January 19, 2019 at 1:12 am #84462
I guess I should start out with a bit about myself.
My name is Tim, I’m a guy in Georgia with a background in CNC machining, CAD, programming (in progress computer science major). I feel like I’ve done a little bit of everything but in reality I’m a complete noob compared to some of you.
I’d been wanting to get into CNC machining since like 2011 and looked at converting a mini-mill, but I didn’t have anywhere to put it and realized even though I could afford to do the conversion I’d end up spending just as much money on tooling and other.
Sooo, fast forward to college, and I got involved with the FSAE team. I ended up basically teaching myself to CAD, CAM, and machine in 6 months and made the vast majority of the parts for the entire car myself. Including all of the composite body molds on a 5 axis C.R. Onsrud with a work area of like, 8’x8’x5′. Drool. (The process looked very similar to this [not my video])
Took some time off from school and got a full time job doing CAD/mechanical design/CNC work (thanks for the experience, FSAE!). About a year in, I decided I missed machining so I started to look into routers. Of course, I was looking at Shapeoko 3/XCarve. In fact, I bought one–a Shapeoko 3 XXL (with the DWP611). It’s sitting in the box, unopened, because I’ve been really busy at work and it’s too cold in my shop (propane heater coming this week). Somehow ended up here and decided I couldn’t pass up the idea of a $300 standalone machine.
Of course, my printers have been running ever since. The parts are a joy to print–tolerances are great, you can really tell they were designed for 3d printing, etc. I ordered the hardware from McMaster (only because I’m not sure what I want to do with the electronics yet, and I’m anxious to start building) and once it arrives I’ll start assembly. Oh, and funnily enough, my girlfriend’s father started to build his own router a few years ago, but once he got to the electronics he gave up–so I got all of his hardware for free. Some 80 bearings, a bunch of belt, pulleys, lead screws, etc. Haven’t measured anything but the bearings yet–and they are 608 2-RSs. If nothing else that will allow me to get the mechanical side done ASAP.
My impetus for getting these machines is that I’m going back to school this fall and I’d like to be able to make a few bucks here and there to help with that (and if you didn’t guess from this post, I’m kind of a machining/creator junkie). Of course I have some particulars in mind (having my own machines has been running through my head for years along with millions of project ideas) but if you want to see that, you’ll have to follow along 🙂 Sizing of the machine is based off of what I want to do with it. If I need more (or less!) I can always change out some conduit and belt easily enough.
Pictures coming soon! Not a lot to see yet. Hoping to build my table Sunday.January 19, 2019 at 12:42 pm #84514
Well you have my attention. Can’t wait to see what you think. since you have an xxl I recommend building this one smaller for some metals perhaps?
made the vast majority of the parts for the entire car myself
Kinda crazy with school, you get out what you put in, sounds like you had a cruddy group. Once I found a good group we just had such a great time and wonderfully divided workload.January 19, 2019 at 8:29 pm #84583
since you have an xxl I recommend building this one smaller for some metals perhaps?
Ugh. I was typing a reply to this earlier and my power went out (never happens) 9.5 hours into the XY piece print. Fail.
Anyway. I have considered this and I think it’s something I’ll explore depending on how this build goes. I did design a VMC style machine earlier this year that is entirely built out of 2D plates. I designed it this way because I still have access to a big daddy Trumpf fiber laser through my old job… but never got around to sending over material/DXFs. The design was a mix of those plates and epoxy granite to add strength and of course to help with vibration damping (for those who don’t know, epoxy granite is roughly 10x better at vibration damping than cast iron, which is used on pretty much every machine tool out there). I basically designed the plates as molds that stay in place after casting the epoxy granite. I specced it out with linear ways, ClearPath servos, and had a counterweight system designed for the Z axis. I’m no machine designer but I think it would be a decent little machine.
What I’m getting at is it might be interesting (albeit a lot less easy) to cut those pieces on a MPCNC. So yes, this is something I’m going to consider doing. But for now, cutting metal is not as high of a priority. Since these machines are so cheap I would have no problem building a second smaller one to play with machining non-ferrous metals.
