Jump to content


Photo

DeskTop CNC

  • Please log in to reply
22 replies to this topic

#1 BoriSpider

BoriSpider

    Vanguard

  • -----
  • Posts: 2150
  • Joined: 10 Aug 2004
  • Loc: S.W.FLA

Posted 06 May 2013 - 04:51 PM

Here is a nice lil' CNC called The OtherMill on KickStarter.

Looks cool.

#2 Pinbout

Pinbout

    Cosmos

  • *****
  • Posts: 8022
  • Joined: 22 Feb 2010
  • Loc: nj

Posted 06 May 2013 - 05:45 PM

did you ever see this one. seems very easy to put together.

http://www.youtube.c...WzPj8e6TjWJb_lQ

#3 sunktank

sunktank

    Lift Off

  • -----
  • Posts: 16
  • Joined: 30 Oct 2011
  • Loc: Switzerland

Posted 07 May 2013 - 04:48 AM

That first CNC is small - good for etching PCBs and some engraving work, but you're not going to be cutting OTA rings with it.

The second one has a charming hand-made look but I would have doubts over the repeatable accuracy of it.

I recently built a CNC from a kit (a European variant of the Shapeoko called eShapeoko). I'd recommend it, but with some caveats:

1. Building a CNC kit is easy. A week of spare time - you follow the instructions exactly and you have a machine; nothing to think about. The electronics are pretty well documented and discussed on the relevant websites. I'm a ham fist with mechanical stuff but was nevertheless able to build the machine to a good standard. The hardest mechanical part for me was truing-up the frame (getting everything at 90 degrees) and adjusting all the screws and bolts just so for smooth movement. The electronics are no more complicated than a few easy soldering jobs, depending on your controller.

2. Learning milling techniques isn't complicated but it takes time and a methodical approach. There is a good reason a machinist's apprenticeship takes years. "Machinery's Handbook" is the standard reference and runs to over 2700 pages. What endmill do I need, at what feedrate and spindle speed for material X? Climb or conventional milling? How many flutes on the endmill? Forums like CNCzone can help here but expect to turn a *lot* of scrap wood into sawdust during the testing phase. Earplugs should be near the top of your purchase list!

3. The software toolchain *is* complicated. I'm an IT network engineer by trade and I found the software learning curve daunting.

(a) you need to learn at least some G-code (offsets, limits, that sort of thing). Generating G-code (i.e. the code that says "move the spindle 3.574 cm in X, 2.789cm in Y..."), from your CAD design is a complex process, mostly handled by software. But you need an understanding what it is doing in order to diagnose problems; a trivial example would be the order in which it cuts features into your stock.

(b) The controller software alone can be a complex application in its own right. Just look at some screenshots of Mach3 for an idea.

© depending how far you want to go, CAM software ranges from free to tens of thousands of dollars. As always, you gets what you pays for.

All that to say, if you're dreaming of turning your nice clean Solidworks designs into neatly milled parts within a month of purchase, think again. It is a very deep subject, there is no substitute for learning by doing; be prepared to make lots of mistakes.

Having said all that, I am very pleased with my machine, I can machine plywood, plastic and aluminium to less than 0.1mm accuracy, and after 5 months of variable-rate spare time effort, I'm about ready to start milling an enclosure for the electronics. YMMV.

#4 careysub

careysub

    Surveyor 1

  • *****
  • Posts: 1906
  • Joined: 18 Feb 2011
  • Loc: Rancho Cucamonga, CA

Posted 07 May 2013 - 02:10 PM

Very interesting. A tabletop CNC machine is the competition with the 3-D printer it would seem. Printers appear simpler to operate, but produce semi-porous ABS parts of rough dimensional precision.

A CNC machine can produce more precise parts out of a high quality material (factory produced polymer blocks, including fiber reinforced, high perfomance composites, Al, and wood, etc.) but are complex to operate.

Which will succeed in overcoming their key limitations first?

I am betting on the CNC. Overcoming the materials problem in a low cost printer seems difficult, possibly insurmountable, if usable engineering parts are the desired result. For the CNC it is a matter of software development and integration, which is a problem that need only be solved once. Many very sophisticated software applications that were one pie-in-the-sky (voice and image recognition, automated route planning, etc.) are now embedded in low cost consumer products.

