DIY-ish CNC Mini Mill

Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum

Help Support Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Joined
Jul 24, 2006
Messages
14,260
Reaction score
786
Location
Southwest
Ok, I admit I haven't been brewing much beer...but I thought you guys might like to see what's been "brewing" in my garage. It isn't exactly beer related for the moment, but I can think of a few brewing related machining projects for it already.

After much research on hobby machinist forums and thoroughly reading Daniel Kemp's website, I decided that buying a relatively expensive tool from Harbor Freight was a good idea. I think I was right!

After more than doubling the initial cash outlay plus a few weekends of tinkering, I have a working CNC mill in my garage. It's freaking awesome! Here are the results.





There are a few more videos under my username if you care to surf to YouTube. I'll be happy to share more details or make another video showing a particular detail or process if there's interest.
 
Last edited by a moderator:
"Off topic Delete", that's what happened to me with a few past postings by one moderator.
JK.
Now seriously that project is great thanks for sharing.
I looked into a CNC for the Bridgeport, quit once I saw the bill, manual with power feed will have to do.
I see cold winter indoor projects are being created.
 
Hey, I plan on machining a belt driven mash stirring motor thingie. It's beer related!

Thanks, man. I've had this one going for a few months, but I didn't want to post anything until I was sure it would work!
 
Very Impressive Yuri! Now that your have the time, you can go back to work on your plutonium upgrade on your brew rig. Don't forget to post picture.
 
please make me this:

3290796442_9f53d5b47b_o.jpg


that is all

:p
 
Don't accuse me of posting off topic! :D



This took about 5 minutes to set up and only two takes. The first attempt resulted in lifting the bottle off the table.

No beer was harmed during filming.
 
Last edited by a moderator:
Wow! And I thought you were just another run-of-the-mill moderators. You actually have skills.:)
 
Yuri, I have read most all of your projects on here and you never cease to impress.

I want to be Yuri when I grow up.
 
Sorry, I'm a purist. The cubes must be independent and free to float.

Still jealous. I've been eyeing a little 1-1/2 HP hobby knee mill but unfortunately it's a ways down on my list of toys.
 
Beautiful! strange enough Ive been on break from homebrewing for the same reason (actually due to limited space my cnc machine is blocking the entrance to the closet containing my brewing supplies). I will eventually post pics/videos of mine in a separate thread, but for now can you share the technical specs on your machine? I assume you used steppers rather than servos so full-step/half-step/micro-step, stepper motor torque and full steps per revolution, x/y/z travel limits, and more about the software. I see you're using mach3 for the controller but what about CAD and CAM processor I havent been able to find ones I like so im curious what you're using.
 
I really want to make one of those... if only I had the space and an extra 1200 bucks or so.
Are you on CNCzone.com? Lots of good stuff going on over there.

What kind of tolerances can you hold with that. I've always been tempted by those milling machines, and little lathes, but I'm a little suspicious of anything HF sells (not to say I don't buy it, I just have low expectations).

I don't think I've ever seen beer opening tooling before, well done! I'm glad you didn't try to cut it off with an end mill... though that would be cool, just run it around the rim of the bottle about 0.010 from the glass... but that's just asking for a mess.
 
+1 to more detail!

I'm looking at selling my Porper cue lathe to finance a machining lathe and an end mill. For some stupid reason a Proper lathe is worth $2k+ used, and I've been looking at the Grizzly mills for around $1k - A coworker has one and likes it.

I'm VERY interested in converting the mill immediately to CNC, but have done none of the research to see what that would cost yet.

I vote for a build thread! If not on HBT I'd love to have a link to a build thread somewhere else! :ban:
 
I do have an account at CNC Zone, but only so I can download pictures and attachments. I've never posted. There is a ton of info there.

