[pippin88]

The larger tube will be much much more rigid. Bolting should not be difficult.

I'm on my phone so can't do a diagram, but the long beam (currently sandwiched between the angles) would have holes drilled and tapped where it crosses each of the perpendicular (cross table) beams. The cross table beams have matching holes in the top. You just insert the bolts and tighten from the open ends of the tubes.

[jerseyguy1996]

Looks great! Your SketchUp skills make it easy to see how the details come together. Nice work.

I think that style of frame will work out really well. Here are a few thoughts rattling around my head. Feel free to take them or leave them, or maybe they will just spark some discussion:

• The steel tubing might not make the best rail support. Rectangular steel tubing is hot rolled, meaning it is formed at high temp, and as it cools it warps and the tolerances are poor. If there is slop in the distance between the rails, your gantry will not move freely. If there is slop in Z, the machine is hard to level. So, a lot of people use aluminum extrusion for rail supports and for the gantry crossbeam, which has much better tolerances and keeps everything truer. My way of thinking is that, if you're going to use the steel tubing, why not plan so that you could upgrade the frame later if you want. 80/20 (series 10) 2040 is 2"x4", uses 5/16" T-bolts and nuts and I believe has slots at 1/2", 1 1/2", 2 1/2" and 3 1/2", so if you can drill your horizontal holes into the 2x4 and angle at 1/2" or 1 1/2" for a 5/16" bolt, you leave open the option to just switch it out for the 2040 extrusion later if you wanted.
  
  
  

In terms of the distance between the rails, I was planning to make the holes in the linear guide rails big enough to make them adjustable. Then I can keep the screws loose while I slide the gantry up and down the rails, tightening the screws as I go. That way the gantry itself will move the linear guide rails to the appropriate distance. In terms of vertical (z axis) warp in the rails....shims??? I don't know. I probably won't have the measurement equipment to be able to tell how much error I have in the vertical direction. How much of a bow would you expect to see in a 60" long steel tube. Does the software have the ability to correct for some of that, for instance if I'm at the middle of the rail (maximum vertical error) the software will add x'thousandths to the Z position to equate to the correct position at the ends of the rail?

I love your idea of drilling the screw holes in the angle to line up with aluminum extrusion slots for future upgrades!

[jerseyguy1996]

The larger tube will be much much more rigid. Bolting should not be difficult.

I'm on my phone so can't do a diagram, but the long beam (currently sandwiched between the angles) would have holes drilled and tapped where it crosses each of the perpendicular (cross table) beams. The cross table beams have matching holes in the top. You just insert the bolts and tighten from the open ends of the tubes.

How much flex would you expect from a 2X4 inch steel tube supported every 10 inches on 2X2 inch steel tubes that are themselves supported on a flat surface?

[jerseyguy1996]

By the way.....thank you everyone sooooo much! This is a great community and it has been so helpful! I still have to come up with the drive mechanism for this thing so I am sure that I will have lots more pages of questions.

[DIYedInTheWool]

The larger tube will be much much more rigid. Bolting should not be difficult.

I'm on my phone so can't do a diagram, but the long beam (currently sandwiched between the angles) would have holes drilled and tapped where it crosses each of the perpendicular (cross table) beams. The cross table beams have matching holes in the top. You just insert the bolts and tighten from the open ends of the tubes.

Well, the inside dimension of the cross tubes is probably 1.625" wide, so that's fine for putting in one bolt right at the end. But, as anyone who has put a pipe extension on a breaker bar can tell you, 40-something inches of leverage will move a well-torqued bolt pretty well. So you probably want to put another bolt further in, and that's where it gets fiddly and you have to come up with work-arounds.

Of course, there's lots of designs that will work great. Here's my thought, though. When you weld, you get your steel and position, clamp, level, measure, and check *beforehand*, then you weld it in place and it doesn't move. Bolts will move, so the design has to take that into consideration. And, any bolt-together assembly I've made I worry about marking my pieces and drilling and tapping and assembling and then, whoops, things are off, and I'm hosed. So when I'm suggesting the extrusion and the strut channel and all, it's more about leaving as much room as possible to make fine adjustments *after* it's assembled then just tightening down the bolts when it's perfect. Other ways to get at that, too--the OP mentioned putting some slotted holes in the angle for that purpose.

