584,808 active members*
4,857 visitors online*
Register for free
Login
Page 1 of 2 12
Results 1 to 20 of 28
  1. #1
    Join Date
    Dec 2013
    Posts
    6

    CNC router plan eval please?

    Hi people!

    I've been lurking and reading for a month or so now and finally made an account so I could ask for insight into my planned CNC router design. I'd like to make sure I'm not overlooking something stupid, or that there aren't any obvious improvements I could make before I begin sourcing parts.

    I'll be using Aluminum extrusion (8020 series 15 most likely) for the structure. For the gantry uprights I'll hopefully be able to make them from 3/8" 6061 plate aluminum. For linear guidance and motion I'd like to use hiwin/nsk style rails. I'll likely start by sourcing these parts since they will determine the measurements for the 8020 structure they sit upon.

    Electronics will be last.

    I'm hoping, for now, that a few people could glance at the images I'm including to see the basic idea of the build. The measurements in the last image are just rough. The Z-Axis is not my actual plan, it's just a stand-in for now. I'll be mounting a Bosch PR20EVSK router as the spindle.

    Are there any obvious improvements I could make (moving locations of ball screws, or rethinking how something is configured, etc...)?

    My only real ambition is to be able to route hardwoods with accuracy and precision. I don't have numbers for how accurate or precise. Let's say that I have ambitions to make solid body electric guitars and need to make snug fitting neck pockets reliably. I also have ambition to make signage for profit on the weekends if I find a market for it in my area.

    If it so happens that I create a machine that is capable to routing aluminum, that would be a sweet perk, but I'm not shooting for that right now.

    Thanks!

    Attachment 215444

    Attachment 215446

    Attachment 215448

    Attachment 215450

    Attachment 215452

  2. #2
    Join Date
    Nov 2010
    Posts
    520
    I bought my linear bearings and rails from Automation Overstock. Great people to deal with. Here is a link to some linear bearings that I used.
    automation products from Automation Overstock: LGW15CAZ0H, 15mm Bearing Block, Flange, Hiwin LG 15 Blocks

  3. #3
    Join Date
    Feb 2005
    Posts
    829
    That machine would be able to route aluminum. The only issue I can see with your design is the gantry sides, 3/8th may be a bit thin and wobbly for that tall of a gantry side.

  4. #4
    Join Date
    Aug 2011
    Posts
    388
    Nice clean design. Agreed with nlancaster on the gantry legs--they will be flexy in the parallelogram mode. Adding some lateral reinforcements to form a "T" section (viewed from above) would fix that.

    Another improvement for more stiffness is to increase the vertical spacing of the rails on the gantry. With a Z-clearance of ~8", the vertical spacing should also be around 8" (although for profile rail only cutting wood, ~5" would probably be enough).

    With 2 screws driving the long axis, usually the screws are placed near the rails, near table height; that keeps the forces more in line with the cutter. That also eliminates the need for the cross-member underneath the table, which allows the table's underside bracing to be made more rigid. Not that your design wouldn't work--it would, just a bit more flexy than some alternatives, but for wood it may not matter.
    David Malicky

  5. #5
    Join Date
    Dec 2013
    Posts
    6
    Quote Originally Posted by RicknBeachcrest View Post
    I bought my linear bearings and rails from Automation Overstock. Great people to deal with. Here is a link to some linear bearings that I used.
    automation products from Automation Overstock: LGW15CAZ0H, 15mm Bearing Block, Flange, Hiwin LG 15 Blocks
    Thanks RicknBeachcrest. I've been looking at that site, as well as automation4less which also seems to be well regarded. If I can't find something worthwhile on eBay, I'll likely end up using one of them.



    Quote Originally Posted by dmalicky View Post
    Nice clean design. Agreed with nlancaster on the gantry legs--they will be flexy in the parallelogram mode. Adding some lateral reinforcements to form a "T" section (viewed from above) would fix that.

    Another improvement for more stiffness is to increase the vertical spacing of the rails on the gantry. With a Z-clearance of ~8", the vertical spacing should also be around 8" (although for profile rail only cutting wood, ~5" would probably be enough).

