Machine alignment driving me crazy

I’ve had my Journeyman for a couple of months. And I have a depth of cut issue I can’t solve. I need the 1F Superfriends :grinning:

  • built a torsion box on top of Kreg frame. Frame was and is level in all directions. Torsion box wasn’t dead flat but within a few thousandths.

  • using a v-bit and the joystick I traced the max cutting area. I had an uneven depth of cut mostly in the X direction but some in Y too. No massive differences but noticeable to my eye.

  • I machined the Meyers wasteboard (the one with a few hundred tee nuts) and noticed the countersink holes were shallower on the left side than the right.

  • flipped that over and flattened it. Didn’t notice a huge difference in depth of cut but I did have to take off a few millimeters to where the bit contacted the entire surface. Tram looked good with no ridges I could feel.

  • I’m vcarving American flags and I see the stars on the left side of the union are much smaller than on the right. I figure that must mean there’s a low spot in the wasteboard or stock in that area. I reflatten the stock and two more stock pieces and get the exact same result.

  • before I go through with flattening or making a new wasteboard I check for Z depth consistency across the X axis by measuring from the table to a spot on the Z gantry and find it’s about 1mm higher on the left than the right. I also see a difference of about 1mm when zero Z on the left and again on the right.

  • I figure it’s my torsion box so I pull that off. Check level of the Kreg frame (no change). Then I put a new piece of 3/4” Baltic birch on top. I know not as stable as a torsion box but I’ve seen so many setups with just a sheet that I figured I’d try that before building a new torsion box.

  • I meticulously disassembled, cleaned, and reassembled the Journeyman. Checked for square. Checked for flush on the X-Y rail interface. Made sure the Y rails were placed exactly the same distance back the front factory edge of the plywood. Mounted the Z slider using 123 blocks below to rest it on to be as close the perpendicular as I could.

  • now I trace the max cutting area and again I have variation in depth of cut in the X and the Y and it’s much more noticeable. Looks OK on left side of front line, disappears in middle and looks deeper in right side. Similar pattern in Y. I went deeper until I could trace a visible line all the way and can see the differences in depth from the v-bit cut.

That’s where I’m at now and I’ve spent hours wrestling with the geometry of this and now I’m a couple hundred more in the hole for new plywood and MDF. I legit have no ideas left for how to figure out how and why this is off.

Before I go any further I would be so grateful for some help. I am just bewildered by this.

1 Like

Very odd. If you surface the stock at one z value then changing to a v-bit and making a constant z cut should give you a constant depth from the surface. If it doesn’t then what is causing the change? Seems like it could only be either the stock or the work surface flexing between the two cuts.

1 Like

Perplexing. I haven’t run into anything like this myself but I do find the assembly instructions from Onefinity to be a bit light on some details( taking it for granted that the machine will be square just by bolting it together). Three things occur to me, have you measured corner to corner for square? Have you strung fishing line from corner to corner and checked the overlap to check for co-planar? Is your Z slide on the bottom holes giving you maximum extension? If so you may be getting some flex when actually cutting that you aren’t getting when just lightly scoring with your V bit. After that you’re left to check whether the rails are perpendicular to the base and each other and then I run out of ideas. Best of luck and let us know how it works out.

2 Likes

A couple of thoughts:

I assume you flatten your new flag surface?

How are you clamping your piece down? If your flag piece is large then Sometimes squeezing from the sides cause the piece to flex up in the middle. And the flex might respond differently to the cutting of a flattening bit vs vbit.

2 Likes

Thanks for reminding me of the fishing line method. Checking rails perpendicular to table … great idea. Will try that today. The extra check for squareness and co-planar is worth it even to just check that off as “not it”.

A few more facts about my setup.

  • Z slider is mounted on the middle holes

  • I flatten my stock either with my planer and then doing a “hover pass” to double check or I flatten with a flattening bit.

  • I have been clamping my stock from the sides, but with hold down clamps that sit on top. Quite possible those exert enough inward pressure to cause bow, especially for a softer wood like pine. I saw this on a piece of 1" thick cherry I was carving too. Light on left. Proper depth on the right. I will include double-stick-tape method in my experiments once I am back up and running.

1 Like

Hey Terry,

the fishing line method is one of the methods for checking if your machine is twisted. With a twisted machine (which is usually caused by a twisted table or tabletop), your Z axis has not the same vertical angle when the X gantry is on the right than if it is on the left. But it may show up perpendicular to the part of the tabletop that is below it when you measure it with a square or with the surfacing bit ridges method, which can be confusing.

So beside checking for perpendicularity of Z axis to tabletop, often also called “tramming the router”, it is also necessary to check if your machine is twisted. People who have a table with casters and no torsion box use casters with height adjustment, and check for twist every time they moved their table and adjust the height of the casters. The same is done by using the Any Surface Leveling System in case you have the QCW Frame.