Here are some more details:

HF Mini Mill #44991 - www.harborfreight.com
Stirling Belt Drive - www.littlemachineshop.com
Fully upgraded 3-axis ball screw kit - www.cncfusion.com
Home/limit microswitches - www.surpluscenter.com
Mini ITX computer - homebuilt with parts from www.newegg.com
Mach3 software - www.machsupport.com
Ridgid laminate trimmer - Home Depot

www.kelinginc.net supplied the electronics:
48V power supply
381 oz in, 200 steps/rev stepper motors (one for each axis)
Geckodrive G540

I love the G540 for its convenience and simplicity, but its microstepping output is fixed at 10. I don't need that kind of resolution. I wish I could 1/2 or 1/4 step instead and take advantage of increased speed and/or torque.

I get about 9" of X travel, 4" of Y travel, and about 8-10" of Z travel, depending on the collet/chuck/tool. Rapids are set to 100 ipm for the X/Y and 50 ipm for the Z. I've actually used up to 200 ipm on all axes without missed steps, but I think I set myself up for failure by increasing the rapid speeds that much, particularly since I keep the gibs pretty tight.

My biggest complaint is the standard X2 owner's blight - it's not sturdy enough. If I take the time to tram the mill, adjust the gibs, measure the backlash, and re-adjust the Mach3 backlash compensation before each job, I can achieve accuracy to within 0.001". However, that is impractical, and I think I realistically see about twice that (if not more on occasion). The machine quickly falls out of adjustment due to a lack of rigidity.

When the gibs are adjusted, backlash is between 0.001" and 0.003", with the Y axis typically being "loosest" (its gib is the smallest and only has two set screws vs four for the X and Z axes). I'm likely going to beef up the column to help the Z axis to stay in alignment. The "engineer" that decided to fit a tilting column to such a small mill was an idiot.

If I had it to do all over again, I'd have gone with a BF20 or its clone (the G0704 that SweetSounds mentioned). I just didn't want to sink an additional $500+ into a project that had a high potential for failure/problems. The G0704 gets great reviews and is supposedly much higher quality than the Sieg stuff. It is obviously more rigid, since the column is fixed and fairly large for the size of the machine. Now that I'm a bit more confident with CNC electronics, I'd spring for a bigger and better machine in a heartbeat if space and funds allowed. Someday...
 
I use CamBam almost exclusively for CAD/CAM. It's decent for 2.5D CAM, and I basically force it to be a CAD program for most of the parts I've designed and machined. It will read DXF/STL, and it's supposedly capable of 3D, though I haven't done any profiling with it.

I've used Google Sketchup and an STL export plug-in for some 3D playing, but it outputs everything as straight polylines instead of arcs, lines, and circles. The resulting G code usually sucks - it's either too blocky or has WAY too many lines, depending on the arc segment settings.

I've played with FreeMill a little bit. It looks cool, and I managed a simple 3D profile with it, but I've yet to do any serious 3D machining.

Stickfont is free and works great for single line font engraving, especially when the letters are less than 1/4" tall. With some tweaking, the text/engrave functions in CamBam can look nicer, but it only does "looped" TrueType fonts. I've used Deskengrave (also free), but I've come to hate it.

Mach3's LazyCAM is ok, but I don't think it's supported anymore. CamBam is similar, supported, and more fully featured.

I've used Blender in the past for 3D modeling, but I think it would have similar issues as SketchUp when it comes to exporting DXF/STL files.

I've yet to find affordable CAD software, and the open source stuff has all failed to impress me. In fact, I've yet to even bother downloading any open source CAD software because it's all been so underwhelming.
 
BricsCAD is pretty much identical to the AutoCAD of 10 years ago. I think the install file is pretty small, and since they are out of business (can't buy it anymore), I wouldn't feel bad about sending you a copy. It's actually very nice for a bare bones CAD program.
 
Thanks, klyph. I'll check it out. I used AutoCAD 10 years ago at school. I'd love to use it again, but it's horribly overpriced.

