Can you spot what I've done wrong?

[i-make-robots]

Hi, CNCZone!

Thanks in advance for listening to my ramble.

Long time listner, first time caller. I've been designing a machine and I'd really love it if you could take a few moments to look over what I've got and tell me where I'm going most wrong (since no machine is perfect).

Those of you familiar with DIYLILCNC will recognize a lot of this, I stole a lot of ideas. One major reason for designing my own is that I couldn't source most of the DIYLILCNC parts - McMaster Carr actually refused to fill my order.
So I decided to design something that didn't require ugly belts and clamps.

The design will allow me a 110cm*110cm*11cm envelope and cost <$500.

All the non-metal parts in this design are made of .5" MDF.

http://www.marginallyclever.com/samp...020-12-19a.JPG
http://www.marginallyclever.com/samp...020-12-19b.JPG
http://www.marginallyclever.com/samp...020-12-19c.JPG
http://www.marginallyclever.com/samp...020-12-19d.JPG

I'm trying to get as many parts locally as I can (Vancouver, Canada). I've got some skate bearings and the rest of the stuff I picked up at Home Depot. There are a few parts I can't/haven't sourced:

• (1) Emergency stop button. Easy to find.
• (3) Cranks (?) attached behind the motor to manually adjust a thread
• (6) Limit switches. Easy to find.
• (3) Shaft couplers. This stumps me.
• (3) Lead nuts (?) The part that attaches a sled to the thread. My threads are 3/4-10
• (1) catepillar track wire organizer thing. No idea what it's called. Not even sure how I'll fit it in and get it to stay out of trouble.
• (1) laser. I've already got a Dremel 300 and I'd like to make an optional laser cutter instead. Not sure what's involved.

If you have any advice or questions on anything I've posted here, please don't hesitate. I'd like to send my designs off to be laser cut in early January with an eye to completing assembly before Feb 1 and samples made in time for the MAKE Faire in July.

I should probably add that the minimum distance from the z-sled to the bed top is 17cm, which leaves me plenty of room to build up and add grooves that I can use to clamp pieces down. I'm so ignorant I don't even know how to properly USE the damn thing once it's built.

Thanks again!

[dpgoldberg]

I see a few potential problems.

It looks like your X axis relies on gravity only to hold it down? You'll need to capture the rails top and bottom to keep it from bouncing around. One thing you'll discover is the forces involved are a lot more than you would expect.

1/2" MDF seems a little on the light side. It's pretty flexible. You'll notice most people use 3/4" for on there machines.

Your torsion box table needs to be skinned both top and bottom to be truly rigid. It's the same concept as an I beam just going in two directions. If you removed the bottom flange from an I beam you'd have a T beam, not nearly as rigid.

A 110 cm gantry is pretty wide to be driven by a single lead screw centered under the table. When the router is cutting off to the side the wracking forces will be substantial. Suggest you consider dual screw drives to drive the gantry.

As to the parts you're having trouble sourcing (not sure why McMAsterCarr would refuse to fill an order?)...

Not sure why you need cranks to manually adjust motors. The motors can't be moved at all by hand when the system is under power.

Shaft couplers - do a search for LoveJoy couplers. Pretty easy to find.

Lead nuts - more commonly called anti-backlash nuts. Try DumpsterCNC.com or Roton.com... or McMasterCarr!

Caterpiller track - More commonly called Cable Chain, Cable Carrier or Energy Chain. Manufactured by IGUS. Try, um McMasterCarr(!) or MSC, Grainger or Ebay.

A Dremel seems a little underpowered for use as a spindle on such a big machine, you might consider using a router instead. Lasers are a whole different animal and EXTREMELY DANGEROUS in open machines. Have you ever noticed that all commercial laser cutters are built into enclosures? This is for two reasons. First, the vapors created can be VERY TOXIC and need to be exhausted away and out of the building. Two, if the laser beam reflects off the material and into your eye... well you'll make a good pirate next halloween!

Good luck.

[i-make-robots]

Thanks, those are some great tips! I'll add a bottom to the torsion box and some rails on the underside so that it can't lift, and make the bottom beam an I instead of a T.

I didn't plan to use the cranks when the machine is on. I saw it in other designs and thought "there must be a good reason I can't see at the moment."