You can use the fishing line method to check the machine for twist by placing the lines diagonally over the tubes, from upper Y tube on one side to other upper Y tube on the other side (move X gantry fully to one side first), or you can use this method to check the coplanarity of the corners of your table without the machine. Remember that if the two lines touch each other at their crossing point in the center, you got to swap the two lines so the other is on top. There must not be a gap either, in the end they have to touch each other in both dispositions, to finally be sure that the four corners are in one plane (coplanarity achieved)

Another possible cause of one Y rail being higher than the other, or of a twisted machine, can be if one rail is twisted itself (video). Onefinity assemble their rails on a flat granite table, but sometimes, twist can happen during shipping.

Finally to come to and end with enumerating what generally can be checked, but what seems not to be an issue with your machine, is if the X rails are on top of the other (and this over their entire length). This method using vertical winding sticks can show you whether your two X tubes are aligned to each other.

1 Like

I think I found the root cause … I blew it with my torsion box assembly. Bigly blew it.

If I did the fishing line method right

  • I put nails on all 4 corners
  • stretched fishing line taught across the diagonals
  • set the height of fishing line at each nail at 1/2” using a known 1/2” piece of wood
  • have a sizable hump from left-back to right-front
  • have a huge gap where the lines cross

I’m gonna have a couple gin & tonics and regain my composure then figure out my next moves.



Ouch, that’s a big gap. Building my table I was so intimidated by other makers search for perfection, I had gone through every procedure I could find to check and double check each alignment, square, co-planarity,you name it. So far, so good. Didn’t you say that you had a similar problem with the 3/4" panel?

1 Like

It’s flat out awful. Get it? :slight_smile:

I’m going back to zero. Scrapping the table top. Disassembling the Kreg base. Ground up rebuild watching for all the gremlins.

Thanks all for the ideas and suggestions.

My advice … don’t let the excitement of getting your machine get in the way of building the foundation correctly.

3 Likes

If you are going back to ground up, consider grabbing the QCW for your machine. I think that could solve your issues, but comes at yet another expense.

Seriously considering that. With the cost of plywood it may not be that far fetched financially.

A solid foundation is always best practice. But if cost is a serious concern, I would argue that your original table top may not need to be scrapped. It’s more important that the top be stable (resistant to bending, warping, movement, etc.) than flat. If it is stable, all you really need to do is shim the feet of the CNC to get them all on the same flat plane. Then, flattening the spoil board will take care of the rest and no one will be the wiser as to your imperfect table top.

3 Likes

Yeah unless your top is moving with the weather (did you use dimensional lumber for the frame), that’s part of the point of the spoilboard…when you flatten it it is in plane with your machine. (If your machine isn’t twisted…which you can fix with shims).

I’m suspicious of tortion boxes built with dimensional lumber. I’d advocate for plywood only for that. I think movement with weather would likely make me insane. It that depends on what you are doing with it too.

If I remember correctly, @terry156 built his torsion box from 1/2" MDF.

Torsion box is all 3/4” MDF on the faces and 1/2” MDF for the ribs on an all steel Kreg frame. I get that stability is more important and this thing is a behemoth of stability. Problem is I think I didn’t build my reference surface flat and level enough to produce a flat top. Looking at the fishing line there’s a diagonal bow that’s close to 1/8” height difference center to corner. I must have built in slight twist and I just can’t trust that top anymore.

I want to build a dead flat torsion box for the challenge of it alone but the next one will be for an assembly and outfeed table for my table saw. I need to get up and running with the CNC soon to fulfill some orders. To that end I decided to stop chasing these gremlins and ordered the QCW with leveling feet. It’s a chunk of change but with my work schedule and hours it will take me as long to complete a new box as it will to get the QCW.

I’m gonna keep the 3/4” Baltic birch I cut for a top and put some leftover 3/4” MDF on top of that. Get it as flat and level as I can and put the QCW on that. I’ll bet I can get that flat enough to not need the QCW but this is going to buy me peace of mind and take a lot of environment and climate factors out of the equation.

Thank you all for the suggestions and jumping in. This is a truly great community and I owe you all a virtual drink.

Hey Terry,

that is always an important thing when assembling a torsion box.

The other day we were discussing how to ensure that the table top you glue/screw together will not be twisted, and what I suggested was:

You may still use a torsion box that is twisted by mounting the Onefinity on it, making sure the feet are all in one plane e.g. by using shims (I would use metal sheets cut out to match the machine’s feet including the holes), and then surface the entire workarea. If you eliminated the twist in the machine by using the shims, you will get a non-twisted workarea, and for the remaining area outside of it, which will remain twisted, you could either leave it this way or remove the material with a hand tool and by using a long straight edge, with the non-twisted workarea as reference.