    With 2 screws driving the long axis, usually the screws are placed near the rails, near table height; that keeps the forces more in line with the cutter. That also eliminates the need for the cross-member underneath the table, which allows the table's underside bracing to be made more rigid. Not that your design wouldn't work--it would, just a bit more flexy than some alternatives, but for wood it may not matter.
    I see what you guys mean about the flexing on the gantry sides. I could use some u-channel to stiffen them up. I'll look into this and post some design updates when I've done some changes.

    I'm curious about your recommendation for the spacing of the Y-axis rails. It makes sense that the higher rail and the cutting bit should be equidistant from the lower rail, in order to apply equal and opposite rotational force about the axis that the cutting force would be working on (the lower rail). I'm guessing that this is the logic behind this design? I like it, but it does add quite a lot of height to the overall design. I wonder if anyone could help with the math to make this decision. As you say, 5" would probably be enough, but I wonder where the line is... =)

    I also see what you're saying about the X-axis ball screws. I originally had it designed the way you suggest, but moved them under the table to cut down on the outside dimensions and to try to get some parts out of the "dust zone". Maybe this was done at the expensive of rigidity. I could move the X-axis rails down on the sides and place the ball screws where the rails are now. I'd probably have them run on the outside of the gantry uprights.

    While I'm looking at doing this, not much benefit would I be getting out of having two screws run the X-axis vs only one centered under the table? I see a lot of designs that only use the one, so maybe I should go that way. It would be cheaper and simpler. Is it worth the extra expense of a 3ft screw, and another motor, or would that money be better spent elsewhere?

    Thanks for the input so far! Updates to come...

  6. #6
    Join Date
    Jan 2008
    Posts
    1523
    I built a single centre screw machine. It allows movement at each end of the gantry. They are suitable for moving table machines or machines where there is no real load on the tool (like drag knife etc).

    I would never build a single centre screw machine again for a moving gantry router. Dual screws also allows you to support the long x axis beams along their whole length, preventing sag under the weight of the gantry in the middle.
    7xCNC.com - CNC info for the minilathe (7x10, 7x12, 7x14, 7x16)

  7. #7
    Join Date
    May 2005
    Posts
    3920

    Nice start on your design.

    First let me say that if you intend to do musical instrument parts on this machine you will need better accuracy and control than most routers built for wood working. Routing a sign is trivial compared to musical instrument parts. In fact luthiers regularly work to one thousands of an inch. As such you need a machine built a bit better than an average sign making router.

    You have already gotten good comments but I will run though some of my own.
    1. It has already been mentioned that the uprights will be too flimsy if made out of 3/8" stock. This is very true and I'd suggest going either with stout extrusions or tubular steel. Flexing will not only impact accuracy but also achievable surface finish.
    2. In the same manner I'd stongly suggest a beefier extrusion for the Y axis. Further how the uprights nad the Y axis beam are put together is important. Think Gussets or reinforcing angle brackets.
    3. I'm trying to get a handle on the size of the extrusions to be used but they look thin in your drawing for the table. Think about it as if you where building a building, you need to start with a good foundation. In this case that is the rails that supports your gantry mechanism. These should be robust and further should have plenty of lateral members to further stiffen the structure. Think about the appearance of a ladder here.
    4. Don't forget that you need to support this structure some how. Long legs / short legs, some how the gantry mechanism needs to be free to move. It doesn't look like you have any provision for such in your drawings.
    5. By the way the member running underneath the table (crosshead) that supports your uprights needs to be stiff also or you fall into the same issues with vibration or flex.
    6. Others have already pointed out the probable need to spread the linear bearings out a little bit on the Y. Along with this is a similar problem on the X, the further apart the saddles (bearings) are on the linear rails the better they will handle twisting motions. The most obvious problem here is that adding spacing to the bearings results in a longer table dimensions overall to get the same X travel. It is a trade off you will have to work with. However I'd be more concerned about stiffness here then at the Y axis bearings as you have more leverage workign against you.


    Lots of stuff. If I read your drawing right you are likely right on the edge of practical size for a moving table design. However if room isn't a big issue you might want to think about it. You should be able to realize a stiffer design with fewer challenges getting it all to work reliably. The big issue with a moving table design in this size of a machine is that you need room around the machine for it to operate or a really big frame, depending upon which design course you follow.

    In the end you want the stiffest machine you can realize within your budget as that will directly impact the quality of the work the machine can do and the amount of post machining work to be done. You might also want to consider servos to get the speeds that will be ideal here and at the same time maintain good precision. You want to do all of this to get a machine that can do work to a fairly high degree of precision which honestly is not what a lot of router machines do.

  8. #8
    Join Date
    Aug 2011
    Posts
    388
    More good advice above.

    Quote Originally Posted by bluntend View Post
    I'm curious about your recommendation for the spacing of the Y-axis rails. It makes sense that the higher rail and the cutting bit should be equidistant from the lower rail, in order to apply equal and opposite rotational force about the axis that the cutting force would be working on (the lower rail). I'm guessing that this is the logic behind this design? I like it, but it does add quite a lot of height to the overall design. I wonder if anyone could help with the math to make this decision. As you say, 5" would probably be enough, but I wonder where the line is... =)
    Yes, you're seeing the forces and mechanics right on. Good question on the math -- I was curious so crunched the numbers. Short answer: if using profile rail and cutting wood or even light cuts in aluminum, a wide spacing probably isn't needed, if the other parts of the machine are stiff enough as well.

    Long answer (and it would be great if someone could check my math):

    Here's a Free Body Diagram (FBD) of the gantry-car, z-car, spindle, and cutter:
    Attachment 215588

    Sum of Forces in x-direction: Fa - Fb + Fc = 0
    Sum of Moments about point B: Fa * AB = Fc * BC
    Solving:
    Fa = Fc * BC / AB
    Fb = Fc + Fa = Fc * (1 + BC/AB)

    For some numbers, let's assume Fc = 100 lb:
    Example 1: AB = 4”, BC = 8”
    --> Fa = 200 lb, Fb = 300 lb
    Example 2: AB = 8”, BC = 8”
    --> Fa = 100 lb, Fb = 200 lb

    In words, the lower rail is the fulcrum, and it gets the sum of both the cutter load and the upper rail load. The upper rail load is equal to the cutter load times the ratio BC/AB. That is, the bigger of the ratio BC/AB, the more leverage the cutter force has on the gantry.


    Now to look at the deflections resulting from those forces, here's a simplified stick model of those same parts:
    Attachment 215590

    This shows the more important disad of a narrow rail spacing: bearing deflections get amplified a lot at the cutter. Any bearing slop gets amplified, too.
    Point O is initially unknown, as it depends on the deflections Da and Db. We can get OB with some geometry:
    OA + OB = AB
    Similar triangles: OA/Da = OB/Db
    Solving: OB = AB * Db / (Da + Db)
    Now we can get the deflection at the cutter:
    Similar triangles: OB/Db = (OB + BC)/Dc
    Solving: Dc = Db * (1 + BC/OB) / OB = Db * [ 1 + BC / (AB * (Db / (Da + Db) ) ) ]
    Substituting: Dc = (Fb/k) * [ 1 + BC / (AB * ((Fb/k) / ((Fa/k) + (Fb/k)) ) ) ]
    That's rather long, so attached is an Excel file to do the calculations.

    k is the stiffness of the pair of upper bearings (or lower bearings). A single profile bearing (Hiwin LG20, Z0) has a k of 26 kg/micron, or about 1,450,000 lb/in. Two bearings would be about 3M lb/in. But mounting to an alum surface with a slot will soften that some. Let's say 2M lb/in as a pure guess.

    Continuing Example 1 (AB = BC/2):
    Da = 200/2M = .0001” and Db = 300/2M = .00015”
    Dc = 0.00065” (amplified 4x over Db)
    So Stiffness at Cutter = Fc/Dc = 150k lb/in (only accounts for flex in bearings)
    As 1 of about 10 components, 150k lb/in is targeted towards an aggregate cutter stiffness of ~15k lb/in, which is plenty for wood and probably pretty good for aluminum. (For this to work, all the components need an effective-stiffness-at-cutter of ~150k lb/in or more.)

    If k = 200k lb/in (from some other calcs, I think this is the approx stiffness of a pair of 608 skate-bearings with point loads on their outer race.)
    Bearing stiffness at cutter = 15k lb/in, targeted to an aggregate stiffness of 1.5k lb/in -- ok for wood but not great for heavy cuts.

    For Example 2:
    k = 2M lb/in gives a bearing-stiffness-at-cutter of 400k lb/in --> aggregate stiffness target of 40k lb/in -- probably good for steel.
    k = 200k lb/in gives a bearing-stiffness-at-cutter of 40k lb/in --> aggregate stiffness target of 4k lb/in -- good for wood, light cuts in alum.
    So increasing the spacing AB to be equal to BC improves stiffness by 2.7x.

    The main disads to a tall AB are
    - A single gantry tube may become heavy. But, a single large tube has fabulous torsional and bending stiffness, and so it's usually needed anyway for a high aggregate stiffness.
    - The lateral spacing of the bearing-pair must also increase to maintain stiffness under lateral cutter loads. This is less critical for profile bearings since they both push and pull. It's very important for skate bearings since they lift-off under tension. So, a common compromise is to make AB = BC = the lateral bearing spacing. For lateral loads, that gives an aspect ratio of 1:1 for the 4 bearings, as seen in most commercial routers.

    These same analyses also apply to the vertical spacing of the Z bearings.

    So, yes, a narrow spacing can work, in and of itself. But there's substantial gain and only a few disads with a taller spacing. The next thing to check is whether that small gantry tube is sufficient, as Wizard pointed out.

    (EDIT: Bearing Stiffness, k, is saved as 200k lb/in in the xls, which is my estimate for 2 skate bearings. 2 profile rails/bearings would be about 2M lb/in or more.)
    David Malicky

  9. #9
    Join Date
    May 2005
    Posts
    3920
    Quote Originally Posted by dmalicky;

    So, yes, a narrow spacing can work, in and of itself. But there's substantial gain and only a few disads with a taller spacing. The next thing to check is whether that small gantry tube is sufficient, as Wizard pointed out.
    Thanks for the spreadsheet, I imported it into Numbers with only a few format errors. I'm not really qualified to check the math but in playing around with it initially I'm surprised to find that you get deflection of some sort (>0.002") for almost any rational spacing of the bearings. At least for the starting values you had. When I get a chance I will play with it some more.

    This reminds me of my days in my youth (fresh out of high school), I was working on a huge boring mill with some mechanics and was a bit fascinated with deflection charts on the side of the machine for the quill and spindle. The charts where done in brass no less. This is way back in time before personal computers where a normality and all I could imagine is an engineer labor in for days with a slide rule to calculate the points required to generate the graph. In my youth that was boring desk work. It took awhile (years really)to realize that sometimes that work can be very interesting in its own right.

    In any event charts like that, riveted to the side of this massive machine, burnt into my mind that everything bends or deflects some under load.

  10. #10
    Join Date
    Aug 2011
    Posts
    388
    Quote Originally Posted by wizard View Post
    Thanks for the spreadsheet, I imported it into Numbers with only a few format errors. I'm not really qualified to check the math but in playing around with it initially I'm surprised to find that you get deflection of some sort (>0.002") for almost any rational spacing of the bearings. At least for the starting values you had. When I get a chance I will play with it some more.

    This reminds me of my days in my youth (fresh out of high school), I was working on a huge boring mill with some mechanics and was a bit fascinated with deflection charts on the side of the machine for the quill and spindle. The charts where done in brass no less. This is way back in time before personal computers where a normality and all I could imagine is an engineer labor in for days with a slide rule to calculate the points required to generate the graph. In my youth that was boring desk work. It took awhile (years really)to realize that sometimes that work can be very interesting in its own right.

    In any event charts like that, riveted to the side of this massive machine, burnt into my mind that everything bends or deflects some under load.
    You're welcome--glad it's interesting. I just noticed I saved the xls with a k of 200k lb/in (estimate for skate bearings). A k of 2M pulls the deflection down nicely. But yes, everything deflects! I like the brass chart.
    David Malicky

  11. #11
    Join Date
    Jun 2013
    Posts
    1041
    I see mention of automation overstock and automation4less. I don't know if its common knowledge or not but they are the same business with multiple websites. I live about 40 minutes from there warehouse and when I place orders I just shoot over and pick things up. They are really nice and helpful people. I needed some t-slot extrusions about a month ago. I called and told them what I wanted and even though they were busy they told me to come on and they would take care of getting it cut. If you've ever been quoted on shipping for extrusions you know the savings by just picking it up locally.

    Ben

  12. #12
    Join Date
    Dec 2013
    Posts
    6
    Holy crap! I've spent a lot of time on forums in my lifetime, with subjects crossing the spectrum from guitars, to photography, to computers, to target shooting, to wood stoves... and I have NEVER gotten a response like dmalicky's! Extraordinary. I feel like I owe you money or something. =)

    To your point of spacing the Y-axis rails, consider it done. I've actually lowered the clearance from about 6" to about 4" for the Z-axis. I realized, I'm never going to need a depth of 6" (uh oh, have I jinxed myself now?). I also raised the upper bearing rail above the ball screw. I would post new drawings, but I have to go through wizard's post and rethink some more things.


    wizard's post is also very illuminating, and brings up a lot of good points. I'll comment on his points and answer questions in the order he presented them:

    1. Right now I was thinking of reinofrcing the uprights with Aluminum u-channel. I Could probably go with steel u-channel or square tube instead, but I'll have to look into cost and workability with what I have (not much!). I don't think I would have the ability to construct them completely from tube steel. I'll have to think about it.

    2. In my drawings from the original post, I have a 3" x 3" extruded aluminum (80/20 fractional series 15 3030). My new version is using an additional 1.5" x 3" piece above the ball screw that the top bearing rail is mounted to. This separates the rails to a distance of 5.25", and keep in mind that I've lowered the Y axis so that the bit will have about 4" of clearance over the table.

    3. As mentioned I'll be using 80/20 Fractional Series 15. The sides of the X-axis are 1.5" x 3" (model 1530), the cross members are 1.5" x 1.5" (model 1515). As mentioned above, the Y-axis uses a 3x3 (3030) and now also another 1530. They also make a 1.5" x 4" (1540) and a 3" x 6" (3060). Currently the cross members are spaced 11.5" on center. I could add another 3 cross members between the current ones. Part of the decisions here are financial, I can only dedicate so much money, and i'm already about 3x more than my plan when I first thought of making a CNC router. The original plan was to just make it from wood. =)

    4. This is probably my favorite thing you brought up! I forgot all about this. I'll probably extend the two sides of the X-axis to accommodate legs.

    5. How stiff are we talking here? I'm not an engineer, so I'm sure other's know better than me here. I was thinking that this probably the least important (not unimportant) part of the Y-axis assembly. I could replace these two 1515s with a 1545 maybe.

    6. Currently the design has the bearings spaced 4 3/8" on center. This was completely made up. I could give them a bit more room. I'll mull this one over and see how much space I want to sacrifice.


    Now, with all of that said... I wonder if I should change my design completely. Now that I have lowered the clearance from 6" to 4", I could more practically construct something with 80/20 that has the rails above the table and no uprights, MechMate style. Sigh... more thinking... =)

  13. #13
    Join Date
    Jun 2004
    Posts
    6618
    You can actually do away with the concern on the gantry supports with high X rails. You can support such a design as extensively as you want or need. Build it kinda like an overhead crane. Those things support ridiculous weight with very little flex. My Z axis rides on such a thing, however it is still gantry riser supported. When I do it again, the X will be raised. Not only that benefit, but the table on such a design can allow incredible support. Enough so with the right spindle, you could machine even steel. Two spindles and you have an extremely nice machine for different purposes.
    Lee

  14. #14
    Join Date
    Dec 2013
    Posts
    6
    Leeway, the more I'm thinking about it, it seems like this might be the way to go. What are the downsides I'm not considering? In my non-engineer mind it seems like it could be overall cheaper, more robust, and easier to construct. There must be bad points too...

    I don't follow how the table is able to support more with a MechMate style machine (what's the real term for this kind of high X-rail design?). I would guess that the table could be built to support an equal weight with either design.

  15. #15
    Join Date
    Jun 2004
    Posts
    6618
    Not sure what the actual term for this style router might be. High rail or C channel system might be okay. It allows zero leverage from the Z axis, which makes it pretty rigid from the start.
    The downside as I see it would be material loading. It would not be as easily accessed. Certainly a big concern on a large machine, but not so much on a smaller one.
    Lee

  16. #16
    Join Date
    Aug 2011
    Posts
    388
    bluntend, Thanks for the kind feedback. I like figuring out CNC questions like these, and had some free time and curiosity, so there it is. And maybe it will help others, too.

    Yeah, it's tough to predict the needed Z clearance. 4" may be a little on the tight side. Most people settle on 6" - 7", which would be no problem with profile rails and good rail spacing.

    That's great you put the ballscrew between the rails -- that's the stiffest placement for it.

    On the other points,

    1. Most any reinforcement would help. The main idea is to get a bunch of metal spaced away from but still connected to the plate. Tube or solid bar would be better than alum U-channel -- most of the channel's cross-section sits next to the plate, so only the tips of the legs do most of the work. Steel is 3x stiffer than alum, and less $, so a steel square tube is a good reinforcement option.

    2. Rather than adding a 1.5x3 on top, it would be much stiffer (and a bit simpler?) to use a single larger extrusion. How about the 3x6 3060? From my FEA models, of all the components in a CNC, the gantry tube benefits the most from being BIG. Regular alum or steel tubing also works great and is less $, but also less convenient.

    3. Actually, I think wood would probably be fine for the table -- the loads here are easy to control, there's plenty of space for a thick torsion box made of BB ply or MDF, and heavier is better here (so not the best place for alum, if $ is tight). In general, for high stiffness at low $, the important places for using metal are the Z-car, gantry car, and gantry tube -- those are the challenging parts to make stiff.

    6. 4 3/8" sounds pretty small and may lead to some tipping of the gantry. The main challenge with the 'feet' bearings is to prevent tip from two forces:
    - acceleration loads acting at the CG of the gantry assembly -- so, about the middle of the tube.
    - cutting forces acting at the tip of the cutter -- so, as low as table height.
    So the ideal X rail height is in-between those two -- typically about 3-4" above the table. At that height, *neither* force has a very big lever arm to tip the gantry, so the bearing loads stay low. (Also to save $, the X rails are the least important for cutter stiffness, so SBR20 would be fine here.)
    Putting the X rails very high (like Momus) is great for accel loads but poor for controlling the cutter. Putting the rails a bit below the table is great for table access, ok for cutter loads, but poor for accel loads. Very high and very low rails can work, but extra long feet are needed to control tipping.
    David Malicky

  17. #17
    Join Date
    May 2005
    Posts
    3920
    Quote Originally Posted by dmalicky View Post
    You're welcome--glad it's interesting. I just noticed I saved the xls with a k of 200k lb/in (estimate for skate bearings). A k of 2M pulls the deflection down nicely. But yes, everything deflects! I like the brass chart.
    Thanks for the update!

    I'm terribly embarrassed by all the auto correct mistakes in that post. Usually I catch the iPad being bad but it was late at night.

  18. #18
    Join Date
    May 2005
    Posts
    3920
    Quote Originally Posted by bluntend View Post
    Holy crap! I've spent a lot of time on forums in my lifetime, with subjects crossing the spectrum from guitars, to photography, to computers, to target shooting, to wood stoves... and I have NEVER gotten a response like dmalicky's! Extraordinary. I feel like I owe you money or something. =)
    I liked it too! I've worked on machines all my life, literally since leaving high school, sometimes you get an immediate feeling that something is wrong in a design. It is nice to have an engineering verification of ones feel.
    To your point of spacing the Y-axis rails, consider it done. I've actually lowered the clearance from about 6" to about 4" for the Z-axis. I realized, I'm never going to need a depth of 6" (uh oh, have I jinxed myself now?). I also raised the upper bearing rail above the ball screw. I would post new drawings, but I have to go through wizard's post and rethink some more things.
    YIKES

    Yes I think you have jinxed yourself here. You talk a lot about musical instrument parts which are very much 3D as opposed to sheet goods work. As such you will need to cover the entire thickness of a piece of stock with a cutter long enough to reach from the top of the wood blank to the table. So if the stock is 4" thick that means at least 4" of cutter stick out. You will need to be able to drive the Z axis high enough to clear a tool that long while at the same time getting close to the table with a shorter and stiffer cutter.

    Obviously the phrase "it depends" comes into play here. What you intend to machine and how place limits on your minimal dimensions. Also consider that most people work on top of a spoil board and at times fixturing. You need to visually imagine how you will fixture something in the machine to get an idea of how much clearance you need.

    The good thing here is YouTube, there are endless posts of CNC machines doing musical instrument parts on there. Many of those machines have more than 6" of clearance


    wizard's post is also very illuminating, and brings up a lot of good points. I'll comment on his points and answer questions in the order he presented them:

    1. Right now I was thinking of reinofrcing the uprights with Aluminum u-channel. I Could probably go with steel u-channel or square tube instead, but I'll have to look into cost and workability with what I have (not much!). I don't think I would have the ability to construct them completely from tube steel. I'll have to think about it.
    How did you intend to cut the aluminum plates? I ask only because if you was about to have somebody cut them out for you there is little difference in having a tube or extrusion cut. U channel might help but if you cut a piece of extrusion to fit that part of the machine you have a simplified solution with half the mounting requirements done for you.

    Having a limited home shop is something all of us (well many) suffer from. This is where either a friend or a local machine shop can help out. Yes another expense but done once right is better than doing it over and over until it finally works.

    By the way I have a tendency to prefer steel tubing mainly because it is cheaper and far easier to weld. That is not to say it is perfect just that I've had better experiences with it.

    2. In my drawings from the original post, I have a 3" x 3" extruded aluminum (80/20 fractional series 15 3030). My new version is using an additional 1.5" x 3" piece above the ball screw that the top bearing rail is mounted to. This separates the rails to a distance of 5.25", and keep in mind that I've lowered the Y axis so that the bit will have about 4" of clearance over the table.
    You are talking the Y axis gantry here right? Honestly what ever material you choose, go for a one piece solution. I've seen gantries for a different sort of machine make use of two welded square tubes side by side. It is a nice design but the builder also finished the gantry in a mill providing flats to locate the linear rails. I just see more trouble than it is worth if you don't have access to a machine shop.



    3. As mentioned I'll be using 80/20 Fractional Series 15. The sides of the X-axis are 1.5" x 3" (model 1530), the cross members are 1.5" x 1.5" (model 1515). As mentioned above, the Y-axis uses a 3x3 (3030) and now also another 1530. They also make a 1.5" x 4" (1540) and a 3" x 6" (3060). Currently the cross members are spaced 11.5" on center. I could add another 3 cross members between the current ones. Part of the decisions here are financial, I can only dedicate so much money, and i'm already about 3x more than my plan when I first thought of making a CNC router. The original plan was to just make it from wood. =)
    I know the feeling.

    The only thing I can add is that aluminum extrusions are expensive. As such you might want to consider other material for parts of the machine. I have a different theory than some here, I prefer to see a robust base to the machine that allows improvements latter. We do this at work where you can see a basic machine in three or four guises as the production line evolves. So from my standpoint I'd put money into a strong and robust table first, even making the gantry out of seet goods if you have too. The intent is to new able to upgrade it easily down the road.
    4. This is probably my favorite thing you brought up! I forgot all about this. I'll probably extend the two sides of the X-axis to accommodate legs.
    Something always gets forgotten in these sorts of projects.

    Speaking of forgotten, many builders like to design their machines so that the gantry can run over the edge of one end of the table. The idea is to support the machining of tenons and other joints on the end of a board. This capability might be of interest to you.
    5. How stiff are we talking here? I'm not an engineer, so I'm sure other's know better than me here. I was thinking that this probably the least important (not unimportant) part of the Y-axis assembly. I could replace these two 1515s with a 1545 maybe.
    I'm not an engineer either just expressing experience from working on lots of machinery over the years. I don't like the idea of least important, everything is important as it all has to work together. Frankly I'm not sure we are on the same wavelength here.

    There are two issues here. The first is the plate that holds the uprights and the second is the frame members (the table) that support the X axis. For the table the 1515's strike me as being rather thin to support the X axis rails. The plate supporting the uprights should be fairly robust.
    6. Currently the design has the bearings spaced 4 3/8" on center. This was completely made up. I could give them a bit more room. I'll mull this one over and see how much space I want to sacrifice.
    That seems narrow to me. Luckily you now have a spread sheet to help you understand how much you will suffer at various dimensions. There is always a trade off between expense, size and functionality. You need to decide what parameters are important to you.

    Now, with all of that said... I wonder if I should change my design completely. Now that I have lowered the clearance from 6" to 4", I could more practically construct something with 80/20 that has the rails above the table and no uprights, MechMate style. Sigh... more thinking... =)
    That might not be a bad idea, however you don't have me convinced that 4" is enough. Don't dismiss rectangular steel tubing for this approach for the main rails. You could bolt the steel to aluminum extrusion cross members (ladder style) for a somewhat simplistic assembly of the machine. The "rungs" would need to be long enough to support gussets on the outside of the steel tube.

    As for you and changing of the design, welcome to the world of custom machine tools. The possibilities for router designs is almost endless. Eventually you will settle on something.

  19. #19
    Join Date
    Dec 2013
    Posts
    6
    Ok Everyone, I've returned with a completely different design, as I alluded to. High X-axis rails, no gantry risers to worry about, bigger structural components all around! Of course, I'm sure there are things I'm missing again. Please let me know your thoughts.

    I know the Y-axis rails are probably not as far apart as I would like, only 4.5" on center. I imagine this will probably be fine for routing wood. Going much farther apart than that is going to require more aluminum framing being attached to the Y-axis structure which can introduce other points of weakness and movement that may prove to be less helpful than leaving it alone.

    The other thing I'm toying with is the location of the ball screw for the Y-axis. Ideally it would go between the rails as shown in most of the images. The last image shows another option I was thinking about at the start of this design. My concern is that in order to locate the ball screw between the rails, I'll have to use some kind of spacers to get the Z-axis assembly out far enough to make room.

    The parts colored blue are steel. As wizard pointed out, steel is stiffer than aluminum, and I think in the places I've used it, the stiffness will be needed. I've just guessed and used 1/2" plate on the X-axis components, and 1" sq tube on the Z-axis to stiffen against the cutting forces pushing on the aluminum plate that the router mounts to. That plate is affixed to a linear stage on the top half of the plate.

    The Y-axis and the two X-axis sides are made from 80/20 Fractional 15 3060 (3" x 6"). The cross members underneath are 1530 (1.5" x 3").

    I found that the Z-axis clearance could remain 6" with this design using the larger structure pieces.

    I need to put legs on, but do not need to worry about moving structure to avoid underneath this time. =)

    Attachment 216962 Attachment 216964 Attachment 216966 Attachment 216968 Attachment 216970 Attachment 216972 Attachment 216974 Attachment 216976 Attachment 216978 Attachment 216980 Attachment 216982

  20. #20
    Join Date
    Feb 2005
    Posts
    829
    That is a great looking machine!

    One place you could save money, is Build a Table to mount it on out of steel. That much alminum extrusion of the size you have joinging the 2 x-rails together will not be cheap. That being said it would be easy to build. you could probably get away without useing so many triangle joiners on the 3 central cross peices, especially if you used a piece of angle iron to join the, what I assume is, MDF bed with the x-axis extrusions. Then it would be using the MDF as a structural member to keep the entire system square. And you would not lose any actual cutting surface as the extrusion would be in an area the router can't reach anyway.

Page 1 of 2 12

Similar Threads

  1. Want to buy a good CNC Router Plan
    By ryanvu in forum Want To Buy...Need help!
    Replies: 5
    Last Post: 02-12-2011, 06:32 AM
  2. Easy to make cnc router '.skp' plan
    By YIORGJX in forum DIY CNC Router Table Machines
    Replies: 3
    Last Post: 10-31-2009, 07:46 AM
  3. Plan to builld a CNC router
    By bothunter in forum DIY CNC Router Table Machines
    Replies: 0
    Last Post: 02-13-2008, 12:51 AM
  4. Plan for DIY Router
    By jensn in forum DIY CNC Router Table Machines
    Replies: 5
    Last Post: 01-27-2008, 10:04 PM
  5. 3-day eval
    By saturnnights in forum SheetCam
    Replies: 2
    Last Post: 02-12-2006, 09:39 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •