Our LoboCNC machine info

[ted.m...@gmail.com]

Nova Labs received a Lobocnc kit a couple months ago and we've been eager to get it running. At the end of March we made our first cuts and we're installing the machine in our metal shop in April. I can't praise Jeff enough for his attention to detail and superior design of this machine!

Our equipment page for the lobocnc will be updated periodically - maybe there's some info other users will like. http://nova-labs.org/wiki/equipment/lobocnc

Ted

[Jeff]

Great to hear you've got everything working, and also thanks for the kind words.  Keep us posted as you use the machine more and start training new people to use it.

-Jeff

[ted.m...@gmail.com]

Actually today I had a problem with a G02 arc command. This was generated from Phlatboyz script (Sketchup plugin) and can be imported in two other CAM programs (Mach3 and CNC USB Controller). Would someone try it on PSCNC and see if they get the same 'Invalid arc data' error? Thanks,

Ted

----------------------------------
G17 G2 X0.7563 Y1.2126 Z-0.1000 R0.5804
X0.7724 Y1.2058 Z-0.1000 R0.5804
X0.7883 Y1.1987 Z-0.1000 R0.5804
X0.8041 Y1.1913 Z-0.1000 R0.5804
X0.8198 Y1.1836 Z-0.1000 R0.5804
X0.8352 Y1.1755 Z-0.1000 R0.5804
X0.8505 Y1.1671 Z-0.1000 R0.5804
X0.8656 Y1.1584 Z-0.1000 R0.5804
X0.8805 Y1.1494 Z-0.1000 R0.5804
X0.8953 Y1.1401 Z-0.1000 R0.5804
X0.9098 Y1.1304 Z-0.1000 R0.5804
X0.9241 Y1.1205 Z-0.1000 R0.5804
X0.9382 Y1.1102 Z-0.1000 R0.5804
X0.9521 Y1.0997 Z-0.1000 R0.5804
X0.9658 Y1.0889 Z-0.1000 R0.5804
X0.9792 Y1.0778 Z-0.1000 R0.5804
X0.9924 Y1.0664 Z-0.1000 R0.5804
X1.0054 Y1.0547 Z-0.1000 R0.5804
X1.0181 Y1.0428 Z-0.1000 R0.5804
X1.0306 Y1.0306 Z-0.1000 R0.5804
X1.0428 Y1.0181 Z-0.1000 R0.5804
X1.0547 Y1.0054 Z-0.1000 R0.5804
X1.0664 Y0.9924 Z-0.1000 R0.5804
X1.0778 Y0.9792 Z-0.1000 R0.5804
X1.0889 Y0.9658 Z-0.1000 R0.5804
X1.0997 Y0.9521 Z-0.1000 R0.5804
X1.1102 Y0.9382 Z-0.1000 R0.5804
X1.1205 Y0.9241 Z-0.1000 R0.5804
X1.1304 Y0.9098 Z-0.1000 R0.5804
X1.1401 Y0.8953 Z-0.1000 R0.5804
X1.1494 Y0.8805 Z-0.1000 R0.5804
X1.1584 Y0.8656 Z-0.1000 R0.5804
X1.1671 Y0.8505 Z-0.1000 R0.5804
X1.1755 Y0.8352 Z-0.1000 R0.5804
X1.1836 Y0.8198 Z-0.1000 R0.5804
X1.1913 Y0.8041 Z-0.1000 R0.5804
X1.1987 Y0.7883 Z-0.1000 R0.5804
X1.2058 Y0.7724 Z-0.1000 R0.5804
X1.2126 Y0.7563 Z-0.1000 R0.5804
X1.2151 Y0.7500 Z-0.1000 R0.5804

[Jeff]

Hi Ted,

The PSCNC G-code interpreter is pretty minimal - it only supports the I,J,K format for specifying arcs.  (The I,J and K parameters are coordinates of the arc center relative to the starting point.)  I believe most post processors will allow you to select I,J,K mode for specifying arcs.  I suppose I should add the R format (something I've put off for years.)  More gory details of the PSCNC program can be found in the PDF included with lobocnc.com/lobopscnc.zip.  

-Jeff

[Ted Markson]

Gotcha. Ill get the arcs in I J K with another processor. I just RTFM'd and found the code list! 