#5 Pinbout

Pinbout

    Cosmos

  • *****
  • Posts: 8022
  • Joined: 22 Feb 2010
  • Loc: nj

Posted 07 May 2013 - 03:01 PM

All that to say, if you're dreaming of turning your nice clean Solidworks designs into neatly milled parts within a month of purchase, think again. It is a very deep subject, there is no substitute for learning by doing; be prepared to make lots of mistakes.

Having said all that, I am very pleased with my machine, I can machine plywood, plastic and aluminium to less than 0.1mm accuracy, and after 5 months of variable-rate spare time effort, I'm about ready to start milling an enclosure for the electronics. YMMV.




or I can take my drawing over to the shop and let them cut it. :grin:

but my shop would charge me $95/hr plus $75 setup if I wasn't an employee.

so I went to a waterjet guy who would charge me $40 to cut this, I have to supply the materials.

Posted Image

#6 sunktank

sunktank

    Lift Off

  • -----
  • Posts: 16
  • Joined: 30 Oct 2011
  • Loc: Switzerland

Posted 07 May 2013 - 04:41 PM

A tabletop CNC machine is the competition with the 3-D printer it would seem.


I see them as being complimentary; it depends what your needs and methods are.

Of course, I'm biased, but: bear in mind it's pretty easy to convert a CNC into a 3D printer, but not the other way round. In my case, all I would need to do is replace the spindle with an extrusion head, and depending on my material, add a heated bed. In fact the guy that makes the kits uses his exclusively as a 3D printer. I've frankly no idea how quick you can make this change but with some creative engineering I'm sure it would be possible to mill out a quick-change sled to do it in minutes. A laser cutting head is of course also possible, and with that you can cut steel.

3d printers aren't made to handle more torque than is required to move the print head so a milling conversion is impossible, but a laser would be possible. Frankly I find my 1kW spindle scary enough, a CO2 laser able to vaporize steel is further than I want to go: astronomy is best enjoyed with functional eyes.

I could also extend both X and Y axes up to around 700mm each by simply replacing the gantries before flexure of the aluminium extrusion becomes a big problem, but I don't currently plan on making anything that big. If you need bigger than that, you really have to start looking at a steel frame.

During my research phase, I looked at both printers and CNC, and decided on CNC for exactly this flexibility. As has been mentioned in several previous threads recently, 3D printers aren't really there yet in terms of turning out robust parts, although that may change soon. If and when it does, I'll be ready.

That's not to say 3D printers don't have a place in the toolbox right now, if you are a visual thinker you can rapid prototype a whole bunch of different designs that are not much more than sketches very quickly and see what works/fits. There are times when the rigorous CAD process feels like a roadblock to your inspiration and you know you can just whip up a non-precision model in 10 minutes in 3dsmax or Sketchup that would do the job fine.

- Barry.

#7 don clement

don clement

    Vendor (Clement Focuser)

  • -----
  • Vendors
  • Posts: 812
  • Joined: 02 Feb 2011
  • Loc: Running Springs, California

Posted 07 May 2013 - 11:12 PM

The learning curves are quite different for using a CNC mill vs. using a 3D printer. Basically if one is proficient in 3D CAD (e.g. Solidworks) then the learning curve to competently using a 3D printer is very easy. For me the learning curve to competently using a 4-axis CNC mill was not so easy and depends on factors such as how much one’s experience with manual machining. A big part of using a CNC mill is knowing what tool to use for a given process and what feeds/speeds to use for a specific material. Another big factor is how to hold and fixture parts. For me fixturing and how to hold a part to be machined can be the most difficult and time consuming part of the machining process manual or CNC. Then there is how one gets those 3D CAD models into G-code to run the CNC mill. Simple stuff can be coded by hand. I highly recommend “The CNC Programmers Handbook” by Peter Smid http://www.amazon.co...31133473/ref...
CAM programs can generate the necessary G-code from a CAD 3D model and also simulate tool paths. One still must decide on the tool type and the proper speeds/feeds. Be prepared for yet another learning curve in using CAM software. Then there is the cost of good CAM software. The industry standard in the US is MasterCAM http://www.mastercam.com/ . Be prepared to pony up $10K+ for MasterCAM that can do 3D with a 4th axis. I went with Russian SprutCAM http://www.sprutcam.com/ that has the same capabilities as MasterCAM but is $9K less. However SprutCAM has its own very steep learning curve.

