Got TurboCNC to work with the FET-3 controller from Stepperworld.

[General]IniDescription=TurboCNC 4.0 DistributionNumberOfAxes=3Start Inhibit=YESBlock Inhibit=YESConfiguredAsImperial=YESImperial Places=0Metric_Places=0Default_Feed=10.000000000000Home_Feed=10.000000000000ArcFactor=1.000000000000StopOnIllegalCodes=YESMonochrome=NOMouse Hide=YESLoad_Tooling=NOVerbose=YESSync Unit Indices=NOSync Axis Indices=NOListDir=c:\maxnc\belt\ListExt=.CNCToolDir=C:\TOOL\ToolExt=.TLPoint Filename=.\POINTS.DATReverseDelay(ms)=0TurretIndexTime(ms)=2000ColletOpenTime(ms)=2000ColletCloseTime(ms)=2000SpindleTime(ms)=7000RelaySettleTime(ms)=100ClampDelay=15000.000000000000MachineType=NoTypeUsePentiumTimer=NOClearMDIPrompt=NOHomeIsLimit=NOM6JogUpdateDistances=YES

[Editor]First_Block=Last_Block=M30Add_First_Block=NOAdd_Last_Block=NO

[Dialect]LinearRapids=YESG04Letter=PG04InMS=NOArcIJKMode=1G8xReleaseLetter=RG8xDwellLetter=PG8xDwellInMs=YESG8xPeckLetter=QG33IsLead=YESJump_Target=ORewindOnM30=NO

[AXIS1]Designator=XStepIncrement=0.000625000000IsLinear=TrueIsStep/Dir=TruePortAddress=$378StepPin=3IsActiveLow=TruePulsewidth=150DirPin=2LowIsPositive=TrueAcceleration=2000.000000000000StartSpeed=150.000000000000MaxSpeed=350.000000000000Fast_Jog=350.000000000000Slow_Jog=50.000000000000InvertJogKeys=NO

[AXIS2]Designator=YStepIncrement=0.000781250000IsLinear=TrueIsStep/Dir=TruePortAddress=$378StepPin=5IsActiveLow=TruePulsewidth=200DirPin=4LowIsPositive=TrueAcceleration=1500.000000000000StartSpeed=150.000000000000MaxSpeed=350.000000000000Fast_Jog=350.000000000000Slow_Jog=50.000000000000InvertJogKeys=NO

[AXIS3]Designator=ZStepIncrement=0.000952380952IsLinear=TrueIsStep/Dir=TruePortAddress=$378StepPin=7IsActiveLow=TruePulsewidth=150DirPin=6LowIsPositive=FalseAcceleration=1750.000000000000StartSpeed=500.000000000000MaxSpeed=700.000000000000Fast_Jog=500.000000000000Slow_Jog=50.000000000000InvertJogKeys=NO

[DRIVE_ENABLE_1_(OUTPUT)]Enabled=TruePin=14PortAddress=$378ActiveHigh=NO

[DRIVE_ENABLE_2_(OUTPUT)]Enabled=TruePin=16PortAddress=$378ActiveHigh=NO

[DRIVE_ENABLE_3_(OUTPUT)]Enabled=TruePin=17PortAddress=$378ActiveHigh=NO

[SPINDLE_INDEX_(INPUT)]Enabled=TruePin=12PortAddress=$378ActiveHigh=YES

[Imperial Jog]Axis 1 Index=1Axis 2 Index=1Axis 3 Index=1Jog Element 1=0.001000000000Jog Element 2=0.001000000000Jog Element 3=0.001000000000Jog Element 4=0.001000000000Jog Element 5=0.001000000000Jog Element 6=0.001000000000Jog Element 7=0.001000000000Jog Element 8=0.001000000000Jog Element 9=0.001000000000Jog Element 10=0.001000000000

[Metric Jog]Axis 1 Index=1Axis 2 Index=1Axis 3 Index=1Jog Element 1=0.010000000000Jog Element 2=0.010000000000Jog Element 3=0.010000000000Jog Element 4=0.010000000000Jog Element 5=0.010000000000Jog Element 6=0.010000000000Jog Element 7=0.010000000000Jog Element 8=0.010000000000Jog Element 9=0.010000000000Jog Element 10=0.010000000000

[Persistent]Halt_On_M01=NOExecute_Slash=YESShow_Backlash=YESJog Mode=YESWorkInImperialSystem=YES

[POSITION]1=0.3750000000002=-50.6656250000473=0.000000000000

[LASTBACKLASH]1=12=-13=0

[Speed_Map]

First off, CONGRATULATIONS STEELERS!!