Another thing I thought of that you can find on some CNC machines, is to mount one of the machine’s feet on a base with adjustable height. Since three points are always in one plane, it’s always only one foot that remains to need to be adjusted.

The QCW Frame

The QCW Frame meets the demand to have a machine that when assembled is immediately ready to use. With the separately available adjustable feet¹, an existing table may be twisted or uneven and if you do not have a table at all, you can also set up and operate the machine on the floor.

However, the QCW frame is rather expensive, and the target audience for a Onefinity is actually people who can build a base for the machine themselves as desired and needed, and therefore do not necessarily need the QCW Frame. For example, many people here want a machine base that allows vertical clamping. And if you think of the money that the QCW Frame costs, there are many things that you could find more urgent to buy, for example a professional dust collector or a vacuum table.


1.) The Any Surface Leveling System provides these inserts with threads that are not part of QCW Frame:

1 Like

Thanks for all the tips fellas. I will keep all this in mind for the next one I build. I think the QCW will simplify a lot of this and be one less thing on my mind.

1 Like

My QCW frame just sits on my cabinet that I built. One corner, of the QCW, did not touch my cabinet top, so I just shimmed it with a wooden wedge until all was solid. I then did the fishing line test and all looked good.

I put some anchor blocks to stop the frame from moving.

2 Likes

Thanks for all the input. I appreciate the detail you go into. You have me thinking twice about the QCW. Reading more posts I am wondering how rigid that QCW is and maybe I am trading one problem for another. I was assuming it was ultra rigid but if it can go out of coplanar I am right back to my root cause … a coplanar and level base that stays that way.

I have the Kreg table base, so getting that square, coplanar and level, adding crossbars of 2x4 for extra rigidity, that would give me the platform to construct a new box. Perfecting the reference platform to build on is the bigger challenge …. At least for me.

If I were to make another go at the torsion box, what specs have worked well for people? Plywood vs MDF? 1/2” vs 3/4”? Rib spacing?

2 Likes

A few of my thoughts on torsion box design and building, I am sure that some will laugh and tell me I am wrong and will be along shortly, but I can only say what worked for me
My woodworker is mounted on a homemade torsion box the torsion box is supported on wall brackets on the rear and left-hand side, there is a support leg made of steel box section bolted to the underside of the torsion box on the front right-hand corner with an adjustable foot at its base, steel being used instead of wood or aluminium as more stable

In my opinion the best material for a torsion box is moisture resistant MDF the best quality that you can find. Plywood is not stable enough it’s a natural product that is prone to warping.
For my base I used 25mm MR MDF when you buy the base check it with a straight edge as if you start off with a bow or warp you will never get it bang on lay the bottom of the box down on some trestles and leave it for a day or so MDF does not move much but everything you can do to help makes for a better outcome, don’t be tempted to lean or stand the top on its edge.
Decide on the height of the torsion part of the box I am sure there may be a calculation that you can do to optimise this my logic says that less is best, so I settled on 100mm
I used the same thickness of MR MDF for the outer and also the torsion grids the strips of MDF need to be cut as consistent as possible if they are more than about .25mm out then they are no good, I had mine cut by a local mill who use a CNC table saw, if you have to store these or move them try to keep them stood up as they will be in the completed box.
Next is torsion grid size I went for twice the height of the grid i.e., 200mm centre line
I assembled the grid using Titebond 3 and pocket hole screws using full length strips for the X axis direction of the grid.
With regards to the Y axis direction cut the length of the noggin’s as accurately as you can and secure them in an offset pattern with the next row of them set halfway between the previous row similar to a brick wall pattern again secure using glue and pocket hole screws
You now have an open box.
Check for flatness and take some time to sand any minor high spots and any glue squeeze
For the top I again used 25mm MR MDF having checked it for flatness and treated it the same as the base I took the decision not to glue the top down onto the torsion box but to only secure it with screws I drilled and countersunk clearance holes for the screws along both the X and Y axis at 100mm centres and secured it down, I mounted the OneFinity checking for level and squareness and spindle tramming.
I secured down a 18mm MDF (not MR as our MR MDF has a layer on each side) sacrificial waste board and the entire cutting area cleaned up with a 0.2mm cut in a single pass which I took as being fairly flat
I have replaced the sacrificial board twice since the reason being a lot of my work is multiple components profile cut so I am a bit hard on spoil boards and I am happy to say no issues with regards to flatness squareness or tramming.

As I say on a lot of my posts, my way may not be the best, it may not be textbook it may not work for everyone, but it’s worked for me, and it’s done with what I consider to be logic and good practice even if it is my logic!!!
D

2 Likes