Just found out I have to print 120 pieces of something for work, so all 3 of my printers will be busy with that for a couple days. Which means I will indeed probably build my table tomorrow. Will be sure to post pictures of that.February 28, 2019 at 5:46 pm #91221
I wrote a long post here a couple days ago but I edited it immediately and I guess I deleted it or it got caught in the spam filter. The basic gist was that I didn’t get much done because all 3 of my printers were taken to the office to run 24/7 (along with myself, lol) to make an upcoming order… Been working really hard to catch up with other jobs since that order shipped and preparing for this bad boy to get installed in my office…
Now that it’s installed, and I brought one of my printers home, and I’m not working 7 days a week anymore… progress on the MPCNC can continue. 🙂 In the meantime I have ordered everything (thank you for the quick shipping, Ryan!)March 4, 2019 at 8:18 pm #91822
Making progress. Slow progress, but progress! I’m sick, too, so double whammy. 🙂
Realizing the only thing holding me back is the table so I got home from work today and glued up my top. I don’t have a good flat reference surface to work from, so I figured my best bet to get a flat top was MDF. Pretty simple, just 2 layers of 3/4″ MDF with glue between the two. Screws from the underside to provide clamping pressure for the glue. Sitting on my saw horses curing it’s as flat as any straight edge I have. As flat as the conduit is straight at least. So I think that is accomplished… I just have to move it into my workshop after the glue cures and build some legs/undercarriage for it now. Side note, 55″x40″ is a lot bigger than it sounds on paper. 🙂
I also started to assemble some of the 3d printed parts. And I gotta say–Wow! They fit together like a dream. My printers were well dialed in before I started but I didn’t expect everything to fit so well without any tinkering, sanding, drilling, etc. Well engineered 🙂 I’m very impressed and happy with how that’s coming. I’m still waiting on maybe 10 more hours of prints to finish. I can’t wait to start assembling the center section/Z axis.March 9, 2019 at 12:03 am #92349
Seems like all my posts are disappearing as soon as I make them… so we shall see if this one makes it through the spam filter.
Felt like I got a lot more done till I looked at this picture. lol! I’m still sick though so taking the scenic route…
X/Y seem to be very perpendicular but not sure on the Z yet. My holes for those C tension screws were a bit undersized so it seems like even with the nut loose it’s still exerting more pressure on the assembly than it should be. Also, anyone know if the z axis leadscrew nut is supposed to be on top or bottom? Instruction pics I saw showed it on top but also say the screws do nothing but keep it from rotating. If it’s on top the screws would also be holding the Z up or no?
Still on track to be moving this weekend.. as long as I can find my soldering iron 😛
But for now I’m going to bed since it’s 3am…
1 user thanked author for this post.March 9, 2019 at 4:34 am #92357
Leadscrew but on top. Screws just keep it from turning, and should be slightly loose.March 9, 2019 at 8:57 pm #92470
Apparently I’m inventing new problems. I don’t have any zip ties that fit in the corner pieces. They are all just slightly too big. So I guess I’m going to walmart.. at midnight 🙂
Other than that, I got all the wiring extended and hooked up. Powered on the mini rambo and all steppers move as they should… Z axis moves perfectly. Happy!March 10, 2019 at 7:22 am #92504March 10, 2019 at 10:24 am #92525
It looks kinda awful but I think it’s possibly due to the sharpie I used.. maybe? hopefully. The fact that the starts and ends match up as they should gives me some hope. Still some fine tuning to do obviously…
Edit: again with a pen…
Hmmm. Dunno what to think. Using the pen holder and pen is zip tied tight. no movement between pen and holder. flex between pen and tool mount, obviously. I did notice one of the bearings isn’t touching the y rail so there is slight movement in the middle assembly if i push on it with some force, but not enough where the pressure from the pen is causing it to move. I’m kind of stumped, it looks like the actual movement from the machine is not smooth. Could it be a belt tension issue?March 10, 2019 at 12:07 pm #92540
Could just need to wear in the bearings on the conduit. You may want to also look at the tension bolts in the middle assembly and make sure that they aren’t too tight. Also check the X and Y rollers to make sure the compression bolts (the 2.5″ bolts that run through the Roller Plate) aren’t too tight.March 10, 2019 at 3:12 pm #92567
Could just need to wear in the bearings on the conduit. You may want to also look at the tension bolts in the middle assembly and make sure that they aren’t too tight. Also check the X and Y rollers to make sure the compression bolts (the 2.5″ bolts that run through the Roller Plate) aren’t too tight.
