need that kind of speed...but neat to see!) My only concern is the frame still seems like it would take a good bit of space. I guess I need to take a closer look at the phlatprinter (he says for the 20th time) since I just assumed it was more of a pinch roller type of setup like desktop printer. Or like this music box punch I ran across today:

But...the nice thing about the frame is you can get full edge to edge coverage easily. With the pinch roller you're going to loose some of at least one edge where it's not supported enough to feed. Which combined with my office being uncomfortably hot right now is why I haven't tried putting anything together to test my ideas yet

I don't have first hand knowledge, but I've listen to the back catalog of the "Crash Cast" twice. Crash had all three versions and Mark and Trish were regular guests on the show. The Phlatprinters used full length grip rollers. They were cutting fan-fold at the time, so the benefit was being able to cut 4 ft stock using a small machine. However, you need more space to operate it than a fixed bed since the stock sticks out both front and back. You need good tabs on all your pieces so nothing falls out and jams the works. This wasn't as big of a deal for the guys who built in and out feed tables. It was also a challenge to get an even grip from the rollers. My impression is that Mark gave up on the design after he bought his big machine to cut the kits and found how much more reliable a fixed bed, moving gantry is and switched his efforts to the Ox and Openbuilds.

If David can figure out how to make the frame removable, you would have the compact storage. Maybe you only need the side rails and make them fold in when not in use.

You will also need some sort of hold down for this to work well on warped material. A spring loaded foot would work, or maybe a full width top roller.

 

[jhitesma]

Yeah, as long as the frame is easily removeable that makes a lot of sense. I knew you needed the space when operating the machine but was mostly thinking about storage. And pinch rollers always seem like a pain to me (it amazes me that printers don't jam more than they do.)

 

[Balu]

Finally you guys are starting to think in a direction I have been doing for a year or so now . The only problem is I just keep thinking about it when I can't sleep while you can't sleep, because you are tinkering with it in the real world.

 

[dkj4linux]

Quite the past couple of days...

Forgive if I seem addled/emotional at the moment... I haven't gone away. I lost a great friend and golfing buddy late yesterday afternoon to a heart attack... a shock for sure but not really unexpected. He was 81 years old and had a history of heart related problems. But he loved his golf (and had a lot of "game" for his age) and he and I had just played a very pleasant, unhurried, and relaxing round... he'd even birdied the 1st hole and parred the 18th! He returned home and, as he sat in a chair by the door changing into his slippers, he slumped and passed... no more than 30-40 minutes after playing his final hole. Bittersweet... and just the way he would have scripted it! He'll be sorely missed.

I've rethunk/reworked my carrier frame drive system and decided I need to reorient the motors and racks 90 deg from where they were. This has the advantage of putting the pressure and outrigger bearings in line and in the same plane and also moves everything out of the inner frame area... where I plan printed clips of some sort to hold the DTFB (probably two sheets... bottom one for the wasteboard). I think I've got the drive system down to just a couple of printed parts on each side... the motor mount and a support bar for the pressure and outrigger bearings.

Here's the general idea... motor with pinion, pressure bearing directly below and riding against the back of the rack, and outrigger bearings at the ends (about 8" apart)... the red spacers are to keep the bolt heads from rubbing bottom of frame.





I'm printing the new support bars right now and hope to be ready to demo as soon as I can get them assembled and installed.

-- David

 

[randyrls]

My condolences on the loss of your friend David....

 

[moebeast]

Good idea David. It looks like the frame will come right out for storage as well.

Sorry for your loss.

I am putting my new machine together, but the printer I was stripping for parts has a NEMA 23 for the z-axis, so I am one motor short.

 

[dkj4linux]

Thanks, Randy and Mark, for the sentiments.

Sorry about the white parts not showing much detail. For some reason, making the bar clip on to the conduit makes sense to me for now... time will tell as to whether the PLA will withstand the strain for a prolonged period. The fit is actually very nice and snug while still allowing some sliding adjustment for the width of the carrier frame...



Shims are needed to adjust the pressure bearing to pinion spacing... and need to get a shorter bolt for the pressure bearing...



The whole drive mechanism is quite low profile... looks a though the lower rails could sit on, or very close to, the bottom of the machine...



One more bearing bar is printing and then I should be able to test... probably tomorrow sometime.

-- David