I know its not 3d but since you mentioned AutoCAD have you considered using DraftSight? Its Dassault Systems (read SolidWorks) free alternative to AutoCAD. I have used it some for work, although we use AutoCAD natively as well as Gerber Cutworks I cant afford to use it all the time with patterning, but its essentially a clone of AutoCAD. 99% of the commands transfer over too.

http://www.3ds.com/products/draftsight/draftsight-overview/

If you are looking for a 3D modeler please ignore this reply. I have also been a bit underwhelmed by the free 3D CAD programs. I bought Alibre a little over a year ago and the older version which I have is manageable, not as nice as Solidworks but I get the impression that the two new versions that have been released since I bought mine are a bit more polished.
 
Thanks for the point out. I've been playing with Solid Edge 2D lately. It seems pretty user friendly and will import/export most common formats.

I was looking at Alibre and ViaCAD as the cheapest options for 3D. Rhino and Dolphin are slightly too pricy. Inventor and SolidWorks are totally out of the question. 3D isn't a requirement for the moment, but I would like to utilize modeling software in the future.
 
Yeah Alibre is pretty reasonable if you are OK with getting the personal edition. However, if you want the cool tools i.e. FEA, rendering, toolbox, sheet metal and things of that nature you are looking at a significantly larger investment. I haven't played with Alibre much lately as we use SolidWorks at work and I have a home license of it through them. They aren't as free giving out AutoCAD licenses though, so for any basic stuff I will use DraftSight at home. For the essential/critical and large files, I stick with AutoCAD simply because we aren't 100% sure of the way that DS will interact with our cutting software - I have no desire to loose 1 weeks worth of patterning because DraftSight and AutoCAD store 2 different things slightly differently (or 2 different things the exact same but after dxfout in AutoCAD, since the file isn't saved as an AutoDESK approved file, it throws a hissy fit.)

I also played with Rhino3D for a while on a trial basis, but didn't get too far because I am used to the parametric modeling of SolidWorks and Alibre and didn't put in the necessary time to get used to a NURBS based modeler in the 1 month I had to play with it.

On that note, I believe there is a free trial for Alibre if you are interested in trying it out as an inexpensive modeler.
 
I'm ok without some of the cool toys. What I really want is 3D design software that will output a DXF with arcs instead of splined polylines that approximate curves. If it doesn't do fancy rendering with all sorts of lighting options and surface effects, I'll survive.
 
I just went through and did a quick test - for circles and Arcs in parts Alibre will export the drawings 1:1 into DXFs maintaining circles and arcs. As you get into more complex geometries I cant guarantee that this relationship will be maintained. Let me know if you want me to shoot you a sample dxf or if you have something you want me to model and dxf out.
 
Awesome, thank you. I don't have anything specific in mind at the moment, but it's good to know that at least one bargain priced bit of software doesn't "cheat" with curved sections.
 
Now all you need is a waterjet.

I made a cube paperweight once, has the typical pocket on all faces and then you go in with a dovetail and machine out the ID at the bottom of the pocket and the remaining material falls.

Fun little object of conversation.
 
The router attachment you are using could be used to make homebrew pub signs :)
 
Ah hell, I've been looking at mini mills for a while, now to see you can CNC it, I had better pretend I never saw this thread or my want will outpace my need.
 
OK, this is pretty cool. Was it difficult to build?

I have a pile of steppers pulled from some old laser printers waiting for a project like this.
 
It wasn't very difficult to build at all. This is one problem that definitely has money as a solution.

I did some of the work myself, but I used a few kit components that made the build fairly easy. For another $1,000, I probably could've sourced all the parts from one supplier, ensuring that the project would be easier to build than some Lego kits. For another $3,000, I could've had a complete machine.

On the other hand, I could've spent half what I did, but I'd have pulled my hair out doing the engineering by myself and manually machining parts on a substandard mill.

I highly recommend using parts from CNC Fusion to get a workable foundation upon which to build.

For something smaller but far more homebuilt, look up RepRap. The work they've done translates well to CNC applications beyond 3D printing.

Also check out www.buildyourcnc.com for some ideas on homebuilt CNC routers.
 
Back
Top