I probably won't have the measurement equipment to be able to tell how much error I have in the vertical direction.

Probably want to epoxy level as pippin said. Then there are lots of ways to measure--here's one with a dial indicator once it's assembled:

https://www.youtube.com/watch?v=_MtngIb_7gk

You don't need the electronics for that, probably better just to push the gantry around manually.

[jerseyguy1996]

Probably want to epoxy level as pippin said. Then there are lots of ways to measure--here's one with a dial indicator once it's assembled:

I did a quick google search and had trouble finding a video on epoxy leveling. What's the process to do this?

[jerseyguy1996]

How heavy can the gantry be? I'm trying to add some additional structure to prevent twisting but I'm worried about the weight. Currently based on the materials that I am using I am up to about 50 lbs...not including the router/spindle.

[DIYedInTheWool]

How heavy can the gantry be? I'm trying to add some additional structure to prevent twisting but I'm worried about the weight. Currently based on the materials that I am using I am up to about 50 lbs...not including the router/spindle.

It depends on the strength of the motor and drive components you are using to move it, the rails you are supporting it on, and the strength and rigidity of structure that is holding it up.

Look at the load rating for your rails. It can be different in different directions. Also the preload.

Also this is a useful resource for sizing the motor based on gantry weight:

http://www.geckodrive.com/support/choosing-a-drive.html

[jerseyguy1996]

It's been a few days so I wanted to post some updates to the design. I'm mostly trying to figure out the gantry. Weight may be an issue but I figure I will cross that bridge when I start working out how to move things around. Right now I just want to work on putting together a rigid structure.

First off I changed the gantry supports to 1/4" walls so that I can drill and tap right into them. I guess a question could be, would I get enough strength tapping into 1/8" walls?

Gantry Support Detail by jg1996business, on Flickr

Second I changed the horizontal beam on the gantry from plate steel to a 2X4X1/8 inch steel tube. The linear rails will be screwed right into tapped holes. Again, the 1/8" thickness question comes up in terms of tapping the holes. I figure with M6 1mm thread screws, I should get 3 threads in 1/8 inch walled tube (.125" = 3.175mm). Am I working that out right?

Gantry by jg1996business, on Flickr

Finally I changed the Z Axis risers from plate steel to 1X4X1/8 inch tube.

Z Axis Risers by jg1996business, on Flickr

Hopefully this won't be just a sketchup project for too long. I do want to make sure everything is perfect before I start buying materials though.

[DIYedInTheWool]

Looking good. Looks like fun. Here are my thoughts:

The drive components should be an integral part of the design process, not an afterthought. You're going to need to bring the sides of your gantry down to meet your leadscrew (or gear rack for rack and pinion). X drive will probably go along the outside of your rail supports. You want to push the gantry down low.

You need to leave some space in between your z rails for the leadscrew.

Your Y rails on the gantry seem a little close together, I would find a way to space further apart, and again, you need to have a motor mount and either leadscrew or rack and pinion integrated into the gantry design.

I think it would really help your design to study some other similar machines and compare. Here's a few:

http://www.cnczone.com/forums/joes-cnc-model-2006-a/

CRP4896 4' x 8' CNC Router Kit | CNCRouterParts

[jerseyguy1996]

Looking good. Looks like fun. Here are my thoughts:

The drive components should be an integral part of the design process, not an afterthought. You're going to need to bring the sides of your gantry down to meet your leadscrew (or gear rack for rack and pinion). X drive will probably go along the outside of your rail supports. You want to push the gantry down low.

You need to leave some space in between your z rails for the leadscrew.

Your Y rails on the gantry seem a little close together, I would find a way to space further apart, and again, you need to have a motor mount and either leadscrew or rack and pinion integrated into the gantry design.

