Benchy

[jvkmakerspace (JMS)]

A pen won’t tell you if you are level on all sides though.

1st have to be sure you are level, getting level was not easy on all corners. Took a few tries and foam board. Level and spacers said I was dead on but a cutting bit and foam board proved I wasn’t

2nd have to be sure it is not skewed

crown shows it is working but that is mostly all it does

Cutting squares, circles and stars will prove accuracy

then a V1 logo benchy to show it off and promote the machine. Cut in half and flipped is not a bad idea

we need several designs for this to simplify the setup.

[vicious1 (Ryan)]

@jvkmakerspace

Cutting squares, circles and stars will prove accuracy then a V1 logo benchy to show it off and promote the machine.

A logo part with careful dimensions can tell you a lot (most everything about accuracy and precision) in one part. Outer dimensions just for accuracy, diagonal measurements for skew/square, height and depth at multiple points can tell you about your Z and level. From there if you have dual endstops you can power down, run the same code again (without moving the part) and verify no material was removed and all dimensions are still within your expected/needed results. From what I can figure there is not other tests needed? After that it become material and job specific, example would be PCB’s and a trace width test.

@jvkmakerspace

crown shows it is working but that is mostly all it does

Shows if an axis is flipped, dimensions are close, build is semi flat, and step rate and speed is close. Using a plot as the first real test makes sure things are close enough to have fun. If you need more accuracy (most surprisingly do not) a cut out part is needed.

[vicious1 (Ryan)]

The crown is a fun plot and a great first test. Some people really do enjoy calibrations beyond what is necessary. I will add a more accurate plot with dimensions to the page here soon with dimensions to give a little more usefulness to the plotting, before a milling test.

 

I think we are close to a good test piece but I think more can be had in one test.

[tlumppy (Tim)]

Thoughts

do we need tabs on the cut out?

double sided tape ok for hold down? If the benchy is run early, people might not have table set up/drilled

a couple of facing operations on the corners?

would most of you do the cut out in one pass?

T

[kd2018 (Kyle)]

@jvkmakerspace

A pen won’t tell you if you are level on all sides though.

Could you not just use a sheet of paper as a feeler gauge to check all around for level in the same way you would for leveling a 3d printer bed? Either way I don’t want to stray from the theme of the thread, a milled part is more fun.

[jvkmakerspace (JMS)]

@vicious1

JMS wrote:

Cutting squares, circles and stars will prove accuracy then a V1 logo benchy to show it off and promote the machine.

A logo part with careful dimensions can tell you a lot (most everything about accuracy and precision) in one part. Outer dimensions just for accuracy, diagonal measurements for skew/square, height and depth at multiple points can tell you about your Z and level. From there if you have dual endstops you can power down, run the same code again (without moving the part) and verify no material was removed and all dimensions are still within your expected/needed results. From what I can figure there is not other tests needed? After that it become material and job specific, example would be PCB’s and a trace width test.

JMS wrote:

crown shows it is working but that is mostly all it does

Shows if an axis is flipped, dimensions are close, build is semi flat, and step rate and speed is close. Using a plot as the first real test makes sure things are close enough to have fun. If you need more accuracy (most surprisingly do not) a cut out part is needed.

Correct, that is what we want to achieve without a lot of hassle
The crown shows that it is working as it should, mostly. But not entirely.
It absolutely is the first test you should run.

But beyond that, we need to get into finer details.

No, it does not have to be within a millionth of a pico meter.

It can save time over all in the straightening of the gantry over all.
Some designs that can be duplicated across every machine built.
Saving individuals from having to figure out how to get setup for best accuracy possible.
Some feeds and speeds that can give people a starting point with a specific type of material (foam board).

The final result would be a wood milled V1 Logo that is close to perfect if not perfect in alignment.

I am on board with the idea of a benchy.

but you have to crawl before you can run.

@kd2018

JMS wrote:

A pen won’t tell you if you are level on all sides though.

Could you not just use a sheet of paper as a feeler gauge to check all around for level in the same way you would for leveling a 3d printer bed? Either way I don’t want to stray from the theme of the thread, a milled part is more fun.

The sheet of paper idea I did not think of cause I was in the middle of testing on a foam board that I already had down.

 

[jvkmakerspace (JMS)]

@vicious1

A logo part with careful dimensions can tell you a lot (most everything about accuracy and precision) in one part.

True, but it also can take longer and if it’s not correct, you have to start over.
This should be a final test.

[barry99705 (Barry)]

How about something that connects to a 3d printed part? Along this idea, but not necessarily this part.

[jvkmakerspace (JMS)]

Corrections made and new ZIP file uploaded. This one is 2 hours 25 minutes.

---

I like Barry’s idea also.

But here is what I came up with for a final test.

I have not run this yet, but may try tonight or this weekend

PLEASE keep in mind that I have very little idea if my settings for the Bit is correct or not.

You will find attached, the Estlcam g-code and project file in a zip file below.
I also included the DXF and SVG.

You use them at your OWN RISK

Please let me know if any changes need to be made. Thank you.

This Benchy should be the final test of your MPCNC with Dual End Stops
This should be very very repeatable, so you can make a set of 4+.

