GME's New 80/20 CNC Build - My Design

[GME]

Gary, I am glad your making progress, getting done just in time for Christmas gift giving!!

Jingle bells, jingle bells. It's just around the corner. I get the fun of making as well as giving. Can't beat that.

Gary

[wmgeorge]

So what are you making? I have a line on some red cedar (aromatic ) slabs with live edges that are about 10 inches wide, just about right for a nice sign over someones fireplace. I will need to use my 2 inch spoil board bit to surface and level. They are priced at $3 per board foot and they are one inch thick.

[Nicksch]

So what are you making? I have a line on some red cedar (aromatic ) slabs with live edges that are about 10 inches wide, just about right for a nice sign over someones fireplace. I will need to use my 2 inch spoil board bit to surface and level. They are priced at $3 per board foot and they are one inch thick.

Where are you located? That would be expensive in my neck of the woods. I am paying $1-25-$1.50 a board foot for eastern red cedar/aromatic cedar in South Carolina/Georgia

[wmgeorge]

In Iowa, not that much cedar around here.

[GME]

So what are you making? I have a line on some red cedar (aromatic ) slabs with live edges that are about 10 inches wide, just about right for a nice sign over someones fireplace. I will need to use my 2 inch spoil board bit to surface and level. They are priced at $3 per board foot and they are one inch thick.

Hello, Bill,

I'd be more inclined to resaw it down in size and use it to line chests. To my oh so delicate olfactory senses, it can get overwhelming pretty quickly when in open air. So, are you buying, or just looking?

I just finished up the spoilboard. I'm covering it in the next thread.

Gary

[wmgeorge]

Going to use for Cabin Signs. Been busy lately decided to buy a used camper and take the Wyoming, Montana , Yellowstone park then up to BC, Canada to see my farming / trucking cousins. Thought they might like a 4 - 5 foot long red cedar rustic sign. Last time we visited 2013 he had just killed a bear that was getting into his BBQ. So fixing up the camper and keeping busy.

[GME]

I'm long overdue for an update. I've finished my spoilboard.



This picture shows much of the beginning layout. The base layer is 18mm Baltic Birch (BB). I then measured the width of the t-tracks and cut strips of 1/2" MDF to space the t-tracks 4" on center. I positioned the first piece of MDF so the center of the MDF roughly aligned with the center of the table. I put the first t track next to the MDF and through bolted the t-track to the BB. I could have screwed the track to the BB, but through bolts ensure that that the track will never pull out. The exception was screw holes that fell over an extrusion cross brace. For those few, I used regular screws. After placIing all of the MDF and bolting down the t-track, it was time to glue the MDF to the BB.

I have a rubber platen for quickly spreading glue when veneering, but I couldn't find it. I tried a couple or different spreaders (one for auto body filler and a Rockler silicone spreader. I found them difficult to get an even spread. Finally, I tried a high density foam paint roller. I worked great, although it soaked up quite a bit of glue before it was adequately saturated. There are different types of foam rollers. This was the type I used:


The photo above depicts the method I used to clamp the 1/2" MDF to the BB. I already had some pieces of hard maple cut as clamping elements for my previous machine. Because of t-track spacing, I had to make a variety of sizes to accommodate different sizes of stock. They sure came in handy for clamping down the MDF.

For anyone who might be interested, made my own clamping screws for the t-track. I bought a supply of 1/4" t-nuts and threaded rod. I used a cutoff wheel on a grinder to cut the screws to the lengths I needed. After cleaning up the threads, I put a drop of CVA glue on the threaded rod pieces and screwed them into the t-nuts until the end of the threaded rod was even with the bottom of the t-nut. Viola! Works like a charm, All the tightening is done on the other end against a washer and very little rotational torque is applied at the t-nut end. I've never had one break loose. However, if one ever wanted to remove the threaded rod, a quick soak in a little acetone and it readily comes apart. Acetone is the principal agent in CVA remover - like what's used when one accidentally glues a couple of fingers together.




Here's another shot showing 3 more strips being glued into place.




