Small CNC plasma table build DIY

[Plasma-art]

I like the mesa 7i76 combined with the THCAD.
When ever I recieve notification that they are available again I am ordering immediately.
For now I am fine with running it through parallel connection, it's a solid way to control the machine and a good chunk of the industrial machines that are being built still use it.

Not super worried about any EMI issues, there is a 50ft grounding rod put in the ground and tested by an electrician for the table.
It was testing ok at about 20 feet but I already bought the extra rods so he drove them in.

[Plasma-art]

Update time
Been working on belt routing and some prep work towards getting the cable chain in place.
Still need to figure out the exact way I am going to tension the belts but I'll figure something out.
For now it's just tyraps as mockup and I didnt drill and tap any holes for now.

20221026_153406_compress90.jpg

20221026_153419_compress48.jpg

20221026_153429_compress51.jpg

20221026_140149_compress72.jpg

Should be working on making a Z axis but I am kind of putting it off because I'm dreading building it......

[acourtjester]

always moving forward is great, making sparks is getting closer.

[Plasma-art]

making sparks is getting closer.

end of next week maybe but its depending on some parts that are in the mail like the correct cable chain and idler parts.

Only thing I can work on is the Z axis, hopefully MGN12 rail and 8mm acme rod do the trick.
Just need to get some steel for it.

[adbuch]

You are making great progress with your build. I have never seen that doe with a timing belt before, although I have seen it done with chain and sprocket drive. Keep going!!
David

[Plasma-art]

You are making great progress with your build. I have never seen that doe with a timing belt before, although I have seen it done with chain and sprocket drive. Keep going!!
David

Thanks

I took inspiration from different builds and there are a few people that build their table in a similar fashion.
This build for example (might look familiar )

The MPCNC also uses the same sort of belt routing around the nema17's
https://docs.v1engineering.com/mpcnc/intro/
Is it perfect? No, does it work yes.
I can not move the gantry when I tie all the stepper leads together so it brakes, the table starts tilting when I do that
There is the option to upgrade to a gear rack in the future but I dont see that as something that would give a dramatic improvement with the way the machine is built.

But enough text, time to head out and get some steel and build myself a Z axis, tried to postpone it long enough and its time to bite the bullet.
Took the measurements I need and have a picture in my mind, once it's working I'll order some aluminum and do it properly.

20221027_090631_compress37.jpg

[Plasma-art]

As expected the Z is giving me al sorts of issues and doubts.
Spent the whole afternoon measuring, a scribing lines tripple checking, making mock ups and the end result is probably not going to end up on the machine.
And the best part, I cant drill the holes because i can't chuck up the drill size I need, to small for the chuck.
Even better, the hole pattern I used and was able to drill for the stepper was for nema17 and not for nema23

20221027_130412_compress20.jpg

I know I need to shorten the rails and acme rod, that's the last thing I wanted to do.

Going to think about the different options and leave the Z axis for the time being.
If I had 350 bucks lying around I would order to the spec I need.

I'll do something simple tonight and draw + 3dprint some brackets with enclosure for the homing switches, hope I dont stuff that up

[Plasma-art]

Have been measuring, sketching, drilling and tapping a little bit.
Starting to look like something that might actually work
Probably going to use one of the rail cutoffs for the floating head part, it's long enough for that.

20221029_155646_compress12.jpg

Those larger holes on the square plate is where it attaches to the X axis carriage
Managed to get it fitted to the carriage, still need to adjust some things but I can do that when I take it apart for paint.
Also need to fab the motor mount but that's not a big deal.
But it's starting to look like a plasma table now.
The Z axis reaches from about 10mm above the watertable up to the point it's higher then the underside of the carriage.

20221029_181247_compress19.jpg

20221029_181131_compress79.jpg

Todo list

• weld the motor mount to the Z axis

• Mount the last 2 steppers.

• Get the belt routing and tensioning sorted.

• Making cable chain U profile and fab that into place.

• Make floating head

• Mount homing switches.

• wiring everything up

And then it's pretty much ready for testing, taking it apart to weld and paint everything.

[Plasma-art]

My stack of PCB's came in today for the extra inputs.
Soldered one up with the good old soldering iron and some lead free solder.
Forgot why I hated the lead free stuff but I remember again.
I'll solder another one with some good old leaded solder and the reflow station.
Gives another fully optical isolated 13 inputs and 4 outputs over parallel.
Ordered them in china with slow shipping with a 3 day lead time, took 7 days from order to delivery.
Only missing the screw terminals because I didn't have them in stock.

20221101_202543_compress64.jpg

20221101_202441_compress67.jpg

20221101_202415_compress76.jpg

Also got the first cable chain in position, still need to make a U channel for it.

20221031_165253_compress18.jpg

[weldguy]

Coming together very nicely. I enjoy these DIY builds and their unique design elements. Good work

[Plasma-art]

Coming together very nicely.

Slowly but surely

Had a small side project I needed to finish first but got some stuff done on the table today.
Side project was a heat treatment oven, I make knives as hobby and the plasma is also going to be used to cut out knife blanks.

20221102_141421_compress74.jpg

20221102_141247_compress43.jpg

But back to the table, after a couple different design options I settled for this belt tensioning setup.
It's not perfect but it's good enough.
Does need a whack with a hammer to make it square but thats for later.
I can remove the bolt after tensioning the belt so it's not sticking out.
Also made the U channel for the cable chain and gave it a quick paint job
So Y axis is pretty much done, X axis is going to be next.

20221103_164859_compress68.jpg

[Plasma-art]

Time for a small update.
Accidentally broke my plasma torch hose, didnt notice a piece of hot steel on it so it melted a hole in it.
Was hand cutting some parts with it so next time I know to keep it out of the way.
Need to sort out a replacement for it but that shouldn't be an issue.

Used some 1,5mm sheet metal and made up a guide rail for the cable chain on the X axis and also figured out the belt tensioning for it.
Wired up everything and finished the Z axis.
Really coming together now.

20221112_164642_compress42.jpg

20221112_164624_compress20.jpg

It sounds like a bucket of bolts on slow speeds and still needs alot of fine tuning with drivers and software but moves.
Probably need to play with the microsteps, have it set at 800 pulses per revolution right now (4 microsteps)
Got a video but not publicly sharing it, send a DM for it.

Only burned out 1 driver due to a wiring fault, had 2 wires switched, glad I chose cheap drivers to start with.

[plasmanewbie]

Coming together nicely, nice work, not many guys would attempt something like this

[Plasma-art]

Coming together nicely, nice work, not many guys would attempt something like this

Thanks
I treat it as a big 3d printer build and with my experience as a welder/fabricator I dont really have alot of issues building it.
Yes I am cutting some corners but that's because I know it's going to be replaced in about a year, maybe 2.
On the positive side, i can cut my own parts with ease for v2.

Small update while I am at it.
Took the whole table apart today and welded up everything that needed to be welded.
First part has been given a coat of primer today, decided against powder coating because I know I probably need to weld some more in the future.

[Plasma-art]

After a burned out driver, bad weather for painting and being busy at the day job I finally had the time to start doing some work on the table.
Did a bunch of painting and cleared out the spot where the table is going to live but more importantly I started final assembly.
Already had the table, carriages and the small parts painted before it became to cold so did the rails and the outside of the watertable today.

20221227_180938_compress83.jpg

Still alot of work that needs to be done but the finish line is getting closer.

[plasmanewbie]

Looking great, it's really coming together. must be very satisfying to see it at that point

[Plasma-art]

must be very satisfying to see it at that point

It really is and especially when you consider this has been a year since I started planning.
The table is going together nicely and the todo list is getting really short.
Must say @acourtjester has been a great help getting things figured out not to mention give me upgrade ideas
Put another 8 hours into it today, finished up the carriages, adding homing switches, E-stop, put the wiring loom and steppers on and made the floating Z.

20221228_175514_compress61.jpg

20221228_175457_compress96.jpg

Todo:
Paint and assemble the Z axis.
Add adjustible homing switch for the right rail.
Wiring
Fab slat holders and add the slats.
So except for the software I might be done with that tomorrow
Only routing the belts are left at that point, that's part of the software configuration just in case something is wired incorrectly.

[acourtjester]

Happy to help just one of many here that do it too.

[Plasma-art]

Getting very close now.
Almost all the wiring is done, it does need cleaning up but I do want to test everything before I button everything up.
Only thing that remains on the table side is soldering 1 aviation plug, I dont have any 4 terminal aviation plugs left but I do have a bunch of 3 terminal ones left.

20221230_161750_compress66.jpg

E-stop acting as junction box for.

20221230_113558_compress97.jpg

The controller does need a little bit more work.
Add my custom BOB and wire that up to some aviation plugs but might do that this evening.


As it stands right now, pretty proud of it and it looks like a plasma table now
The question becomes, does it cut tomorrow.

20221230_161712_compress81.jpg

20221230_161724_compress73.jpg

Also what to do with the 6 inputs and 4 outputs I have left
Push buttons to move the gantry and some status lights maybe?

20221230_195038_compress80.jpg

[Plasma-art]

Have been fighting to get my second parallel port working in LinuxCNC but got it working and do want to make sure someone can find this for future reference.
Wanted to write this down before I forget how I did it.

It's a Digitus DS-33010-1 with Sunix SUN1989 chip

It did not show up at all except with:

Code: Select all

sudo lspci -vv

20221231_124358_compress15.jpg

This shows the different addresses of the card: 1018, 1010 and 1000, needs trail and error to figure out what adres to use.

However using the parralel diagnostic script found here it was not showing. (It's still not showing but the HAL-scope shows it working)
http://www.vdwalle.com/Norte/Second%20Parport.htm

20221231_100315_compress89.jpg

Now for the fun part that actually made it work.
Using nano make a new config file

Code: Select all

sudo nano /etc/modprobe.d/alsa-base.conf

Added the following into the config to add the IO as parport1 (use parport_pc instead of 1 to add parport 0)

Code: Select all

alias parport_low level parport_1
options parport_1 io=0x1018 irq=10,auto

Save and exit the file and reboot

QtPlasmac is still complaining that it can not find a parport at 0x1080 but halscope is showing the inputs working when on output mode, still need to get the input mode working.
The thing that made it click was this website and it was some trail and error after that to get it working.
https://www.linux.org/threads/solved-un ... ort.24355/

Also got my din housing in I ordered from china.
Don't want high voltage lines not isolated at all so this should work perfectly.

20221231_115401_compress35.jpg

20221231_115634_compress68.jpg

20221231_115620_compress85.jpg

[Plasma-art]

The table is ready and cutting.
Still need lots of fine tuning but so far I am impressed with the results.
Did not fill the table with water entirely because the outside tap is shut off due to frost risk and buckets of water takes to long.

No THC, 1.5mm scrap at 1500mm/m (I think)

This is the bottom side up, almost no dros.
A little bit of swerving going on at a straight piece that I cant really explain yet.

20230101_141406_compress32.jpg

Got myself a sheetcam license and grabbed a random deer SVG to test out and that also went pretty good.

Screenshot_20230101-155240_Video Player_compress4.jpg

20230101_153751_compress16.jpg

20230101_154214_compress2.jpg

Did have some issues getting linuxcnc up and running due to some error messages I was getting but got it sorted in the end.

Todo:
Add the THC.
Alot of fine tuning and making a cut chart.
Square up the gantry using software squaring, it's off a tiny amount.
Level the table because I needed to move it with setup.
Make a computer monitor, mouse and keyboard arm so that can all hang out if the way instead of being on a trolley in front of the door.

[tinspark]

Very impressive. I have always had an itch to build one before I just bought one. Still have a little itch to take that journey .

[acourtjester]

This shows hard work pays off, nice job.

[adbuch]

Congratulations on your success! It look like you are getting it dialed in nicely.
David

[Plasma-art]

I tried to do some 2mm stainless and had a slight issue with the gcode that made it stop for some reason it cut out reasonably well with minimal dross.
Hence the extra hole pattern exists.
Do have an issue with rigidity of my X carriage that I was afraid of.
As you can see there is a little bit of swerving going on and that's because my carriage can rotate ever so slightly due to a design flaw.
Atleast that's what I think is happening because it only happens on the X direction cuts.
You can see it at the top left at the straight and also in the hole roundness.

20230101_182157_compress7.jpg

And yes I am testing while making parts for the table, this is the 100mm vesa mount for the monitor

Very impressive. I have always had an itch to build one before I just bought one.

Buying one is probably cheaper when you take in consideration the hours it takes to build one.
I didn't pay close attention to it but assembling and setting it up alone took over 50 hours.
However I do enjoy the fact I built my own table.

This shows hard work pays off, nice job.

It definitely does, thanks again for the help getting me to this point.

Congratulations on your success! It look like you are getting it dialed in nicely.
David

I am choosing some random values for feeds and speeds, if it doesn't cut through up the amperage works like a charm.