|
Anyone here have any experience with the Shapeoko 2 desktop CNCs? I was seriously considering the MyDIYCNC Sprite but the Shapeoko 2 has a 12"x12" cutting area compared to the Sprite's 7"x13" for only $50 more. The Shapeoko just seems to look a little more polished compared to some of the other CNC kits that are currently available.
|
# ? Nov 18, 2013 05:51 |
|
|
# ? May 28, 2024 16:14 |
|
Just saw this today. A desktop 5-axis mill with a projected price of $3000.
|
# ? Nov 21, 2013 20:02 |
|
bung posted:Just saw this today. A desktop 5-axis mill with a projected price of $3000. sounds neat but I have no idea why they're trying to make it weigh less also having a hard time imagining that it'll actually cut steel with any effectiveness. regardless, sure hope it works. a 5 axis for 3k sounds sweet as hell.
|
# ? Nov 21, 2013 20:34 |
|
rotor posted:sounds neat but I have no idea why they're trying to make it weigh less Yeah, trying to constantly reduce the weight of a mill seems to suggest a lack of understanding of the concepts at work. Then again maybe the designers only work in foam and wax, and saying it'll cut steel well is wishful thinking. I don't think I'd try to do steel on a CNC without an automatic cooling system anyway. Realistically a 4-axis machine can do about 90% of what a 5-axis can with some clever setup, but I'd like to have a 5 axis yea.
|
# ? Nov 21, 2013 20:44 |
|
I'm not very clever though so it would suit my needs
|
# ? Nov 21, 2013 20:58 |
|
I'm not sure why you'd need 5 axis unless you like impellers because thats all i've seen them make
|
# ? Nov 21, 2013 21:00 |
|
quote:The mill uses stepper motors and they can cut dependably to a tolerance of +/- 1/5000th of an inch I totally believe this 100%. Loving Africa Chaps posted:I'm not sure why you'd need 5 axis unless you like impellers because thats all i've seen them make People who don't understand machining think more axes = better. 5 axis is great for holes and fittings at odd angles and off centerline. Transform plane milling is so easy. oxbrain fucked around with this message at 21:12 on Nov 21, 2013 |
# ? Nov 21, 2013 21:07 |
Yeeeeeeaaaaaah, they clearly have an understanding of where error comes in: it's all at the steppers.
|
|
# ? Nov 21, 2013 21:10 |
|
There's no way that thing can hold 2 tenths with a 25 lb mass, some dodgy looking ways, and a table supported on one side only. The spindle looks surprisingly decent yet I doubt its runout is even nearly that low. edit: Really don't think I trust a machine tool designed and assembled in a shop that consists of a computer and an incorrectly supported surface plate. sixide fucked around with this message at 03:21 on Nov 22, 2013 |
# ? Nov 22, 2013 03:15 |
Yeah, they seem to be not at all shy about taking theoretical max accuracies and using them as actual, real-world values. "These steppers with this gearing means I can get .00000001 mil accuracy! I did the math!"
|
|
# ? Nov 22, 2013 03:22 |
|
Loving Africa Chaps posted:I'm not sure why you'd need 5 axis unless you like impellers because thats all i've seen them make honestly? because they look cool as hell when they're running.
|
# ? Nov 22, 2013 05:46 |
|
also I think you don't have to be as smart about how it's milled but idkwtf I'm just some dumbass
|
# ? Nov 22, 2013 05:47 |
|
You can use the A and B axes to rotate the part so you can do 3 axis milling at different angles, that's not too difficult. Doing true 4th or 5th axis profiling is a gigantic pain in the rear end and the only CAM software that will do that is expensive as hell and most machine controllers won't do it well anyway. For $3k you could get a 3 axis mini mill, a cnc conversion kit, and a mini 4th axis. Another $2-4k will get you a copy of OneCNC, which will do all your 3 axis and some 4 axis stuff. Or you could drop $20-50k on something more industry standard like mastercam.
|
# ? Nov 22, 2013 07:13 |
|
What's the difference between a $30 chinese controller and a $300 one? I'm considering reverse engineering a good one and cloning it. I also probably have access to a cheap one I'm told it's slow and has a lot of backlash(doesn't software correct this?).
|
# ? Nov 22, 2013 23:23 |
|
Speed, power, and build quality. Cheap controllers start skipping steps when you run them too fast or when they're too hot. Now your computer thinks the machine has moved further than it has and everything gets progressively more hosed up. Bigger controllers can drive bigger motors too, which lets you use less reduction from motor to axis for even more speed. The only backlash I could see is if it can't stall the motor, which even the chinese special can do no problem.
|
# ? Nov 23, 2013 03:04 |
|
bung posted:Just saw this today. A desktop 5-axis mill with a projected price of $3000. This person did a neat thing, knows jack poo poo about machining quality or machine tool building.
|
# ? Nov 25, 2013 00:13 |
|
bung posted:Anyone here have any experience with the Shapeoko 2 desktop CNCs? I was seriously considering the MyDIYCNC Sprite but the Shapeoko 2 has a 12"x12" cutting area compared to the Sprite's 7"x13" for only $50 more. The Shapeoko just seems to look a little more polished compared to some of the other CNC kits that are currently available. I preordered the Shapeoko 2 but I don't think they've started shipping any of them yet. I can tell you about it once it shows up though. It looks fairly well polished and they designed it to be easy to expand, it just requires longer Makerslides and some longer g2 belting. Being able to expand it was one of the biggest draws to me.
|
# ? Nov 28, 2013 04:59 |
|
What electronics will you be putting on it? I've been thinking about making a laser machine with a shapeoko 2 base, a tinyG control board and a couple nema 17 to run it. Definitely post updates please.
