Waste board flex

I’ve read topics about the waste board (mdf) flexing under the force of the router when cutting. Very reasonable concern for working tight tolerances. I don’t use the leveling feet because I built my enclosure into the corner of my workshop and poured a leveling concrete for a floor in it. Absolutely flat and level as I can get it, Than I put a 1/2" sheet of exercise rubber matting on top to rest the Shapeoko on. Total support under the entire machine, BUT the mdf base board doesn’t really touch the floor under it (if I didn’t have the rubber matting). Because the angle iron front and back “rails” are mounted under the mdf base board. SOOO… (see picture) I bolted an angle iron under and along the sides of the mdf base board and since I had a scrap piece of sheet steel, I bolted it under the center of the mdf base board. NOW the entire machine completely rests on the rubber matting totally supported. The bolts in the center are recessed like the front and back base board bolts are, The sides bolts are under the rails and not in conflict so I didn’t recess them. This also may help with the rigidity of the machine in whole since the base board side angle iron now creates an end box with 4 sides rather than before with 3 sides (the 2 end vertical supports and the rail) maybe the mdf won’t be able to vibrate if it really vibrated before. I learn a lot from this forum and simply hope to enhance other user’s intelligent comments. This effort cost me $1.16 for the bolts but (imo) increased the value of the machine.


Just on this, are you saying the angle iron is just on the sides? If so surely there will be some flex in the centre where the middle of the board is not in contact? When I was looking at my base I did consider using a thicker sheet of MDF or wood to put the machine on - make small cut outs for any bolts.

No, the entire bottom of the machine now rests on the 1/2" padding which is on the concrete floor of the enclosure, under the base board (mdf) I bolted a 1/8" thick sheet of plate steel that is about 8" wide to stiffen the center section of the mdf base board, the holes in the mdf are recessed. I have t tracks on order and will use them with 3/8" mdf (t tracks 3/8" height) and than a 1/8" waste board under project stock. I am mounting the router so that the maximum depth of the bit will be 1/16" above the t tracks and 3/8" mdf. That way the bit can not hit the t track and any cutting thru will be into the 1/8" waste board. I’m a lousy photographer (no extra lighting) using my phone for pictures so my description probably is better than any picture I take.
Carbide3D no longer provides the center straps under the mdf board (imo) just a cost savings because strap steel was an insignificant stiffener for the mdf. The front and back are angle so I decided to use angle on the sides which is a definite stiffener over the mdf.

They welded “nuts” on the bottom side of the front and back angles that the machine sits on without the leveling feet. So I used the 1/2" pad to compensate and absorb those projections plus total vibration isolation compared to about 1 square inch of isolation with the leveling feet. The concrete leveled itself and provided a solid platform, I may drill holes thru the floor and mount heavy angle iron brackets at the 4 corners to incorporate the machine into the entire structure of the shop. Than I’ll have a 3 ton cnc machine which would classify it as a heavy machine, but that is definitely over kill for a belt drive cnc.

1 Like

I can attest to the improvement in bed leveling without the leveling feet:

I hadn’t considered the elevation of leveling PEMM nut vs. rolled edges of the framing being at disparate elevations?.. I took a quick look just now and they appear even/flush with each other. This would be a tough measurement to make.

EDIT: I neglected to check the flex on this last depth gauge go-round, but it has to be vastly improved since my initial level assessment. I could easily deflect the wasteboard when the machine was sitting on its leveling feet

1 Like

I FINALLY got my Shapeoko 3 into the enclosure and tested it with Carbide Motion. Homed perfect. Jogged perfect. So I checked my base board flatness by jogging to the four corners and the middles (9 points) and lowering the z and trapping a piece of paper.
The values in the top square are really negatives and show the working square on the base board from looking from the front. I called the center value “zero” and calculated variations from the center point below. These are mm values.so in inches I’m 0.01496 inches high on the right front corner. ( 0.38 / 25.4 = 0.014960 ) How does that fare up for you machinists, did I do okay??? For woodworking I don’t think I’ll surface the waste board, I don’t think I can see 0.015 inches. I’m wondering how all that effort to stiffen the center worked out in your minds. Please comment, thanks.


Here’s exactly how mine went:

1 Like

Thanks Mike, I feel good but I’m still 0.011 of an inch behind you, not good enough for the Olympics but just fine in my backyard. I checked the paper I used and found one hole in it, maybe I’m closer but which point did I poke thru the paper?? Just noticed I’m farther than that, -0.22 plus 0.38 equals 0.60 or (0.60 / 25.4 = 0.0236 ) wow Mike , I’m only at half speed. The base board is stiffer but maybe I still need to surface the waste board.

I couldn’t make enough adjustment to get it all out, surfacing is quick and pretty easy. I used a dial indicator mounted to the carriage with a mag base and ran it directly on the surface. You could also put a heavy copper wire in the collet, and use that as your pointer, and the flat end to avoid going through the paper, or you can use an end mill and a feeler gage for it to touch down on (then you also know exactly how much to add - it’s printed on the gage)

Your results sound pretty decent, and I agree that wood-working larger pieces won’t be much issue. I have prepped to undertake this final assault with bed leveling too (Inserts, T-nuts, Leveling bit, T-track, etc.), but I’m still setting up my controllers and electrical arrangement and I’ve committed to complete that first. I wouldn’t trust myself with paper and relative pressure and have succumbed to a depth gauge.

Thanks Jim and Mike, T track goes on tomorrow with the waste board. The waste board will be 18" wide as I have a 3/4" straight router bit to surface with. Might as well, the t track on either side and one in the middle off center some. The widest material I could cut from would be 16 3/4" so an 18" wide waste board with a surfaced area 16 3/4" wide leaves me a 3/8" edge on either side of the material to clamp down. I guess I’m saying that surfacing leaves a higher edge outside the cutting area of the Shapeoko. It’s easy for me to confuse myself, how am I doing ??

Picture is what I mean…here’s a closeup of the higher ledge,
maybe I’m over thinking this?? Thanks

what did you use to model this up? @grumpa


Proper prior planning prevents (alliterative rude word for urine)-poor performance.


Yes, the higher outside lip of surfacing is a pain — at one point I was actually trying to work up a way to make a jig to cut tempered Masonite panels to use as insets into that area — my new plan is to try to take advantage of it — just always do a surfacing pass on whatever fixture I’m using, and ensure that it perfectly fits the stock along two edges square to the machine.

One concern with your image is you’re not getting the best mechanical advantage on the clamping — the clamped part should always be lower so that it is held acutely and the other end of the clamp should be higher — yes, this makes keeping the clamps from interfering with the works more difficult, but it makes the work less likely to raise difficulty.

I only draw in sketchup make 2017. SVG files are lousy, always broken and scattered into Carbide Create, but it’s what I’m used to. 30 years ago I was an AutoCad expert but I will NOT work in a cloud enviroment so Sketchup works fine for me. That SimpleCad of the other member looks amazing for carving pictures and I’ll look into that soon, but still within sketchup.
Oh, the Shapeoko 3 model is from Edward Ford on the wiki, It is exact down to the screw threads, wonderful but very intense drawing wise.

1 Like

Hi Will, Why the language warning ?? I try t keep all my texts family friendly. But thanks for the clamping tip, haven’t thought of that cuz of the possible interference with the rail assembly. Trial and error will prevail, maybe I’ll make some styrofoam clamps to test with first.

The language warning was in reference to myself — I edited out a word in my post out of an abundance of caution for delicate sensibilities.

Shouldn’t’ve used a quote block there, and have removed it

Your post was fine.

1 Like

I have a standard-size SO3 but that isn’t going to stop me from making some observations.

I arrived at this point at about the same time as I noticed this thread. I’ve replaced the stock spoil board (the previous owner had chewed it up pretty good) and then I wanted to add a second 1/2" board on top of that, so I can clear the front and rear brackets of the SO3 (so I can use longer pieces).

I wanted to make sure my board was fairly flat relative to the Z so I made a DI holder that goes where my dust boot would normally mount.

When I purchased my SO3 (used), I noted that there were no feet on the bottom of the unit. Having the unit sit on my workbench, I didn’t want the waste/spoil board screws that were protruding from scratching-up my bench top.

So I visited my local Ace hardware store and purchased some vinyl bumpers. They’re 3/4" square and 1/2" high. I placed nine of them on the bottom of the SO3. I had figured these bumpers would be made from hard enough vinyl/rubber that they would not compress.

Checking w/ the DI, I think I can compress them by about .004 or .005" by applying pressure to the spoil board with my thumb. It is also possible my bench itself is moving that much. Could be a combination of the two, as well.

In testing various areas of the spoil board, I’m about +/- .003 or .004" across almost all of it, save for the front left corner which I think is about -.009.

My intent all along was to shim under the feet of the SO3, or under the corners of my top spoil board, to get everything to +/- .003" or better.

