First build: 5-axis. Could use some advice!

[giz]

Like so many others, I've spent countless hours here on CNCzone, Mechmate forum, Joe's forum, MachSupport, etc reading everything I can about CNC and dreaming of someday owning my own machine. Well with the new year I've realized that at some point I have to bite the bullet and start a build, so here it is!

I'll preface this with the understanding that many people dream of 5-axis builds, plan them, start them, and sometimes even complete them. Unfortunately it seems like not many are completed or converted to 3-axis. I hope I don't fall into this category.

One of the first questions asked with any 5-axis build log is what CAM software the builder plans on using. I'm fortunate enough to have access to CAMWorks through my work, so I'm set there.

The work envelope I am aiming for is 4'x4'x2' though my design now is slightly above that in the X and Y, and just below in Z. The side trusses are made up of 4"x4"x.25" welded square tube topped by 80/20 2040 extrusion. The cross members are 2"x4"x.25" tube and they are bolted so I can disassemble the machine. As it is, the gantry is a mixture of 4"x4"x.25" tube and two pieces of 80/20 3060 extrusion. I have done this since I have heard it is unwise to try and mount linear rail directly to steel tube? The moments of inertia or much higher for these extrusions anyway, I believe they would do a better job withstanding torsion in the X-direction.

25mm linear rails lines are used on the three linear axes. I purchased these as a set of 62" rails w/ blocks and 120" rails w/ blocks. I plan to cut down the 10' rails to make up the long axis and Z axis (6' and 4' respectively).

I have rack and pinion set up on X and Y using CNCrouterparts NEMA 34 setups. I was planning on using a linearmotionbearing ballscrew for the Z, but now I'm thinking I'd prefer R&P if I can do it. Opinons? ~75lbs is weight that will be moved up, down, and held during power off. I don't know much about ballscrews - would 20mm diameter be sufficient or would I need to step up to 25mm? With either R&P or ballscrew, I plan to use a brake on my stepper motor to keep it from falling when I power down.

The B/C head is designed around a pair of Harmonic Drive SHG-25-80 gearheads I purchased off eBay. The 80:1 drive ratio is much higher than I would prefer, so I have a belt reduction designed in (the details of this are not set, but it will be at least 1:2). So we are looking at most 40:1 ration on the B and C axes. I plan to use smaller NEMA 23 motors to drive these, as I don't need the extra torque and help reduce the volume of the head unit.

The B/C head model won't be complete until I know belt reduction in place, center-to-center distance between the gearhead and stepper, and what stepper I will be using.

That brings me to electronics, where I admittedly have little solid plans. My original idea was to use dual G540's to cut down cost, but now that I'm looking at a little over 300lbs for the gantry+z-axis, I realize this is just a pipe dream.

My backup plan right now is to use one of the dual BOBs from CNC4PC (C23 or C32). I have not looked into the difference between them. Along with that, I will need some Gecko drives, power supply, and compatible NEMA 23 and 34 motors. I'm hoping for some suggestions in this area.

Depending on the final cost of the stepper setup, I *might* consider a DMM servo package. If I am going to see a great improvement for a few hundred dollars more, that would be hard to pass up.

The spindle I have modeled up is the 2.2kW fan-cooled spindle sold by love-happyshopping on ebay. I have asked him for dimensions of his new square bodied 3kW but I have yet to receive them. I have to draw the line somewhere and I think I may just go with the 2.2kW, though I'm sure I would prefer the water-cooled (noise concern).

I think that covers most of the basics. Another thing I need to figure out are home switches for the B/C head.

If anybody can spot any pitfalls I'd appreciate you pointing them out. Comments and concerns are welcome, I would love to discuss this build.

[giz]

Also, a big thanks to Master of nun for posting his 5-axis build as I used his machine largely as the basis for my own. He has also been helpful when it comes to information of the Harmonic Drives.

As of now, the only parts I have in-hand are the rails and the gearheads. And yes, some parts are missing from those renders! They are not complete by any means, but they are getting there.

[giz]

Here's a shot of the Harmonic Drives - these things are gorgeous.

[CarveOne]

That's a rather ambitious project and I hope you see it through to finished and making things. It's way out of my league. (But so was my first wooden CNC machine and that didn't cure me of this either.)

CarveOne

[DIYaholic]

giz,
Quite the ambitious build. But as they say, "No guts, no glory".

Having not built a machine of my own yet, I doubt that I will be able to offer anything other than (im)moral support.
With that being said, it looks to me that the Z-axis may be a little "light" for a 2' length and subject to deflection in the X & Y directions. Of course, you did say the rendering is not complete, so perhaps you already plan on addressing the issue. Just my initial observation.
Good luck with your build. I'll be looking forward to following your progress, ALL the way through to completion.

Randy,

[CarveOne]

There have been a couple of builds I've seen that are using R&P on the Z axis. They are off to one side of the dead weight center line. I would worry about binding if it was not center driven somehow. There are braking systems of various types in use for this.

CarveOne

[lgalla]

You do not have to worry about binding if the Z axis is driven from the side and not down the centre.The drive rack or screws only move the weight.The rails keep it straight and take the load,moment.
Larry

[691175002]

To be honest, based on your rendering I highly doubt the machine will be rigid enough to cut much beyond plastic and wood.

Normally the fourth and fifth axes are part of the table. You are going to have a hard time getting power to the spindle as most software requires full 360 degree rotation.

[CarveOne]

You do not have to worry about binding if the Z axis is driven from the side and not down the centre.The drive rack or screws only move the weight.The rails keep it straight and take the load,moment.
Larry

A lot of mechanical things work fine this way, but I think of it as pushing a loaded wheelbarrow by one handle. How well it works is dependent on the loading, the type of guides, and how far off center the push/pull force is applied.

CarveOne

[giz]

Thanks for the comments guys - I agree it is an ambitious build, but if I think of it less as a whole, and more as smaller systems that come together it doesn't seem so bad.

I don't foresee a problem with binding in a R&P setup on the Z. I'm not sure what the preload on the rack would be, but I can't imagine it's much more than 20lbs or so. With two 25mm rails guiding the Z, and I don't think they'd even notice that kind of force on them. If they did, I'd be in a world of hurt when I try to make cuts 30" below the gantry.

The Mechmate uses RP on the Z without any problems, and it uses steel plate and V-bearings as guides.

As it is the Z-axis is a 48" piece of 3060 80/20. With the 6" width and rails bolted to either side of that, I'd be surprised if I saw flex in the Y direction. The X - yeah I can see it being possible. I think the fix there would be to bolt 1/4" steel plate to one or both of the faces of the 80/20. I won't know what to expect until I start cutting.

To be honest, based on your rendering I highly doubt the machine will be rigid enough to cut much beyond plastic and wood.

Normally the fourth and fifth axes are part of the table. You are going to have a hard time getting power to the spindle as most software requires full 360 degree rotation.

Luckily, I'm not aiming to cut more than foam, wood, and plastic

Where do you see potential rigidity issues? I would say the weakest part would be the 1/2" aluminum plate I have in the head unit. I have thought about bumping this up to 3/4" but am waiting until I have a complete model and have run some basic FEA to do so. Also, the render is a bit misleading as the "side arm" will not extend out as far as it is there - it's maybe 1/2"-1" outside of where it will be (I just hadn't designed the bracket that fits the CNCrouterparts 80mm spindle mount to the HD gearhead).

I'm not too sure about the 4th/5th axes "normally" being part of the table... there are dozens of commercial units using B/C head units. I'd would say this setup is actually by far the most common for 5-axis routers intended for wood and foam. I wouldn't even be able to machine what I wanted using trunnion table.

The HD gearhead is hollow-shafted, meaning I can run any wires for the 5th axis (motor, switches) through the 80/20 and directly through the 4th axis gearhead. The power cable for the spindle will lie in free air and have enough slack to allow +/- 180 degree rotation in both axies. I don't need full 360 degree rotation and the software doesn't expect me to have this.

[DIYaholic]

giz,
So, you have a plan for deflection in the X & Y direction. Good to hear. It's tough to retrofit after construction. Then again, that is what CAD and simulations are for, to sort out design weaknesses so they can be solved prior to constuction.
Hope you don't mind me looking over your shoulder. This build will be fun to follow. I promise to stay out of the way!

Randy,

[giz]

Hey Randy,

I appreciate the feedback for sure! It's the only reason I started this log, so I welcome any and all opinions. There will be times when I defend my design or decisions, and others times when I concede to changes.

I'm gonna go ahead and thicken the plates in the head unit when I get into work tomorrow. I haven't measured these timing pulleys that came with the gearheads, so I'm about to do that and see what kind of reduction I can design into the head (based on center-to-center distances and diameter of the motor-side pulley).

My preference is to build up the frame first, and then move onto electronics, but the winters are just too cold to work without a heated garage. So I'm focusing my time on the model first and perhaps purposely delaying choosing and purchasing the electronics haha.