Is your router frame a parallelogram? Mine was!

I’m on day 2 of CNC setup and I just discovered something that I do not recall seeing on this forum so I thought I’d share.

I mounted my machine per the instructions in the manual and YouTube video and it seemed to go well. Everything was jogging smoothly and no binding was evident. I expect most other people have the same experience (good job on design Onefinity!).

I was in process of making my spoilboard (t-nuts) and got ahead of myself. I mounted the board to the table and surfaced it, trammed the router (my tramming jig) and surfaced it again 90 deg to the first pass. It came out lovely…and then I remembered the counterbores for the t-nuts need to be made on the other side :man_facepalming:

I figured it would be best to fixture the spoilboard in some way before I flipped it so I zeroed the X axis at 408mm (centerline) and offset +/-395mm to bore some fixturing holes. I did this at the full Y extents front and back. So now I’ve got 4 holes, one at each corner of my spoilboard. If everything is square and true I should flip the spoilboard over and all holes will line up…as you may have guessed from the post title, it didn’t. The holes at the front of the machine and the holes at the back lined up independently but not together. When I pinned the front holes the back were shifted almost exactly 0.100".

I luckily had some wiggle room in the way I mounted the rails and was able to slide the back of the machine over by 0.050" and now I have a square machine!

I hope this may help someone else! I can share pictures if it helps explain things.

7 Likes

I was wondering the same thing when I set it up and was unable to come up with a highly reliable measuring method. I would love to hear exactly how you measured it. I know the diagonal measurements but for what point to what point?

Hey Dave, thanks for asking. I am not going to imply that my method is simple or reliable, but it seems to be fairly accurate (at least by the way I am measuring). I mocked up some pictures to hopefully help explain.
It may also be worth clarifying that when I drilled the fixturing holes mentioned in the earlier post, I drilled into my tabletop. That way I could drop a 1/4" bit into the thru-hole of the spoilboard and fixture off of the blind hole in my tabletop.

The image below is an overhead view of the machine with spoilboard in MDF brown/orange. The green dots at the corners represent the fixture holes that I drilled in the 4 corners. At this point I was working under the assumption that everything was square and therefore the green holes are presumed to be in alignment. The CNC was in the same X/Y position for each pair of matching holes, so why shouldn’t it be?

When I flipped the spoilboard over I saw the below condition. When fixturing the front two holes (now in blue) the back two were shifted to the left (almost exactly 0.100").

My spoilboard had not moved and the CNC was in the same Y location for each pair, so the only logical conclusion was that my back two feet needed to be shifted by half of the measured hole offset. The below picture shows the condition as the holes were drilled to dramatic effect. You can see why the back holes were shifted left when the spoilboard was flipped over and hopefully you can understand why the back feet only needed to be moved half of the measured discrepancy.

I set up an indicator touching the side of one of the back feet and loosened the mounting hardware. I used 1/4-20 bolts to mount the feet so there was a little wiggle room to knock the back over the required 0.050" to straighten things out.

2 Likes

I just took this picture of the actual way I measured the offset. The hole is the fixture hole drilled with the machine. The traced circle is me putting a same sized endmill into the hole of the spoilboard after I flipped it and giving it a twirl.

2 Likes

Yes, this is a good subject to bring to light. I’m guessing that there are many people unwittingly running their machines slightly out-of-square. While the assembly method is self-paralleling front to back, it’s not necessarily self-squaring. This condition only manifests itself under certain circumstances, like when you flipped over the wasteboard. In spite of the machine being very easy & straightforward to assemble, it still takes a good bit of knowledge & understanding to really understand the dynamics of how it can still end up out-of-square. I used to try explaining this on the x-carve group. WOW. You want to talk about a machine that can be out-of-whack six ways to Sunday. The more you understand the x-carve on an accuracy level, the more you understand how many ways it can be built with compounding errors.

4 Likes

Good points Bill. I have more than a passing familiarity with industrial machining centers and decent understanding of the importance of starting and staying square and I was still caught out by my sub-par measuring when I first mounted the machine.

1 Like

Nick
Thank you so much for sharing your experiences and taking the time to illustrate the problem. I am just setting up my machine now and will backtrack a little to try and square up the unit x and y first.

I would definitely like to see how you square up the x and y axis.

1 Like

@Techrise do you need more pictures or better explanation? I’ll admit my description is a bit confusing

1 Like

Hello, I would think you would really see any discrepancies in your machine when you do a two-sided job. Maybe doing a simple toolpath one flipping your job from front to back and one flipping your wood from right to left. Then checking your cuts for accuracy.
Marty

1 Like

Thanks Nick. I am more of a visual learner. Just trying to see how you fixed this problem when I set this machine up. Thank you.

This method would be fairly down & dirty, but reasonably accurate for most applications. I don’t have the woodworker, so I don’t know the span between the front & back mounting blocks. If less than 36", you could use an aluminum yardstick (depicted in red). If over 36", you may have to buy a 4’ piece of aluminum sold by most decent big-box home improvement stores. This, along with a large framing square (depicted in blue) will give you a reasonably accurate square edge to work with. Ideally, you would use a test indicator to run along the square and adjust the x-axis as required. Optionally, you could simply use a cutter shank to check at each end of the x-axis travel.

image

5 Likes

I had the exact same problem. I followed the video instructions word for word, but mine wasn’t square. I measured the diagonal dimensions between the mounting feet inside corners - my dimensions were 5/8" off!!! I kept adjusting and checking the dimensions until the diagonal dimensions matched up and called it ‘good-nuff’.

1 Like

Same here. Between measuring the distance between the Y rails, the diagonal distance between the Y rail corners and feeling the X to Y rail connection for flushness (dig this), I think I got it as square as I can without special tools. Should certainly be close enough for my purposes.

In theory the mounting blocks and how the X axis contacts them should square the machine. Does that mean that the blocks are not being accurately squared from the factory? The surfaces are not that wide so the “lever arm” magnifies any inaccuracy, therefore holding those tolerances would not be an easy task … but isn’t that one of the main benefits of this CNC design that we should assume is delivered?

I wonder if the recently released mounting frames will make this problem worse. They look extremely fiddly and unlikely to be easily squared.

Mine was out only 0.050" over nearly 4ft. There is easily that much slop in the mounting feet + mounting hardware. I believe the machine itself is fairly robust, but with the wide range of ways that people are mounting their machine to the table could easily introduce the error.

1 Like

I believe it has more to do with the weight of the machine parts - simply tightening down the mounting bolts is not enough to twist the frame into perfect alignment (unless I was overly cautious and wasn’t putting enough torque on the bolts, which is entirely possible - I didn’t want to risk stripping out the aluminum block threads). But I agree with your first sentence and that is/was the intention of the design - to be self aligning.

1 Like

The process I used… push the two y-axis ends up against a qualified reference ledge on my machine table top, push both y-axis carriages up against their end stops, install the x-axis, transfer mounting holes and drill thru table top, install mounting fasteners… easy-peasy.

2 Likes

I am still in this situation and can’t seem to get things square. I noticed it with two sided jobs as well as my tiled cuts not lining up correctly. Not sure what else to try… is it possible that my waste board is not square and is what is causing me issues??

I must have lucked out. I read this thread before mine arrived, so I was aware.

I squared my 1F to the front of the table as instructed and checked the diagonals with a laser distance measure (hard to do with a tape measure)