This shot shows the third layer (3/4" MDF) glued in place. The top layer overlaps the t-tracks, but leaves the full width of the slot exposed. The clamping technique was the same used above for the 1/2" MDF.

Note that the MDF strips overhang at the front. That was by design. Note also a thin strip at the left in the photo, with a piece of t-track even further left. Also by design.

To finish it off, I made up profile toolpaths to cut along the left edge and along the front at right angles. That way, I can reference parts off the front or left side. The t-track to the extreme left, which is outside the cutting envelope, gives me a platform to use a hard maple block with a square end. The end of block butts up against the left edge of the MDF and gives a hard stop reference edge. I did something like this on my first machine and it worked great.




Here's a closeup of the 3-layer spoilboard arrangement after the top layer was cut back. Note that I chose to not cut it all the way back flush with the 2nd layer and neither the top or intermediate layers are flush with the BB base layer. Also by design. I wanted to have room to make further cut backs, if I ever felt it was warranted. Also, I had to keep both layers a little out beyond the extrusion, because I will be cutting dovetails off the front end. To accommodate a jig I'll be using, it was essential that the jig sit out beyond the extrusion.

Gary

[wmgeorge]

Beautiful work Gary, as usual.

[GME]

Here's a continuation of the previous post.




This photo depicts what I'll call "side curtains" I installed on the spoilboard. I considered two designs. This one, and one PL Jack used in his build thread. Jack Used magnets to hold pieces against the frame. Easier for him, because his frame is steel. I thought his design has one possible shortcoming though. His curtains were lower, so dust had a better chance of getting over the top and onto the linear rails. In contrast, on mine, the Z axis mounting plate just clears the tops of side curtains, and the curtains sit out away from the linear rails. Yet, they are way outside the cutting envelope. Both conditions will hopefully do a better job of confining dust.

If you look at the pics in the post above, you will see an uncovered bed area to the extreme right and left sides of the machine. In this photo, you will see that the gap is closed. Here's how I did it:

I cut 2 strips of 18mm BB. I created tiled toolpaths to cut counterbored holes to line up with slots in the extrusions. On the underside, I cut counterbored holes to insert 5/16" t-nuts. After inserting the t-nuts, I bolted the BB birch to the extrusions.

The upper piece is also 18mm BB with a piece of 1/4" ply glued to it. I created a tiled toolpath to cut slots in the BB to line up with the t-nuts. Then, it was a simple matter of gluing on the ply, starting 5/8" bolts into the t-nuts, pushing the assembly tight against the 1/2" MDF and tightening it all down tight.



Here's a shot down the right side. You can see how the top assembly mounts.

Gary

[GME]

Also included in the set up process was surfacing the spoilboard and doing a final tram of the spindle. It's at this point I encountered a problem.

When I surfaced the MDF, I felt some minor (about .001") ridges. Not big, but I could feel them. They were pretty separated, because I used a 60mm diameter surfacing bit. The ridges were parallel to the X axis, because I cut side-to-side.

At that time, I thought, 'no big deal. I can easily take care to that with tram.' WRONG! First off, I had a devil of a time setting up the glass plate I use for tramming. I should have just been able to set the plate on the table and go. Nope, not this time. Oh well, I fussed with it until I was satisfied. I got tram set side-to-side and front-to back, but when I moved the spindle either right or left, the tram changes. What? That shouldn't happen. So, I started over and redid it. Same issue. So, I did it all again. Same thing. After repeating the exercise a dozen times, I finally came to the conclusion something else was off.

Next step was to use a dial indicator front to back and measure the ridges. Left side first. Off about .001". Went to the right side. Also off about .001". Wait a minute, this is crazy. What gives? Okay, moved to the center. Not off. No ridges. What's going on? This isn't making sense.