Don Clement

#8 Jim Curry

Jim Curry

    Surveyor 1

  • *****
  • Posts: 1700
  • Joined: 29 Oct 2007
  • Loc: STL

Posted 08 May 2013 - 06:19 AM

While I've used Solidworks in a work environment I don't have the $ to play with it at home. It's a multi-thousand dollar program. I use the "hobby" version of Alibre for solid modeling. They have a great tutorial CD package to get over the learning curve quickly.

As far as CAM is concerned, same thing. Mastercam, etc are expensive for home use. Mach4 is a reasonably affordable system to converting your files to Gcode. There's a bunch of videos on youtube. Look up Ron Ginger, I know I've watched some of his intro videos at a live steam show.

Jim

#9 sunktank

sunktank

    Lift Off

  • -----
  • Posts: 16
  • Joined: 30 Oct 2011
  • Loc: Switzerland

Posted 08 May 2013 - 08:54 AM

Mach4 is a reasonably affordable system to converting your files to Gcode.


I could be wrong but it was my understanding Mach3/Mach4lite doesn't convert files. It just reads G-code generated by the previous CAM step and outputs pulses on the parallel port for the controller to read - that's how I'm using it anyway. So an example toolchain in the case of Solidworks:

Solidworks part -> SLDPRT or export to another 3d file -> CAM package -> G-code -> Mach3 -> hardware controller -> CNC

You can do some basic importing of 2D DXF/JPG/BMP into Mach3 via LazyCAM. I tried this but gave up as the g-code it produced was iffy. Not sure if my DXF was all that great to start with.

The reason that going from a 3d file to good g-code (i.e. the CAM package) is so expensive is because once you are trying to calculate the toolpath for a >3-axis machine, it is an NP-complete problem. You can optimize for cutting time, or tool contact time (for max tool life), or not crashing the tool into the stock or holder, or any other parameter, and there are a very large number of possible solutions, so how does the CAM package choose the optimal one? A traveling salesman problem in 5 dimensions...ouch :help:.

Check out a 5-axis mill in action: http://www.youtube.c...h?v=RnIvhlKT7SY

Now think how complex it was to calculate the g-code for that!

But hooo boy, that thing could probably mill an entire 16" dob out of a single block of Al in one go: focuser, spider, most of the mirror cell, included. And probably make you a coffee while you wait. :lol:

- Barry.

#10 Mary B

Mary B

    Vendor - Echo Astronomy and Electronics

  • *****
  • Vendors
  • Posts: 3070
  • Joined: 21 May 2010
  • Loc: Minnesota

Posted 08 May 2013 - 02:51 PM

CO2 lasers typically aren't used for steel cutting, I have a 60 watt and it won't even mark metal unless I use a dye process or it is anodized or painted first

A tabletop CNC machine is the competition with the 3-D printer it would seem.


I see them as being complimentary; it depends what your needs and methods are.

Of course, I'm biased, but: bear in mind it's pretty easy to convert a CNC into a 3D printer, but not the other way round. In my case, all I would need to do is replace the spindle with an extrusion head, and depending on my material, add a heated bed. In fact the guy that makes the kits uses his exclusively as a 3D printer. I've frankly no idea how quick you can make this change but with some creative engineering I'm sure it would be possible to mill out a quick-change sled to do it in minutes. A laser cutting head is of course also possible, and with that you can cut steel.

3d printers aren't made to handle more torque than is required to move the print head so a milling conversion is impossible, but a laser would be possible. Frankly I find my 1kW spindle scary enough, a CO2 laser able to vaporize steel is further than I want to go: astronomy is best enjoyed with functional eyes.

I could also extend both X and Y axes up to around 700mm each by simply replacing the gantries before flexure of the aluminium extrusion becomes a big problem, but I don't currently plan on making anything that big. If you need bigger than that, you really have to start looking at a steel frame.

During my research phase, I looked at both printers and CNC, and decided on CNC for exactly this flexibility. As has been mentioned in several previous threads recently, 3D printers aren't really there yet in terms of turning out robust parts, although that may change soon. If and when it does, I'll be ready.

That's not to say 3D printers don't have a place in the toolbox right now, if you are a visual thinker you can rapid prototype a whole bunch of different designs that are not much more than sketches very quickly and see what works/fits. There are times when the rigorous CAD process feels like a roadblock to your inspiration and you know you can just whip up a non-precision model in 10 minutes in 3dsmax or Sketchup that would do the job fine.

