5 Axis CNC Plastic Cutting Router "The CutaRouterer"

[LeeWay]

Looks excellent from here.

[ChristianLL]

Manofmanyhats:
Is it possible for me to move the thread to the DIY section?

Mcphill:
Thank you for the encouragement, and thank you for the suggestion. I have been just uploading straight to the website, but I've always used photobucket.com for image hosting, so I've put my images there for this message. Hopefully that works out better.
As for your planned build, I encourage YOU to start one and show us your journey too, I'd love to see you build the granite parts in particular. I spaced the bearings further apart on my X axis (left to right) and possibly made the gantry beam a bit more of a square cross section rather than that taller rectangle that mine currently is, so that's my advise to you - nice square beam and good spacing for the bearings if your machine will be tall like mine.

mkd:
Thanks man!

A_Camera:
Here's some of the action from the last post to the state of the machine today. Latest at top, oldest at the bottom.


Z axis


Back of machine


This is how I figure out the general wiring for testing - take a photo, then draw the wires with Photoshop (or any other image manipulation software).



Electrical box with castors. This is detached from the machine frame to avoid the vibrations and all. The box is 760mm x 760mm x 310mm, it will contain all the electronics controller electronics and the PC in one neat unit.



The electrical panel.



I managed to snag a nice big 24" NEC touch screen monitor for AUD$265 (USD$201). I've seen them brand new for over AUD$1000! I've completely rearranged all the Mach4 screens to suit a portrait orientation. Why? Because I can.






Z axis motor mounting.



I had to remake the Y axis (forward and back) driven end ballscrew mounting plates. They are also thicker, I really don't like the way I've mounted the X and Y ballscrews, but I think it will work for a the materials I'll be cutting (plastics).











Hold the plate on the end of some flat bar with two small magnets...



And weld inside tube for rigidity and to keep the sand inside.



This was a bit of a side project where I needed a thin 17mm spanner to lock the spindle while I undo the collet chuck, so I take some time out to make this with basic tools only (not that I have many fancy tools anyway... Besides the CNC mill). I use an angle grinder to cut the basic shape out of the some 5mm x 50mm flat bar, then hand filed to finishing spec, and a nice 'old, handmade' look was the aim with some sanding to finish it off. Now I'm paranoid of it rusting, I should give it a lacquer coating.

[A_Camera]

Very nice work and excellent images. I like your idea of doing the electrical wiring design.

[marting]

Nice Build there Chris!!! Can't wait to see your work finished. Well done.
Rgds
Marting

[skippy]

I also like the way you do your electrical planning! Hey what city are you in?

[ChristianLL]

Thanks A_Camera and Marting!

skippy, I live an hour north of Melbourne.


More photos just to keep you all updated:




I thought people might like to see these custom T-nut stops that I use on my custom T-slot welding table to make two identical welded parts. In this instance, they were used to make the sliding doors and the frames for the front panels.




Here's one of the sliding door frames being welded up with the stops and clamps holding everything square.








The control panel kind of almost complete... Sort of.


It took me a while to decide how I want the mechanics and specific user experience to go for the interface between user and computer. I wanted to encourage touch screen use, but I still need a keyboard and mount occasionally (particularly the keyboard). So I decided on a small, minimal surface to keep the keyboard and mouse at easy access, but still small and out of the way. Only the monitor swivels in towards the machine, this helps encourage swift, easy touch screen use when setting up work pieces or other stuff where the operator has to constantly stick their head in and out of the machine.




The very keen eye would notice that the piece of hollow section that meets the machine frame is shorter in this shot versus three photographs above. I didn't allow enough clearance for the sliding door to fit in that narrow gap bellow it.

That's it for now. Full steam ahead tomorrow morning.

[ChristianLL]

Another quick update:


This is my first ever attempt of a shin sheet metal box. It's got some rough spots, but I'm pretty happy with it.


The only sheet metal tool I have is hand snips, so it was all a stitch weldment, no bends anywhere.








Squaring up the mounting flange on the end of the moving Z axis column. This flange is welded close to square, then will be outsourced for machining to get it accurate. The surface to be machined will be the mating surface for the B/C head which should be getting machined pretty soon (my mill is far too small and incapable, so I'm outsourcing the B/C head parts)






The welds on this side are a little cold. I can't turn my welder up higher or it trips the breaker 9/10 times. (Yes I have it on a 15A circuit, the inrush current seems to be too much still)



Here is an animation I made to help my machinist understand what's going on. I'm happy to give out the 3D model for the head, but right now Fusion 360 won't allow me to export the entire assembly, only individual parts. So unless you have Fusion 360 and I can share the model with you in their cloud service, I can't get the 3D model to you (please correct me if I'm wrong).



The head uses a 25-80 harmonic drive units with NEMA 24 steppers for actuation and THK 90x60x13mm cross roller bearing (RB6013UU) for support on both axes. Both axes also have a friction clamping mechanism for impact resistance. It's kind of like a disc brake, except there's two aluminium rings clamping down on the circular spline of the harmonic drive. If the tool receives an impact that is dangerous to the machine, the entire harmonic drive and stepper motor shaft will rotate first. To recover, just relocate the 0° angles for the affected rotary axes, set to zero in Mach and continue on.

I will elaborate more on the workings of the head once it has been machined.

[ChristianLL]

Okay, time for another update! This one's exciting because the machine is nearing completion and it's been mostly painted!
First primed.

Then urethane coated.



It's been a bit of a battle to get the machine shop that's making the parts for the rotary head to actually get on the same page as me and give me a quote, he's honestly fairly inefficient at communication until I go give his shop a visit, then he's great to talk to... Still haven't got a quote, but he's got all my material sitting there. Most of it is 65mm thick 6061 T6 with one of them being 5083 H112. I'll be sure to make a big deal when all that is machined and being assembled.



The control panel is largely compete and painted. Quite a fun and cool weldment. The keyboard and mouse surface is grey because it will obviously get a lot of dirty fingers touching it, so blue would end up looking bad, the the front plate of the little switch box (not installed in this picture) will also follow the same logic.





Here I'm test fitting the small bracket that holds the limit switch. It switches with presence of the aluminium linear bearing even though it's an inductive sensor. I was given to understand that they only sense ferrous metals, I'm okay with that though, it made it very easy to mount them in an effective spot.



This is it's current state as I post this update. I'm half way through setting up the cable chains, I'm quite pleased with the aesthetic and fit and finish that these folded sheet metal runners give the machine. Obviously need to be painted still, white or grey?
The X axis runner (the one mounted on the gantry) needed to give clearance for a floating ballscrew support that's yet to be made, you can see how I've done that with the mounting screws, it works really well. Sorry the second photograph is out of focus.



And here's the Y axis cable chain.




That's all for now. The next update will likely include the ballscrew supports to stop whipping on the Y axis (1800mm x 16mm ballscrew didn't do well at high RPM in testing), I think I've got a pretty neat and thorough solution to the problem. Also there might be a video of the machine running the X and Y axes. Z, B, and C all depend on how quick the machine shop gets the job done.

Of course, I love answering questions or just discussing, I'll be sure to get back to you.

[mcphill]

Looking GREAT!

Can I ask you to clarify your comment here:

Both axes also have a friction clamping mechanism for impact resistance. It's kind of like a disc brake, except there's two aluminium rings clamping down on the circular spline of the harmonic drive.

I don't think I see these parts in the video, but I confess I watched it with sound off, so you may give the perfect clarification verbally and I just didn't hear it...

[TTalma]

I'm looking forward to seeing the head in action. Keep the posts coming!

[Khalid]

wow.. Great job..i am waiting for more updates...

[ChristianLL]

mcphill
Don't worry, my demonstration video in the last update doesn't have sound anyway.

I'll try explain this clearly and concisely - firstly, if you aren't totally familiar with how a harmonic drive works, watch the video below.


The 'circular spline' is most typically the rigidly fixed part of the harmonic drive unit. Instead of bolting it in place with no chance of moment, I am simply clamping it there with a finely adjusted amount of force so that it will not move under standard cutting forces, but if the tool ever encounters a bad cut or I simply crash the head into something, the circular spline will slip in its clamp, thus relieving the flex spline and other breakable parts of damaging stress. Refer to the pictures below.

Here you see it stripped back a bit to make it clear what parts I'm talking about.



Here it is more assembled to show how it all comes together.


There is one issue with this system - if the tool experiences heavy force in the direction of the arrow shown below, there's no more rotary travel on either axis to absorb the force, so the machine just has to cop whatever comes its way. The only saviour there is my reflexes to the e-stop, and considering I'll be walking away from the machine while to works 5 hour shifts, I doubt I'll get there on time. This is just something I have to be careful of, but really, isn't that basically the case with any CNC machine? They're all easy to crash and break.


Hopefully that clears it up. It might help you when I've got the system physically built and I'll be able to show how it works.


TTalma and Khalid
Thanks for the comments fellas, might be a week or two before any more updates. Then I'll be unleashing a secret weapon of mine into the CNC community for all to enjoy and expand their possibilities! Stay tuned.

[Khalid]

Again thanks for such nice work and lot of pictures.. I am still learning and i admire your gr8 work...i will be waiting for the SECRET WEAPON you going to unveil...

[ChristianLL]

Hi again,

I'm having some really ridiculous limit switch problems. There's been a million and one of them, but I'm out of alternatives and ideas now. Please help me!

I changed from proximity switches to standard lever mechanical switches because the proximity ones were driving me insane with totally random behaviour. Now I'm having problems with the mechanical switches, I can get the YL (Y axis Left side) home switch and negative limit to work (one switch for both functions), but when plug in the YL positive switch it jams up the ESS and does some outrageous stuff. Please watch the video below to see what I mean.



I guess it might be getting EMI issues as it gets nearer the stepper motor wires, but everything is shielded and the wire I'm holding is earthed (grounded). I even tried running the wire so that it always stays quite far from other wires and it doesn't help at all. What on earth is going on? I'm about ready to just ditch limit switches and rely on soft limits alone! I've spent 80% of my time in the past 3 weeks pulling my hair out over damned limit switch. LIMIT SWITCHES, OF ALL THE THINGS IN THIS BUILD!

PLEASE help.

[Devastator]

If it's just one switch, I'd check that switch with your ohm meter to make sure it is working ok. I got a few NC switches in a batch of NO's and vice versa.

[ChristianLL]

If it's just one switch, I'd check that switch with your ohm meter to make sure it is working ok. I got a few NC switches in a batch of NO's and vice versa.

I haven't checked that switch with a multimeter, but the markings on it say NC. Also, if the switch I'm plugging the wire into was NO, then it would activate the Y++ LED on the Machine Diagnostics tab.

[bob83]

Don't know if this will help, but I had this issue for a long time. This might be more of a NO problem not an NC limit switch problem, but I know how you feel and I remember being willing to try anything.
Did you change your debounce rate?
In Mach 3 you go to
Click / Config / General Config / top right corner you will see de-bounce, try a 5 or 100 or 500 in each column. The longer the cable the higher I would set the debounce rate.
"Debouncing routines sample the switch's state at a fairly high rate. When a change in the state is detected, the routine will count anywhere from 1 to whatever the debounce number is set to. Mach samples to make sure the transition was not a glitch and that the switch has settled to its new state. The higher the number the longer it waits to verify that the switch has indeed been triggered and is not just noise from servo,spindle, etc. "

I also added ferrite suppressors on the wires as they approached the board, I stole them from old computer cables like these http://i.stack.imgur.com/CejVV.png
You probably have a couple of these laying around and could test it for free.

lastly I would order a new switch, a coupe bucks is worth spending to save a couple days of headache. I will say.... I have never actually hit my limit switches while cutting anything in the past 2 years(fingers crossed) With probably just over 100 projects cut.
By the way! The machine looks great.

[MegaMoog]

Hey…
Shouldn't there b a BEER HOLDER somewhere by the keyboard??? or is there a fridge nearby??? Gotta have a FROSTY ONE when tinkerin…!!!


Machine is lookin GREAT…!!!

Keep dem pics comin!!!

[ChristianLL]

bob83:
I have turned the debounce rate up to a full second without any improvement. I've noticed that the switch inputs on the breakout board work reliably with short wires. I also suspect that the really far away switches are suffering from voltage drop. They're running on 5V and the longest wire is about 5 meters. So for now I'm putting the limit switches on the back-burner. But when I get to it, I'll put all the switches on either the 72V or the 48V supply that also runs the stepper motors. That high voltage circuit will switch relays in the control box that that activate the 5V input signals to the breakout board. For now I'm just waiting for someone to make my rotary head. I've moved on to a different machinist, actually it's these guys H&H Machine Tools. I'm so pleased to know there's people so close to me making these amazing machines, truly cutting edge machines with 15m long linear motors and more! Nice people to talk to as well. And now I need to get a quote from them.

Also, some FYI for anyone with some cash and wants to build a 5 axis machine, HSD USA have a large range of rotary heads; still out of my price range though. Assuming I get quoted $1500 for the machining of my rotary design, it will still have only cost me $2600.

MegaMoog:
Haha I don't drink beer man, rum and vodka are the only way. My father once got a custom made cake with boobs, it looks just like that beer holder. :P

And thanks for the kind words.

Everyone else:
I'll be getting an update out pretty soon with a video of the X and Y axes running.

[warrenb]

Swap the inputs / outputs for X and Y that are defaulted in Mach and see if it does the same on the other axis. DO the troubleshooting with mechanical switches. (I am assuming you have tested them all.)

Not sure what it would tell you. Maybe the break out board is bad?