Before proceeding, though, I think I will replace the bumpers with some discs I’ll cut from plywood, so I can remove the compression of the bumper from the equation.

The thing is, if you guys are setting your entire machines on top of foam rubber padding, it doesn’t surprise me you can get quite a bit of movement.in your spoil boards. I assume with those larger units you’re using foam to deaden sound coming from the larger panel, sort of acting like a drum head?

If that is the case, the solution may be to use some panel damping stuff normally used in cars and speakers to reduce noise, while also using some sort of solid/hard feet that can accommodate shimming. So the SO3 ultimately sits on hard feet.

Or I may have completely misread the problem in which case, never mind.

FWIW, the DI shows quite a bit of action ( a few thousandths is easy to achieve) when pushing the Z-carriage to and fro. And I can push the rails and get .001 or .002". My point being, if I’m going to start quoting machining jobs for NASA (kidding), I’m going to need to fix any variability I can actually fix.

1 Like

To suppress the dirchilet (sp) effect (drum) I followed Mark Bellum’s advice and mounted an uprite 4x4 under the center of my enclosure base to prevent vibration. The foam duct wrap I used to act as cheap dynamat was nil to minimal effect, but the homemade MLV (alum flashing w/ rubber tape on each surface) removed 20 decibles on it’s own. MLV is very expensive and I had that stuff laying around. Couch cushion foam does nothing but acoustic ceiling panels added roughly 5 more decibels of reduction. A solid door is much quieter than one with a window. All these items are credited to Mark Bellum and remember not to [quote=“mbellon, post:5, topic:5016”]

I will yield to Mark.

1 Like

To suppress the dirchilet (sp) effect (drum) I followed Mark Bellon’s advice and mounted an uprite 4x4 under the center of my enclosure base to prevent vibration.

The fundamental resonances in a Dirichlet problem are the full and half tone. Placing a support half way - in the middle - prevents a huge number of resonances. If the skin material is stiff (furniture grade plywood is an excellent choice), this is often sufficient. The best results come from TWO cross pieces - split the drum head in thirds - or fifths.

My CNC table, design to handle a huge amount of weight and vibration, is split in fifths (5 segments) for near perfect suppression; the deck material is screwed to the frame and the cross pieces. Admittedly, this is extreme and has to deal with conditions FAR more than any Nomad or SO3 will have to deal with.

Couch cushion foam does nothing but acoustic ceiling panels added roughly 5 more decibels of reduction.

This is EXACTLY why I suggest acoustic ceiling tiles for people on a budget! Acoustic ceiling tiles are CHEAP and they work! Using MLV or more exotic materials offer 10 dBA and up… 30 dBA is possible. A bit of care is necessary to make it work out.

We can talk about wall design for noise suppression separately. Inside the enclosure, we need LAYERS. Closest to the machine is light, more open material to diffract high frequencies and prevent resonances. Next comes heavier materials, typically MLV. Next, if one can swing it, should be an air space filled with open sound suppression insulation (it must not be under compression (loosely fitting). Finally comes the walls. MASS is your friend (MDF, pressure treated plywood, furniture grade plywood, steel!).

There are some AMAZING materials available out there… but they cost. Creativity can go pretty far; the rest is the tough part of Physics.

A solid door is much quieter than one with a window

Sure is! Yes, It’s possible to design a window that can suppress sound fairly well… but nowhere as well as a solid wall. A single pane of polycarbonate or tempered glass is largely sound transparent.

With todays costs so low, leaving out the window and installing a high definition camera is well with reach of many.

The foam duct wrap I used

Ductwork can be a significant emitter of sound. In large shops special care is taken to acoustically insulate (or at at least isolate) ductwork.



I didn’t mean to miss quote Mark here, He advised me about the center beam under the floor (base) and I added the vertical 4x4 on my own thoughts, so I do have a center cross member and a vertical 4x4 to reduce vibration. I get carried away trying to respect and credit others as I see their intelligent observations. He also advised me NOT to use the couch foam but I did because I did. The homemade MLV , when I installed it in the vacuum cleaner enclosure did reduce 20 decibels as per my decibel meter (actual meter not phone app).reading before MLV 101 and reading after MLV 81 decibels. Their are many members here on this forum whom are due their proper credit even for their humor. Thanks Mark for all your advice.

1 Like

Should likely avoid dividing lengths/areas in perfect halves, as each will likely result in equal frequencies which are additive. Probably better to go 60-40?