- Barry.



#11 sunktank

sunktank

    Lift Off

  • -----
  • Posts: 16
  • Joined: 30 Oct 2011
  • Loc: Switzerland

Posted 09 May 2013 - 06:30 AM

My bad...don't know much about lasers...I always figured they were the holy grail of CNC machines, potentially the most versatile and precise of all.

For one thing I seem to spend a lot of time vacuuming my workspace, the spindle produces clouds of really fine wood dust that gets onto/into everything. I plan to eventually make a spindle shoe with a vacuum-cleaner attachment. Presumably lasers don't have this issue?

I'm curious, are there any drawbacks to laser cutters? I love the idea of them but I'm perhaps irrationally scared of a chance reflection going near my eyes. They seem to leave scorch marks on wood but presumably this is only a few mm and can be sanded out?

#12 careysub

careysub

    Surveyor 1

  • *****
  • Posts: 1906
  • Joined: 18 Feb 2011
  • Loc: Rancho Cucamonga, CA

Posted 09 May 2013 - 07:36 AM

... Presumably lasers don't have this issue?

I'm curious, are there any drawbacks to laser cutters? I love the idea of them but I'm perhaps irrationally scared of a chance reflection going near my eyes. They seem to leave scorch marks on wood but presumably this is only a few mm and can be sanded out?


They do vaporize and/or combust stuff. Inhaling the vapors produced by laser cutting is probably not recommended.

A consumer laser cutter is probably not going to offer you the opportunity to "see" the beam (CO2 lasers are infrared) through the use of shields, chamber walls and such.

I can't operate my food processor unless the Lexan lid is locked into place.

One interesting variation of the laser cutter is the laser water jet hybrid. Saith Wikipedia:

"A laser microjet is a water-jet guided laser in which a pulsed laser beam is coupled into a low-pressure water jet. This is used to perform laser cutting functions while using the water jet to guide the laser beam, much like an optical fiber, through total internal reflection. The advantages of this are that the water also removes debris and cools the material. Additional advantages over traditional "dry" laser cutting are high dicing speeds, parallel kerf and omnidirectional cutting."

This removes the problem of vapor control and cooling (often water is needed for cooling anyway in non "microjet" systems).

#13 don clement

don clement

    Vendor (Clement Focuser)

  • -----
  • Vendors
  • Posts: 812
  • Joined: 02 Feb 2011
  • Loc: Running Springs, California

Posted 09 May 2013 - 11:39 AM

There is also EDM. http://en.wikipedia....harge_machining

Don Clement

#14 Achernar

Achernar

    Cosmos

  • *****
  • Posts: 9126
  • Joined: 25 Feb 2006
  • Loc: Mobile, Alabama, USA

Posted 09 May 2013 - 12:01 PM

Lasers that vaporize metal ARE extremely dangerous. Not only will they burn, and slice through skin and the flesh below, they are a grave threat to the eyes. Radiation reflecting off of the surroudings will cause permanent damage to your vision without protective eyewear. Just because the beam is invisible to the unaided eye, does not mean it's not harming your eyes.

Taras

#15 careysub

careysub

    Surveyor 1

  • *****
  • Posts: 1906
  • Joined: 18 Feb 2011
  • Loc: Rancho Cucamonga, CA

Posted 09 May 2013 - 12:27 PM

Indeed, that's why a secure shield interlock to prevent operation of the unshielded laser would be incorporated into any such product (unless they really want to get sued out of business) - it looks like consumer laser engravers already have these features.

#16 sunktank

sunktank

    Lift Off

  • -----
  • Posts: 16
  • Joined: 30 Oct 2011
  • Loc: Switzerland

Posted 09 May 2013 - 02:21 PM

Indeed, that's why a secure shield interlock to prevent operation of the unshielded laser would be incorporated into any such product (unless they really want to get sued out of business) - it looks like consumer laser engravers already have these features.


That's certainly true of commercial products, and I'd be happy to use one. There are also DIY/open-source options that are a whole lot cheaper and probably don't include such a feature unless you add it in yourself:

http://www.buildlog....-net-2-x-laser/

...my point being, wiring up a 40W laser yourself is unforgiving to errors, and outside my personal comfort zone.