The total size is actually 1500mm on each side, with a 1100mm envelope. Is there a good way to keep the two screws synched without using a belt?

[dpgoldberg]

Is there a good way to keep the two screws synched without using a belt?

Actually a belt is not a good way to go anyway since it can slip and you need to be able to individually adjust each screw to fine tune the squareness of the X and Y axises.

The best way is to drive each screw with it's own stepper motor and slave them together. Mach 3 makes this pretty easy. One motor is on the X axis channel and the other is on the A axis channel. There is a setting in Mach 3 to slave the A to the X so they both move at the same time.

[ger21]

I agree with everything dp has said. In addition to being dangerous, a laser is also very expensive. They typically use a large tube, about 1 meter long, and use mirrors to get the beam to the cutting area. I've seen the tube mounted to the gantry, and also along the side. Either way, it's a far more complex project than the machine itself.


And McMaster Carr doesn't ship to Canada (or other countries), and hasn't for years.

[i-make-robots]

Actually a belt is not a good way to go anyway since it can slip and you need to be able to individually adjust each screw to fine tune the squareness of the X and Y axises.

The best way is to drive each screw with it's own stepper motor and slave them together. Mach 3 makes this pretty easy. One motor is on the X axis channel and the other is on the A axis channel. There is a setting in Mach 3 to slave the A to the X so they both move at the same time.

Of Course. Slave it! Staring me in the face. Thanks!

[i-make-robots]

Please,

I've been working on the improvements you've suggested and I'm a little stuck.

http://www.marginallyclever.com/samp...010-12-23a.jpg

Clearly I've got to move the threads (pink) somewhere else. I'm thinking away from center and upwards so that they would be attached to the exterior of the gantry (green), below or at bed height. (near the horizontal green line.)

My challenge is securing the "car" that now runs along the bottom of the bed. It's supposed to prevent the gantry from raising itself when the drill is pressing down. Now that I don't have a moving beam running across the underside of the table I can put supports under the center - say, 6" from the edge - and build as much as I like to secure the side. I'm pretty sure it would be better to have a single point of contact on the underside rather than two because it would be simpler to keep adjusted. Any thoughts?
My struggle is to visualize the correct solution and I have yet to find any reference material I can compare to.

I'm also a little concerned that the support I've built for the end of the thread is all wrong. I'm centering a bearing with a couple of adjustment screws and there's nothing to stop the thread from sliding right out the end, should it desire to do so. Am I worried over nothing?

Thank you! I'm learning so very much. It's the biggest thrill I get all day.

[ger21]

bad link

[i-make-robots]

Fixed

[ger21]

Clearly I've got to move the threads (pink) somewhere else.

Not necessarily. Mine are in a similar place.

I'm also a little concerned that the support I've built for the end of the thread is all wrong. I'm centering a bearing with a couple of adjustment screws and there's nothing to stop the thread from sliding right out the end, should it desire to do so.

It will slide out the end immediately upon movement. You need some type of thrust bearing to hold the screw in position. There are several options.

Here's 1 method.
CNCRouterParts

[dpgoldberg]

I would be concerned that the gantry leg running down alongside the table might flex, being just a single sheet of MDF. It's also a good idea to make each mechanical system, or subsystem, independent of the next, at least as far as adjustment goes. In other words, try to avoid running the lead screws thru the the bottom bearing blocks as it is now. If you go to adjust the bearing block, it might change how the lead screw is adjusted, and vice versa. I'd recommend trying to attach the antibacklash nut for the lead screw to the gantry leg itself, ewither above or below the table.

[i-make-robots]

ger21, wow... i feel so dumb. maybe i've been completely miunderstanding something, please educate me: the anti-backlash nut is what goes at the opposite end of the thread, to keep the thread from "sliding out"? The what do I call the nut in the middle (the part I'm trying to move)?

dpgoldberg, you make a very good point about the legs. in fact those aren't the legs, the table will rest on a set of legs inset from the outer edge. I only understand half of what you're saying about the independent adjustment. Can you illustrate your point? I don't grok your description and I'm a very visual person.

You guys (and girls?) are awesome! I'm learning so much here. Best xmas ever.