Can you please post a photo? I just want to be sure I understand as It sounds like a good idea that I should implement.
I could not find any photos and I would have to do some serious disassembly to get at them. Basically, on my shaft couplers there are two set screws set at about 90 (or maybe 120) degrees off from each other where the lead screw goes in and where the motor shaft goes in. I just ground two small flat spots at the end of the lead screw where these set screws come into contact with the lead screw. This is similar to what is already done on the stepper motors at the factory. I also replace the tiny set screws with longer M3 socket cap screws so if the socket gets stripped, I can use a plyers or vice grip to get them apart. This method II just described is way easier than what I did on the lead screws on my first ever CNC build. I drilled and tapped holes in the screws. Those too have never slipped.
Oh I get what you mean - that is a good idea and my couplers do have the threads for a set screw at the lead screw end.
While browsing around the net looking for inductive limit switches, I found these 5v+, 2 wire units. So I ordered a few front sensing versions in the NC spec. https://a.aliexpress.com/_mKB83HS I guess I just couldn’t let go of the idea to get rid of mechanical switches. I’m also going to experiment with grblHAL today as I’m finding FluidNC and openbuilds control don’t play very nice together.
Im really happy with the accuracy now - those TR8x8 4 start screws and AB blocks, etc work really well. I just punched in the default 200 steps per mm for X & Y and now I get 10mm when commanded, every time, in both directions. Happy days! And with some help from @Misterg I got grblHAL working on my MKS DLC32 - Im still finishing the config, but so far so good. grblHAL on MKS DLC32 - can't move any Axis? Oh and apparently I can't let go of the inductive probe idea and while browsing the web, I found these Front sensing, NC, 5v+ inductive limit switches. Interestingly they are 2 wire, not 3, should be fun to test. https://www.aliexpress.com/item/100...der_detail.order_detail_item.4.4292f19cKKadee
My CNC lives again! I think that is by far the nicest spoil board surface I've ever produced! I ran my machine at 1500mm/m and 20k RPM with a 15mm surfacing bit.
I finally got around to wiring up my probes, including my 3D Touch probe which has been sitting in a box for many many months. They connect with 3.5mm stereo plugs. The normal touch probe is only 2 wire and the 3D Touch probe is 3 wire - so very easy to hook up to my DLC32.
If anyone was wondering if my ghetto 3D printed dust separation system works, the answer is yes! This is just from my spoil board surfacing. I did find some build up on the top of my bucket reinforcer, so I enlarged the slots.
Umm, Mistakes were made, but v1 of my front cover looks pretty good … even If I stuffed up the side holes location. I need to rotate the panels and do one side at a time due to my machine work area - Im definitely going to make the next version of my CNC work area 400x500m if possible. Alignment oops #1: Mostly fixed: Alignment oops #2 (on both sides for extra points ): But it does work, even if the side holes out by 10mm from where they need to be. At least I know what to fix now.
Why on earth do windows laptops have such bad trackpads? Arrgh, I hate them so much! I got so fed up with it being so consistently inconsistent, that I pulled my old wireless Trackball out of retirement and made a little side shelf for it … soooo much better!
I finally got around to designing the next iteration of my CNC machine. I was thinking of moving to a flying gantry design, but I think staying with the current concept will result in a more rigid machine. So essentially I’m making the X axis 30mm wider to get back the 20mm I lost to the bigger X axis. The other objective to to beef up the machine. The front and rear plates, and the X axis side plates are all going to be made out of 15mm thick aluminum. I’m also making allowances for Nema 23 motors in case I ever decide to go down that path. And I’ve built in pockets at each end of the lead screw for 688-2RS sealed bearings. These can support both axial and radial loads. Also, all 2020 aluminum extrusions are going to be replaced with 2040 extrusions. I have almost everything I need already, I just need to buy the aluminum stock. I should be able to make all the parts on my machine too. Oh and one other little detail, I designed it so that the stepper motors can be mounted to whichever side or end of the machine someone wants as the parts are mirrored.
So I thought it might be fun to make a DIY Tool Setter. So I ordered a16mm metal momentary switch with a green LED ring light. My plan is to design and 3D print a case / mount for it in ASA. For extra cool factor, I plan to make the LED turn on when the switch is activated Server Busy
I got some of similar ones for my momentary switches on mine too. But, I was dumb and ordered 12V and I have a 36V power supply. I guess I can wire mine in series to be on all the time. they have 5 terminals so signal can be separated.
i intentionally bought one that uses an external resistor so i can just step down the voltage to 3v if I need to. That said, I’m going to be stealing an input and 5v from the EXP1 header on my DLC32, so it should work just fine with a small value resistor.
I still didn’t have quite the Z accuracy that I wanted - I was seeing about 0.1mm of backlash which is very high for a ball screw. So I did some testing with my dial gauge and different shaft couplers. I had been using a jaw style coupler on Z, but after replacing it with a dual helical slit coupler, backlash is now down to less than 0.01mm which I’m quite happy with. I’m also using these on my X and Y axis’s and have found them to be excellent.
I haven't made any major changes to my CNC machine lately, although I do have a small pile of external stepper motor drivers and a proper 36v Mean Well PSU sitting on my bench waiting for me to use. I also have a small box full of GX16 connectors and plugs, some din rails etc waiting for me to build a proper electrical cabinet - the only thing missing is the metal cabinet. I know which one I want, but it'll have to wait a bit before I get around to buying it and building it out properly (too many projects as usual). In related news, Jon very kindly sent me his latest GRBL_SPD v4 board to test - it's a nice upgrade from the v3 I've been using:
Thank you for sharing - very impressed with your transformed 3018. Can't wait to see your new rugged design coming to fruition.
Cheers, I think I finally have the new cabinet design nailed down. High voltage cabling will run to the right, and low voltage to the left.
Looking good! Yes, keeping the two voltages rails away from each other is a good practice, it will aid to prevent EMF interference which in the cnc space has been known to lead to missing steps...
When I studied electronics many years ago, they drummed this into us pretty hard. It’s bitten me most times I’ve ignored the advice!
Yes, I too was trained as an electronics engineer in the 80's and was involved in the design and fabrication of PCB's. I was using cnc machines for drilling the different sized tiny holes of the PCB's. But that all died away as the production moved to the Far East. I aways had a keen interest in cnc machines and now that I have take retirement - I thought it was time to dabble in the cnc world.
Very good! REading your posts as I prepare to upgrade my own CNC as well with simillar gear. Question, you have only 2 MGN carriages (1x per rail) on the X axis right? (I guess the longer ones, H version) Do you notice any play in it, if it was now would you have done with 2 per rail? Thanks
Howdy, thank you. Yes that’s correct, I’m using 2 MGN15 carriages, but they are the larger H version. I’ve had no issues at all with play. The main reason I went with a single carriage per rail, is space. I’d have lost even more X axis travel than I did by going to the ball-screw version of the Z axis. Most of the info about my X axis is here if it helps. Printables
Thanks, that's my concern as well. Doing some measurements I noticed I can't really push it much or I'll be loosing a noticeable mount travel on the axis. So I'll go with the same single carriage per rail on the X.
For the 2020/2040 extrusion frame, I think even the 12mm rails and blocs would suffice. That’s just my two pence worth, but I think it could work equally well too at a reduced cost. That being said I don't know what the price differencial is nowadays.
True, but the 15’s are a little more stable IME than the 12’s - I run the latter on my 3D printers. The 15’s are more expensive, that is true.
When I started looking into this, I was undecided between 12 or 15mm rails. The 12 mm are slightly cheaper but not that it would make it the deciding factor, size was. (2x) 12mm vS (1x) 15mm carriage
Fun is about to start! Still a few parts missing tho. Went with 350 and 400mm lenght rails. For the X, 400mm will be cut down to 360mm (yes I am that picky and I don't want to see that 10mm gap). For the Y. the 350mm will be used. Ordered also 350mm 4040 profiles to extend it. Idea is to fully use the Y axis and to allow the spindle to reach slightly outside for a ouch probe. Z is a lit more tricky, forgot that I could order longer rails and cut them as needed. So ended up having to order another pair for the Z. It will use profile as well and the top/bottom plates, idea is to cut from the frame of my old prusa style 3D Printer, that is aluminium.
