Cutting foam sheets... with a needle!

[jhitesma]

I'd also advise to skip the dremel idea, learned that lesson the hard way. They're too weak, too inaccurate, and burn up way too easy.

Getting a 3D printer first and using it to build your own CNC is a great path for sure....but definitely adds more to the learning process and will delay having a functioning CNC. Though much of what you'll learn with 3D printing applies directly to using the MPCNC (at least the way most people use it with marlin for control.) so it's not a bad route to take at all.

Yes, the MPCNC can be used as a 3D printer...but swapping tools, making sure things are leveled to get a good first layer, tying up your CNC for hours waiting on a print. It's a less than ideal solution. Having both a CNC and a 3D printer really is empowering - some things are just better done with subtractive manufacturing and some with additive. You can usually come up with a work around if you only have one method or the other, but compromises are never a lot of fun.

If you can afford it then the official Prusa machines are probably the best bet going for a 3D printer. I've yet to talk to anyone who bought one and regrets it whether it's their 1st or 10th printer.

If you're on a budget there are a lot of lower cost options - but they all come with compromises. Some like the anet and creality use customized version of Marlin that somewhat limit the machines and their upgradeability. In the case of the anet the firmware can be downright dangerous as apparently it disables temperature overrun protection - combine with a machine made from flammable acrylic and you've got a potentially dangerous setup. There are also kits like the Folger that many of us have - but they aren't super quick or easy to assemble and often require a good bit of problem solving and creativity to get up and going...but are very affordable.

So while I do think getting a 3D printer first and then printing your own MPCNC is a great path - it's also not for everyone. Those 12hour plus center section parts are very demanding prints - getting a cheap 3D printer tuned and running reliably enough to create them definitely adds an additional challenge.

Buying the printed parts and electronics from Ryan at v1engineering.com is definitely the quicker/easier way to go. There are several of us who'd be willing to help with printing a needle cutter if that's the route you want to take (over making one from wood or metal) or as mentioned earlier you can put an extruder on the MPCNC and use it. Though I'm not sure I'd want to use a MPCNC built large enough to cut DTFB as a 3D printer. I'd probably want to build it smaller first and then expand it later like I did with mine.

 

[Shurakair]

In the case of the anet the firmware can be downright dangerous as apparently it disables temperature overrun protection - combine with a machine made from flammable acrylic and you've got a potentially dangerous setup.

Maybe that is true with the stock firmware, I'm not sure. I have an Anet A8 with Skynet3D firmware and it has thermal overrun protection. Its a custom version of Marlin and it allows quite a bit of expansion. If you're looking for a plug and play solution, I would not advise the Anet A8. If you want a super cheap, good printer and you're not afraid of making modifications then I would recommend the A8.

Shur

 

[dkj4linux]

I have built two of the FolgerTech 2020 I3 printers and both served me well... I used them to build/rebuild several MPCNC's and derivative machines along the way. Then, while trying to sell the school's technology guys, who I used to work with, on a first 3d printer purchase for the school... I did such a good sales job I sold myself on the idea of buying a Prusa I3 MK2S, though I was already 3d printer "rich", And I have never regretted it a moment. The quality of both the machine and its prints, the ease of use, and all the self-calibration stuff makes using it such a pleasant experience that my perfectly usable Folger machines now sit idle. The school managed to drag their feet long enough that they were actually able to buy the MK3 kit, months after I got my MK2S built and running... and they did manage to get it running without issue. Like Jason, I'd highly recommend the Prusa machine if you can afford it.

And speaking of Dremel tools... a blast from the past, another iteration of the needle cutter for your viewing pleasure...

 

[ironkane]

Don't forget the OpenBuilds V-Slot alternative to the MPCNC. Costs are comparable enough that you'd want to look at both before making a decision. They all boil down to rails, motors, belts and a controller to rule them all. All you want is a frame that is true enough and rigid enough to move your tool head around accurately and repeatably.

My biggest complaint with the MPCNC is the zinc plated electrical conduit. Tolerances at the mfg is relatively straight. You're better off sourcing the stainless steal tubing that Ted himself upgraded to. And then you get to have some quality alone time with your conduit, some emery cloth and a bucket of water as you sand that bumpy zinc coating smooth. And even then it's still pretty loud and rough.