Thanks BT. The tension bolts in the middle assembly are completely backed off with only the nylock engaged. Though the wear pattern on the conduit is much bigger on the middle X/Y rails than the outer X/Y rails…. Which may make sense since the two outside rails on either side only hold 1/2 the weight…
The other tension bolts are only engaged enough to get all bearings in contact with the conduit… though I did try with them looser and it made no noticable difference.March 13, 2019 at 6:49 pm #93006
Did my first cut tonight!
Realized that the endmills I ordered were 2 flute, not 1 (they were the top rated item on Amazon when I searched for “single flute 1/8″, shouldn’t shop while tired I guess) but they ended up cutting decently in this red oak scrap I had laying around. Did make a chip, albeit very small, at 17mm/s. DOC 1.5mm and optimal load of 1/6 tool diameter. Helical ramp of 2 degrees at 6mm/s. I obviously have a lot of room to go deeper and faster (lol). I was quite happy with the surface finish, feels smooth to the touch. Almost forgot, this was climb milled…
Before I cut any more I need to sheathe all my wiring and put some holes in the table. Have not fully decided if I want to raise the feet up 3/4″ so I can have my 4″ of Z travel or just put my spoil board on top of the table and lose 3/4”.
Thanks to everyone who helped. It turns out most of my issues were caused by having the feed rate override at 300% somehow. And by somehow, I mean I was messing around with repetier host and forgot I touched that slider. 🙂March 14, 2019 at 8:24 am #93107March 14, 2019 at 9:57 am #93130
And by somehow, I mean I was messing around with repetier host and forgot I touched that slider.
Shhhhh, it was a bug for sure, none of us make mistakes. Well at least you know your machine can move real fast if it has too!
Yeah I’d like to point out from my end it looked like the issue was the pen mount not being able to keep up with the machine at that speed. So that’s a good sign!March 15, 2019 at 11:38 pm #93371
Cutting good, other than the fact that I can’t get repetier-host to keep the steppers energized when the job finishes, so the Z starts to drop into the work.
I experimented a bit with feeds and speeds and got a lot more aggressive with my cuts. Still feel like I have a lot of room to grow, though. But I wanted to cut something more than a few holes, so I pulled out a file I made that I’ve wanted to cut for a while… I think it turned out pretty good. Obviously I wouldn’t use mdf for the final product but it was scrap I had laying around. It wasn’t too awful with my upcut bit but there were quite a few fuzzies to remove. I’ll definitely be ordering a few downcuts just to experiment with, if nothing else.March 16, 2019 at 8:09 am #93383
There is a setting in repetier to kill steppers after job completion, or something like that.March 16, 2019 at 9:17 am #93386
There is a setting in repetier to kill steppers after job completion, or something like that.
It does not do anything for meMarch 16, 2019 at 9:52 am #93389
There are only two way they will shut off, the last few lines of your gcode M84, or the repetier toggle. If it not an issue with repetier then how do you generate your gcode and is there an m84 present?March 16, 2019 at 10:39 am #93398
There are only two way they will shut off, the last few lines of your gcode M84, or the repetier toggle. If it not an issue with repetier then how do you generate your gcode and is there an m84 present?
