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

[jsheerin]

I think you're right. A flex shaft would really reduce stiffness, considering they're just a big spring. The correction would be pretty simple (something like B_rot=B_rot_desired-C_rot, but implementing it automatically in your post might be the complicated part. I forget if you have a software package picked out, but if you do you might ask tech support if that would be possible. Otherwise you could write a script to process your gcode after you posted it to find any incident of C rotation and adjust the B rotation on that line as well. If the B and C rotations were incremental this would be pretty straight forward. If they were absolute I think it would get a bit more complicated but still definitely doable.

But you would have to deal with lash in the bevel gears.

[giz]

I'm using CAMWorks because I have access to it through work. I talked to the guy here (I work for a reseller) about the proposed motor setup and he said it is possible, but didn't seem to enthused about the idea. We could possibly get it to work, but at the end of the day it won't make a huge difference. The look of the head is improved, and the moment arms are reduced, but it introduces backlash and a can of worms with the post.

So I'll keep the B motor mounted horizontally like it is, but I'll add the second arm.

[giz]

A weekend of pondering netted a list of changes I'm looking to make this week. They're not all completely thought through, but overall I think they're improvements.

1. Larger steel all around. With my horizontal bandsaw, I can cut up to 4"x6" so this is what I've change the uprights in the frame to. Up from 4"x4". The X-rail tubes are 6"x6"

2. Switched the gantry to a single 8"x12" tube setup much like spoiledbrat's machine (though mine will mount differently). I think this will make construction much simpler and I like the look better. It is stiffer, but heavier. How much heavier I don't know yet.

3. The wider X tubes mean I can mount both the rails and the rack to the top of the beam, allowing me to setup the motors vertically and get them out of the cutting area. I'm going to ditch the belt reduction drives I designed in favor of planetary gearboxes. I felt like I was wasting the servos running them at only 4:1 anyway.

DMM Tech has a partner company that sells the gearboxes and I'm assuming they'll be close in price to the custom belt drives. They have precision and standard versions, running <=5 and <=6 arcmin backlash respectively. DMM's 900W servos are supposedly getting new drivers that will run them up to 1350rpm, so I was thinking a 7:1 ratio with a slightly larger pinion (the 1" diameter pinion has to be bored out to 14mm and the remaining wall will be dangerously thin in my opinon). Seems like a very good balance between power and rapid speed.

4. Looking into switching the Z-axis to a ballscrew. It reduces the length of Z column considerably, is tidier, lets me ditch the gas spring (assuming I can find a servo brake...), and again is cheaper than the R&P. I would mount the motor upside down in the Z column and use a belt like nun has done. I need to read up on ballscrews because they are one thing I haven't ever paid much attention to. So right now I don't have many details.

6. I'm still undecided on what type of bearing I'll use opposite the B-axis harmonic drive. A simple shaft and needle bearings? A cross roller bearing? I don't know yet.

7. I'm trying really hard to implement at least some sort of dust collection system. I think something is better than nothing. The big problem here of course is designing something that won't collide with the workpiece in 5-axis mode. If I could literally surround just the bottom portion of the spindle and have the vacuum come up through the BC head it would be perfect. Say up into the spindle chamber and out the side arm through a cross-roller bearing. I think it could work...

That's all for now. I got my Z-axis rails in today. Too bad it looks like I might be cutting them down.

[giz]

I didn't adhere to all my ideas above, but the design has undergone what I consider a lot of good changes in the past week.

It didn't take long to realize the ballscrew/no gas spring was not the best idea so I ditched that one, and also the vacuum system. To help guard against debris, I'll enclose all my racks, pinions, gearboxes, motors, and rails in a combination of sheet metal and bellows.

I got the quote back for the gearboxes and the prices was very reasonable. The only downside is that the shaft is 22mm diameter and I can't find a 1.5" pinion to fit that...

A lot of the details are missing from the model right now, but I wanted to show you guys the direction I'm heading. I'd love to hear your guys' opinions.

You can see that my pinions aren't spring loaded, which as far as I know is ok. You can also see the damping strategy in the Z column. There is an inner steel tube, steel rebar, and an epoxy granite mix. Missing are the thin damping sheets that will divide the mix into 4 sectors. I plan to employ the same strategy in the big gantry beam - but not the full length since the motors sit in the beam. The rest of the frame will likely be sand or concrete filled.

I wanted a right-angle gearbox for the Z drive but they were double the cost than the inline gearboxes with slightly more backlash.

The BC head is designed to allow the water and power lines to route up the side arms and up into the Z column. I decided on a press-fit cross roller bearing opposite the harmonic drive. I'm a little worried about alignment here between the two bearings.

That's it for now. Again I'd love to hear any criticism or advice. Also if anyone knows of a good deal on an engine hoist, I'm looking to buy one... gotta move these beams somehow!

[jsheerin]

It looks like you're making all your moving parts a lot heavier - have you gone back and looked at your calcs for the motors to make sure you can move that weight around with your motors and drive system?

The mounting bolts on the rear of the motors in the gantry look like they could be tough to get to - do you have adequate clearance to get a wrench on them?

I got the folding engine hoist from harbor freight with a coupon from the internet. It's worked fine for me so far. Just remember that the legs have to extend under whatever you're lifting. It's not like a gantry crane where you can bridge across the top of something. It makes it somewhat less convenient to use.