I think it would really help your design to study some other similar machines and compare. Here's a few:

http://www.cnczone.com/forums/joes-cnc-model-2006-a/

CRP4896 4' x 8' CNC Router Kit | CNCRouterParts

I am starting to work on the drive components. I wanted to get the gantry complete so that I could estimate the weight of it for sizing of the drive components. The Z axis uses 1X4 inch tubes. I would prefer to use 6 inch tubes but I can't seem to source 1X6 inch tubes. The best I can do is 2X6 inch which in my mind sticks the spindle way out in front of the horizontal beam which will cause twisting. I'm trying to build in a lot of travel in the Z axis....you know just in case I want to machine something really thick...i don't know. This is how I was trying to do it. I notice that on some of those machines that you linked to the Z travel tends to be somewhat constrained to the width of the gantry horizontal beam. Any thoughts on the method that I am showing below?

Z Axis Travel by jg1996business, on Flickr

[DIYedInTheWool]

I am starting to work on the drive components. I wanted to get the gantry complete so that I could estimate the weight of it for sizing of the drive components. The Z axis uses 1X4 inch tubes. I would prefer to use 6 inch tubes but I can't seem to source 1X6 inch tubes. The best I can do is 2X6 inch which in my mind sticks the spindle way out in front of the horizontal beam which will cause twisting. I'm trying to build in a lot of travel in the Z axis....you know just in case I want to machine something really thick...i don't know. This is how I was trying to do it. I notice that on some of those machines that you linked to the Z travel tends to be somewhat constrained to the width of the gantry horizontal beam. Any thoughts on the method that I am showing below?

Well, you're putting your finger right on it with the "twisting". Just thinking it thorough, with gravity and especially with the forces of cutting these things will make the problem *worse*:

• the lower the z axis is extended down while cutting
• the closer together the gantry rails are
• the further out the z axis components extend from the gantry (e.g. 2x6 is worse than 1x6)
• the higher-profile the rails are--same issue and those round rails contribute a lot to the height

So there are tradeoffs with the amount of travel/clearance you build into Z and the stress you put on the system when cutting when fully extended down

So, why use rectangular hollow tubing?--it seems like it just adds bulk. You could use aluminum plate. If you're all about steel what about 6" flat bar, I don't know, maybe 3/8"?

The round rails are adding a lot of depth as well. Profile rails like THK or HIWIN or whatever would be good, but expensive. If you're looking for a good, less-expensive low-profile rail you could try the cncrouterparts carriages that ride on the edge of any 1/4" cold-rolled flat bar:

https://www.youtube.com/watch?v=BRbpQCQqhgw
Extended Linear Carriage with ABEC 7 Bearings - CRP102-00 | CNCRouterParts

A little work and I bet you can get your Z axis hugging much closer to the gantry. Here's a more low-profile example using 1/4" CRS rails and those cncrouterparts carriages:

http://farinastudio.files.wordpress....0/dsc_0103.jpg
http://farinastudio.files.wordpress....0/dsc_0123.jpg

Also you are going to need a leadscrew for Z, which will probably go in between your Z rails, so you probably can't have them right next to each other like that.

[handlewanker]

Hi, big router......my own build is teeny compared.......one thing I did notice, the bearing blocks on the base rails for the Y axis are too close together....space them at least half their spacing apart more.......also I would fit three bearing blocks as they're so cheap but will give heaps more rigidity to the gantry sides as they individually track along the base rails.

This will add a bit more to the base length for the added spacing, but are worth the few mm more.

If you can get the gantry crossbeam directly down on the bearing blocks without any side frames, even if you have to have the table set down deeper in the frame work, the main factor causing lack of rigidity in the side frames will cease to exist.

I think the frame work for the base needs to be two tubes one above the other with the gantry crossbeam riding on the top tube.......and a couple of vertical tubes between the two tubes to keep them from bowing.

The end on view of the router body looks like a U shape, and although the work piece is now down on the table and inside the framework, the end result is that the moving parts will now have less parts to flex.

You still need two drive screws for the gantry and I would by choice have a single central stepper with two toothed belts to each screw to prevent slewing of the gantry.

BTW, using the round SBR linear rail adds a lot of height to the X axis rails and bearing blocks and this pushes the spindle further away from the face of the gantry crossbeam, which cause more twisting leverage from the cutter forces against the gantry cross beam section.

The thing that must be understood.......you can't get the same value for quality with less value in part quality.....SBR rails may be cheap, but they add problems in increased cross sectional depth when the various slides cross one another as in the ZY junction........making a big router means more cost anyway you look at it.