Oops. Yeah you’re right, apparently the fusion post does an m84 by default. I’ll have to fix that..March 16, 2019 at 10:45 pm #93420
Modified my F360 post to do more of what I want it to do…
Took a trip to this place that’s literally 1 mile away from my house (I say this because I live in the middle of nowhere, nothing is ever 1 mile away from my house), think “pickers”. They go all over the place buying antiques and re-purposing them, but they also sell a lot of reclaimed wood for decent prices. So I picked up some nice looking 10″x3/4″x8′ pine boards… For testing purposes
Right off the machine…
Trying my hand at painting
I actually kind of like the tooling marks on the orange parts. I think they look sort of like tree rings…
Also ordered a 1/8″ downcut bit, a couple 1/16″ endmills, and some 45/60 degree endmills.
Still waiting on my wire sheathing and threaded inserts from McMaster. Should be here Monday.March 17, 2019 at 9:23 am #93456
Wow, the tool marks are great, sharp edges and thin walls. That seems like you found some really good settings.March 17, 2019 at 9:30 am #93461
Wow, the tool marks are great, sharp edges and thin walls. That seems like you found some really good settings.
Not to tell on myself, but in defense of the machine… the only bad areas (you can see inconsistency in toolpath depth in a couple spots) are where I didn’t have it held down well enough and it could flex upwards resulting in a deeper cut…
Still rocking the 2 flute BTW 🙂March 18, 2019 at 9:39 pm #93659
Tonight I got a couple things done…
Raised the feet up 3/4″ using a 3.5″ wide by 35″ long MDF strip… then got a piece of MDF to fit between those 2 strips for my replaceable spoilboard.
Then I set to work on my wiring mess. Got all the wires sheathed in some purdy expandable polyester sheathing. Got the box for the minirambo mounted on the side of the table… Went well but I need to figure out what to do with my Z wiring/router power cord. For now it will be fine…
Also measured my actual usable X/Y area and it came out just a bit higher than 24×42. Who’da thunk it?
Tomorrow I plan to drill holes in the spoilboard and maybe surface it. I plan to just plunge the 1/8″ endmill into the spoilboard maybe 1/4″ as a pilot hole then drill the rest of the hole with a hand drill. Since my table top is now 2 1/4″ thick… haha
Depending on how long that takes I have some other stuff I want to do. Will post some pictures tomorrow when I’m bored at work, lol.
I want to do some testing with fusion and see how hard I can really push this machine and the tooling I have. I will do it as scientifically as possible (changing only one variable at a time) and post my results… lots of features in HSMworks that have been added that I want to try out.March 19, 2019 at 9:16 pm #93834
Got my holes drilled and inserts put in..
Then started playing around with HSMworks a bit.
My base settings are as follows…
Tool: $1.40 each 1/8″ 2 flute carbide FEM https://www.amazon.com/gp/product/B010NI39WO
RPM: 30,000 (supposedly, also less with vacuum on haha)
Cutting feed: 1200mm/min
Rapid feed: 2000mm/min
Ramp feed: 900mm/min
Optimal load: 2mm (I could definitely go full width but I feel like the chips evacuate better this way)
Retraction policy: minimum (IMPORTANT! Since our machines move so slowly in the Z axis, you can save a LOT of time if you configure your toolpaths to keep the Z as far down as possible as often as possible! This is especially important when using an adaptive strategy where the z will often retract and traverse in order to continue climb milling, more on that later.)
Using these settings with the 3d adaptive strategy (aka trochoidal milling) leads to a cut time of 2:02 (test #1) for the test model I used.
Since I mentioned adaptive chooses to climb mill, I will point out a feature that was recently added to HSMworks that is great for routers. That is the “both ways” check box. As you might guess, this allows adaptive to work not only in climb milling but also conventional. IN THEORY! This would be great for our machines, since we don’t have blazing fast rapids, and we’re cutting wood where it isn’t as important to stick to climb milling… so instead of lifting the Z and traversing back to a spot where it can climb mill, it would simply turn around and keep going.. but in practice, on this particular model, it was actually *slower*, at 2:25 (test #2). This is because it was lifting the Z axis all the way to the top of the stock to go to the next point in the tool path, because it had to do that to get a A-B linear path. It would have been faster if it stayed in the pocket and made a L shape, but of course it is not that smart. On some models, both ways will speed things up TREMENDOUSLY… more time cutting, less time traveling. It really just depends on the situation.