Having the R format in PSCNC would be helpful since Phlatboyz seems to only export in that. At least I don't see an option for I J K mode in their Ruby script or parameters. 

thanks for the quick reply Jeff!

[Jeff]

In thinking about implementing the R mode of specifying arcs, you have to be careful when the two end-points are close together.  Slight variations in one of the end points can cause the location of the calculated center point vary considerably.  This doesn't matter so much for short arcs, but if you are trying to specify an arc of, say 350 degrees, a thousandth off in the location of one of the end-points can make a big difference in the path cut.  (You'd have to draw a picture to convince yourself.)  The pathological case, of course is in trying to do a 360 deg. circle - if the end point and the start point are the same,  specifying just the radius doesn't tell you where the center is at all.  Therefore, any complete circle must be specified as at least 2 arcs of less than 360.  (I use CamBam, and it seems to to want to break complete circles up into three 120 deg arc sections to avoid any ambiguity.)

The I,J,K specification is much better behaved mathematically.  Also, it turns out specifying a complete circle is easy.  In PSCNC, if your cutter is at the perimeter of the circle on the X-axis, the statement:

G02 I-0.5

assumes the end-point is identical to the start point, and executes a complete circle.

-Jeff

[Ted Markson]

Thanks for the info Jeff!

I just downloaded Cambam and got the I J K format from it. And I've found a new favorite cam processor! It's seriously nice software. I'll give it more trial time before purchasing but it's on my short list. 

My CAD file process procedure looks like this:

 1) Export from Sketchup as DXF file (using Sketchup DXF plugin. http://www.guitar-list.com/download-software/convert-sketchup-skp-files-dxf-or-stl)

 2) Open the DXF in Cambam. Add toolpaths. Export Gcode

 3) Run on Lobocnc! 

Pretty simple and it lets me work in Sketchup, which I find to be easier than any other 3D modeler I've used. And it's free!- ted

[Ted Markson]

Some updates on our machine.

-------------

We found a decent option get 1/8" shank tooling on the Makita compact router. We use a "Rotozip chuck adapter" with a 1/4" shank and 1/8" inner bore plus set screw. See my post on G+ for more. https://plus.google.com/u/0/109842191190000189028/posts/61YLv2tffRc

Just today another guy at Novalabs found the optional 1/8" collet for our Makita router on ebay - of all places. I couldn't find it at Makita or anywhere else. It's $16 while the chuck adapter is $10. Jeff - I think you have the Makita RT0700C? Did you get the 1/8" collet? I'm curious where if you did. 

-------------

Jeff, do you plan to add support for G81 in PSCNC?  Cambam outputs that code for canned cycle drilling. Do you know of a way to change what gcode is used for canned cycle? Maybe I'll read up on the 'Custom script' option.

Thanks, 

Ted

[Jeff]

Regarding 1/8" adapters, I suspect the Rotozip one is going to have horrible runout.  I tried a similar sort of adapter and the runout was 0.010 - 0.015"!  A better bet is the sort of split sleeve type - a simple cylinder with a 1/4" OD and 1/8" ID.  This one is a nice one.  This split sleeve type should have more like a 0.001 - 0.002" runout, but that's on top of the 1/4" collet's runout.  I've actually made my own collets on my lathe, which have about a 0.001" runout.  I'll let you know if I'm going to be making more. (They're a little bit of a pain to make.)

Regarding canned drill cycles, I should figure out how to implement that in PSCNC.  In the meantime, I have been using the custom script with CamBam.  I usually just use something like:

G00 X$x Y$y

G01 Z-0.25 F5

G00 Z0.125

for a custom script.  I'm sure this could be done more elegantly, but it gets the job done.  You can also extend the script to add pecking.

Hope this helps.

[Ted Markson]

Jeff - Thanks for the script. It gets the job done but let us know if you get to add drilling support to PSCNC. Is there an issue / milestone tracker for PSCNC?  Any reason not to put PSCNC on Github? 

I added the Z variable to the drilling script and another feed speed to the lead out. 

G00 X$x Y$y
G01 Z$z F5
G00 Z0.125 F10

I don't see more than 2mil runout on the adapter we got. Then again I am using a digital indicator from harbor freight..... I choke the adapter all the way in the 1/4" collet and run at low RPMs. Seems to work

If I do get a split tube adapter, it'll be the Stewmac part. They're the only guys with multi-slit adapters in stock. Single slit adapters are useless in routers