For example, an SBR 15mm round rail and bearing block stand 39.5 above the surface, whereas a HIWIN 20mm rail and block stand only 30mm.......cost more but are heaps more precision and longer lasting.

You still have to allow for the 1605 ball screw nut mounting block height of 40mm to clear the inside face between the rails even if you have to sink the block into the back face a bit to get the screw to clear the frame faces.
Ian.

[jerseyguy1996]

Well I've done a lot of work on my Z-Axis so I figured that I would check in. The Z Axis plates are 1/4 inch steel plate. I did a couple of different iterations including one using the linear slides supplied by CNCRouterParts. Trying to make the ball screw nut fit was going to be messy so I switched back to the SBR rail. Certainly part of the reason is because the SBR rails are relatively cheap and I can buy a bundle of parts off of ebay and then just design around them...which is what I have done here. I was thinking I could use the sketch up export scaled 1:1 as a drill template.

Z Axis Front by jg1996business, on Flickr

Z Axis Back by jg1996business, on Flickr

and this would be the template for the Z axis plate:

Z Plate Template by jg1996business, on Flickr

[handlewanker]

Hi, just my opinion as I'm also building a router and came on the Z axis stick out from the gantry problem along the way.

In photo 2 you show the moving part of the Z axis with the rails mounted on it.......make the saddle longer as you can go high without affecting anything and mount the rails on the face and have the bearing blocks on the moving plate that holds the spindle.

When the spindle is at the bottom of it's travel you won't have so much thin plate hanging down to flex and cause leverage to the gantry.

I would recommend you go to Hiwin rails and bearings as you win on the height, but you might need to sink the ball nut housing into the plate as it's quite high compared to the height of the Hiwin rails and bearing blocks.....using SBR round rails does give you more clearance for the ball nut but at the cost of more stick out from the face of the gantry X axis.

Here's a photo of a full size model I built for a 3025 bridge mill/router to get the proportions right prior to the actual build and shows the Z axis arrangement I will be using.

The clearance under the lowest part of the Z axis when it's down is going to be 130mm as the photo shows with the dummy spindle on a block to indicate the amount of overhang you can get with HIWIN rails etc.

In this design the rails are on the XZ axis saddle and the bearing blocks on the spindle platform, shown at it's lowest travel point.

Even for a small model like this the proportions still scale to give minimum hang down of the slideways.
Ian.

[jerseyguy1996]

Hi, just my opinion as I'm also building a router and came on the Z axis stick out from the gantry problem along the way.

In photo 2 you show the moving part of the Z axis with the rails mounted on it.......

This is what I'm thinking:

I will probably never need that much depth in a machine....as I will probably spend almost all of my time routing sheet materials no more than 3/4 inch thick. I'm building it with all of that Z travel as a "just-in-case I need it" kind of thing. I figure that under normal circumstances I can just lift the work piece up to the router head from the bottom. That way the router doesn't have to extend down very far and will mostly be routing from the level of the gantry cross beam. If there does end up being a situation where I need the additional travel I can lower the work surface down however much I need. Think layers of MDF stacked up with the work piece on top so that the router vertical travel can be minimized for whatever I am working on. I am trying to make this a machine that can be built using nothing more than some clamps, a good drill or drill press, and some screws. Having to mill out a spot for the ball screw nut to go doesn't really fit that goal ( I don't have a mill and I'm not going to pay what most machine shops would want to mill the Z plate and router mounting plate for me) which is why I am kind of stuck using the SBR rails. A 2 inch by 6 inch steel tube is basically a box beam which should have very little twist unless I am putting an incredibly heavy object on the end of the router mount or if I am really pushing hard through some material. That is my thoughts at least.....certainly tell me if I am thinking about this wrong.

[Nickhofen]

Nice thread, thanks for sharing.

[wizard]

It's been a few days so I wanted to post some updates to the design. I'm mostly trying to figure out the gantry. Weight may be an issue but I figure I will cross that bridge when I start working out how to move things around. Right now I just want to work on putting together a rigid structure.

First off I changed the gantry supports to 1/4" walls so that I can drill and tap right into them. I guess a question could be, would I get enough strength tapping into 1/8" walls?

In the machine tool building world, it is commonly accepted to have at least one diameter thickness in tapped material of the same type as the bolt. So if you use 1/4" bolts you would want at least 1/4" thick material. On a web site, I can't remember where now, I guy did some FEA work on gun barrel threads in this case he found five threads of engagement are pretty good with results rolling off fast to seven threads. This more or less confirms the rule of thumb. For example 1/2-13 pretty much implies 1/2" thick material in both cases. I suspect that this falls apart when you start to deal with really large diameter threads but that isn't important here.

I say at least above because there are some other considerations. For example the tapped holes should be counter sunkenked, chamfered to in some way dealt with to keep the first thread from pulling up proud of the rails mounting surface. This can be a problem as you end up loosing at least a half a thread. Another issue is aluminum which does not hold screws well compared to steel.

Also don't dismiss the possibility of through holes with a tapped bar behind the wall of the tubing. This can be a big advantage in small tubing sections where you can't get your hands in there. It does require a lot more accurate drilling. You can also go the nut route with access holes lined up with the through mounting holes.

Second I changed the horizontal beam on the gantry from plate steel to a 2X4X1/8 inch steel tube. The linear rails will be screwed right into tapped holes. Again, the 1/8" thickness question comes up in terms of tapping the holes. I figure with M6 1mm thread screws, I should get 3 threads in 1/8 inch walled tube (.125" = 3.175mm). Am I working that out right?

That is a fairly good sized tube though I might go with 4" square myself. As for wall thickness I'm going to suggest that it is on the thin side for tapping.

Finally I changed the Z Axis risers from plate steel to 1X4X1/8 inch tube.

Z-axis? Do you mean the saddle that the spindle attaches to? For that I'd go with a thick plate, probably aluminum.

If you mean the supports for the gantry beam (Y-axis), you probably aren't doing too bad there though I'd prefer something 2x4.

Hopefully this won't be just a sketchup project for too long. I do want to make sure everything is perfect before I start buying materials though.

Lesson for wise, no machine is perfect. In most cases they are works in progress.

[jerseyguy1996]

@Nickhofen - No Problem! Thanks for reading it!

@wizard - Thanks for the great information! I've done quite a bit more work and changed up some stuff. Check out post #34 for some updates on the Z Axis. Like you said at the end of your post, I think I am ready to go ahead and buy some materials to at least build the Z - Axis. That seems like it will be the most complicated part so if I can build that I will feel quite a bit more comfortable purchasing the rest of the materials.

[handlewanker]

Hi, don't be put off by the build or your lack of engineering skills or equipment......we all had to learn at some point and a hands on learning curve is the shortest way to gain knowledge.....far more powerful than book reading.

I noticed that you changed the Z axis design to have the rails on the back of the spindle mounting plate.......it has drawbacks in that mode.

BTW, if you use a thick aluminium plate for the spindle saddle and drill the cavity right through to clear where the ball nut housing is to go.....clearance is not a problem as long as it is clear of the sides, then you can attach the ball nut from the other side with a secondary plate bolted to the face where the spindle attaches.

This gives you an effect of having a ball nut housing with a flange to bolt from the back face.

On the Z axis that I have on the model I get 130mm from the X axis bridge face to the spindle centre.

The biggest problem I came across was where the X axis and Z axis ball nut housings cross one another....that gives you 40 + 40mm of height to accommodate.

BTW, I would aim to get as much height travel in the Z as possible as this will allow you to mount drills and get the spindle high enough to remove them as well as digging into a job.

I will be catering for a possible parallel spindle body, approx. 65mm diam. and this will in itself allow a further up and down positioning of the tools if/when the job warrants it.

I think it would be wiser to build the machine from the ground up and add the parts as you go, as there may be some modifications along the way as the build progresses.........you get a better idea of what is wanted when you see the machine in the flesh and can get your hands on it........ one big reason why I went to a complete full size model build in polystyrene foam before committing to the steel buying......I changed a few design points as the model took shape.

With the build you need to decide on what you are specifically going to use it for otherwise the tendency is to make a sledge hammer to crack a walnut.
Ian.