Another quick way to speed up a toolpath is simply upping the cut feedrate. Since my spindle is spinning at 30,000 rpm I can not feasibly max out the tool with a MPCNC, so I could pretty much just up the feedrate until I start to lose steps.. but I don’t want to do that, so I just went up 1/3, leaving me at 1600mm/min. This gave me a cut time of 1:42… (test #3)
For test number 4 I went up to 2,000mm/min, leaving everything else the same. The machine sounded fine and motion was smooth, but I could see a difference in the results. However, adaptive is *supposed* to be used as a roughing strategy, so that should not matter. Just finish it off with a 2d contour/horizontal profile (actually, I pretty much always run with 0 axial stock and only finish the vertical walls. Just don’t have issues with floor finish.) This, surprisingly, only sped up the path by 1 second, leaving me at 1:41 (test #4)
For test 5 I decided to increase my stepdown. So, I went back down to my default feedrate of 1200mm/m and then upped my stepdown from 3mm to 5mm. Got quite a good increase in speed here, with a time of 1:12 (test #5)
The last test I did was the same as #5 but with a higher feedrate (1600mm/m). Time was 1:03 (test #6)
The keen eyed may have noticed there was a test before #1… this was actually the same gcode as #1 but I, again, had my feedrate override set on 300%. Cut fine. I took a video of that and I might post it tomorrow while I’m at work (my internet at home is awful, it took me 10 minutes just to upload those 6 pictures to imgur)
I think the machine still has a lot left in it, I could up my rapids (this I will almost surely do) and up my feedrates and increase my DOC some. At the cost of cut quality. I would rather spend more time on the machine and less time hand finishing so I will play around to find my happy medium.March 23, 2019 at 2:16 pm #94252
Got a bit curious on the accuracy, cut out a 1″ square…
It’s as good as my calipers (which I built the entire racecar on the back of, and which are pretty accurate/precise according to my gauge blocks) which is good enough for my needs by far.
1mm tabs in f360 are also great, very easy to remove with a couple swipes of sandpaper but enough to hold the mdf in place during final contour cutting.March 23, 2019 at 9:11 pm #94277
DAMMMM!!! 0.005″ Uhhhh, breath on it for a second to make it swell just enough to zero that bad boy out!March 24, 2019 at 12:55 am #94293
Cutting out the arcade cabinet I designed, shown in my profile pic/avatar. Coming together just how it should, everything just comes right off the machine dimensionally accurate.
Wasn’t planning to make it out of MDF but I have soooo much scrap from building the table for the MPCNC. Table dimensions were just too big to fit 2 on 1 4×8 sheet and my top is 2 layers thick + 1 layer removable spoilboard… so I have like 1.5-2 entire sheets worth of scrap.
Can’t go any further with it for now because I don’t have the buttons/joystick (didn’t actually have them in my CAD file either, hope the ergos work out somewhat decently, if not I know someone with a CNC machine who can whip up some new parts haha) but I might go to Microcenter tomorrow to remedy that situation. My screen/screen mask will be done on my laser engraver at work most likely as well.
Really pleased with how this thing is working out. Also really pleased with the amount of abuse and neglect my $1.50 endmills are taking without complaint. I am still on my first one and it’s showing no signs of degradation.March 24, 2019 at 6:52 am #94299
I was at microcenter a couple weeks ago and was surprised with the amount of build your own arcade stuff they had. That’s fairly new.
As for the mdf, they make a penetrating epoxy for wood that will stabilize it.March 24, 2019 at 1:37 pm #94344
I was at microcenter a couple weeks ago and was surprised with the amount of build your own arcade stuff they had. That’s fairly new.
As for the mdf, they make a penetrating epoxy for wood that will stabilize it.
Yep, I like microcenter.
I’ve heard of all sorts of stuff to seal mdf… from wood glue diluted with water to pretty much any type of sealant. Not sure what I’ll do yet… definitely open to suggestions, I’m no woodworker.
You must be logged in to reply to this topic.