Making a flat table for a router

You can use the 1/2" it will stiffen the box just as well as the 3/4" I’m with you on the weight reduction.

1 Like

Oooh, or maybe hardboard. Thoughts?

It would probably work out ok to use 1/8" hardboard, providing it is glued and screwed. But if it were me, I would use the 1/2" . My first OX was a torsionbox design with 3/4" cabinet grade plywood on top and 1/2" cabinet grade ply on the bottom with all the ribs made of the 3/4" material. I actually bolted the Y axis rails to the side of the torsion box so I did not need any of the cross bracing and made the rails stiff as can be.

I came to the same conclusion and just finished loading the 1/2" MDF into the car, along with pieces to make the legs for the table.

1 Like

Turns out what I loaded into the car is what will become (part of?) a spoil board, because I got the minor dimension too small by slightly less than an inch, and it won’t quite reach to both edge pieces. :sob::blush::roll_eyes::money_with_wings::rage:

The plywood saw at Home Depot or Lowes cuts straighter than I can with the saws I have, at least easily, and getting as many cuts as possible there reduces the amount of MDF dust in my workshop. (Once this OX project is done, I want to hang a dust curtain between the wood and metal sections of the shop to contain the MDF dust.)

I got that board wrong because I remembered that I had rounded up 1 meter to 40 inches to have a little extra room (the stores have never heard of metric, anyway) and 1.5 meters to 60 inches. However, I forgot that I took 4" cuts and made the table the remainder of the 49 inch wide sheet of MDF, so the real size of the table top I got was 41 inches wide minus two saw kerfs. It turns out that one saw kerf was 5/32, so my tabletop was actually 40 11/16" wide.

So far, the box I’ve made isn’t a true torsion box, just most of a torsion box. I have a box of 4" wide 3/4" MDF with a 3/4" MDF top and a 1/2" MDF bottom, and two 4" wide 3/4" MDF stringers across the width. I can’t figure out that I actually need anything more to support an OX…

For the legs, I’m using fir 4x4 pieces 36" long, with pieces of MDF screwed to the sides to stabilize and stiffen them.

To level the table, I bought 2" long 1/2" stainless steel bolts and 1/2" nuts, and 1/2" coupling nuts, one of each for each leg. I drilled 13/16 holes as deep as the coupling nuts are long into the base of each leg, and seated the coupling nuts in the holes. I threaded the nuts onto the bolts, and then threaded them into the coupling nuts on the bottoms of the legs. I will level the table with them essentially the same way as leveling a lathe.

1 Like

Since you are just resting the OX on the table, and not using it as part of the machine structure like I did, I think you will be fine. Good idea for the leveling feet. You may want to rethink the leg height, with the addition of the machine on top the spoiler board height may feel a little high for working on, depending how tall you are. I set my Sphinx on top of a 33" table and it feels just perfect to 6’1" of me.

Much easier the cut the legs shorter than to lengthen them. I tend to prefer substantially higher work tables than normal people.

I didn’t take pictures as I went along, too much in the moment. I’m not sure it’s a design anyone else would want to copy anyway, I kind of made it up as I went along.

If it isn’t rigid enough, I can gusset it. But I think I’ll move on to building the OX before I worry about the table any more. 16 more screws and it’s ready to level and build on!


Have fun with the build. That is my favorite part, the actual building of the machine. I tend to sell them off and build another one after a few months time.

1 Like

It’s upright and somewhat level. I had a problem (didn’t set a clamp right?) at one corner so that corner is too high, so I’ll probably take the sander to the area around that corner until the top is a plane.

I wonder if I should fill the top partway with sand to make it stable and make the adjustment feet more effective at taking out twist? Might absorb vibration too.

1 Like

I do not think that sand would be required. You will never get the level of precision out of a large OX build where a small amount of unevenness of the table would affect it.

You have a couple inches on me… My table is about 41" high, and I think 39 or 40" would be about perfect, but I don’t care enough to do the work to take off an inch.

If you find it bothers you, make a riser platform to stand on in front of the table.