Before the game, I took another crack at the “raster-like” CNC photo-engraving this weekend.  In my first attempt described in a previous post I got less than satisfactory results by using about a 30-degree angle V-bit in the CNC (engraving on right side of photo above).  There just wasn’t enough change in width of the V-bit as it advanced deeper into the wood to show the carving relief needed to make the image visible.

For this second try, I used a 90-degree angle V-bit (made by Proxxon) because I couldn’t find a 60-degree bit from my usual sources.  The 90-degree bit worked MUCH better and the image is very legible in the wood (engraving on left side of photo above).  I think some of the detail of the image was lost because a 90-degree bit is too wide for this application, but it was a big improvement.  It looks to me like 60-degrees would be perfect, but I can’t find anyone yet that makes a 1/8″ shaft bit with a 60-degree angle.  This is going to take some Internet research.

This weekend I made my first attempt at a "raster-like" photo-engraving into wood using my home-built CNC mill.  My previous photo-engravings with the CNC used the "vector" method of using software to define the "edges" of color zones in the image and only engraving these edges into the wood, sort of like a black and white line drawing.  For the "raster-like" method, every inch of the image gets "cut" into the wood by the CNC’s cutter using lawn-mower-like back-and-forth passes.  The darker the color of each photo pixel, the deeper the cut’s plunge.  Therefore, if you are using a "V" style bit in the CNC, the deeper the cut also translates into a slightly "wider" cut as more of the "V" enters the wood with depth.  This method results in a more photo-like result with much more detail than the vector method, but it is also harder to do and slower.

For this first attempt, I used a 9"x14" piece of mahogany ply mounted to the CNC’s table, and a carbide Dremel 1/4" V-bit.  I generated the g-code of the image of the "Big Ben" clock tower, using the trial version of the Vectric PhotoEngrave software that I download from their website.   The software was intuitive, and produced a g-code file that easily imported into TurboCNC on my CNC’s PC with no errors. 

The CNC ran for almost 3 hours, and engraved the photo very well, but most of the detail of the image was not visible because I apparently used a "V"-bit that did not have enough width at the top…i.e., the changes in depth did not cause enough changes in width of cut to be easily visible.

Therefore, I will buy a bit with either a 60-degree or a 90-degree point and try again….

So, my first attempt at raster-engraving met with mixed results…time to try again.

I finished the RoboAngus EAS (Etch-a-Sketch) side project this weekend.  I got the idea from Jeff Epler and Chris Radek’s website and figured that it would be easy to adapt the controller and software from my Wood Turtle project to make my own version of the EAS CNC…it was…it only took me two hours for the complete build.   It runs using TurboCNC software and two stepper motors.

My innovations to this project were:

• the removable rubber cups that attach the motors to the knobs on the EAS using only friction.  Use of these rubber cups for the attachment allows you to remove the Etch-A-Sketch easily without tools…just pull.  I just drilled a 1/4″ hole in the middle of a standard rubber chair leg end bought from Home Depot.  I then pushed a 1/4″ by 2.5″ bolt through the hole and used a nut to tighten it on.  Then I used a 1″ length of automotive vacuum line to attach them both to the stepper-motor axis.  Works great and pulls off in a second.

• the use of a stuffed monkey as the operator (wink)

The above photo shows my son’s name “Mason” drawn by the RoboAngus EAS. You can see the device is really only stepper motors mounted in plywood and attached with the bolts and rubber cups.

The photo below shows the rubber chair end mounting device described above.

