New to me CNC

[ddmckee54]

Got a little more accomplished yesterday, I've added the base and the main frame to the CNC assembly and determined the "safe" locations in the X & Y 2040 extrusions, places where I can poke the holes that will let me start bolting this thing down. The holes through the X 2040 extrusions will let me bolt the main frame down to the base. The holes through the Y 2040 extrusions will let me bolt the 2 Y extrusions together into a sort of truss. That should really help to stiffen up the gantry. I determined that 50mm from the ends of all the 2040 extrusions was a safe location, far enough away from the end that it wouldn't affect any of the threaded holes in the end of the extrusions. (I just keep forgetting to export the main assembly as a JPEG file before I to post - on different computers dontchaknow.)

That let me determine the locations of the 6 through holes, and the 4 counter-bores that are 10mm deep in the base. I initially was going to tap the holes in the base, I've talked myself out of that. I'm going to give myself a little wiggle room, I plan on using four 1/4"-20 bolts in 5/16" holes to bolt the main frame to the base. The bottom spacers that will support the X 2040 extrusions will be 6mm thick. This will give me 1mm of clearance between the "feet" of the endplates and the base - to try and eliminate any twist that could be caused by the high spot I know exists in the base. I designed 3mm thick spacers for the tops of the extrusions to give the bolt heads and washers a flat spot to live. Totaling up the extrusion, spacers, washers, base thickness, and then subtracting the 10mm counter-bore depth I'll need about 71mm, a 1/4-20x3" bolt should work.

For the 16.5mm spacers that I'll need to bolt between the Y 2040 extrusions, I'll probably use something similar to the bottom spacers for the X 2040's. it'll just have an alignment tab on the top and bottom. That could make printing this part interesting, we'll see how the P1S handles it. I may try print this on an angle for strength, like Clough42 suggested in one of his videos. So far with what little I've printed the results have been outstanding. And I am far from being comfortable with either the printer or Orca - just because they're still so new to me.

I also used Alibre to print a 2D drawing of the base, so I'd be able to lay out the holes that need to be drilled in the base. It took me a while to figure out where, and how, to dimension the model so that I'd get a usable 2D drawing. It took me almost as long to find where to change the drawing units from inches to mm. I plan on laying out the hole locations on masking tape first, as a sanity check, before committing to actually drilling the holes. I also plan to test the counter-bore depth on some scrap to see if I'll need to go deeper than 10mm to be able to use a 3" long bolt. A max of 4mm of threads might get a little dicey, bolt threads might not get to the nylon lock. I've still got 20mm of thickness left in the base after the counter-bore, so I can always go a little deeper.

[ddmckee54]

This time I remembered to export the new CNC assembly as a JPG file.

The magenta colored bits are the parts that will be 3D printed. I added all the spacers to the 3D model that will be used to bolt the main frame to the base, and the spacers that will turn the two Y extrusions into a sort of truss. Coincidently, all those spacers also got printed today, and the center-points for the holes in the base got laid out. I've got to get the holes drilled in the X and Y rails so I can verify that the locations are correct before I drill the base.

[Blender]

My apologies for somehow missing this thread. Looks like the machine will add some real capabilities to your shop. Combing through your list of planned improvements all looks solid and is mostly solved. What will be your CAM software?

Once it's up and running I'd focus on high speed cutting with light DOC. My little sherline which is probably similar rigidity would usually take .3mm passes, but I'd run it as fast as I could with a 10k spindle. Always upwards of 300 mm/min

[ddmckee54]

Don't think that I included in my evil plans for this machine, the MKS SERVO42K kits for all three axis - and a 500W spindle. When all the changes are finally said and done, the only original parts left on this machine will be the two end plates and the bed extrusion. I would like to get results similar to Awesome CNC Freak's results in this video at about minute 23. (3mm DOC, 1mm stepover, and 20mm/mm feed-rate@about 1300 spindle RPM - in steel) Awesome CNC Freak 1310 modifications

CAM software is up in the air yet, probably will be FreeCAD because the price is right. I don't mind paying for a license, but I detest having to rent one. Especially one that has to phone home every few weeks to keep your license active - or worse yet needs to run on the worldy-wide-webb.