Comatoast fucked around with this message at 18:15 on Nov 28, 2013 |
# ? Nov 28, 2013 18:12 |
|
So far the plan is to use an Arduino with a Grblshield for the controller,and I want to use Nema 23 steppers on the X axis and Nema 17 steppers on the Y and Z axis. I'm really interested in being able to expand it out to 4x4 or 4x6ish (since the only options on Makerslide are 1m or 1.8m). I know they like to bill it as a cnc mill but it's really more of a cnc router. Although I think it would work great as a laser cutter, and I've seen a few Shapeoko 1 builds that were used with a laser.
|
# ? Nov 29, 2013 06:38 |
|
Chickenbisket posted:So far the plan is to use an Arduino with a Grblshield for the controller,and I want to use Nema 23 steppers on the X axis and Nema 17 steppers on the Y and Z axis. I'm really interested in being able to expand it out to 4x4 or 4x6ish (since the only options on Makerslide are 1m or 1.8m).
|
# ? Dec 7, 2013 06:40 |
|
I got my Shapeoko 2 about a month ago, so it's all put together and I've had a chance to do a few test cuts. I also had the Shapeoko 1. If anyone has any questions about them I might be able to answer. Overall I really recommend the 2. They fixed a lot of the sturdiness issues from the 1 and it's a pretty capable machine, at least for hobby level stuff. You definitely want to upgrade the rotary tool to something better than a Dremel though. I have a question about creating files though. I've got the hang of creating 2D paths at various depths, but now I want to try some 2.5D stuff. So far my workflow is Illustrator-> export dxf -> import into makercam -> export gcode. I just got MasterCam, so I'm guessing I'll use that to make tool paths. It seems kind of janky for actual modeling though. I'm curious what you guys use for a 3D workflow.
|
# ? Feb 27, 2014 17:14 |
|
MasterCAM? Janky? Well, I never! don't even try to use it for modeling. it's a cam software and that's it. Rhino is the most reasonably priced program that's truly professional-grade if you ask me. Download the trial -- it's unrestricted except that it can only save or export 25 times. More than enough to get you started.
|
# ? Feb 27, 2014 19:43 |
|
SomethingLiz posted:I got my Shapeoko 2 about a month ago Did you get the full kit or did you BYO electronics?
|
# ? Feb 27, 2014 20:25 |
|
I used to use the gently caress out of CASMate and I swear after 6 years fulltime of using and abusing it I could basically use it as CAD software. I was way more fluent in it that AutoCAD or anything and there were some nice things about it, since it was sign software it had some great tools for tracing raster images to vector and for tidying them up. The templatey features were amazing and I think I made that program do far more than it was ever designed to do. If it wasn't for that program and my ability to use the gently caress out it that business would have been in the shitter
|
# ? Feb 28, 2014 08:23 |
|
Thanks, I'll take a look at Rhino and CASMate. The learning curve for this CNC stuff is pretty high, but it's so cool when it finally starts making sense and you can cut out your own designs. CrazyLittle, I did the mechanical only kit. One lame thing about the mechanical kit is that they consider the belts and pulleys to be electrical I guess, so if you go that route you have to order belts and pulleys separately.
|
# ? Mar 1, 2014 00:02 |
|
CASMate is pretty old now, you have to find a hacked copy because the legit copies rely on a dongle which the program accesses directly so NT windows can't use it, ie, you have to use Windows 98. It was a program I used when I did this poo poo for a job years ago but I'm sure you can find something more appropriate. When I did CNC I didn't have to bother with tool paths or any of that bullshit. Depends how close to the bits you want to get and how close to real 3D (vs 2.5D)
|
# ? Mar 1, 2014 00:30 |
|
SketchUp DXF -> CamBam -> Mach3 is what I use. Cheap, easy and versatile. CamBam has an import-from-image wizard for 2.5D so you can skip CAD entirely. I've used that to make wax positives for cookie/chocolate molds.
|
# ? Mar 1, 2014 01:59 |
|
This guy does some minor taig lathe/end mill work around air guns. I grew out of guns when I was maybe 12 but I do appreciate his machining.
|
# ? Mar 2, 2014 06:26 |
|
I had no idea what you all were talking about so I looked it up. Holy poo poo: https://www.youtube.com/watch?v=xZOiNdkJ8SU How long does it take to program that up??
