Is your router frame a parallelogram? Mine was!

[jfleser (Jim Fleser)]

That’s a good article for those of us working with wood especially. I spent quite a while on Tram in an attempt to get a pocket made with a 1/2" flat end mill that would leave very little sanding (because you really can’t get into those). But for area that are going be sanded, I need to be a little less picky sometimes and get rid of those jobs that take over an hour.

[TheyCallMeJohn (John (Huntington Builds))]

Additionally 2 sided pieces and inlays probably wouldn’t work well.

[ConvenientWoodwork (Nick)]

Excellent point John and incidentally that’s exactly how I discovered I had a parallelogram

[Techrise (Noble Design)]

I saw this method done and like the simplicity of it. Just wandering what the community thinks about it? Any potential flaws to consider?

Squaring up a Lead CNC machine

[Aiph5u (Aiph5u (not affiliated with Onefinity))]

Hey Techrise,

this method was already discussed above. I think there exists even simpler procedures. At the moment, when asked, I link to this method:

@Aiph5u

to consider a machine ready to start work, I would make sure, especially with the Onefinity CNC machine, that it is

1. rectangular (“squared”) and that it is
2. coplanar (not twisted).

For checking the rectangularity, there were different methods suggested, but as a first check for rectangularity (“squaring”), I would take a bar gauge:

@Aiph5u

and stick its ends diagonally on the machine right where the chrome tubes go into the black aluminum block, first as a diagonal between left front and right rear, then right front to left rear. Since in a rectangle the two diagonals are identical, these two distances should be the same so the machine is square. If you can move the bar gauge to a greater length in one of the two positions than in the other, then the machine is not a rectangle but a parallelogram, that is, the angles are not right angles. This can lead not only to workpieces where right angles are not right, but also, in extreme cases, to blocking of the Y-sliders.

To adjust this, you need to loosen a little bit the screws of at least two adjacent machine feet (say, e.g. the left and the right front feet), as well as loosen the screws that attach the X-axis to the Y-sliders a bit, so that you can move the feet so that the machine becomes right-angled, i.e. both diagonals can be measured as equal with the bar gauge.

For a more accurate test for squareness, you should do a test with the machine milling a test rectangle. Create a rectangle as large as the workarea as g-code and run it on the machine using a v-bit (with minimal cutting depth), and then use a bar gauge or a ruler to check whether or not the diagonals are equal. If you take a ruler, its accuracy or its scale does not matter at all, just make a mark on it when you check one diagonal of the milled rectangle, and check if the mark is identical when you check the other diagonal of the rectangle with it.

The so-to-speak “official” thread in this forum about rectangularity of the machine is “Is your router frame a parallelogram? Mine was! ”

For the second check, the check for coplanarity (that the machine is not twisted), you can use the […]

[CandL (Carl and Lynn)]

Reviving an old thread, before the foreman gets here. Along the lines of “Machinist” (Bill Blades] the return of the 3-4-5 triangle … but the big boy.

I have one of these: (it really is quite handy)

Folding Layout Asquare-3’x4’x5’ 90° 3Pcs. - Carpentry Squares - Amazon.com

And it claims an accuracy of:
" The tool is accurate to within 1/32” (0.03125") across 4’ which means that the angle tolerance is .0373 degrees to that same length. (by manufacturer)"

[WaywardWoodworker (Derek Lokee)]

I adopted this approach for squaring the machine with an old pair of calipers now dedicated for this verifying square from time to time. It removes any error induced by the alignment of the feet (twist) or Y rails that are different lengths etc.

Squaring a CNC Gantry Machine with Simple Tools and Math

[jfleser (Jim Fleser)]

Wait, I was told I wouldn’t need to use math any longer…

[TheyCallMeJohn (John (Huntington Builds))]

Moooooooore math. Always more math.

[DainBramage1]

A two-sided piece was when I figured out I was out of square

[DainBramage1]

While wrestling with how to get my parallelogram X-50 square (and the MANY possible challenges), I think I will attach the feet to aluminum (or possibly wood) plates that can somehow move for adjustment if necessary. I’ll have to figure out how to accomplish that, but I think it’s doable. This will have to attach to my torsion box with a 1/2" top plywood skin. Any thoughts?

[makerjace (Jace)]

Hi @DainBramage1
Not sure if you’ve seen this post of an all aluminum table, but you could use small pieces of this aluminum extrusion or entire rails of it for each Y-rail.
Cheers to square and coplanar machines!

[DainBramage1]

I did it out of what I had, which was wood! Works pretty well!

[OnefinityCNC (Onefinity Support)]

Another thing to consider is it could be a trapezoid or otherwise. See diagram. Left is correct, right is not and can cause Y rails to bind.

[stuperduck]

I’m using the older controller and there is no calibration for X and Y. But, if you can calibrate for linear distance then why not just add a calibration routine for squareness too!!! Getting these machines perfectly square is a pain in the BUTT!

[CalgaryD]

I am not sure it was discussed here. I have recently realized that there are two kinds of square we have to worry about.

1. The Y-axis rails have to be the sides of a rectangle with the same diagonals. This is one.

2. Both the moving blocks of the Y-axes must register exactly at the rail ends at the same time. If one of the block is touching the end of its rail and the other one is not, the stall homing will still find the home, but with one rail off. This will keep the X-axis under angle to the Y-axis when both Y-motors move it.

I have watched some videos on youtube how to square 1F machines and the guys did not talk about it. If you use a v-bit in the spindle to check your square and it is different from the one you get by physical measurements, this homing issue may be the reason.

[Aiph5u (Aiph5u (not affiliated with Onefinity))]

Hey CalgaryD,

• Machine has a gap on the back side after third setup attempt
• How to realign the Y rails (x35/x50 machines)

[CalgaryD]

These are really good links. And I have not seen them yet. Thank you.

[jfleser (Jim Fleser)]

I think that “Rhombus” would be the correct term although it is a subset of Trapezoid… Ahem…

[Aiph5u (Aiph5u (not affiliated with Onefinity))]

Hey Jim, hey onefinitycnc, hey all,

a rhombus is any flat four-sided object with four sides of equal length. It is a special case of a parallelogram and of a kite.
A parallelogram is a four-sided object with two pairs of parallel sides. Its opposite sides are of equal length.
A trapezoid has at least one side pair parallel.
A rectangle is a flat four-sided object with four right angles.
A square is the same as the latter but with all sides of equal length.