I am sorry if this has been asked a thousand times, but I have spent several hours trying to navigate the site and really can't find an answer to my questions. I am thinking of building a new machine. I built my last one ten years ago when all this excellent lightweight modular extrusion didn't exist. There are several designs that I would be excited to try. The only issue is I can't seem to get a good understanding of the Z axis travel and clearance on the existing circulating plans. I need a work envelope of 18" x 24" x minimum Z clearance of 5" and travel of 10". On one regular job I work with a 6" long reach bull nose bit that extends roughly 5" from the bottom of the spindle collet. I need to clear a 3.5" lip of the work and reach down to cut through the bottom of the blank and 1/4" into the hold down fixture, which is itself 3/4" thick. So the gantry itself has to clear the 3.5" blank and the .75" hold down to be no less than 4.25" above a spoil board. Of course that is cutting things way too close, so really we need 5" of clearance for the gantry (or fixed bottom point on the z axis, whichever is lower). The Z axis has to retract the spindle so that the bottom of the 5" of exposed bit will be level with the lowest fixed point on the gantry. Of course I don't always use a 6" XL bit. The spindle has to also be able to travel down to the spoil board. The travel therefore has to be close to 10". I needed all that travel to retract the 5" bit, but if I have a more typical 2" LOC (or even a small bit with just 1/4" of LOC) the spindle has to be able to travel down almost to the spoil board itself. So, 5" up + 5" down means 10" of travel and 5" of clearance above the spoil board. Whew!!! Sorry if that is long winded, but I've had several arguments lately with people who insist, INSIST, that what I need is a ShapeOko with 3" travel and about 1.5" of actual workable Z. The OX seems like the best option because the spindle is mounted to an extrusion that theoretically can be any length. But I can't find what the existing plans give you for clearance and travel. Can anyone tell me off their actual built machine? What is the clearance from the lowest fixed point on the gantry to the spoilboard? The riser plates for the gantry look very low. With a spoil board mounted between the Y axis extrusions as called for in the plans, it doesn't look like the gantry would clear my workpiece and its fixture. I see that there are plans where these plates have been extended 50mm, but no where do I see what that has been extended to. If the clearance was 50mm to begin with and these "extended" plates give you 100mm, that still isn't even 4". With a lot of time and effort, I can probably design my own risers and Z, but I don't want to re-invent the wheel if it has already been done. Can anyone point to a design or build that would have the Z travel and clearance I am looking for? Thanks!
What you seek is possible but there are two questions. What material are you cutting and what size router are you using?
Maple, mostly. Occasionally other hardwoods like Cherry and Walnut. Here is a picture of my current setup:
Concept shown below. It's actually a fairly simple design using stock parts. Design is based on two platform levels: a pit with 5" clearance and a drop in platform/work surface which gives 2.6" clearance. It was designed to keep things low and lean which will provide for best stiffness. Work area is approx. 21" x 32". The larger/heavier router you use hasn't been fully integrated but that's a fairly minor change. Stock gantry plates shown on Y and Z axes are probably not the best option. You'll probably want to cut something a little more robust. Tried to attach the Sketchup file but the system is having issues right now. Will try again tomorrow.
what brand is your current CNC? Someone was trying to sell me one that looked exactly the same but couldn't give me a brand.
@Rick 2.0 I like your idea of having the C-beam lifted up by the 20/40 extrusion, but leaving a well for the actual bit travel. Ingenious! Do you think that the Y-axis gantry plates riding inside the C-beam on mini wheels is enough rigidity, or should I make a gantry plate that will allow full size wheels above/below the C-beam as well as inside to "grab" it better (like the OX, just larger for wheels below as well)? Perhaps even "box" it in with an inside plate connected to the top/bottom full sized wheels. As for the z-axis, the OX plates with their gazillion wheels seems really sturdy, but I've never seen one in person. Do you think I would need more than that? @McFuddlebutt are you asking me or Rick 2.0? My CNC is something I made for myself about 10 years ago from a collection of parts and ideas floating around the net back then. The cold rolled steel and aluminum bearing blocks with ABEC wheels were from a company called cncrouterparts.com. The idea for a plywood torsion box bottom/sides came from Momus CNC. I don't know if cncrouterparts.com sell those bearing anymore. I think they moved on to selling complete machines with Hiwin bearings and ballscrews.
Actually on the Y-axis plates I was thinking more along the lines of the Sphinx build where the smaller interior wheels were left but a row of larger wheels was added above the C-Beam. Using the same basic layout of the Double Wide C-Beam Gantry Plate, I would bring the front edge an inch farther forward and take the top edge up an inch. This allows the front wheels to be moved farther forward to better balance the router weight and two large wheels to be added above them to help support the router weight. This gives 4 wheels supporting the weight of the router and the two center wheels preventing uplift. On the Z-axis, I would make the plate about an inch taller and move the center wheels down to the bottom. The latter move puts more wheels to grip where the majority of the force exists. Increasing the plate height reduces the forces on the wheels caused by the eccentric load of the router. Photos attached (click to enlarge). Still having issues posting drawing files. Current photos also show the vertical members extended 50mm. At the Z-axis this allows a 6" plate 6" of motion. Increasing the sides provides a little better leverage on the tie back plate on the Z-axis.
Very Interesting. What is the resulting lengths of the c-beam gantry legs and the z-axis vertical c-beam?
@Rick 2.0 Thank you a million! Inspired by your design of using a second stabilizing gantry, I came up with this. I don't know why I am fixated on having the Z Axis move as a whole rather than just move a router mounting plate. My assumption is that gives me more rigidity, but that just comes from thin air. I saw the use of the C-Rail XL Gantry where they backed them together to form a two axis plate. Taking that idea, I modeled a mini back-to-back gantry that would use standard V-Rail on 20/40 and a set of inside wheels for the inside of the C-Rail as the second stabilizing gantry. All told, I lose 212mm of Y-Axis travel with my plates and 125mm of X-axis travel for the plates. The key measurement however is 187.5mm of lost Z-Axis. So If I want 125mm of clearance, the lower edge of the lower gantry plate has to be that height off the spoilboard. The top edge of the top gantry plate has to be ~300mm above the spoilboard. I will need 187.5mm + my desired travel for my Z-Axis. In my case that puts me at 300-350mm travel. Not unmanageable, I think. Does anyone see anything glaringly wrong with my design/calculations? Please save myself from myself before I start cutting plates! First picture is without the Z axis, second has a stock 500mm axis from the drawings library. Obviously it is too big, but I will have to construct my own at 350mm.
You're better off keeping the stronger structure closer to the cut. Right now, the 2040 is being supported by the 4080. It should be the other way around, unless you plan on removing it.
This thread is helping me with my new build, so thank you for that. I might be mistaken in this, but I thought that moving the z axis plate and router would be easier that having that AND the beam, the motor, and the threaded rod being moved up and down. All that weight seemed like it might accelerate wear on the nut block material. Interestingly, the cavity in the waste board idea, helps with a way to do work on pieces that the CNC router might be sat ON. Might be a nice way to add some decoration to a 7 foot harvest table we just finished this summer.
@Kevon Ritter I'm not sure by what you mean, "Right now, the 2040 is being supported by the 4080". Do you mean the X-Axis? I don't know that the order of the cross beams matters, but the 2040 is holding no weight. It is only there as an anti-racking support. It interfaces with the Z-Axis only by two mini wheels that ride in the inside channel of the C-Beam. The weight of the Z-Axis is entirely on the C-Beam. The idea for the second beam is just to keep the torque from cutting from pushing the Z-Axis out of plumb.
@mikework You are right about wear. The issue is that with a fixed actuator for the Z-axis the most you can lower into the work would be the distance from the bottom of your spindle mount to the tip of the bit. The spindle mount would travel to the bottom of the fixed Z, and that would be it. It wouldn't matter how long the actuator was above, you are limited in your travel down. I am trying to maximize Z travel (within reason) even to the expense of a nut wearing out. I believe that the OX and its sibling the R7 both do it this way, so I assume it is tried and true.
I agree with you for two reasons. 1) No matter how you set it up, your leverage point will remain the same. But one method attaches more weight to the Z plate, while the other attaches that weight to the X gantry. I would rather have that extra weight on the X where it can be distributed better (wider plate, more wheels). 2) Let's say you want to be able to machine 3" deep. Your gantry has to clear the material. Your end mill has to be able to reach 3" and clear the material. Your required travel is now 6". Having the Z axis in one position allows your end mill to travel above the gantry bottom. This means that your gantry only needs to be able to clear that 3" material. If you have the entire Z assembly moving, you have to raise the gantry an additional 3". You are now compromising your rigidity way more than you need to. You are never going to be cutting above your gantry so having that extra travel ABOVE your work area will not create any negative effects. So the pictures in post #10 are very wrong, while post #6 is correct. That's my perspective. I can't really find any counter argument to that, but please say something if you think my thinking is flawed in any way.
You say it holds no weight, but weight is not the enemy. The enemy are the forces created during cutting. once you apply a load pushing or pulling along the Y direction, you'll be pivoting on the 4080 and pushing on the 2040. My point is that you should build around the main structure (4080), not the other way around. As I mentioned in my previous post, please say something if you think my thinking is flawed.
@Kevon Ritter I'm not sure if you really want my opinion, but I think I addressed your points earlier in this thread. But taking you at face value, here goes: 1) Yes, the issue is torque from an extended Z-axis. However, unless my engineering and physics is failing me there is no difference between the setup you say is "right" and the one you say is "very wrong". Both have two gantries. Both carry the weight on the C-Beam gantry and use the 20/40 as a brace against twist. Both tie the C-Beam and the 20/40 together and both would have a rotational axis somewhere between the two. It doesn't matter their orientation. Both would "push" on one beam while "pulling" on the other. Regardless of which orientation, the issue correctly identified is the deflection of the two beams: C-Beam and 20/40. The C-Beam alone is very rigid. It is unlikely to bow in response to cutting forces. The bit would likely break first. However, in a "belt and suspenders" way @Rick 2.0 came up with the ingenious idea to add a second 20/40. For the purposes of resisting this force, the C-Beam is two 20/40 beams with a webbing to hold them in place. @Rick 2.0 has added a third. With the axis of rotation somewhere between the C-Beam and 20/40 beams when the Y advances one will bow one way and the other in the opposite way. Doesn't matter which way, but we get the cumulative effect of both beams resisting a bowing deflection. The other force to contend with is the rotation along the plane of the X-Axis. Honestly, I think this is harder to account for. The stock C-Beam gantry plates gripping from above and below provide a lot of resistance to twist. An additional four wheels 60mm below will help even further. Will it be enough? I don't know. But which are above and which are below will make no difference. Again, the axis of rotation is somewhere between. The force will have to be countered by the wheels in the V slots. 2) As I mentioned in my first post, the point of this design is to create something that can reach 5" into a work piece requiring 10" of travel. However, it also has to cut when I am engraving shell or making lettering. I understand that is a challenge, and maybe the C-Beam/V-Rail platform isn't the right tool for the job. But it makes no sense to change the design because the design goals are difficult. From the bottom of my router mount to the bottom of the collet is 2.75" My smallest bit is .625" long from stop collar down. I never use it to through cut material, but in theory my MAXIMUM reach with a small bit and a fixed Z-Axis is 3.375". I want 5". Therefore whether the Derlin nut will wear out, or I'll need a larger stepper to lift the weight of the whole axis, is precisely the design challenge. Simply limiting myself to 3.375" because it is easier does not meet my design goals. Now, perhaps with the use of an insert table I can cut using the really small bits (lettering bits, shell cutting bits, etc.) and take it out to work with bits with 2.5" or 5" of exposed bit. However, I am trying to find a solution that does not rely on that. My dream is not to be limited to what bit I'm using, and not having to swap out tables. I do that already with a series of torsion boxes that lift the work for small bits. I like having threaded inserts in the table at known machine coordinates for fixtures. Every time I have to swap tables I have to calibrate the location of the table vs. the machine. It is inefficient and introduces error. Therefore I am looking for a solution where the gantry has 5" of clearance but I can still lower the spindle down very close to the table so I can use everything from a 6" long XL 1/2" bull nose bit to a .625 long .020" fishtail shell cutting bit on the same machine. Again, maybe that is not doable with these components. I note that the OX is built with the Z-Axis as I propose, and that has been a popular and robust design. That gives me confidence to try that design. 3) Quite honestly, my biggest concern with the @Rick 2.0 design that I am incorporating is racking between the two gantries. Perhaps I will widen the 20/40 gantry to be the same size horizontally as the C-Beam gantry and hope that is enough. It is a lot of money to experiment with these designs, so I have to consider carefully if I am ready to move forward. I think the design tweak proposed @Rick 2.0 should work. Now I have to decide whether to pull the trigger.
For what you described as a goal, I firmly believe the Rick's design is a much better option,which is actually what I was trying to reinforce in ALL of my comments. With a sliding C-Beam for Z, you have to raise the X axis C-Beam higher, which increases the sensitivity to flex. The design that Rick posted did not have this issue. You say you want to use different length end mills. We all do. In the end, you're either going to have to lower the spindle in int's mount or raise the table. There is no one size fits all (which you even said). You completely missed my entire point, but I'm not going to go any further than this: You have to pick your compromises, but I personally would not be picking the ones you chose.
@Kevon Ritter, I have tried very hard to respond to all your various points in good faith. I seriously doubt I "missed your entire point". Nonetheless, thank you for your input.
