[shootingarts]

This is a rough sketch of the machine I am planning. X axis is planned to be 50” travel, Z axis 3” to 6”, Y axis 16”. However the design may be scaled to suit scrounged components and dollars!



I am inclined to go with a strongly supported cantilever design instead of a moving gantry at this time. As the image indicates, the X Axis is in a vertical frame instead of the more common horizontal design. The Z Axis attaches to the X axis to control the vertical movement, and the Y Axis attaches to the Z Axis to provide cross travel and support the spindle.(router)Some of the work for this machine will require a rotary axis also but the converted 4" rotary table and a tail shaft should be adequate for that.

It is possible that I may have to add a light vertical support and roller going from the free end of the Y Axis to the outside table support however I hope that the design will not require that.

Hopefully this design will give me a high level of precision near the ends of the X Axis travel and with the Spindle near the attached end of the Y Axis while giving a reasonable level of accuracy for 2.5D carving and sign making throughout the travel range.

I am debating making a substantial iron frame to mount the components on or just a heavy base/tabletop mounted combination. Unless someone sees obvious problems the basic design configuration is finalized and it is now a matter of chasing components. The 3x3 and 2x2 square steel tubing is easy to come by and the aluminum will be basic shapes too. That leaves all of the working parts, many of which I will have to find before beginning construction. The X Axis ball screw looks to be the priciest of the parts needed to move the axis so I’ll chase it on E-Bay first. My minimum travel requirements are X=32", Z=3" and Y=12" so I do have some room to work with when shopping for components.

I also plan to retrofit a 14x40 lathe and Bridgeport J-head mill clone in the future so I am inclined to learn how to deal with servo's now. Good idea or should I just use steppers with this design?

Hu

[CNCAddict]

I'm just curious why you aren't going with a gantry design? If you ever plan on cutting any large aluminum sheets, the consistency of the cut will vary a great deal depending on where in the y-axis you are. I am also going through the metal struggles of designing a machine, and I'm consistenty coming up with a gantry style machine driven by a center located ballscrew underneath the table. The ballscrew will be fixed at the ends with a spinning nut. The only problem i may have is skewing. But a very stout gantry should prevent that.

Also consider the loads on your linear bearings when making a cut way out at the end of the y-travel. If you're worried about a lack of z-range with a gantry design, then make the y-axis movable. So when cutting large foam blocks the gantry height can be raised, but then when cutting aluminum everything can be lowered to increase stiffness by a large amount. Does that make sense?

Edit: I didn't see your dimension up top, I think at that size it should be fine. If you scale up you may want to go with the proven gantry style machine.

[shootingarts]

I'm just curious why you aren't going with a gantry design?

The gantry design may be best, particularly with the movable Y axis like you mentioned. That was an option I strongly considered. There were a number of reasons to eventually choose this design: (unless someone hits on a major drawback!)

One reason was that this machine may well be outfitted with a plasma torch someday and I liked the idea of the most expensive ball screw being out of harm's way. The other reasons for this design come down to weight and leverage. With the fixed Y axis gantry, it seems like the design actually creates compound leverage when the spindle is down. The spindle weight works through leverage on the gantry and then the height of the gantry provides more leverage for the load forces to place on the X axis bearings and ball screw. The second issue is weight and more importantly the location of the weight. With my design the motors for all three axis are located very close to the bearings and supports, meaning I don't have a z axis servo or stepper motor out there with the weight working through leverage. Another plus is that the open front will make it easy to swap out the rotary axis and a plain work table as I will need to do fairly often.

The obvious benefit, of my design is that there is very little leverage and only the weight of the spindle plus the arm structure itself to manage when the spindle is close to the X Axis. However the flaw, as you noted, is that the further out the spindle is the more leverage and weight has to be handled by the cantilever assembly.

Of course if this design doesn't work there are also a couple of options for "plan B". I can build down from the free end of my arm with either a light weight support or even the other leg of a gantry. Another option is to come off of that side of the table base with a heavy full length raised beam and add linear bearings so that what I find myself with is an L-shaped gantry style design. That or basically a flat gantry section in a raised frame with the entire worktable being dropped below all of the moving parts also had a lot of appeal to me.

To be honest, I probably consider far too many options and the standard gantry design would be the least risky design to build since many others have walked that path already. However, if a little real world testing shows that all of my theory doesn't amount to a hill of beans, most of the materials can be used on a standard gantry style router.

Hu

[CNCAddict]

I'm sure your design can be made to work fine, it just may take a lot larger gantry in height to distribute the load over the x-axis bearings and give enough space for a substancial z-axis bearing carrier. So the machine will probably cost more and weigh more than a standard gantry setup for the same stiffness. I plan on building a machine with at least 4ft of y-axis travel. So in my case I would need a gantry that's VERY tall to support the stresses of machining while still giving enough z travel.

[shootingarts]

My design is pretty much self limiting. I was wondering about the wide gantries, four feet wide and even more. Do you run dual ball screws or does a single center screw work OK?

I considered a dual drive design with a ball screw on both ends of the gantry, however I was concerned with slack in the drivetrain driving two screws from a single motor or synchronizing two motors. Are these usually issues?

Thanks,
Hu

[skippy]

I assume you have looked at Mike F's build log of his fixed gantry design entitled "Not sure I'm ready for this - I've started building!"
http://www.cnczone.com/forums/showthread.php?t=6840

[CNCAddict]

My gantry will be about 20" x 60" which is unstable. However I don't want to use dual screws for cost and complexity reasons. A center mounted screw does work fine and has been used in many commercial and homebuilt machines. Check out the link below for TONS of info on a similar machine to the one I'm getting started on. The trick is to keep the gantry ballnut mount system very very stiff diagonally by either using a truss system or just making it very beefy. So during a off center cuts everything remains square to prevent binding. I'm planning on spinning the nut instead of the screw so I can get very high rapid speeds without using a huge diameter high lead screw (which cost a fortune). Even with the standard rolled 5/8x.2" screw I should get 600in/min rapids. I also plan on having the y-axis slide up and down for different height materials. If I want to machine an entire foam kayak plug for example, I just slide everything up and it would have a very huge working volume at relatively low stiffness. Then when I work on an aluminum part, I can slide the y-axis down, lock it into place and get much higher stiffness. It's really the best of both worlds.

I had origionally planned on going with dual belt drive and move the table up and down to change the working volume. But I later found out belts have terrible stiffness which isn't at all constant over the x travel. I've also considered gear racks. But I am not to sure about the spring loaded anti backlash mechanism I see (like the ones on the shop bots). This just doesn't feel right when I want to cut hard materials. I still haven't calculated how much sag the ballscrew will have, but 8ft is a very long way so I may have to go with the rack in the end. But I really don't want to do that. As you can tell, I'm not real firm on anything yet. I have the servos, drives, and linear rails. The rest of the design can still change at any time.

http://oneoceankayaks.com/madvac/madvac_index.htm

[ger21]

I've also considered gear racks. But I am not to sure about the spring loaded anti backlash mechanism I see (like the ones on the shop bots).

If you download the current shopbot assembly manual, it looks like they use a turnbuckle to keep the the pinion tight, not a spring.

[CNCAddict]

From what I've seen it's a turnbuckle with a spring on the end. It can be clearly seen in the video PRTalpha_mel.wmv