[lgalla]

to get the weight of your tubes and an idea of cost go to
OnlineMetalStore
Larry

[giz]

Yes the weight has gotten a bit high. I'm trying to get a complete model so I can do some proper harmonic and transient sims, but this is what I found the acceleration to be with a quick F=ma and assuming the gantry weighs 400lb and the Z column/head weighs 300lbs. The weight estimate for the Z/BC head are high as I'm working to lighten them.

The first number is at the servo rated torque, the second is at the peak torque.

X axis = 22 in/s^2, 54 in/s^2
Y axis = 25 in/s^2, 63 in/s^2

The X is moving 700lbs across with 2 motors, and the Y is moving 300lbs with 1 motor. I could bump the gearboxes up to 10:1 ratio and get ~40% higher values, but even ~20 in/s^2 seems ok to me?

[jsheerin]

I run my router somewhere between 10-20in/s^2 with 425oz*in steppers. So yeah, it will work, but higher is definitely nicer. I would classify my machine as low performance in that area. I am targeting 400in/s^2 on my mill - a bit over 1g - which is very high performance from what I've read. Also remember that you'll need cutting force on top of what's necessary to accelerate the axes. So if you want 50lbf to cut with, you'd need to set your accelerations lower so you didn't use all the available torque from your motors. Efficiency of your drive trains would also eat into those numbers.

So is your Z acceleration about the same with the spring counterbalancing the mass of the column?

[giz]

I run my router somewhere between 10-20in/s^2 with 425oz*in steppers. So yeah, it will work, but higher is definitely nicer. I would classify my machine as low performance in that area. I am targeting 400in/s^2 on my mill - a bit over 1g - which is very high performance from what I've read. Also remember that you'll need cutting force on top of what's necessary to accelerate the axes. So if you want 50lbf to cut with, you'd need to set your accelerations lower so you didn't use all the available torque from your motors. Efficiency of your drive trains would also eat into those numbers.

So is your Z acceleration about the same with the spring counterbalancing the mass of the column?

I saw your discussion in spoiledbrat's build - that's quick! I took a second look at his jog video and it wasn't moving as fast as I remembered, but still faster than most hobby machines I've seen. He was running 50 in/s^2 in that video.

If I bump the ratio up to 10:1, those values increase to:
X axis = 31 in/s^2, 77 in/s^2
Y axis = 36 in/s^2, 90 in/s^2

10:1 + 50lbs cutting force give you:
X axis = 28 in/s^2, 75 in/s^2
Y axis = 34 in/s^2, 84 in/s^2

10:1, No cutting force, 50lb lighter gantry/30lb lighter Z/BC
X axis = 35 in/s^2, 87 in/s^2
Y axis = 40 in/s^2, 100 in/s^2

7:1, No cutting force, 50lb lighter gantry/30lb lighter Z/BC
X axis = 25 in/s^2, 61 in/s^2
Y axis = 28 in/s^2, 70 in/s^2

Cutting forces don't seem to have a huge effect on the accelerations, but weight does. I'm confident that I can still enjoy damping effects of E/G with a thinner wall, and with the second support arm in the BC head I'm sure I'd be ok sizing the AL plate down to 3/4" from 1".

Because of the gas spring, the Z motor is pushing almost no weight so the max acceleration values end up an order of magnitude or 2 higher than the X and Y axes. I could certainly go with a smaller servo if I wanted to but likely will not.

I took into account the friction of the rails (mu=0.003 according to THK so it's negligible anyway) but not the efficiency of the gearboxes (97%), servos, or losses between the gear/rack.

Looking at those number, I'd say it might not be a bad idea to go up to 10:1. I don't know if I'd miss the ~200ipm in the rapids I might lose over the benefit of greater torque/acceleration. I don't see them costing much more, and although they have lower load ratings they're still way above my application. I sure wish those servos weren't limited to 1350rpm though.

I'll have to look into the radial acceleration on the B/C axes later. It's too hot in my office to think straight...

Larry, thanks for the link. I've been using Metal Supermarkets as my pricing guide since I have one local. eBay seems to be a good source for MIC-6. I know a laser guy so I'll prob get any steel plate I need cut through him.

[giz]

Another big factor in acceleration is pinion diameter. The smallest 22mm bore pinion I can find is ~2.25". I'll try to bore out a 1.5" 18mm pinion or something similar.

What do you guys think of picking up an older Bridgeport clone (roughly $1k) so I can do most of my own work? I could mill almost all of the aluminum, and with some work I think I could even face mill the beams. The purchase hinges on the ability to mill the beams.

I think I'd come out ahead AND add a manual mill to my shop. Thoughts?

[ad_bfl]

That is the path I am following, except I started with a Series II CNC which makes the conversion a lot easier since it already has ballscrews, limit switches, auto oiler, ....

The lessons I learned with my conversion to new electronics and the rebuilding of the mechanicals has greatly influenced the design of my system, plus the ability to do a lot of the machine work is invaluable in containing costs.


Given the level of build you are contemplating I would suggest getting a mill and doing the conversion.

Best Regaards
Al

Another big factor in acceleration is pinion diameter. The smallest 22mm bore pinion I can find is ~2.25". I'll try to bore out a 1.5" 18mm pinion or something similar.

What do you guys think of picking up an older Bridgeport clone (roughly $1k) so I can do most of my own work? I could mill almost all of the aluminum, and with some work I think I could even face mill the beams. The purchase hinges on the ability to mill the beams.

I think I'd come out ahead AND add a manual mill to my shop. Thoughts?

[jsheerin]

I'd get a bridgeport for $1k even if I didn't have an immediate use for it as long as it was in decent shape (as otherwise it's just adding another project to the pile to refinish it).