Now that my first CNC machine is completed and working, over the holiday weekend I continued to try and refine the design of my other experimental machine, the "CNC Lathe Attachment" based on John Klienbauer’s Wood Turtle design.  That machine is now also complete, as John designed it, and it works successfully as a CNC-lathe duplicating machine, but that was not my purpose in building it.  I wanted to make a machine that could be used to cut creative spiral patterns in wood spindles…that was the original goal of my whole Rube Goldberg venture into CNC!

In the meantime, I’ve learned a lot about CNC, and I really expect that my "7th Sojourn" milling machine is the one that I will use the most.  That’s good, because I’ve hit a snag on the "Wood Turtle" spiraling machine design (I’m calling this modification of John K’s design the C-Turtle).

Over the weekend, I completed a more solid mount for the C-Turtle to the lathe.  It is now rock-solid and has no wiggle at all in its mounting to my Nova lathe.  This will allow it to work better even in its original duplicating-machine functions…but I’m shooting for spirals.

Next, I designed and built a mount for a stepper-motor which attaches to the "outrigger" on my Nova lathe.  This outrigger in its folded position allows me to line-up the stepper-motor with the outboard hand-wheel on the lathe headstock.   I then mounted essentially a jamb-chuck into the cup of the hand-wheel.  I installed an 1/4" axis on the jamb chuck and linked it to the stepper-motor with a 1/4" coupler.  Now I had a stepper-motor that could turn the lathe’s main axis under control of the computer.  So, basically, I could carefully coordinate motion of the C-Turtle’s cutter head with the very slow rotation of the lathe’s axis (much slower than the lathe’s motor could turn).

Everything works great.  The problem is that my 60-oz stepper-motor can just barely turn the lathe’s axis and if any pressure at all is applied to the wood, the motion stops…the stepper-motor is just not strong enough to work like this.

So, I’ve proved two things:

1.  My design for the C-Turtle works in concept, but
2.  I will need a strong stepper-motor to make it work.   The stepper-motor on this axis does not have the mechanical advantage that the 1/4" threaded rod drive screw provides on the other axis.

Since my FET-3 board will require a bigger power-supply and a resistor to drive a bigger-stepper, a cascade of upgrades occurs in order to take this experiment to the next level. 

I may put this experiment aside for a while before I invest more money and time into it.  I am happy to  have completed one of the CNC machines and may move to another project for a while before returning to CNC design school.   I am now confident that this contraption would work if I spend some money on parts…but that may wait for a while.

Over the weekend I successfully used the “7th” to engrave a photo of my daughter onto a piece of wood.  Turned out very well.  This was just a test cut, so I just used some scrap hardwood plywood.  I’ll do it again on real wood later.

I used a carbide v-bit as the cutter and attached the plywood to the y-axis table with double-sided tape (after cutting through a clamp with the cutter on an earlier attempt!).

This engraving used the raster-to-vector method.   Essentially converting certain highlights and edges on the photo to vectors using various freeware and shareware software routines.   I took a digital photo of my daughter and loaded it into WinTopo (the freeware addition).  In Wintopo I converted the raster photo to vector and then eliminated a lot of the “noise” vectors that were created to get a fairly clean drawing.   I then saved the vectors out of WinTopo as a DXF file.   I then used Acer (freeware DXF to g-code converter) to convert the DXF file to g-code.  I set the maximum z-axis cut to 0.1 inches in depth. The g-code was then loaded into the my CNC machine’s PC and I used TurboCNC (shareware CNC controller software) to run the machine.

The CNC machine took about 1 hour and 45 minutes to cut the picture (engraving is about 13″ x 9″).  I ran the machine slowly about about 6 ipm for this test run.  Everything worked beautifully.

I rubbed a cherry stain over the finished piece to darken the engraving marks.

For my next photo-engraving experiment I want to try the depth-of-cut shading type (I don’t know what to call this method) engraving.  This method leaves the photo as raster and engraves deeper for darker pixels and more shallow for lighter ones.  I am using shareware called IMGTOGCODE for this attempt.  These engravings take much longer for the machine to carve, so I will do I much smaller image.

Tried a cheap diamond burr bit in the "7th" last night and quickly broke it when the machine moved laterally in a cut.  These bits apparently do not have enough lateral cutting ability for CNC.  Back to regular carbide or HSS bits tonight.

Last night I disassembled the x-axis and z-axis on the "7th" and tightened all of the connections and reassembled the machine.  Now that I am running the machine I realize just how tight the machine needs to be for accuracy.  I also readjusted the y-axis by setting a bit in the spindle at a fixed Z-axis depth (maximum) and then moving the spindle horizontally to each corner of the y-axis.  At each corner I did a fine adjustment of the rails until the bit just touches the table.  It is now adjusted perfectly.

Now, I consider the 7th to be "complete."  I’m sure it will continue to evolve from here, but it is time to use it for awhile before I do any more work to it.

Here it is!  Above is a photo of the finished machine (click here for more photos). Last night,  I mounted it on an old kitchen cabinet that I put on small wheels.  The PC that drives the machine resides inside of the cabinet. Makes a nice portable setup.

I powered the machine up and configured the software.  Everything runs great under “jog” control from the PC.

I began to test a g-code file and got mixed results.  I am clearly making some kind of error in my g-code files, so I need to do a little more research on that issue before the machine real cranks up.

Lots of progress on the “7th” this weekend.  It is now essentially finished except for the wiring of the final home-switch.  The table is fully installed, 2 of the 3 home-switches are installed, all of the wiring is complete.    The machine is mostly calibrated.  I am basically happy with the calibration, although the table may need a final bit of adjustment once the machine is run.  Very little wiggle in the spindle axis…very good.

Testing begins next!  I generated a test g-code file, and am now ready to connect the machine to the PC and crank-it-up….hopefully in a couple of days.

I’m looking forward to getting back to woodworking.  My “experiments” with CNC have taken me away from woodworking for much longer than I intended.

Last night I calibrated the rails, installed the power strip for the spindle, ran some of the wiring,  and drilled and counter-sunk the moving table.  Tonight I hope to make the aluminum "skates" and install them onto the table.

Last night I installed the rest of the stepper-motors onto the CNC mill project.  I also installed the power-supply, on-off switch and the housing for the controller-card.  Next step is to align the rails and install the spindle and on-off switch for the spindle.

I didn’t get a chance to work on the "turtle" CNC machine over the weekend.  Hopefully tomorrow night.

Happy Halloween!

Well, my Wood Turtle tasted wood for the first time over the weekend!

It’s still in an "experimental" mode as I tweak it out. All I have done
so far was to "jog" the cutter head in-and-out and back and forth
along a square stick of wood until the wood turtle turned it into a
round stick of wood. Simple test, but it showed me where I need to work
on the machine. I have not had it run a g-code program yet. Maybe
tonight.

I am basically pleased with this first trial. The Turtle cuts much more
smoothly than I expected. It took the beating of the intermittent
contact with a square of wood (until it is made round by the machine)
without any problem. Strong design. I thought that I might have to
manually round stock first before engaging the turtle, but it can
handle it.

Areas that need improvement to my machine:

1. My NOVA wood lathe is a floor-standing unit unlike the one that John
K tested with, so I needed to come up with a good method for mounting the
Turtle next to the lathe. For now, I built a temporary table out of
saw-horses and wood planks with bricks on top to give it stability and
reduce movement. I then clamped the Turtle to the "table" with wood
clamps….it still moves WAY to much. I need to design a more stable
mount. This is the #1 issue that I need to resolve. I think that I need
to attach it to the lathe somehow (with clamps) rather than to a
separate table. I’m thinking on this.

2. I built the Turtle as a 36" long machine. I’m not sure that it can
handle the full 36". After about 24" I get a lot of flex in the lead
screw. My stepper also struggles past 24". My "carriage" might be too
tight on the rails. Need to play with this.

3. It ripped my carbide cutter out of the key stock. Stripped the tapped
hole that I had screwed it in to. I need to put a screw entirely
through the key stock and put a nut underneath for more strength. Seems
to cut just fine without the carbide too.

My other CNC machine project (the "7th Sojourn" milling machine) also made progress. The unit is now fully painted and has the adjustable rail-blocks installed.  I also installed the Y axis and Z axis mechanisms and mounted the stepper motor on the Z axis.   Next step is to adjust the rails and then build the rolling "table."  My goal is to have the "7th" ready for testing in two weeks.