Knew a guy on another forum that used a 2.5D software that generated the the g-code from a DXF file. He'd draw the toolpath in CAD for each layer, generate the g-code for that layer, then manually splice the files together in a text editor. That had to have been a butt-load of work, but he cranked some amazing parts by doing that.

[Zabco]

FWIW... I use a program called Estlcam to generate my gcode files. Have to buy a license if you don't want to put up with increasing time delays but it's minimal cost. One time pay and it's yours for ever with free updates until a major change. The program was written in german but you can set it for english. Uses .dxf files to generate the gcode. Can also create 3d cut files from a .stl although I don't recommend it. All those straight line used to create curves in .stl really show up when cut in metal. There are a couple of tutorials on the web site but the program has good tool tip information in pop-ups so it is pretty easy to learn. The web site is not https so windows will probably throw a fit but just click through. estlcam.com

[Blender]

Originally Posted by ddmckee54

Don't think that I included in my evil plans for this machine, the MKS SERVO42K kits for all three axis - and a 500W spindle. When all the changes are finally said and done, the only original parts left on this machine will be the two end plates and the bed extrusion. I would like to get results similar to Awesome CNC Freak's results in this video at about minute 23. (3mm DOC, 1mm stepover, and 20mm/mm feed-rate@about 1300 spindle RPM - in steel) Awesome CNC Freak 1310 modifications

CAM software is up in the air yet, probably will be FreeCAD because the price is right. I don't mind paying for a license, but I detest having to rent one. Especially one that has to phone home every few weeks to keep your license active - or worse yet needs to run on the worldy-wide-webb.

Knew a guy on another forum that used a 2.5D software that generated the the g-code from a DXF file. He'd draw the toolpath in CAD for each layer, generate the g-code for that layer, then manually splice the files together in a text editor. That had to have been a butt-load of work, but he cranked some amazing parts by doing that.

Looks like it's being pushed too hard in steel. May be tough to keep everything tight and true. Easy enough to dial the cut settings once it's done.

Almost 20 years ago I did the same note pad Gcode. Lots was hand written, bug the advanced code was from an engraving programs. I'd do the tool offsets in Autocad LT and import the DXF for the engraver to cam. Worked well enough. Things are way more accessible now. Full CAM suite used to cost as much as two cars.

Carry on with the build

[ddmckee54]

Originally Posted by Blender

Looks like it's being pushed too hard in steel.

I thought the same thing about the manual cut that he did, from the way things jumped at the end of the cut. It was throwing respectable chips though, and it didn't sound like a dental drill while doing it.

The current plan is to get the machine running with the new rails and the new ESP32 controller. That new controller will let me use it as a stand-alone unit - no g-code sender drip-feeding the g-code required. When that's done I can cut the new side plates that will allow me to raise the gantry. I will also stiffen the new side-plates by bolting the original 2020 Y axis extrusions to them on the outside. Then I add the MKS SERVO42C's into the mix. Not sure If I want to go to the 500W spindle before, or after the servos.

I've got to get this thing running, there's parts that I want to make for the RT7R out of aluminum, that are currently 3D printed.

ESTLCAM - never heard of it, I have to look into it. Like I said, I don't mind buying a license, I bought a perpetual license for Alibre. I just don't want to rent a license.

[Zabco]

Not many people have heard of it. I found it googling for an inexpensive CAM program back a dozen years ago when I bought a Sherline CNC Mill. The guy who wrote it is a one-man shop. I believe he is an electrical engineer as he also developed a driver and controller for CNC that he also sells. Been using it ever since. Mine is a version 11 copy and he is now on version 12 and I note that he just released an update for it the other day. I can't upgrade to 12 without buying an upgrade license but I really don't see the need to as it works fine for my needs as is.

