You Know You Want a Revolution

[Mooselake]

Got a BobsCNC Revolution 4th axis router, with a Makita trim router spindle.   Great fun!

Picked the package deal with Vectric VCarve Desktop, as recommended by Bob, for it's rotary axis wrapping capability.  This machine has X, Z, and A axes, no Y.  Kinda sorta somewhat like a lathe with a router instead of turning tools, although this lathe will turn a square.  Still learning how to use it, in conjunction with turning one end of a 24x48' barn into a woodshop.  As usual, despite now being completely retired, there's never enough time for everything.  Also got a Sainsmart LE5040 and rotary axis this summer, plus Lightburn, same time issue.  Spending more time working on the shop than in the shop, still, this has been a slow 5 year project that started with building a wall followed by breaking up the old concrete floor (500 drywall buckets of rocks and busted up concrete) only to find another floor a foot deeper.  None of the surviving previous owners (120 year old farm) knew it was there or why the floor had been raised.  Good enough to build an insulated treated wood floor over it, and got enough height for a 2hp cyclone dust collector which still lacks ductwork.  As I said, never ending...

Kirk

[MisterAcoustic]

Hi Kirk,

That's an interesting thing - would you call it a louter or a rathe? . Router, lathe, what's the difference?

I was surprised when I first saw the machine, because it simply didn't match my expectations - I thought it would be a normal router. But it's not quite . It seems like it will be great for a very useful class of things. Do you have a specific project in mind for it?

I haven't been making much of anything - had to rustle up a new job, so spending got put on hold for a bit. If I were going to buy a normal router, right now this one is at the top of my list:

https://sienci.com/product/longmill/

Once I get settled in to the new gig, I might pull the trigger on that.... or, I also got a great deal on some second-hand extruded aluminum pieces, so I could try to throw together my own frame and see what happens. Building a good CNC would be a challenge... maybe I could just do a laser cutter/engraver instead.

I also have a house to remodel... as you said, it never ends, apparently.

Enjoy your new toy - and of course, post any results in the Community Project Showcase section .

Stan

[Mooselake]

It never ends, I've been here for 43 years and found that if you think you're done you've missed that you need to start over again.

I'm not sure what to call it other than 4th axis CNC.    Lathe doesn't really fit (I have too many, both wood and metal, lathes), while a lathe can't turn a square (you can come somewhat close with eccentric chucks or creative mounting) this can't hollow out a bowl.  I think of it as more complementary.  I work mostly with home-harvested wood (have around 70A of woods) that only looks round, so I've been using a wood lathe to prepare stock for the Revo, way faster than 3mm deep cuts with a 1/4" bit even at 100ipm.  Making it straight afterwards is pretty easy, I've been hand typing gcode, but I'm the kind of guy that thinks typing G1 X300 A40000 F2000 is pretty cool (or distance/stepover times 360).  The A axis is effectively infinite, although that can bite you when you do a G1 A0 instead of resetting zero. 

My original motivation for the Revo, and for woodturning, was ornamental turning and rose engines.  Regrettably I still haven't gotten there, but I'm closer.  I found rose engine simulation software on github, it was written for a homebuilt lathe/cnc lashup and will need modification but it's a good starting point.  The biggest problem I see is that even CNC ornamental turners use a thing called a cutting frame instead of a spindle.  It's a belt driven disk with cutting tools stuck in the sides so instead of a single point spindle it cuts a disk shaped path.  Definitely possible to simulate, but are my memories of engineering math classes a long time ago (I went to engineering school with a slide rule) good enough to figure it out?  Until then I've been dinking around on walking sticks (bad knee...) and for grins made a wooden spoon.  Not particularly efficient to start with 3" round stock and turning almost all of it to chips (vectric seems to only be able to start with round stock on a 4th axis) but it was an interesting project.

The Revo is too big to take on our winter journey south so I'm thinking of building one of these while we're down there.   Long enough to do a walking stick without flipping it end for end.  I already have most of the parts so maybe it'll actually get done.  This site has a lot of information on making them.  Turns out that 2" diameter maple, removed a lot of brush this summer and stacked it to dry, isn't particularly straight or round, so I may need to rethink my planned source of material and have a maple tree or two turned into 2x2s

Kirk

[Mooselake]

I know (in the forum membership sense) a couple people with Longmills, they seem to be happy with them. I'm not impressed with the machine design, with that hardware store aluminum angle. If I had the space I'd get a ShapeOKO 4, or build an MPCNC.

[Administrator]

I was pretty sure it wasn't any kind of lathe - I presume that the rotary axis doesn't spin either fast or freely. I just wanted to get the lame word play out of my system .

The target of your link in the phrase "building one of these while we're down there" appears to be messed up. The target appears to contain some of the text of your post - possibly a failed paste? Nevertheless, from the other link, I think you are talking about a 'CNC Rotary', or something to that effect.