Okay, I decided to pay a little closer attention to the ridges. Now, I'm on to something. On the left side, the high point was toward the rear. On the right side, the high point was toward the front. Now, I had my answer. The gantry beam was mounted with a slight twist in it. The left side was angled back by about .001" and the right side was nodding forward about .001", or a combined .002" twist. What caused it? I'm guessing that the triangular shaped mounting plates I milled were each off by .001", but in opposite directions. I took a lot of care when milling the parts, but it looks like I didn't take quite enough care. I'm just a hobby manual machinist and not that much experience, even as a hobbyist. No training. Just a lot of reading and watching YouTube videos. I don't know enough to know what I don't know. Still, the outcome was disappointing.

Fortunately, I should be able to fix the issue with shims. .001" at the lower front edge of gantry beam on the left side; .001" at the upper front edge on the right side. That's next up for me to do. Probably tomorrow afternoon. Of course, that also means another round of surfacing, squaring the gantry, and another round or two of tramming.

I thought I might try loosening up one side and trying to slide in a shim on the left. Loosening up means loosening the 4 bolts underneath and the bolts that screw into the face of the gantry beam. I left the bolts that screw into the horizontal interface plate tight. No way. I couldn't open a gap no matter what I tried. I concluded that it would require loosening bolts on both side, and it would take some experimentation to figure out how many I'll have to loosen. Bummer. I was hoping it would be simpler.

More to follow.

[GME]

Beautiful work Gary, as usual.

Thank you, Bill.

[Arjay]

Going to use for Cabin Signs. Been busy lately decided to buy a used camper and take the Wyoming, Montana , Yellowstone park then up to BC, Canada to see my farming / trucking cousins. Thought they might like a 4 - 5 foot long red cedar rustic sign. Last time we visited 2013 he had just killed a bear that was getting into his BBQ. So fixing up the camper and keeping busy.

You think that bear is cooked by now?

[GME]

I decided to make myself feel better about the twist in the gantry beam. What I'm seeing could well be what came from 80/20. Their stuff isn't perfect, and the folks at 80/20 don't claim it is. If the twist is "baked in," I hope that shims and bolt tightening are sufficient to remove the twist. I guess we'll see.

Gary

[wmgeorge]

But if your gantry is has a .001 twist and the shims don't work, its not really going to make a difference in any work your doing.

[ger21]

If the ridges in the spoilboard are .001", wouldn't that mean that the beam is off by much less?
Imo, .001" steps with a 60mm bit, is perfectly acceptable. That's only .0002" with a 1/2" bit.

[GME]

If the ridges in the spoilboard are .001", wouldn't that mean that the beam is off by much less?

Not necessarily. The beam is 6" high and the bit sets out in front of the beam by about the same amount. So, there is a 1:1. The fact that the bit sets lower when cutting should make no difference. it's just an extension of the gantry beam face.

What you ask about would occur, if the gantry beam height was less than the distance from the beam to the cutter. In that case, small changes at the gantry would maker larger changes at the cutter. Conversely, if the gantry beam height was greater than the distance to the cutter, you would need more shims to correct. With a 1:1, the ridge error=shims at the gantry. If the gantry height was 2X the distance to the cutter, it would take shims equaling 2X the error to correct.

Imo, .001" steps with a 60mm bit, is perfectly acceptable. That's only .0002" with a 1/2" bit.

As noted above, if the bit position is just an extension of the gantry beam's face (and it is), it doesn't matter how large the cutter's diameter. For a given depth of cut, the ridges will be the same, except with a smaller cutter, the ridges will be closer together. FWIW, I confirmed this by running a 1/2" cutter. Exactly the same ridges, only closer together.

Remember, you have to zero the cutter to the top of the spoilboard. When zeroing, the edge that contacts spoilboard first is what sets zero. That means the error will be the same with any cutter - only closer together.

Gary

[ger21]

Actually, I had that backwards.

Say you have a 6" beam, 6" above the table.
Draw a line straight down and you have a 12" leg of a 90° triangle. The bottom of the triangle is your 60mm bit.
Now, rotate the triangle so that one edge of the bit is .001" higher, and the top moves over about .005".
But since the beam is only half that, it would mean the beam is off by about .005"

Thinking again, you don't need to account for the distance to the spoilboard. Just the beam height, and the bit diameter.
It's not 1:1.