[ddmckee54]

I'm not sure Estlcam is compatible with the ESP32 controller that I purchased. Estlcam's controller is based on an 8 bit architecture. That being said it still sounds like I can generate a g-code file, use the Sneakernet via SD card to transfer files - like I'd always planned on doing, and run the file using the ESP32 as a stand alone controller. That being said, Estlcam's only $60... So I'm not out that much if it doesn't work out.

Estlcam has moved on to version 13, but it sounds like that it is mainly a licensing change. This is from the 13.003 release, the video's available on Youtube:

What's new in version 13?

Version 13, to be honest, primarily serves to change the licensing model:
Previously, all updates within the major version number were free, but the cycles between two major versions have become long and irregular.
From version 13 onwards, the license is valid indefinitely for all versions and updates released up to 1 year after the initial entry of the license key.
However, everything published after that will only be fully usable by purchasing an "update code" which extends the period by another year.
Instead of irregular major version changes, I would like to introduce a predictable cycle:
If you are satisfied with the further development of Estlcam, you can support me by purchasing an update code and receive the latest updates for another year.
Estlcam is not a subscription and the license itself has no expiry date:
All versions and updates released up to the end of the 1-year update period remain permanently and fully usable for you!

What applies to owners of older license keys?
License keys for version 12 that were first used less than 1 year ago:
They can also be used directly with version 13.
The free update period ends more than one year after the initial use of the key.

License keys for versions 1 to 11 or version 12 keys for versions older than 1 year:
They can be upgraded to version 13 with a discounted "update code".
In this case, the free update period only starts after entering the update code.

[Zabco]

Originally Posted by ddmckee54

I'm not sure Estlcam is compatible with the ESP32 controller that I purchased. Estlcam's controller is based on an 8 bit architecture. That being said it still sounds like I can generate a g-code file, use the Sneakernet via SD card to transfer files - like I'd always planned on doing, and run the file using the ESP32 as a stand alone controller. That being said, Estlcam's only $60... So I'm not out that much if it doesn't work out.

This is pretty much how I operate. Just use Estlcam to create g-codes, transfer to a thumb drive and transfer to linux pc running my CNC controller board.

[ddmckee54]

If I can't get the ESP32 controller to run as a stand alone unit, I've got a back-up plan.

When I got the CNC, I also got an older pre-owned Tiny PC running WIN 10 - that I planned to use to drip-feed the CNC's original controller with the Universal G-code Sender. When it showed up, I was impressed enough with it that I ordered another, to replace the aging laptop I'm using right now - just haven't set it up yet.

When the CNC arrived, I was less than impressed with Vevor's control board - so I got the ESP32 controller. If I can't get the ESP32 controller to run as a stand alone unit, I can revive the drip-feed plan with the Tiny PC, using either FluidNC or the Universal G-code Sender.

[ddmckee54]

Shortly after I got my Tiny PC's my brother was looking for something he could set up for a couple of grand-kids that are living with him, something that he could load Mint onto. I remembered seeing that Mint was one of the available OS's when I was looking for my Tiny PC's, and I pointed him in that direction. He got one and got it set up a couple of days ago, so far he's loving it - claims it's faster than his WINDOZE laptop.

[ddmckee54]

Well... It's one step forward and two steps back time again. I got the machine mostly disassembled, the Y rail extrusions have been removed, marked, drilled, and bolted back together into a single truss. (Had to run out and get some longer bolts, the bolts I had were too short, or too long.) I also added the grease zerks to the Y axis linear bearings, two straight zerks on the two top bearing blocks, and two 90 degree zerks on the bottom blocks to make greasing them easier.