The table has some twist, I think. I mounted a dial indicator to the spindle body and moved around the MDF surface clamped down to 2040 extrusion on top of the table. The ends are 2040 that the Y c-beam is mounted on top of. The MDF is sitting on a matrix of 2040 in the middle, between those end pieces:

I have a total of about 0.075" (almost 2 mm) variance relative to the movement of the head:


I can’t imagine that the MDF is off by 2mm in thickness, but if the table is twisted and the 2040 end supports are twisted relative to each other, then the path that the spindle takes should likewise be similarly twisted. Perhaps not exactly aligned, but off by 2 mm seems unlikely as well. I’ll put a straightedge across the diagonal and if the top is twisted I can shim up the 2040 pieces to take out the twist, but if I twist both the MDF and the table the same way it seems like I’ll have the same problem. :thinking:

1 Like

Don’t you remove all this variance by surfacing the spoil board? All that matters is where the board surface is to the cutting head?

If the machine is twisted, then surfaced isn’t flat. If the corners of the table aren’t co-planar, the machine will be twisted out of square, and if I surface the spoilboard, I’ll end up with a “potato chip” shape cut into the top.

With a height gauge, I discovered that the gantry beam was 0.1mm higher above the right-side Y c-beam than above the left-side Y c-beam. I loosened the right side screws a bit and gently tapped it down with a rubber mallet until I measured it repeatably within 0.02mm side to side relative to the top of the c-beam. I’ll call that good.

I put the whole frame up on pieces of c-beam above the table, and am most of the way through re-aligning all the joints vertically. I’m clamping the joints top and bottom with extra pieces of extrusion to hold them aligned, while re-fixing the corner brackets. This should reduce the variance across the top. I’ll check by running the indicator down the tops of the bare cross-pieces when I’m done.

The table does have a bit of twist, so I’ll shim that out too. Probably due to not fixing the imperfect side panels well enough. It’s not as flat as I measured earlier.

I finished squaring up Z on all the individual joints across the bed. I put the frame up on stacks of extrusion 100mm high at the corners, and then used a 1500mm 2080 extrusion and various feeler gauges to determine shims to add. Then I set it down on the shims and clamped it in place.

Running a dial indicator across rails front to back is within about 0.01" / 0.25mm. Running a dial indicator side to side on the front rail (haven’t figured out a setup yet to reach the back rail) is about 0.03" / 0.75mm side to side (was 0.04" before I loosened the left side and encouraged it upwards with a rubber mallet). The variance in thickness of the MDF is at least .15mm having measured only along one edge so far. Measuring 9 points across the MDF bed as before, my max variance is down to about .05" / 1.25mm, so I took out 1/3 of the variance by squaring and shimming.

As far as I can tell, about 3/5 of my lack of alignment is in the plates, out-of-center tapping of the gantry extrusion, or some combination of both. I think I have about .012" end to end of twist on the back rail relative to the front that I could shim out, but otherwise what I have is probably about as good as I’m going to get without at least modifying the side plates.

If I use safe Z clearance and extra cutting depth of 1.5mm I should be in good enough shape for most through-cutting jobs in the meantime. I do have one thing I want to cut that has some fairly precise depth requirements but those cuts are localized and I can just use that area of the bed to home Z on my stock and Bob’s my uncle!

Until I fix the side plates to square the gantry with the rails, though, surfacing the bed will scallop it because ultimately the head is out of tram with the bed. So I see an hour or three of disassembly, playing with needle files, and reassembly, probably repeated a few times, in my near future.


Leveling an engine lathe was so much easier because I knew everything was essentially square, and all I needed to do was adjust the feet until the ways were level and square to a test bar. With this OX, I’m having trouble with errors stacking up, so hunting the real sources of various variances isn’t as easy.

I spent some quality time with master precision level, dial indicator, dial test indicator, and various sizes of straightedge. I’m glad I’m using c-beam, because even that deflects enough under the weight of the gantry to make it tricky to use the level to take out twist front to back. I’ve discovered that the v-rail pieces aren’t really as straight as I hoped. So I’m backing off on tweaking the plates for now, until/unless I’m certain that’s what I need to do.

1 Like