|
# ? Mar 2, 2014 07:26 |
|
If you're using CAM software, not super long. A couple of hours maybe. It's a symmetrical part so once you've figured out the general strategy to cut one blade it's pretty simple to repeat it around an axis. Then again, I've seen so many goddamned turbine wheels used as demos for 5-axis machines that I wouldn't be surprised if some mill out there has a "turbine wheel" button that you just push and it makes one. I don't know if it's even possible to program something with that sort of 3D curvature by hand. It would end up being hundreds of thousands of lines of code no matter how you put it together, though, so I sure wouldn't want to try. If you wanna see something really impressive that would actually take several days of concentrated effort to program, even using the best CAM software available, check out these guys https://www.youtube.com/watch?v=7S6NnNm-74A https://www.youtube.com/watch?v=RnIvhlKT7SY Sagebrush fucked around with this message at 08:20 on Mar 2, 2014 |
# ? Mar 2, 2014 08:17 |
|
This is straight up nuts. Wow.
|
# ? Mar 2, 2014 08:29 |
|
thats insane
|
# ? Mar 2, 2014 08:49 |
|
echinopsis posted:thats insane I think that's done in HyperMill not MasterCAM.
|
# ? Mar 2, 2014 13:44 |
|
Oh poo poo, in that case I take it back.
|
# ? Mar 2, 2014 19:11 |
|
Seems likely that it would be possible to make a turbine wheel by hand, or to code a machine to. I remember reading an article years ago about how one of the major Japanese manufacturers (I think it was Hitachi) who received a large fine for selling a 5 axis machine (or maybe it was 6 axis) to the soviet union, suitable for making submarine propellers, in the 1980s.
|
# ? Mar 3, 2014 03:24 |
|
I've gotten my taig mill + steppers + linuxcnc all hosed up. I can't figure out how to zero it. Currently it thinks (0,0,0) is somewhere in the middle of the table about 2" up and it won't let me jog past it. This it putting a serious crimp in my abilities to machine things. he;lp
|
# ? Mar 3, 2014 06:28 |
|
Wild EEPROM posted:Seems likely that it would be possible to make a turbine wheel by hand, or to code a machine to. I remember reading an article years ago about how one of the major Japanese manufacturers (I think it was Hitachi) who received a large fine for selling a 5 axis machine (or maybe it was 6 axis) to the soviet union, suitable for making submarine propellers, in the 1980s. I believe it was a combination of Toshiba milling machines and Kongsberg numeric control (not even CNC!) control systems. I remember seeing a blurb about it long ago in the Red Storm Rising manual. https://en.wikipedia.org/wiki/Toshiba-Kongsberg_scandal
|
# ? Mar 3, 2014 08:12 |
|
Wild EEPROM posted:Seems likely that it would be possible to make a turbine wheel by hand, or to code a machine to. I remember reading an article years ago about how one of the major Japanese manufacturers (I think it was Hitachi) who received a large fine for selling a 5 axis machine (or maybe it was 6 axis) to the soviet union, suitable for making submarine propellers, in the 1980s. Well, it depends how far you want to consider "by hand" and how complex the shape is, I think. From the ground up, the process would be like this 1. design the turbine wheel 2. define the shape of a blade surface mathematically (with eg. bezier or NURBS equations) 3. subdivide that surface along a set of three-dimensional isoparametric curves 4. figure out which isoparametric curves you need to run the tool along to cut the shape out 5. break those curves into three-dimensional linear segments at the desired resolution 6. determine the entrace angle required to reach the first line segment with the tool 7. do the trigonometry needed to work out the x, y, z, a, and b coordinates for the beginning of the segment 8. repeat for the x, y, z, a, and b coordinates of the end of the segment 9. repeat this process for each line segment 10. for good accuracy, there will be roughly 1000 segments per inch of toolpath 11. depending on the size of the wheel, the toolpath could be hundreds or thousands of feet long 12. you must do this for every surface, so multiply this by three for the front, back and edge of the blade, and then by the number of blades on the wheel to make the whole part Possible to do it entirely by hand? sure, if you have a thousand engineers working for a year. You could speed it up a lot by programming a computer to do the trigonometry all those hundreds of thousands of times and write out the values so that you wouldn't have to. But if you're doing that, why not have the computer also do the subdivision of the isocurves into line segments? That's pretty easy for a computer. Then why not have it generate the isocurves from your surface definition automatically? And if you're doing that, why not just start from a digital model of the wheel so you don't have to program the surface into the computer in the first place? And suddenly you have a complete CAD-CAM system from start to finish. Sagebrush fucked around with this message at 10:56 on Mar 3, 2014 |
# ? Mar 3, 2014 10:51 |
|
rotor posted:I've gotten my taig mill + steppers + linuxcnc all hosed up. unless you have some bizarre firmware on the mill controller or something that's gonna be 100% a setting in linuxcnc. The mill doesn't know where its center is and the steppers just spin happily around as many times as you give them a pulse. you could just try resetting the software to default values. also if you don't know how to zero your mill god drat old man, get with it
|
# ? Mar 3, 2014 10:53 |
|
|
# ? May 28, 2024 16:14 |
|
rotor posted:I've gotten my taig mill + steppers + linuxcnc all hosed up. Sounds like it's locked in some tool-change position. Use the EMC2 Stepconf Wizard to write a new .ini-file.
|
# ? Mar 3, 2014 11:25 |