I had taken one of the X axis rails off and as long as I has the kit of zerks fittings out I figured that I'd add the zerks to those bearing blocks as well. All of the set screws in the Y axis bearing blocks were black oxide, the set screws in the X axis blocks were SS. I didn't pay too much attention to that at the time. Not, until I discovered that the Allen wrench that fit the black oxide screws would not fit the SS screws - too big. That's when I discovered that the black oxide screws were M6 - which I knew, but the SS screws were M5. I ain't got no stinkin' M5 grease zerks, M6, M8, M10, 1/8" BSP, and 1/4" BSP I've got - but no M5's. They are now on order. The X rail that is out has been marked and drilled, tomorrow I'll remove the other rail and get it marked and drilled too. I center punched the mounting hole location that was closest to the edge of the base, when I used a transfer punch to mark the other mounting hole in that rail it was about 10mm off from my layout. I remember that when I added the mounting holes to the base, one of the mounting holes at the RH end of the machine was 10mm off - so I corrected it on the base. I then made the 2D drawing of the base that I used to lay out the mounting hole locations on the base. I think I moved the wrong hole in the model, so that's why the RH mounting hole locations were wrong. Good thing I decided to do a sanity check before drilling the holes in the base.

[ddmckee54]

Picture time...


This is what things look like now, doesn't look a whole lot different than this one does it?


If you answered no to that question you'd be absolutely right. But there is one MAJOR difference, the frame is now bolted to the base and living in its' new, and permanent, home.

Next step? Get the bed mounting brackets designed/fabricated, and get the bed mounted to the X axis bearing blocks. If you look closely at the base below the second mounting bolt for the X axis linear rail on the far side you'll see the access hole for greasing the X axis bearing blocks. (There's a similar access hole on the near side that's hidden in the view.) Once the bed is mounted I can get to work on the X axis lead-nut mount. I really need to find mt ToDo list and get it updated, this should check off at least a couple more items.

[ddmckee54]

I found the ToDo list that I had printed out, in the last week or so I've managed to cross over 15 more items off the list. It's only got 9 items left, probably time to modify and expand it again.

I'm working on X axis bearing blocks/bed mount brackets now. The table saw, shop vac, and dust collector are getting a workout cutting the brackets and cleaning up the inside corner radius of the aluminium angle.

[ddmckee54]

I was having a panic attack last night, I had mocked up the bed, the mounting brackets, and rest of the machine... And found that the bed was interfering with the mounting bolts. After the initial OHMYGOD, OHMYGOD, OHMYGOD, whaddamIgonnadooo wailing passed - I sat down and looked at it. I decided that a small notch in the bottom of the bed would be the easiest solution, but that I needed to model the bed first. It was getting late and I called it quits for the day.

Is there a Meathead of the Month award? If there is, then I'm nominating myself!

I was down in the shop today taking enough measurements to allow me to model the bed. I found that I've got 5.28mm between the bottom of the bed and the top of the frame. OK, now how far does the head of the mounting bolt protrude above the bottom of the bed? I measured the top of the bolt at 11.62mm down from the top of the bed. A little head-scratching later and I determined that a 5mm deep notch would give me almost 2mm of clearance. Hokay-fine, now how wide does the notch need to be? Better make sure the bed is centered first. Why is there such a LARGE gap on the far side of the bed, how wide is the bed anyway?

Survey says: Bed width - 300mm, space between frame rails - 400mm. (Insert head smacking Homer Simpson "DOH" moment here.) Wade-a-minnut-here, why is there only about 10mm of the bed sitting on each of the mounting brackets? (Insert another Homer Simpson moment here!) Some dip-stick forgot that the horizontal leg of the bracket needed to point to the center, not the outside, of the machine.

I still need to trim a few mm off the vertical leg of the bracket. But, I can actually gain a couple mm of usable Z height out of this kerfuffle - instead of losing height as I originally feared.

I didn't model the bed because it CAN'T interfere with anything. Despite my original panic, this turned out to be tempest in a tea pot, a much ado about nothing, a Rosanna Rosannadanna "Nevermind" moment.

[Zabco]

That's why we all got into these types of hobbies, for the mental health benefits.

[ddmckee54]

Yup, I keep looking for those friendly gentlemen in the little white coats. You know, the ones that are coming to take me away - for a nice rest. The guys with the jacket that's all sleeves and buckles.