CNC Desktop Router

Discussion in 'Other metal working projects' started by cojo98v6, Nov 28, 2018.

  1. cojo98v6

    cojo98v6 Copper

    So this is a project I started late this summer. I designed this completely around using material that I had on hand. I plan on using NEMA 17 motors and anti backlash lead screws. Trying to keep cost down, but accomplish a fairly accurate machine. I plan on using it to make some intricate foam patterns for casting and some light metal engraving. The full frame is made from 6061 extruded aluminum, bearing mounts and shaft mounts machined from 6061. slides will be O1 hardened drill rod. I plan on using PTFE/oil filled bronze bushings instead of linear bearings. I feel if I make the mounts correctly and adjustable I can hold a fair amount of precision. The work surface I plan on doing a casing of it using LF, this way I can put the T-slots in the pattern. I have the whole thing created in Solidworks and am willing to fully share if anybody would like. I can provide the Solidworks or .stp/.igs files. I also have the dimensional drawings of the individual parts.
    I don't have many picture of the actual machined parts, I thought I took more, but couldn't find them. I will be sure to take more as I progress on the build.

    Frame Assem1.JPG
    Frame Assem2.JPG
    Front View
    Frame Assem3.JPG
    Right Side View
    Frame Assem4.JPG
    Back View
    Frame Assem5.JPG
    Bottom View
    Frame Assem6.JPG
    Top View
    IMG_1273.JPG
    Rough cut to size parts for the bearing slides and slide shaft mounts.
    IMG_1281.JPG
    Some of the parts surfaced and blued up for finalized machining.

    More to come shortly.
     
    Jason and _Jason like this.
  2. Mighty fine!

    You might consider adding an angle to the uprights. You can get a lot of inertial forces starting and stopping the router travel in what I'm guessing is the Y direction. (Assuming the work surface travels in the X direction). I have a traveling gantry so yours might be more stable with a fixed gantry but a lot of my error is in gantry vibration when cutting circles like when boring into a blind cavity. My router is just a zip tool but it is still enough weight to put some load on the gantry during rapid accelerations.
     
  3. cojo98v6

    cojo98v6 Copper

    Thanks.
    Hmm, I'll keep that in mind about the uprights. I figured it would be strong enough as I used 6"W x 1/2"thick Al, but this is my first experience with a CNC router and appreciate the advice.

    I do plan on building a much larger CNC router down the road with the traveling gantry, but wanted to start small first.
     
  4. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    I'm sure you would find it as no surprise that I have one on the drawing board too so I'll be watching with great interest.

    What are the dimensions of the drill rod/journal diameter? I ask because most round linear race than spans any appreciable distance is supported on a rail and the linear bearings slotted to clear the rail. The combination of the aluminum, and unsupported spans might cause some vibration and resonance issues. Foam is such light duty cutting you may get by with it as is. I'm design with V-wheel and track but I want a >30"x24"x 8" active machining space, so a bit larger than yours but same issues still exist for me. My overarm router actuator has 1" diameter journals and a stroke of 5". It was completely insufficient when the journals were 5/8 diameter but it is also a heavy unsupported mass making much heavier cuts than one would see in foam.

    For me the mechanicals are the easy part. Stepper drivers, gcode, software and modeling......will be a steeper learning curve. -Can hardly wait to see yours in operation.

    Best,
    Kelly
     
  5. I didn't see the 1/2" dimension on the uprights. That's a lot of material. You could always drill and tap it to screw a flange on later if there's a problem.

    Kelly:

    I see the acceleration of the router head producing the most loads in my router. The speed of the tool coupled with light cuts makes tool loading light. but if you're doing true 3D milling out curves and surfaces the tool translational velocity has to be rapid or you have 1,000 hour cut times. On small projects I've already had some 8 hour cut times. The steppers accelerating and decelerating the cutter head make for some real loads on the gantry. Light and stiff are king.
     
  6. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    The tradeoffs with run time and part complexity/design, along with the fact I already have a pin router that gives me reasonable capability now, is why I haven't taken on the project yet. The closer you stay to profiles and simple 2D machining, the less advantage there is to me to go cnc compared to my hand guided pattern/pin routing foam patterns, though I could probably pick up some positional accuracy. The cycle times on repeat parts for me are literally a few seconds. I get to what I euphemistically call 2 1/2D by changing bit profiles and cutting height but still all hand guided.

    My real interest in CNC is complex curved and blended 3-space geometries that have no flat surfaces, that is all step over ball end milled surface....think the internal runners and external features of an automotive intake manifold foam pattern, machined in two pieces and glued together. If you used a 1/2" diameter bull nose router bit with a couple inches of straight cutting length you could have a very large step to vastly reduce run times, and then just require some modest hand blending of the foam pattern with abrasive paper. If you were going to run a step over program with the puny <1/8" diameter typical cnc bits.....forget it.

    Even so, I figure hours of run time, liquid cooled spindle, a dust hood and collector (that flippin foam gets everywhere), and remote operating location so I don't have to listen to it.

    No argument there.

    I'll be anxious to see where Cojo gets to.

    Best,
    Kelly
     
  7. Jammer

    Jammer Silver Banner Member

  8. cojo98v6

    cojo98v6 Copper

    Does not come as a surprise at all, lol.

    The drill rod I chose is 1/2", most router kits and plans I looked at was using 8mm shafts, so I felt 1/2" (12.7mm) would be a good step up. I'm undecided on the bearing slides, my first idea was push two bushing from each side into the mounds and put a grease fitting in the middle and fill with a high quality grease. My second idea to make the tighter and adjustable would be to use a one piece bronze bushing and slot it just a small bit, I have not determined the amount yet, but probably around 1/32". Then slot the mounts with screws added to basically pinch the shaft.

    The travels are (table) Z 18" by X 13.5" by Y 5" (looking at the front for the XYZ travels). A bit smaller than what you want. The larger one that I'll build down the road will be very close to the size you want.

    Out of all the different CAD software that I've used over the years I strongly prefer Solidworks, easy to use and can do everything and then some that I want. For CAM software, I've only used Mastercam and this is what I have to use. I took a class many years ago in college, but have not used it since, so it will take some relearning it. The good thing is Solidworks and Mastercam work very well together. The best part of using CAD is I can draw the part dimensionally correct then add the shrinkage factor to it, import into Mastercam, run my tool paths/operations and post it. Then its up to the router to make the pattern. This will help me speed up my castings significantly. Most of my parts I want to make will be multiple patterns that will be glued together, CNC is the way to go in my opinion for what I need. If you do not want to use CAM software, G-code is not terrible to learn for simple machining, for complex shapes, it just comes down to doing the math. I took a class on this as well many years ago.

    Its not going to have a super powerful motor either, most routers of this size use the 775 size motors and this is what I'll start with and can move up from there.

    No worries, I did not put those in the pictures, so you did not miss them. That's what I was thinking, I want to keep it opened up as much as possible incase the part is larger than the table, but will add the flanges if needed as you suggested.

    I did really consider this material as most DIY kits use it, but for the size I wanted I could not find it cheap enough. So I decided to use material I had on hand, so my design might not be the best out there, I'm making it work with what I got.


    Frame Assem7.JPG
    Main frame was made with this U channel.
    Frame Assem8.JPG
    I'm using .25" thick 90degree brackets to bolt the frame together.
     
  9. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    There are a lot of those modular chassis component manufacturers out there. I hadn't seen 80/20 before. That is a mind numbing amount of hardware options they have. I was interested because they have bigger profiles but wow....very expensive for just structure. If you managed $2/lb you must have done well. I'd just make my own structure. Not sure I like 80/20s linear bearings.

    I think I like Open Build's stuff better with respect to the vtrack features incorporated but it's smaller, and looks structurally marginal for the size machine I'd like to build, but their linear motion packages are relatively reasonably priced for what they are.

    https://openbuildspartstore.com/

    I'm with you on this aspect. Unless you are getting the linear track features in the extrusion as well, it's just expensive structure that doesn't really get you much unless you don't have the ability to cut and fabricate the structure.

    I think the 1/2" journal is certainly ample size in regard to bearing surface. I'm just not sure two 1/2" rods will be structurally rigid enough over the distance they are spanning without being tied to your support structure at additional locations across the span. It may be a bit of a trampoline when the table and gantry are centered in their respective spans. If you are going to be machining dry materials, I'd really discourage the use of grease and stick with self lubricating bearing materials. In fact, in this case I think the engineered polymer bearings are an even better choice than filled bronzes. Bearings are spec'd by PV (contact Pressure and Velocity), and the V is so small, just about anything will do. Best to have the lowest friction and best wear characteristics possble. I have used Igus brand but there are many, almost all available in sizes directly interchangeable with bronze bearings, and they are inexpensive.

    There are so many ways to build linear motion schemes, but having built and used many round journal linear bearing and twin rail systems, I would offer one tip about bearing fit. When two rails and four linear bearings are used on a sliding carriage, two bearings on one rail are fit with very low diametric clearance and the pair on the other rail have more clearance. The industry calls this "compensated bearings" which is a euphemism for looser fit. Without, racking & binding problems are more common and you'll find it's very difficult to get rails set to the parallel perfection required without the additional clearance on one pair of linear bearings.

    There are so many xyz table and actuating configurations. I was leaning toward a traveling gantry instead of traveling table, just trying to keep the footprint smaller....but all configurations have their +&- and they all work.

    When the time comes, I'm just going to need to do what I always do....dive in. But, certainly happy to tag along as your build progresses.

    After you get the xyz table, drivers, and software living in harmony, you can hang about anything you want on your mount.

    Keep the posts a-coming.

    Best,
    Kelly
     
  10. cojo98v6

    cojo98v6 Copper

    The thought did cross my mind that the 1/2" might flex, but I do not have any plans on doing heavy cutting with it. I was considering a 5/8" or 3/4" but thought it might be over kill. How its designed it would easily be upgradable if need be. That's a good point on why not to use grease, I was planning on putting some sort of bellow over the shafts to keep debris off if I used grease.
    Here are the bushings I was planning on using: https://www.mcmaster.com/1688k16

    Thanks for the advice on the compensated bearings idea, makes perfect sense. I did a slide rail on a much smaller scale many years ago and had a bit of a binding issue. I was planning on just taking my time and machining it with high precision. I like your advice better and I'll give it a go.

    I'm sure with everything that you've accomplished thus far, you will do just fine. If there is anything that you need modeled in CAD I can certainly help.

    Hope to get out to the shop this weekend and make some more progress on it. I'll be sure to get some pictures.

    Thanks for everyone's input and comments,
    Corey.
     
  11. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    I'm sure those will do just fine. Not sure what the nominal clearance is between those bearings and the O1 rod, but probably a couple thousandths looser on the compensated pair. May as well see how they work as delivered. Probably need to buy/borrow reamer to tune them up.

    As OIF said, in addition to light cutting, the suspended mass matters to so keeping the slew rates down would help too if needed.

    Best,
    Kelly
     
  12. DavidF

    DavidF Administrator Staff Member Banner Member

    One thing I found out in my endeavours is that drill rod will get grooves cut in it pretty quickly from the linear bearings. Better off using tgp rod or a linear bushing. FWITW..
     
  13. cojo98v6

    cojo98v6 Copper

    That's what I was leaning toward, just seeing how they work as is and tweaking from there. Luckily I have a hand full of reams and have access to some if need be.

    That is a good point to keep in mind on the suspended mass, I have no plans on running this thing balls out, I'm all about precision in place of speed as these parts will be for me, not mass producing and selling.

    I did look at TGP, but was trying to keep cost down as the standard drill rod is fairly cheep. This project is a build up to converting my Grizzly mill to CNC, which also includes installing ball screws on it, so trying to save money on this to put towards that project. I might go back and take a look into using the polymer bearings that Kelly suggested, since you pointed out that the bronze will score the drill rod.

    Here are some more pictures of the work I have accomplished so far.

    IMG_1351.JPG
    Shaft supports, only thing left is to drill and ream the hole for the shafts. I based these very closely to the ones that McMaster sells, the cap is thicker than the ones they sell.

    IMG_1353.JPG
    Bushing supports for the main table. Need to drill and ream for the bushings and drill holes for mounting.

    IMG_1356.JPG
    Bushing supports for the X and Y slides. Need to drill and ream for the bushings and drill holes for mounting.

    IMG_1358.JPG
    Main frame U channel cut to length and mounting holes marked.

    IMG_1362.JPG
    Angle brackets cut, just need to drill mounting holes.

    More to come,
    Corey.
     
  14. cojo98v6

    cojo98v6 Copper

    No actual progress on the frame, but been buying some parts.

    IMG_1418.JPG
    Got the electronics/electric parts. Got it as a kit for $71 (not including the motor). So far I'm impressed with it for the cost.
    I'll bend up an aluminum enclosure with fan for it.
    Also purchased a 24V 300W power supply for it.
    Package Contents:
    1 x CNC Shield Board
    1 x Kuman UNO R3 Board
    4 x DRV8825 Stepper Motor Driver
    4 x Aluminum heatsink w/ 3M tape backing
    3 x Nema 17 stepper motor 1.7 A (with bracket and screw)
    3 x Nema 17 Stepper Motor Mounting Bracket
    12x M3 Screws for bracket of stepper motor
    3 x Mechanical Switch Endstop
    3 x 3Pin 70cm Cable
    1 x 1.5m USB cable for R3 Board

    IMG_1420.JPG
    Anti Backlash nuts, ER11 collet holder for the 775 motor and a 1/8" and 1/4" collet.
    I figured those would be the most common sizes I'd use and will cover me for some time.

    I've ordered the lead screws and bearings, just waiting for them to show up.

    There will be a slight change in plan from my original rail system. With a suggestion from Kelly and some more research I figured I'd do it right and go linear bearings and rail system. Originally I looked on Amazon and did not find any of decent quality for the price. Apparently eBay is where I should have looked, thanks for that Kelly. I'm so used to priming things now, I stopped using ebay, lol. I will have some time during Christmas and hope to fully assemble the frame and start adding the new parts to it.

    Still undecided if I should leave it in the bare aluminum or spray paint it or might powder coat it. Open to ideas.

    More to come,
    Corey.
     
  15. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    That is pretty impressive for the price.

    How is that collet holder secured to the motor shaft?

    It really just comes down buy versus build. Linear motion components in small sizes are so inexpensive the difference in buy vs build price is small and the time savings large......and that time can be well spent elsewhere.

    I still do some buying on eBay. I use it more as a search engine and price estimator. Hate selling on eBay, especially big ticket items. The fees are exorbitant and the entire enterprise is geared towards rights of the buyer versus seller which is often severely abused with no recourse for the seller.

    I'd leave it aluminum. Paint just gets scratched up in my shop.

    Best,
    Kelly
     
  16. cojo98v6

    cojo98v6 Copper

    Its pressed onto the motor shaft and has a set screw. I'll notch the shaft where the set screw is and also use red lock tite. Learned that from my R/C days, those dang pinion gears came off so easy until I did the before mentioned.

    I agree, like say making LF patterns. I liked the ones with the attached support rail, they will be much more ridged than my original plan with the suspended shaft.

    Been awhile since I've sold something on there too, its just not what it used to be.

    It will be easier this way. Probably better off leaving it bare to save time and just get it working first. I can always paint it later if I feel the need to.

    Thanks,
    Corey.
     
  17. Petee716

    Petee716 Gold Banner Member

    It's great watching this build progress. There sure is a lot to it!
    I'm using a Dewalt DWP611 router that came with the standard 1/4" collet and use an adapter for 1/8" bits. Very soon I bought a 6mm collet for it. I was told by others that the 1/4" collet should hold a 6mm bit but mine definitely would not. With the variety of tooling options out there, I find myself utilizing the 6mm quite often (I'm in the U.S.).

    Pete
     
  18. cojo98v6

    cojo98v6 Copper

    Yes there is, still lots to do, but it should be very handy once done. I do plan on making a larger one down the road which will use an actual router. I've come across people claiming the 6mm that are becoming a standard thanks to China will work in the 1/4" collets. This simply is not true, its a 0.0138 difference, most collets are not designed to adjust that much.
    1/8" will be my most used with this size router. The collets came as a pair for a couple bucks more, so I figured why not. Might have to look into the 6mm at some point.

    Corey.
     
  19. cojo98v6

    cojo98v6 Copper

    I finally got some time to change the design and update it with the linear rails (green) and bearings. I also started to drill the holes in the main frame and found an error I made. Some how I made the U channel 3"x3"x3", its actually 1.75"x3"x1.75", so I made those changes as well, did not affect the model too much. Also got some push momentary switches soldered to JST connectors, these connect to the CNC shield board. They will control: E-stop, Hold, Resume and Abort. Did get the lead screws, bearings and shaft couplers. Just need the linear parts.
    With the design updated I should be able to get all the holes drilled and start assembling real soon.

    Assem1.JPG
    Assem2.JPG
    Right Side
    Assem3.JPG
    Top
    Assem4.JPG
    Bottom
     
    Al2O3 likes this.
  20. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    Looks very solid Cojo. Will be good to see this one under power.

    Best,
    Kelly
     

Share This Page