I have a MPCNC setup for cutting DTFB that works great at what it does. I haven't decided what I'm going to do with it yet. Options include: Breaking it down for future use, loaning it out to a Jr. High Engineering teacher friend who is starting up a STEM program, selling it or just harvest the motors and electronics for another 3D printer. The teacher is unsure. He has a Shapeko that's big enough and may just mount a needle cutter on it. Anyone interested can message me privately.

I think the MPCNC is better suited to use as a CNC wood router. Especially since it's so cheap to adjust size. Unless you go stainless. But then you really need to compare costs to the V-Slot extrusion.

 

[ironkane]

I am thinking of picking this one up:

https://www.google.com/search?q=Any...UIDSgE&biw=1366&bih=654#imgrc=DNkY3kP_rn34hM:

any thoughts?


Or the CR-10s Upgraded version

Not bad.
Prints 8 x 8 x 8 and can print hot enough to do ABS and other higher temp filaments.
You're at a good price point for a competent entry level printer. I wouldn't go shopping for a lower cost printer.

CR-10s would be a good choice as well. Note that it's built using V-Slot rails

Prusa if you can afford it.

 

[jhitesma]

Maybe that is true with the stock firmware, I'm not sure. I have an Anet A8 with Skynet3D firmware and it has thermal overrun protection. Its a custom version of Marlin and it allows quite a bit of expansion. If you're looking for a plug and play solution, I would not advise the Anet A8. If you want a super cheap, good printer and you're not afraid of making modifications then I would recommend the A8.

Shur

Yes, which is the point I was trying to make. The stock firmware on some of these Chinese clones is Marlin...but with modifications or changes which aren't released back to the public (as required by the license Marlin is released under) and sometimes with poor choices made. It can be fixed with aftermarket firmware - sometimes just flashing stock Marlin is enough - other times the control board is just different enough that you need a customized build. But either way it's not a great option for someone who wants a printer they can "just start printing with" and not have to mess around with.

If they're willing to mess around...well then there are a lot more options. But more complexity. Willing to mess around with hardware? Software? Both?

So printers like the A8 and CR-10 are great budget options - but are not nearly as plug and play as things like the Taz, Prusa, Ultimachine...

Monoprice seems to be a popular middle ground - lower cost but with reliable hardware and software. Not quite as user friendly as a Prusa or Taz or Ultimachine...but capable of producing quality prints out of the box at an affordable price without the drawbacks of many of the ultra budget options.

 

[jhitesma]

My biggest complaint with the MPCNC is the zinc plated electrical conduit. Tolerances at the mfg is relatively straight. You're better off sourcing the stainless steal tubing that Ted himself upgraded to. And then you get to have some quality alone time with your conduit, some emery cloth and a bucket of water as you sand that bumpy zinc coating smooth. And even then it's still pretty loud and rough.

Quality/availability of conduit can definitely vary. Here in my town I found that Lowes has conduit which is super smooth - but finding straight sticks was tricky. HD on the other hand has the stuff you describe with the bumpy overly thick zinc coating which doesn't work nearly as well...but on the upside they had more straight sticks available.

For a "normal" sized MPCNC conduit is great. But I'd agree when going big enough for DTFB stepping up to SS tubing is a good idea. Mine is conduit and I'm happy with it...but as I've mentioned before I have about .5-1mm of variance in Z across the bed. And it's not very consistent. i.e. it doesn't slope in any one direction. I have spots that are randomly high and others that are randomly low. Though it is consistent enough (the high/low spots don't move around) that I could probably enable Marlin's bed leveling routines to compensate for it if I really wanted to. But I also have the older designed center section which isn't as stiff...and I had some major bed adhesion/support issues when printing some of the center...and my widest center bar is that annoying bumpy conduit which also was the least straight piece I used. So...plenty of things I could do better if I really wanted to.

That said...SS is a lot more expensive and harder to get ahold of. My low-rider is still sitting in a pile because I haven't managed to get my hands on any SS tubing for it yet. I know 3 places in town that can get it...but none of them are open during hours I don't have to work. So dealing with them is really tricky for me. Plus with a working machine I just haven't been in a hurry to spend the money for the SS tubes.

 

[bigALUSMC]

Not bad.
Prints 8 x 8 x 8 and can print hot enough to do ABS and other higher temp filaments.
You're at a good price point for a competent entry level printer. I wouldn't go shopping for a lower cost printer.

CR-10s would be a good choice as well. Note that it's built using V-Slot rails

Prusa if you can afford it.

Yea I think I may be going with the Prusa. A little pricey but you get what you pay for. My worry is having an expensive paper weight. Right now I am satisfied buying speed build kits and may just spend 6 to 12 months 3d printing stuff and make the needle cutter next years project. No hurry at this time.

Now the big question.... kit or pre built?????

Thank you jhitesma, IronKane and dkj4linux.... you guys are top notch!

 

[RAGII]

I just purchased the MK2S Kit and have no regrets. I was tempted to wait for the MK3 but at the time I would have had to wait about 2 months for it.

If you have any assembly skills at all I would save the money and buy the kit. I used the online instructions and had it up and running in a weekend.

I have build tool fixtures for my MPCNC (including needle cutter), control horns, firewalls, and a slew of parts for some of my drone friends.

 

[bigALUSMC]

Outstanding.... Yea I may get the kit as it will help in the long run. They seem like a great company with a good following and support.

 

[TEAJR66]

So... In August 17, i ordered my first batch of parts from vicious1.com. Just some wires, LCD Controler, belts, pulleys and a lead screw with brass nut. The box was lost in the mail, no fault of vicious and a postal inquirey produced nothing. I reordered the parts and built the MPCNC. Love it!

Today, the box of parts made their way to me. Seems being lost in the postal system here sent it to some wrong camps and a waiting room.

Now i am thinking about building another machine. Still very undecided and leaning towards not, but there is a small voice saying "Build it, you will use it.".

 

[Michael9865]

So... In August 17, i ordered my first batch of parts from vicious1.com. Just some wires, LCD Controler, belts, pulleys and a lead screw with brass nut. The box was lost in the mail, no fault of vicious and a postal inquirey produced nothing. I reordered the parts and built the MPCNC. Love it!

Today, the box of parts made their way to me. Seems being lost in the postal system here sent it to some wrong camps and a waiting room.

Now i am thinking about building another machine. Still very undecided and leaning towards not, but there is a small voice saying "Build it, you will use it.".

Glad your original box of parts made it to you. I bought a watch one time that got lost for a while. After the company couldn't find the package either they sent me a replacement watch. They both arrived the same day. The company told me to just keep both watches.

On another note, I never minded the sound of the bearings going over the rails, it was rhythmic and soothing to me. They are still quieter than some machine shops that I have been in. I was able to find some nice straight pieces at the hardware store. Although I did receive some odd looks from passing shoppers as I rolled the conduit across the floor and looked down them like I was looking for straight boards.

 

[ironkane]

Yea I think I may be going with the Prusa. A little pricey but you get what you pay for. My worry is having an expensive paper weight. Right now I am satisfied buying speed build kits and may just spend 6 to 12 months 3d printing stuff and make the needle cutter next years project. No hurry at this time.

Now the big question.... kit or pre built?????

Thank you jhitesma, IronKane and dkj4linux.... you guys are top notch!

Definitely the kit. It'll be easier than assembling Ikea furniture.
And definitely get Autodesk Fusion 360 (free) and it'll never become a paper weight. Every orange plastic piece you see in my build is a 3D printed original designed in Fusion. While plumbing my vacuum table with 1/4" copper tubing, I needed to support a valve so that it wouldn't bend the tubing when opening/closing. It probably took me 15 minutes to design and 30 to print a support piece. I have a friend who had a camp table with a broken plastic part that I recreated for him with extra replacements.
Fusion is pretty easy to learn enough to get you going. Thingiverse is chock full of ready to print items, but eventually you'll want to be able to design your own things. The one thing that I print the most from there is the fossil fish skeleton key fob. It impresses the hell out of friends and family.

With the speed build kits, if you separate the parts from the DTFB carefully, you can use them as templates/stencils to cut more parts by hand. Which is way less labor than printing out the tiled drawings and cutting/taping stencils. Laminating your templates with clear Gorilla packaging tape will keep them in good shape.

 

[Duggem]

Yea I think I may be going with the Prusa. [...] Now the big question.... kit or pre built?????

I'll add another positive experience with a Prusa kit. I built my Mk3 a few weeks ago and had zero problems. Hands down the best step-by-step instructions I've ever used. Make sure to use the on-line assembly manual and read through the comments for tips, etc.

Doug

 

[bigALUSMC]

Thank you ironkane.... that is outstanding and thank you for the ideas on making templates. Also can't wait to start learning this stuff. Possibilities are endless for this hobby and like you said just life in general. Had to throw a few things away because of cheap plastic parts with no way to order them or make them.

Duggem, thank you for your input. I will get the kit.

Long term goal.... we home school and hoping my kids will catch this STEAM bug and design whatever they can envision in their minds. Also hoping to start a park flyers club so the needle cutter will be my next project for sure.

 

[ironkane]

Anyone have any tips on how to initialize a BLHeli ESC from GRBL? When I power up, the PWM output is 1.9v and when I send the S1000 command, it goes up to 5.17v. Executing S0 will cause it to do the three beeps, but doesn't finish the initialization. S1000 doesn't make it beep. Any ideas would be appreciated.

 

[randyrls]

IronKane; If you are using an Arduino/RAMPS board, I read somewhere that the servo output isn't enabled by default and needs a jumper on the board. The memory is fuzzy but I copied this down in my notes, "Remember the servo power isn't supplied by default. There is a jumper near the fuses and reset. This must be in place for the servo to work."

Hope this helps...

 

[ironkane]

IronKane; If you are using an Arduino/RAMPS board, I read somewhere that the servo output isn't enabled by default and needs a jumper on the board. The memory is fuzzy but I copied this down in my notes, "Remember the servo power isn't supplied by default. There is a jumper near the fuses and reset. This must be in place for the servo to work."

Hope this helps...

Thank you, but I'm using a GRBL board. In posting this and then running across the page for my T-Motor ESC, I think I have found the problem. I said my voltage at zero was 1.9V. The T-Motor page said this about the Startup Power setting: "Startup power refers to the max power allowed at startup stage, which can be any relative value from 0.031 to 1.5."

So I think I'm 0.4v too high. Now all I have to do is figure out how to reduce my min voltage to 1.5 or less.

 

[jhitesma]

Whoops, pulled a bit of a Mars Orbiter last night

So I have this non RC project I've been working on for about 13 years. It's a MIDI sequencer which I started buying parts for shortly after building a MIDI synthesizer. But after getting the synth working electrically I started building up the control panel for it - and quickly realized that things like knobs, buttons and cases were ridiculously expensive. And since the sequencer has 2 2x40 LCD's, 16 rotary encoders, 52 buttons and a bunch of LED's....the price to get the front panel for it cut was looking like close to $100 and all the knobs and button caps were going to get expensive too. The final straw was that the project at the time was based on a PIC processor and was getting too big to fit anymore - but there weren't any better options yet available.

Fast forward 13 years...I now have a 3D printer and CNC machine so I can make my own knobs, buttons, cases and panels! And the project has moved from the PIC processor to a STM32F4 using the DiscoveryF4 board as the core of it's brain. Which was convenient since I just happened to have a DiscoveryF4 board on hand which I had used on one of my quadcopters for awhile. But which has been gathering dust for a year or two now since we determined it wasn't a good board to run as a flight controller (nothing wrong with the F4 itself - just the Discovery had some issues with how it's power circuits are wired that makes it a very poor choice for what I was doing with it.)

So for Christmas this year I got myself a professionally made PCB for the control panel circuit and for my birthday in March I got the last of the components I needed.

The board is a beast:


But after a few evenings of soldering I had it assembled enough to test...I just had to wait on 2 new LCD's to arrive from China. I had a bunch of 2x40 LCD's I got dirt cheap 13 years ago - but they're extended temperature range versions which require a negative voltage for contrast and generating that is a real pain. So instead I decided to just order new ones - but that meant waiting for shipping from China. While I waited I started working on the faceplate for the control panel.

The standard design calls for some commercially made button caps which are rectangular. But cutting rectangular ID's with my CNC wasn't going to work cleanly (at least not with a 3mm bit and I didn't want to deal with smaller more fragile bits.) And since I was going to 3D print my caps I could make them any shape I wanted. So I redrew the panel in OnShape (https://cad.onshape.com/documents/8...4432f128a825737854/e/4ef1ed22197c361d177a10e2) along with some button caps (https://cad.onshape.com/documents/4...42ab2936c72c70a86c/e/93f6c34597677bfb5540abb0) designed with a minimum radius of 3mm so they'd be easy to machine:



The buttons themselves turned into quite an adventure. I went through a couple of design revisions - originally making them a snug fit on the switches with no shoulder...but finding I wanted them easier to remove so made them a slightly looser fit and added the shoulder so they wouldn't get lost if they came loose. And printing them took some work as well - I really wanted that nice sculpted top...but at my normal 0.24mm printing resolution it came out pretty lousy. Using slic3r's variable layer height I printed most of the caps at 0.24 but tapered down to 0.1mm for the parts that would stick out...which was better but still made for a not very nice feeling button:



Eventually I gave in and pulled out a spool of ABS so I could vapor smooth them after printing. The combination of 0.1mm layers and vapor smoothing made for some pretty nice caps:



Though I am going to have to redo them once I order some ABS in a better color

The front panel I cut on my MPCNC out of 1/8" PVC sheet since it was cheap and what I had on hand - I had a minor hardware/software issue but got it worked out and Easter Sunday was able to start making some noise with it!





Just to clarify for those who aren't familiar with MIDI sequencers - it doesn't actually generate any sounds itself. The sequencer just tells a synthesizer what noises to make. Basically it's a way to program songs and interact with them while they play. So in the photo I'm using a software synth on the notebook to actually generate the sounds that the sequencer is telling it to play.

But Jason - I thought you said you made a NASA sized mistake?

Yep. At this point it was working - but very fragile. Most people who build them use aluminum front panels which give them a lot of strength/stiffness. My PVC panel is even flimsier than the PCB. So the whole thing can flex like crazy. Half the times you press a button the whole thing just flexes instead of the button pressing. So I decided I needed to add a backplate as I didn't want to deal with trying to cut aluminum on my MPCNC for a stiffer face plate. (I could have looked at a wood plate or acrylic - but those would cost extra and would need to be thicker to be stiff enough so I'm going to stick with the PVC for now.)

I fired up onshape again and modeled a backplate. I already need to cut a new face plate (I forgot half the LED holes when I did the CAM and the LCD cutouts are bigger than they need to be) so I also changed the design so the LCD's can mount to the backplane instead of to the front plate which should make for a cleaner look when finished.







I had to add some cutouts to the back plane to allow wires to pass through - and I didn't bother to model the countersinks I designed (which wouldn't be visible in that view anyway) they aren't necessary but let me use shorter screws and make for a cleaner final product even if it will be in a case where no one will see it.

I had that design done on Monday and ready to cut - but wasn't able to get out to my CNC until last night.

Which is where I made my mistake.

I use metric in my designs. But the only ruler I had in my shop last night was imperial. So I measured the width of the backplane in estlcam (in mm) and did a quick conversion to inches. Not sure if I messed up the measurement in estlcam or the conversion but somehow I wound up thinking the back plate was only 14" across not 17"

I thought I had left myself some extra room...but apparently not enough to cover for my 3" mistake



Oops. I knew I should have positioned the workpiece with more material just in case. Well - still good enough to test with. I printed up a TON of 10mm spacers which I then tapped so I could use them between the panels:



Overall it looks like my design will work. I need to adjust the position of the big wire opening just a few mm - it works but it could be better. If I didn't have to recut the part due to running out of material I'd live with it.



The good news is it made a HUGE difference. Wow does this thing feel more polished now. Button clicks are much more solid and the whole thing just feels better. I don't worry about shorting it out on things or the cut off leads snagging on my pants if i put it on my lap.

Unfortunately something went wrong as well and now it's reading spurious input signals Not sure what happened - going to have to take it back apart to figure out tonight.

Plus I still have to cut the proper front panel and engrave the labels on it. Still have to install all the LED's and the big jog shuttle encoder in the center. And I still have 2 more sub-boards to build up that give it 4 pairs of MIDI input/output connections, then have to finish the case and mount the brain board. All of which will take longer than it should because I'm having fun just playing with this thing

Oh - and for the record I don't plan on using all of those potential mount points. I cut them all because I wasn't sure how many I'd need to stiffen it up enough. I found that just 4 or 5 made a HUGE difference so I'll probably only use 6-8 total - doing all of them would be crazy overkill!

For those who are interested full details about this beast are here including some videos by people far more talented than me showing what it can do: http://www.ucapps.de/midibox_seq.html

 

[ironkane]

Well that's pretty damn impressive. The things you can do with a 3D Printer, a CNC router and some Online shopping...