Regarding the longmill, there's a fellow on youtube who seems knowledgeable to me (since I barely know anything about CNC machines), and he posted a video of himself climbing up on to the gantry of the longmill, and riding it for a little while. Of course, there was no measurement of any flex, but it didn't obviously flex either, so it was a pretty good advertisement for the device . It doesn't hurt that the size and the price are right for my situation. I've seen the Shapeoko - if I recall it was higher in price... I periodically check in on X-Carve too, but they are more expensive now too. I'll check out the other one as well. Oh wait, I already have.

I saw an MPCNC someone built at a maker faire. It certainly seems like a viable option, in spite of the fact that my filament printer is trapped behind an exercise machine, and I haven't used it for a while . I actually have a bunch of parts myself... I just need to schedule some time to do things.

I need fewer hobbies... .

Stan

Oops, posted as Administrator - was looking into the email thing. Oh well

[Mooselake]

No idea how that happened, I fixed the demon cnc link so maybe it'll work now. Been tied up with farm and shop work, last minute rush before everything freezes and we head south, or I would have seen this sooner.

Took half of yesterday to dig the tractor mounted rototiller out of where it had settled in, and about 10 minutes to actually till the garden. I'd tried earlier in the summer with the 40 year old Troy-Built but it got away from me in the heavy sod, the garden didn't get planted for a couple years while I was recovering from knee surgery. The sod was no match for 50 hp and a 5' wide tiller. Before that I brush hogged about half an acre of thick small maples, up to maybe 2" diameter. They take over fast when you're distracted. The brush hog (actually a 5' IMCO bought from a neighbor a year after getting the Troy-built horse) handled them just fine, biggest problem was being careful so one of them didn't get me or damage the tractor as it went over them. I backed into a lot of them instead. Leaves are almost gone here, a few trees with yellow leaves but just about all the maples have shed theirs despite an unusually warm fall (well, 47 today, over 70 yesterday), first frost is way overdue.

Without seeing the actual message headers it's hard to tell for sure, but it's quite likely those suspicious emails had spoofed from addresses and didn't originate here.

[Mooselake]

After almost a year I've replaced the Revo's controller (a CNC Shield V3 on an Uno clone with a slightly modified grbl to change Y to A) with one based on FluidNC. FluidNC is the new name of grbl_esp32, which runs on a 32b ESP32, they changed the name after seriously modifying the configuration system to use a text file instead of recompiling. I'm using a Wemos D1 R32 board, an ESP32 in an Uno form factor, with the same kind of CNC shield (wanted to keep the original controller intact), and it's configured for 4 axes, X, Y, Z, and A. About $20 for the board, shield, and DRV8825s, can't beat that. Can't test Y, but the rest work fine. I had X homing issues that turned out to be a combination of a bad switch and it's being located so it was just barely compressed. A change of switch and a screw to adjust the trip location has that fixed.

FluidNC has a built-in web server with a pronterface style jog interface, a grbl command line, streaming log (I keep position status messages turned on for now, they also include home switch status), and access to almost all configuration variables, and you can upload/download to/from the ESP32's onboard file system. It can communicate via USB, bluetooth, and wifi via telnet. SD card support, but my board doesn't have one (couple bucks and some soldering to add, but haven't bothered). It's just enough different that most gcode senders have minor (to major) issues talking to it, mostly because they try to read the firmware variables that are pretty much gone. The latest UGS nightly works, while it's at the bleeding edge I haven't had any problems.

Besides going round and round the A axis is pretty interesting. First infinite axis machine I've ever had, I do blank rounding with commands like G1 X200 A12000 Z35 F5000, the A value calculated to give the right stepover. One command to shave off the entire length of an up to 24" blank, although you really don't want to do a G0 X0 A0 unless you want to see it spin for a while, best to rezero A... VCarve Desktop will do decent 3D carving, but you have to keep in mind that it's mostly 2D wrapped around a round object, and it's not aware that A goes past 360 so it does unnecessary direction changes. The only exception seems to be when doing designs that exceed it's workspace boundaries, for example I've been doing straight diagonal lines that can wrap flutes around the cylinder. Rather than faking it and manually figuring out where it goes off one edge onto the other it'll keep going for you. I was hoping it would raster 3D round and round, but it stops at 360 and changes direction. Oh, well.

FluidNC, other than the homing problem (it does a hard stop requiring a reset and losing workspace offsets if the switch doesn't release when backing off, a fix is in the works) has been solid, and I'll be changing out the controller on my 3018 and converting the ZB Mini from grbl_ESP32 when I get around to it. That machine uses a plug-in DevKitC module so I can pull the existing module and replace it, and still have a viable backup. They're about 10 bucks on Mouser and I have a few on hand. The CNC shield requires cutting and removing it's only through-hole resistor. That pulls the stepper enable lines to +5, and the ESP32 lets the smoke out when that happens, and trashed my first board (it lost wifi but still talks across the USB, might try throwing micro python it). Pretty amazing what you can get for 10 bucks these days... I haven't really pushed FluidNC very hard so I'm not taking advantage of it's much higher step rate and command processing ability. It has a much deeper command look-ahead than 8b grbl, it might be running smoother but that's subjective. It should make a big difference on jobs that use teeny tiny lines for arcs, that's on my list to try when I'm done having fun with the machine and playing with VCarve.

Kirk