Finally got a picture of my CNC if anyone's interested:

Posted Image

yes, those are zipties tensioning the belts, and yes, that is scotchtape securing the e-stop switch :grin:

Total cost was about $1300 for all the HW. It could certainly be made cheaper; I went for a high quality Gecko G540 controller board instead of an Arduino, and a Kress 1050 mill instead of a Dremel tool. It's sitting on an Ikea Bekvam kitchen table (the perfect size and very sturdy).

More info and tech specs:

http://store.ambersp...anical-kit.html

- Barry

#17 polaraligned

polaraligned

    Viking 1

  • -----
  • Posts: 691
  • Joined: 26 Dec 2008

Posted 09 May 2013 - 06:07 PM

CO2 lasers typically aren't used for steel cutting, I have a 60 watt and it won't even mark metal unless I use a dye process or it is anodized or painted first


I have had metal cut out with a CNC laser. It cut 1/4" steel for me to a few thousandths tolerance and left a very nice edge. These lasers are running in the few kilowatt range, and yes, they are CO2.

#18 Pinbout

Pinbout

    Cosmos

  • *****
  • Posts: 8022
  • Joined: 22 Feb 2010
  • Loc: nj

Posted 09 May 2013 - 07:28 PM

Finally got a picture of my CNC if anyone's interested:



cool pic, thanks for sharing.

#19 sunktank

sunktank

    Lift Off

  • -----
  • Posts: 16
  • Joined: 30 Oct 2011
  • Loc: Switzerland

Posted 10 May 2013 - 06:04 AM

Another big factor is how to hold and fixture parts. For me fixturing and how to hold a part to be machined can be the most difficult and time consuming part of the machining process manual or CNC.


I'll second that! Even with my relatively simple 3-axis home-brew system cutting mostly 12-15mm plywood parts, I spend a lot of time figuring out how to hold the stock. With a 4-axis mill like yours it must be even more of a headache.

I found I also need to think about what happens at the end of the cut:

One the first pieces I cut was part of a spindle holder - a simple 2.5D closed profile cut:

Posted Image

Once the mill had plunged through the last layer of the cut, it liberated the finished piece from the stock, which caught in the tool and was projected across the room before I could hit the e-stop. Not fast enough to cause injury, but fast enough to give me pause for thought. Since then, I operate it with my hand hovering an inch above the e-stop switch at all times. Essentially, what I'm doing is giving a home-made robot the control of a potentially lethal tool. In the movies, that usually doesn't work out so well...

#20 careysub

careysub

    Surveyor 1

  • *****
  • Posts: 1906
  • Joined: 18 Feb 2011
  • Loc: Rancho Cucamonga, CA

Posted 10 May 2013 - 06:12 AM

...my point being, wiring up a 40W laser yourself is unforgiving to errors, and outside my personal comfort zone.


Right there with ya. No way would I DIY a laser cutter.

I consider a table saw to dangerous to own or use, and only ever use jigs or sleds to feed material on my router table.

#21 Pinbout

Pinbout

    Cosmos

  • *****
  • Posts: 8022
  • Joined: 22 Feb 2010
  • Loc: nj

Posted 10 May 2013 - 08:07 AM

Once the mill had plunged through the last layer of the cut, it liberated the finished piece from the stock, which caught in the tool and was projected across the room before I could hit the e-stop. Not fast enough to cause injury, but fast enough to give me pause for thought.


your suppose to leave little tabs that keep the parts attached to the material. the tabs should be left off the last pass and thin enough that you can push the part free.

#22 sunktank

sunktank

    Lift Off

  • -----
  • Posts: 16
  • Joined: 30 Oct 2011
  • Loc: Switzerland

Posted 10 May 2013 - 09:05 AM

Once the mill had plunged through the last layer of the cut, it liberated the finished piece from the stock, which caught in the tool and was projected across the room before I could hit the e-stop. Not fast enough to cause injury, but fast enough to give me pause for thought.


your suppose to leave little tabs that keep the parts attached to the material. the tabs should be left off the last pass and thin enough that you can push the part free.


Yep, I know that now :blackeye: :grin:

#23 herrointment

herrointment

    Aurora

  • *****
  • Posts: 4805
  • Joined: 12 Mar 2011
  • Loc: North of Hwy. 64

Posted 10 May 2013 - 07:08 PM

I'll stick with my bandsaw for the time being.






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics