CNC Router Up & Running

Discussion in 'CNC machining projects' started by Al2O3, May 25, 2020.

  1. Al2O3

    Al2O3 Administrator Staff Member

    Dust collection!

    I have a larger central bag type dust collection system in my shop. But, the dust shoe on the Router only has a 1 ½” port. Though my central system had very high flow, my shop vac can pull much higher vacuum/flow rate in that small port size, so I use the shop vac on the CNC router. Problem is the filter will foul in short order with heavy rough cutting so I added a cyclonic dust separator.

    9 Dust Collector Ducting.JPG

    Wow, that thing really works well! I can now run for hours or maybe even days without having to stop to clean the filter and just occasionally have to dump 5-gal bucket under the cyclone when it fills. I can empty and replace that bucket in about 30 seconds. Big time saver and huge improvement in dust collector performance.

    I was going to build my own cyclone separator, but since they were fairly inexpensive, I bought the Oneida Dust deputy and the lid that fits typical 5-gallon buckets and made two brackets to mount it on the post on my CNC Router.

    10 DD and Bracket.JPG 11 Base Bracket.JPG

    It functions well but the plastic and mounting flange was pretty flimsy. I made these mounting clamps because the flanged base on the cyclone deflected 1/8” when tightened against the foam gasket.

    12 Base Clamps.JPG

    Now it’s a nice, tight, flat, and well supported. Since space is always at a premium in my shop it’s mounted out of the way…..and yes, grounded every segment.


    I got tired of turning the shop vac on and off with the router so I added this autostart for the vac.

    14 Autostart.JPG

    I went through several dust shoes before I found one that worked well.

    15 Dust Shoes.JPG

    The machine was supplied with #1. Though it was aluminum, it was a total POS. The hose adapter fell off, and since it was this large oval, it would uncover a large part of the area under the brush when cutting at or near the part perimeter, and when this occurred, would eject and spew chips everywhere. If I increased the vacuum level to move more air it would suck the bristles into the vacuum port opening, attract and retain chips, and quickly plug up the vacuum port and not work at all.

    After a search, I found #2 on Amazon, for $27 delivered. It’s an amazing value for that price and has some great features like being able to remove the front half for tool changes but it was massive had a 4” dust port, and wouldn’t fit.

    So, I tried #3. Since I mostly run small diameter cutters, the smaller circular pattern means the brush perimeter and surface area is smaller, which means higher air velocity, volume, and general performance for a given level of vacuum and flow. It worked very well, and better as I increased vacuum level, but it started to suck the bristles into the vacuum port. So I glued in the little baffle in the vicinity of the port. It extends half way down and supports the bristles near the port.

    16 Dust Shoe Baffle.JPG

    Now I can run very high vacuum without plugging the vacuum port, and have the best of all worlds, with excellent performance and nice soft brush contact.

    Still, occasionally you have a combination of cut depth and pattern shape that doesn’t allow optimal dust shoe setting and you will get some dust.

    It’s a lot of anal-retentive detailed BS but the difference in overall performance is fantastic, and the invested time will return big savings over the life of usage.

  2. I was troubleshooting an ancient 1980's Wadkin CNC router with RS-232 comms problems and interference errors and found the comms cable had no physical earth wire between the PC and the router and there was only the mains earth between the two. Sticking a multimeter probe on the chassis of both at the same time showed an 11 volt AC difference. Once an earth wire was established between the two all the problems disappeared. That cyclone looks great!

    Also while I think if it, some vacuum cleaner systems have a bare earth wire running the length of the suction hose to prevent static buildup, if you're sucking certain materials like plastics and in my case carbon dust, the shower of sparks was non stop, like a Van de Graaf generator in a way. you could try routing a bare earth wire down your hose and maybe a few turns in the cyclone chamber to reduce static build up.
    Last edited: Jan 20, 2022
  3. Al2O3

    Al2O3 Administrator Staff Member

    Before the last move of my residence, the dust collection system in my last home shop originally used corrugated plastic duct. When running machines that produced a lot of chips and fines, the static charge on that duct could make a 3" long splinter stand on end on the outside of the 4" plastic duct! It was in the basement of my home and I grew concerned about the possibility of a powder explosion. Replacing the long straight runs with metal HVAC duct, and then wrapping the flexible sections with a spiral per foot of copper wire, and grounding all with the metallic sections was sufficient to eliminate static cling on the outside of the duct.

    So on my CNC machine, I'm hoping that grounding the support wire in the hose along with interrupting the circuit with metallic sections will sufficiently bleed off and prevent accumulating static (dis)charge.

  4. BattyZ

    BattyZ Silver Banner Member

    I FEEL THIS PAIN! That picture you posted brings back nightmares...shudders. Really happy that you are finding some working solutions. Always guessing whether your 2-hour op is going to finish or not; is not a place to be.

    My vote is the pseudo air curtain, nothing physical to get in the way. One could also plumb in a timed pneumatic dump valve to pulse, so every 30 seconds or whatever, your rails get a blast. I think Red Hat is the industrial brand work uses.
  5. crazybillybob

    crazybillybob Silver Banner Member

    Where did you find the goldy locks dust shoe for the router? I have the same router and have been on the hunt for a good shoe as well.
  6. Al2O3

    Al2O3 Administrator Staff Member


    If you search dust shoe and your DeWalt 611 on Etsy you'll find quite a few. This one does not have a Router body clamp but does incorporate the OE DeWalt height adjustment collar which is actually petty nice to have in use. Others do. Without the clamp it does requires a rather thin section of the PLA print to manage the hose and slew rate loads. I broke it once but CA glued it and added more vacuum hose strain relief and no problems since. If you can 3D print, you can download this model from Thingverse and elsewhere. Let me know if you need a steer on that.

  7. Al2O3

    Al2O3 Administrator Staff Member

    Before I could take on my last intake manifold pattern project I had to address a couple deficiencies in my existing CNC Router. I needed at least 3” of Z-axis travel with a bit and collet above >3” stock height. The machine is advertised as having 3.75” Z axis travel but in reality, it is slightly under 3”. Why? There is enough height under gantry but the router runs into the overhanging Z-stepper mount and the router body is too long to simply lower it in the mounting clamp……then there is not enough height under the collet to clear 3” stock height.

    1 Unaltered MV Z.JPG

    So I had to raise the Z-stepper mount and did so my making an extension for the mount and the acme screw drive.

    2 Z Height Adapters.JPG

    Now it looks like this:

    3 Z Height Adapters Installed.JPG

    After doing so, I had cut 3 of the 4 sides of the two more complex intake manifold pattern pieces, when on the second to last MOP on the last side, it lost steps in the Z direction and cut too deep during the roughing MOP. Arghhh!!! Turns out, the linear bearings on the Z axis had seized, so I spent the morning completely disassembling the Z-axis and the linear bearings, cleaned, relubed the bearings, and reassembled everything.

    4 Linear Bearing Refurb.JPG

    While I was at it, I attended to another very annoying design feature of the machine. The router clamp mounts with screws on the backside of the Z-plate so if you ever want to mount a different router or component on the Z-plate, you need to completely disassemble the Z-axis. I modified the router clamp so the fasteners were accessible from the front, which also helps tramming. Now it looks like this:

    5 Improved Router Mount.JPG

    If I have this machine long enough (I have a new build in process), I’ll make a new router mount so I can use another router that has a 1/2 “ collet, since present one only has a ¼” collet.

    So now I was back up and running, and decided to run the finish MOP on the part to make sure there were no further surprises. I thought I had scrapped the part but discovered the increase depth of cut during the roughing MOP due to the missed step fiasco were very shallow and didn’t break through the part. Being lost foam, I paused the program, glued in some small patches...

    6 Patched Pattern.JPG

    ……and ran the finish MOP which trimmed all of the patches to the intended surface……saving the part….never give in!!

    The other thing that was a hassle with the longer stroke 3D machining was I could not use my dust shoe because it travels with the Z axis stroke and there just wasn’t room for it, which of course made one heck of a mess. Since 90% + of my machining is on boards and planks, I made a fixed height dust shoe.

    7 Hoe Pieces.JPG 8 Shoe Pieces.JPG 9 Shoe Assembled.JPG 10 Shoe Assembled.JPG

    The dust shoe casting is only 1/8” wall so I went a bit crazy on the gating. I’ll let you know how it comes out.

    11 Shoe Gated.JPG 12 Shoe Gated.JPG

  8. Robert

    Robert Silver

    Nice! 1/8" wall? Wow.
  9. Al2O3

    Al2O3 Administrator Staff Member

    Yes, I wanted it to be light and it's probably inviting a fail, but where's the thrill without the possibility of failure? I probably have more than twice the metal mass in the feed system than the actual part but it's my best shot to get hot metal delivered evreywhere. It probably would have been better as a 3D print......but I don't have one those.....yet!

  10. Al2O3

    Al2O3 Administrator Staff Member

    I was reading back over this thread. There's a lot of learning and water under the bridge 3 years on.

    CamBam is an incredibly good value. It does pretty much everything I need to as far as CAM, and as a bonus is even a decent 2D CAD package. For easy 2D work, I'll still use it instead of Alibre. Once I made a tweak to the GRBL post procesor it just reliably spit gcode that worked every time. I'm still not a Gcode cowboy but know enough to get by. I have no plans to change CAM software.

    As far as CAD, I'm happy with Alibre, especially now that I'm over the hump on learning curve. I suspect as far as capabilities and learning time it's comparable to the other parametric solid modeling packages such as Fusion, SolidWorks, etc, but I think it's the hands down winner in cost of ownership. An integrated CAD/CAM package had it's alure, but no way could I accept the Fusion business model......all subscription based and they control your body of work in cloud storage. That's just a preamble to a good rump ride by the future business management.

    The MillRight CNC package is probably a decent value at its price point/feature set. If it had more Z-travel and gantry height, it would do everything I needed for foam patterns. That would be a fairly easy hardware mod but it's really geared more towards wood carvers. Though you can cut soft metals with small bits and very light cuts, it's really just not rigid enough for that. The majority of my tail chasing problems were EMI issues caused by static discharge generated in the dust collection system. Once I discovered and remedied that, I hadn't had a problem until the Z-axis bearings seized up. That was really more my fault because I had never cleaned and relubed them since purchase.

    Along the way I learned enough about 3D motion control to build my own/next CNC machine.......and so it goes.

  11. Al2O3

    Al2O3 Administrator Staff Member

    Small surprise. After dipping and hanging to dry, one of the glue joints on the dust shoe pattern failed. I had used white PVA glue on the joint, and I guess since the wall was so thin, when I spritzed it down with soapy water or perhaps water leaching out of the coating reactivated the glued. Think that's a first for me. No big deal....I just applied some hot melt glue and resealed the joint.

    13 Glue Joint.JPG

    Tops likes this.
  12. Al2O3

    Al2O3 Administrator Staff Member

    Nice weather today so I had a little casting session. Dust shoe came out well.

    As it came out of the sand.....

    14 Shoe as cast.JPG

    Degated and cleaned up.....

    15 Shoe Degated.JPG 16 Shoe Degated.JPG 17 Shoe Degated.JPG

    .....still hard to believe this lost foam stuff works LoL! Looks like I have some light machining to do.

    BattyZ, Robert, Rocketman and 2 others like this.
  13. rocco

    rocco Silver

    I say the same thing to myself every time I see pictures of your work.
  14. Compare Kelly's work with operating a metal powder 3D printer with it's cost to buy and cost to run.
  15. DavidF

    DavidF Administrator Staff Member Banner Member

    Have to hand it to you.. pretty friggin impressive...
  16. Al2O3

    Al2O3 Administrator Staff Member

    Got the castings machined…

    18 Machined.JPG

    Made baffle plates out of ¼” MDF. The run time on each was only 90 seconds so I made four from a pick list of features. They register on the step in the shoe casting and attach with two magnets in the dust shoe casting. I just used washers in the baffles.

    19 Baffles.JPG 20 Baffles.JPG 21 Baffles.JPG

    I made a new bracket for the hose. Even though I use anti-static hose, I still only use hose where necessary and break the hose into shorter lengths and use the support wire to ground each section so static cannot build up and wreak havoc with my controller.

    22 Hose Bracket.JPG

    Here it is installed. Wow, it works great. I liked my old shoe but the inlet was only 1.5” diameter whereas this one is 2”, so that’s almost twice the area. I can tell by the pitch of the shop vac that it’s moving a lot more air. With the old shoe, occasionally the brush bristles would get sucked toward the inlet and collect foam dust. -No more of that. It doesn’t even need a skirt brush. I just set the height close to the stock surface. I used the vented baffle so when the router body gets close to the baffle it can still flow cooling air through its body. The baffle with the larger diameter opening is for when the router body needs to plunge below the baffle surface. The other baffles are just while I was at it since the run time was so short.

    23 Installed.JPG

    Here it is in action…..I really like it. No dust escapes!

    Last edited: Nov 12, 2023
    Mark's castings, Tops and Tobho Mott like this.
  17. Al2O3

    Al2O3 Administrator Staff Member

    I made a couple router clamps.....castings of course. My primary motivation was one for the Porter Cable 690 because it can accomodate a 1/2" collet, but figured I may as well make one for the DeWalt 611 while I was at it.

    24 Clamps.JPG 25 Clamp.JPG 26 Clamps on Router.JPG

    I still have to broach some slots for the 690 and drill and tap some set screw holes for tramming and squaring but pretty close to done.

    Here's the 611 installed with the dust shoe.

    27 Installed.JPG

    ......and with all the old take-offs from the previous dust shoe and temporary router clamp.

    28 The Take Offs.JPG

    That'll be enough upgrades for this machine and should keep me in business for a while. My new CNC machine will get my future attention.

    Rocketman, Tobho Mott and Tops like this.

Share This Page