Men's Shed Router

[electrosteam]

I understand the comments on using proper development tools, but I am trying to convince the shed members that a simple approach can achieve results, with the hope that a spark will be generated that will encourage each member to investigate the development tools.

I think that eventually we will have to provide some formal instruction using the tools.

Attached is the hand code that will be trialled on Wednesday.
'Surfacing 1' is the generic Gcode program, and 'Surfacing 2' is an example for a platen 600 x 800 with a 20 mm cutter.
The blocks of code need to be copied/pasted to make up sufficient passes to cover the length.
The 250 mm/min feed rate will be increased if possible.

At 250 mm/min the time per pass is 600/250 = 2.4 minutes.
The number of passes is 800/5 = 160.
Total time is 160 x 2.4 = 384 minutes = 6.5 hours.

Not very attractive for a shed that meets once per week for 4 hours !
We will see,

Surfacing 1.txt Surfacing 2.txt

John

[Greolt]

I understand the comments on using proper development tools, but I am trying to convince the shed members that a simple approach can achieve results,

I understand your desire to introduce shed members to gcode and what can be done via manual coding.

However experience has shown me that what will get members REALLY excited about the router, is seeing how easily they can design on the computer, and carve a wooden plaque with their grandchild’s name in a fancy font.

Such a project could be coded in minutes with a CAM program, but would be an arduous and mind numbingly boring task to do manually.

I can't think of anything that would more quickly convince members that "this is too hard for me"

[ian]

I understand your desire to introduce shed members to gcode and what can be done via manual coding.

However experience has shown me that what will get members REALLY excited about the router, is seeing how easily they can design on the computer, and carve a wooden plaque with their grandchild’s name in a fancy font.

Such a project could be coded in minutes with a CAM program, but would be an arduous and mind numbingly boring task to do manually.

I can't think of anything that would more quickly convince members that "this is too hard for me"

[ian]

I understand the comments on using proper development tools, but I am trying to convince the shed members that a simple approach can achieve results, with the hope that a spark will be generated that will encourage each member to investigate the development tools.

more likely that the "development spark" will be generated by a realisation that the simple to use tools are very limited in what they can do.

among my past learning experiences is learning how to use a program called SolidWorks with the intent that the skills developed could be used, via an appropriate code compiler, to program a CNC machine to cut 3-D parts for things like toy cars.

I suspect that knowing that it's possible to program the CNC to carve a toy car from solid wood is more likely to trigger the "spark" you mention than the alternative you've proposed.

[ian]

Attached is the hand code that will be trialled on Wednesday.
'Surfacing 1' is the generic Gcode program, and 'Surfacing 2' is an example for a platen 600 x 800 with a 20 mm cutter.
The blocks of code need to be copied/pasted to make up sufficient passes to cover the length.
The 250 mm/min feed rate will be increased if possible.

At 250 mm/min the time per pass is 600/250 = 2.4 minutes.
The number of passes is 800/5 = 160.
Total time is 160 x 2.4 = 384 minutes = 6.5 hours.

Not very attractive for a shed that meets once per week for 4 hours !
We will see,

Surfacing 1.txt Surfacing 2.txt

John

Hi John

some comments if I may ?

if the pattern is MDF, you may want to limit the depth of cut to 0.1 mm so you don't excavate all of the hard surface layer of the MDF.

I think your note
; Surfaces platen flat relative to machine working surface
should refer to the surface produced by the code being relative to the Zero value of the machine's cutter head, rather than the bed of the machine.

if you are aiming for a 25% overlay cut to cut -- which by the way seems way too much -- I think your number of passes calculation should be 800 (long) / 15 (per pass) = 54
giving a machining time of 54 x 2.4 minutes = 129.6 minutes, plus the time required to re-position the cutter between passes, say 6 seconds, for a total of around 2-1/4 hours. Easily doable within the a 4 hour window.

[ian]

I am keen to push hand coding because I know quite complex jobs can be done this way, and most members of a Men's Shed are not going to ever achieve the ability to use CAD to draw, and CAM to generate code, with subsequent manual review.

hi John

I seriously question the assumption that most members of a Men's shed are not going to ever achieve the ability to use CAD to draw.

my father was well into his 70s when he taught himself to use CorelDraw. At a basic user level programs like Corel are just computerised sketch pads that accept typed text. It's very easy to import a vector drawing of a teddy bear with a bunch of balloons, use the key board to add the new grandchild's name, then punch out a vector file that a basic CAM program can turn into machine code that will run through the CNC.
Grandad (or grandma) gets a routered outline which they can colour to produce personalised door labels or whatever they desire.

A tip I saw on a YouTube yesterday -- cover a piece of melamine with painters tape. CNC your plaque or name plate, paint the exposed surface, then remove the painters tape for a clean engraved sign on a clean background.

[foobillious]

If your machine can take it you could also move up to a 25mm bit, which could speed the surfacing process up.

[RustyArc]

Fiddling about with Gcode is good for getting a feel for what drives CNC, but if you want to get excited and get on and make things without too much of a learning curve, something like Vetric Cut2D would be ideal. You'll spend maybe half an hour doing a couple of tutorials and then you're good to go on most routing operations.

When I got my 6040 machine, I didn't bother with surfacing, I just clamped down some MDF and started going through the tutorials. Trying them with different cutters, feeds and speeds and then different materials taught me a heap in not a lot of time.

[electrosteam]

Guys, thanks for all the comments, especially from Ian.

Ian, I understand you are suggesting we change from a 1 mm cut with 5 mm overlay to a 0.1 mm cut with 15 mm overlay.
This is certainly attractive from a time point of view.
We can easily repeat later, if necessary.

Please all, be assured that AutoCAD, LibreCAD and CamBam are currently being exercised to provide some real jobs to run, as soon as we have enough flat working surface.

John.

[electrosteam]

We ran today with several things accomplished.

One member arrived with a program to engrave his name and we went through it correcting a couple of minor errors, and we spent some time discussing how semi-circles can be used to make letters like "O".

Another local guy arrived interested in the router, and it appears he has a metal rated mill converted to wood router at home and had lots of good advice to us all.

It took some time to cut our 25 mm MDF platen to 750 x 750 mm and to screw it down.
There is plenty of Z-axis clearance to the hold-down fixings.
The platen was tested with a DTI (dial test indicator) as about 0.3 mm out of level to the spindle carrier.
We mounted a 20 mm diameter 2-flute carbide surface routing bit for some test cuts.
The spindle motor is 400 W 24-52 V running at 48 V, supposedly at 12,000 RPM.
We soon discovered the spindle tram was poor and it took some time to correct the problem with two thicknesses of paper.

We then spent some time running the surfacing program with a feed rate of 1000 mm/min, cuts about 0.2 mm and step-over 15 mm.
No distress from any part of the machine.

The surfacing program contained one error which took some time to discover - a "," in one line.
Our error tracking and file handling skills are now much improved.

An additional deficiency discovered in the Universal G-code Sender program we are using is the lack of a 'single-line-step' button.
Is such a feature desirable ?

Along the way we seemed to get an additional 3 errors that led us into limiting the Gcode to one instruction per line and no comments.
I am sure this is not necessary, and perhaps a consequential effect of the errant ",".

At 1000 mm/min total surfacing time should be under one hour - perfect.
The plan for next week is to set up and test the surfacing program first thing in the morning, then let it run during the morning tea break.

John.

[elanjacobs]

For such a small cut you should be running much faster, assuming all the motors can handle it. At that speed you'll be generating lots of heat which will wear the bit faster than it should. At 0.2mm, you should easily be able to cut at 10,000 mm/min assuming your axis motors will move that fast.

[foobillious]

I thought UGS had a single line step function - maybe try double clicking on it.

You may also want to check out chilipeppr. There is an offline version available, but I think you need to be internet connected the first time you run it or set it up or something (sorry, don't remember). You can check that out here: ChiliPeppr - Hardware Fiddle as well as simulator your gcode files etc there without being machine connected.

While it can be run online I recommend using it offline.

[electrosteam]

We will certainly give a higher speed a test, I think our machine has a max speed parameter set at 5000 mm/min.

UGS has got a Macro facility.
There are 5 lines of commands that can be selected one at a time, as often as required.
The Help description states that each line can have multiple commands separated by a ";".

Not sure what the size limit is per line, and what the significance of the ";" may be.
I had assumed the whole command line would execute, if selected.
Perhaps it steps along, pausing at each ";".
Something else to check.

John

[epineh]

I am pretty sure the trial version of V-carve pro still lets you output machinable g-code so long as you use their sample vectors. I did the flower one with the corner flourish one for a jewellery box lid for the missus ages ago. Might help with some inspiration for funding...

Russell.

[electrosteam]

Russell,
The flower is inspirational, and I am sure that one day we could get there.

For now, we have progressed a little, once the electrician had restored power, and we recovered from a Fire Alarm.

We got the short name cut into some thin board once a couple of aspects were straightened out, in our heads and on the machine.

The surfacing was progressed, but not finalized - we suffered yet again bad joints on stepper motor wiring.
Along the way we resorted to unplugging the USB cable between the laptop and Arduino to force a program stop.
Must investigate what commands we can issue from UGS (Universal Gcode Sender) to the Arduino to cancel any program in memory.
We have already decided to add a reset button for this task.

Tramming is a problem, the gantry obvious twists and the spindle carrier deflects in proportion with the cut.
Surfacing is going to need several passes to get the platen flat.

The UGS macro tests were put back to later.
It still seems, however, that the Grbl/UGS combination does not like multiple G codes on the same line.

Of interest was the situation when we had a problem, there was a joint discussion. some suggestions made, and a re-start decided on.
Just as the start button was operated, the new inexperienced members clustered in to get a good look, but the two members with machines and experience stepped back !
There is a lesson there.

Although we will be using Cambam in the shed, I have started experimenting at home with the dxf2gcode and Camotics programs to see if they can provide a simple 2D router CAM with simulation.
They both show promise, with some issues with dxf2gcode.

John