Parameters:

The part is 4in (101.6mm) x 4in (101.6mm)

Material Size: 5in (127mm) x 5in (127mm)
Thickness: 1/2in (12.7mm) minimum

place your material 2in (50.8mm) Left and 2in (50.8mm) Up

[attachment file=97391]

Using a 1/8in (0.125mm) 2 Flute Carbide Long bit

1/8″ 2 Flute Carbide Long

[attachment file=“8th inch 2 Flute Carbide Long Tool list.png”]

End Mill 1/8th inch
3.175mm
2.00mm
90*
15.00mm/s
2.00mm/s
24000rpm
blank
blank
blank
blank
5.0%
0.00%
50.0%
0.02mm
90*
0.00mm
0.00mm
0.00mm

In order of cuts

Inside ring:
Hole
Toolpath depth: 5.00mm
Start Level: 0.00mm
Machining Order: 1
Holding Tab Length: 0.00mm
Holding Tab Height: Full (this doesn’t matter cause it’s 0.00mm)
Finishing Tool:
Finishing Allowance: 0.00mm
Pocketing Strategy: Linear
Chamfering tool:
Chamfer width: 0.00mm
Pocket

Carve Logo Outline, V and 1:
Carve
Depth Limit: 3.00mm
Start Level: 5.00mm
Maximum carve width: 2.33mm
Carve Pocketing Tool: End Mill 1/8"
Machining Order: this will actually be 2, 3, 4, 5
Finishing Allowance: 0.00mm

Carve Logo Design:
Carve
Depth Limit: 3.00mm
Start Level: 5.00mm
Maximum carve width: 10mm
Carve Pocketing Tool: End Mill 1/8"
Machining Order: 6
Finishing Allowance: 0.00mm

Engraving Outside Ring:
Engraving at the Left Side of shape
Toolpath depth: 12.70mmmm
Start Level: 0.00mm
Machining Order: 7 or higher should be the last cut
Holding Tab Length: 5.00mm
Holding Tab Height: Full
Finishing Tool:
Finishing Allowance: 0.00mm

[tlumppy (Tim)]

No sure a unproven 20 hour G-code is in the spirit of this post?

[barry99705 (Barry)]

20 hours! 8-O

[kd2018 (Kyle)]

1/8" = 3.175mm

[jvkmakerspace (JMS)]

@tlumppy

No sure a unproven 20 hour G-code is in the spirit of this post?

No, not at all, I had the bit size set wrong. and I did not realize the milling time, never really looked at it.

Remember this is all preliminary. Adjustments and corrections are probably needed.

[jamiek (Jamie)]

When I load the svg in inkscape they appear concentric but in Estlcam they are way off. Am I doing something wrong or an incorrect setting?

[attachment file=97454]

[jvkmakerspace (JMS)]

I had this exact same issue when I started using these tools and it was freaking me out and driving me nuts.

Make sure that when you open into Estlcam you are selecting mm
However, I do not remember how I fixed it. It just stopped happening.

Make sure your Estlcam is configured properly https://www.v1engineering.com/estlcam-basics/

Also note that I am going to possibly be posting a new one soon. I think I might have the milling time down by a lot.
I think some of my bit settings are way off.

I need to fully test the milling of the gcode and design. But I am redoing it in Esltcam to try and figure out exactly what I am doing. I have a lot of testing left to do.

[jamiek (Jamie)]

I used mm, and nothing in the Estlcam settings looked wrong. The DXF file is fine, so I can continue with what I was trying to do.

[jamiek (Jamie)]

This will make a nice coaster.

[attachment file=97465]

I took a quick video of the making:

[billsey (Bill)]

So my suggestion is to cut the logo pretty much as suggested so far, but then move to the side and cut a rotated logo inlay. I think the V1 lettering and logo should be cut to different depths and there be full height tabs on the design. Assume piece thickness at 1/4"/6mm or a similar equivalent with readily available material. If the user chooses a wood or plywood material you’ll be able to see the inlay grain as different than the rest of the piece. Likely not so with MDF…

So, full depth of cut 6.35mm, lettering cut to 2mm, logo to 4mm, inlay down 4mm then full depth except for tabs with design rotated. Outer circle full depth except for tabs. All done with a 3.175mm single flute end mill.

Once the design is finished and ready, Ryan can do the CAM and generate the gold gcode. As with the crown any user should be able to fire up their machine, run Ryan’s gcode and get a usable piece. If the inlay doesn’t fit you’ve likely got an out-of-square issue, if the logo cuts fully through or the outside doesn’t go fully through you have a Z steps problem.

We might want a different design for a carving benchmark… Done with a specific V-bit.

[vicious1 (Ryan)]

That sounds like the perfect, advanced test. I would be very intimidated to cut that myself.

 

[vicious1 (Ryan)]

I purposely cut thing to either align any errors or cancel them, that would be a test of a dam near perfect machine, depending on tolerances. I have been keeping an eye on this thread. It is not an easy topic, but I do think some sort of basic test and advanced test is needed.

The LCD endcap is a super basic run but there needs to be some dimensions involved.