[wmgeorge]

Your spoil board can move X much when humidity changes?

[GME]

Your spoil board can move X much when humidity changes?

Nope. My shop is in my house. We have a heat pump. Totally climate controlled. If there's any movement, I'm confident it wouldn't be measurable.
Gary

[GME]

Actually, I had that backwards.

Say you have a 6" beam, 6" above the table.
Draw a line straight down and you have a 12" leg of a 90° triangle. The bottom of the triangle is your 60mm bit.
Now, rotate the triangle so that one edge of the bit is .001" higher, and the top moves over about .005".
But since the beam is only half that, it would mean the beam is off by about .005"

Thinking again, you don't need to account for the distance to the spoilboard. Just the beam height, and the bit diameter.
It's not 1:1.

I guess I'm just dense, but I'm not following. I don't understand what the bit diameter has to do with the issue.

Full Disclosure: I inadvertently misreported that my spindle set out in front of the gantry beam at about the same distance as the height of the gantry beam. When I wrote that, I was out of town and reporting from a faulty memory. The gantry beam is about 6" high and and the center of the spindle is about 8.5" from the face of the beam.

I believe I did a bad job of explaining the 1:1 ratio thing. It had nothing to do with the bit. I was referring to the ratio of the height of the gantry beam to the distance from the face of the gantry beam to the center of the spindle. If the two distances are equal, there is a 1:1 ratio. With a 1:1 ratio, if I added a shim to rotate the gantry beam .001" forward, the spindle would rotate downward by .001"

In my mind, it's really all about angles. To explain where I'm coming from, let's take the .001" deviation out of it. Rather, let's look at it from an angles perspective. With a 1:1 ratio, rotate the gantry 5 (or any) degrees, and the spindle angle rotates 5 (or any) degrees. Now that I have the true measurements, I dealing with a roughly 1:1.42 ratio (8.5/6=1.42) The angular relationship does not change, but the amount of movement at one end affects the amount of movement at the other end differently. In this case, a .001" deviation at the spoilboard translates to a .0007 at the gantry to get things right. (1" at the spoilboard divided by 1.42= about .0007" at the gantry). Thought of another way, for every .001" rotation at the gantry will result in .00142" rotation at the spindle.

Here's why I'm having trouble relating the bit diameter to the issue at hand. First let's make the assumption there is no error to correct. Second, let's assume that I rotate the gantry by 45 degrees. Finally, let's assume that my DOC is .005". The gantry is at 45 degrees relative to the spoilboard, and so is the cutting face of the bit (in a perfect world, the face of the cutter is parallel to the top and bottom of the gantry beam). Everything rotates the same. Now, I use my touch plate to zero the 60mm bit. Only the lowest edge of the bit can/will make contact with the touch plate. Start the surfacing routine. The bit lowers to -.005" and cuts. Only the leading edge will cut. Most of the bit won't make contact at .005" DOC. We end up with a series of ridges with a .005" deviation. The real deviation will be much more. Now let's change to a 1/2" bit. Same routine. The cut will be .005" deep and most of the bit won't make contact. We end up with a series of ridges with .005" deviation, same as with the 60mm bit. .005" won't tell us the true deviation, because there isn't full contact. Assuming a 60% stepover for both bits, the ridges are exactly the same, except the ridges are closer together with the 1/2" bit.

If we change to a 5 degree rotation, everything works out the same way, except more of the bit (maybe all) will make contact with the spoilboard. Ridges from both bits will be the same, except for the ridge separation distance and a longer taper angle with the 60mm bit.

Note: I ran both the 60mm and 1/2" bits last week. What I observed comported with what I've described. Note also that I had full contact with both bits, which came as no surprise at a .005" cut depth. FWIW, I took a pencil and scribbled lines all over the spoilboard. Surfacing removed all the lines.

So, I'm left not understanding how bit diameter plays into this. I may be overlooking something, but I don't know what that something might be.

Any help you can offer to help clear this up is much appreciated.

Thank you,

Gary