Repstrap Universal CNC

Hello everyone

I decided its about time I did a CNC based project. So recently I have done a lot of research and pulled together a little design.

If it works as well as I want it to I will be able to use it for:

# I’ll start with an Ink plotter: it would be cool to make a font of my own hand writing, then get it to write letters or something.

# Laser engraver: when an old DVD player turns up on freecycle

# PCB mill: set and forget easiest system for homebrew pcb

# Softwood mill: to make creepy wooden portraits of family for presents

# 3D printer: because everyone needs one

Check out my idea’s so far here, tell me what you think I could do better. I want to send off for parts soon

I think it looks good. I would go with a bigger rod than 8mm. The max deflection is a function of the diameter^4. Going up to 10mm will be 2.4x stiffer. Your rods are a bit longer than a usual reprap so expect it to be very springy without bigger rods. Without crunching the numbers I would say go with 12mm or 1/2".

I would not use aluminum rods or pipes. For smooth motion with standard linear bearings you would want to go for ground and hardened steel.

I would beef up the rod that you have the z nut attached to. Why not make it a piece of MDF?

A free end on the threaded rods is really the best way to go on the cheap. People usually get into binding problems and alignment problems if both sides are fixed.

Steppers have end play but the forces are probably going to be small enough you don't have to worry about it. However, it is standard to run the threaded rod through two dueling bearings with thread clamps on each side. In big CNC's you would actually have a big bearing stack. Either 2 angular contact bearings or thrust/radial/radial/thrust.

---

ConceptFORGE
Wally, GUS Simpson, LISA Simpson, THOR Simpson, Sextupteron, CoreXZ

Thanks for the feedback nicholas.seward

I see what you are saying about the 8mm rods. I choose that size because the shaft bearings were the best value, 12mm shaft bearings will cost more than double. But there maybe a win win situation yet, the 12mm bearings are much longer. Do you think I could get away with using 12mm but with only 2 bearings per axis? If that was to work then it wouldn't cost much more than using 8mm.

I know I guy locally who sells everything aluminium, so its not a problem if I give it a go first. If I'm not satisfied with the result it wouldn't be too much trouble to upgrade to the precision rods. It will probably end up that I will use the aluminium when I use it as a ink plotter and laser engraver, then when I want to move on I can upgrade.

Yes using MDF instead of that rod for the Z nut is a good idea, I don't know why I didn't think of that.

I'm glad to hear you agree a free ended nut is best. I was dreading have to by some expensive specialized bearing like that.

I believe that the point contact of the steel balls on an aluminum rod will cause you to get rough motion in short order. Anodized aluminum would be better but I would still be concerned.

I personally would suggest non-ball linear guides. http://www.igus.com/wpck/3665/drylin_r_rjm_01 These are perfect for a dirty environment and will probably work well with aluminum. I can throw grit and sawdust at the guides and you wouldn't even notice. The other option is to install bellows.


I have designed and built a 600x600x300 mill that has 40mm solid steel rods with the guides mentioned above. The rods are 900mm long. You would think that 40mm rods would be stiff enough but there is enough flex that I can't really do metal because of the vibrations that get setup. (As you scale up, cylindrical rods quickly become a bad way to go. The solution at that point is to use rails that can be supported along the whole length.)

Edited 1 time(s). Last edit at 02/20/2014 11:39PM by nicholas.seward.

---

ConceptFORGE
Wally, GUS Simpson, LISA Simpson, THOR Simpson, Sextupteron, CoreXZ

Solid bearings, that's an interesting alternative.

I'll see if I can find some suitable, thanks Nick

Do you think the 3D printed ones work just as well?

For the case of a CNC mill every thing should be beefed up.

8mm rod: DryLin bushing clearance 0.00098" to 0.0024".

Bushings need a load, X, and Y axis maybe you'll get away with it, but bushings are not recommended in the Z axis because they're floating.
You could try to increase the interference fit to get the bushing 1:1, i.e. zero clearance for the Z axis, but I don't know what the friction will be like.

Quote
Quwat
Solid bearings, that's an interesting alternative.

I'll see if I can find some suitable, thanks Nick

Do you think the 3D printed ones work just as well?

Maybe printed from nylon in halves and pressed on a hot shaft. You can do this in a slow controlled way by aiming a heat gun at the rod and letting it conduct the heat to inside the bushing. When the shaft cools down it will be ever so slightly smaller than your bushing. I would design the halves with fins like the Igus ones.

@A2: What do you mean floating?

---

ConceptFORGE
Wally, GUS Simpson, LISA Simpson, THOR Simpson, Sextupteron, CoreXZ

@nicholas.seward:

Quote
nicholas.seward
@A2: What do you mean floating?

No preload, (i.e. zero Hertz stress), coupled with vibrations generated from a cutter on a journal, bushing, plain, sleeve bearing
with shaft clearance in a vertical attitude (i.e. Z axis), with no load to maintain zero clearance is not typically recommended for a machine
that utilizes a cutter that has forces that load, and then unload the system.

When there is clearance between the bearing, and the shaft there is a lack of rigidity.

If there is clearance between the shaft and the bushing I.D., say 0.0005", and an oscillation,
then there exist an opportunity for the mass to float (i.e. it's unsupported).

I'm not addressing tool/machine design, where one set of bearings is fixed, (i.e. the guide),
and the second bearing set is floating (i.e. load bearing).


@Quwat:
To reduce vibrations, and oscillations during CNC milling in the Z axis, you may need to widen the distance between the X axis shafts,
and increase the diameter of the shafts.

The forces on the cutter in effect increase the moment on the Z axis.

Spreading the X axis shafts further apart from each other will help reduce the Z axis shaft flexing.

You want to place the loads as near as you can to the frame supports (attachment points) on the Z axis that your design will allow.

“Binding Ratio”
Demystifying the 2:1 Ratio and the Stick-Slip Phenomenon
[www.pbclinear.com]

Cantilevered Loads and Simplicity® Linear Bearings
[www.pbclinear.com]

Why would increasing the length of the x rails reduce vibrations? It would mean there is more flexibility in the rods, hence more susceptible to vibrations.

I think I will upgrade the rails to 10mm and give these 3D printed solid bearings ago, I don't think my local hackerspace supports nylon prints through.

I realized that only having the one Z stepper motor will cause clearance problems, so I will use the same system but add another stepper hanging off the top brace. That should also help avoid problems with floating as it will insure the two ends of the X axis don't pivot.

What A2 was saying was not to make the x axis longer but increase the separation of the two rods. This will give you a wider stance on the y rails and would make it easier for you to use 1 y axis motor.

---

ConceptFORGE
Wally, GUS Simpson, LISA Simpson, THOR Simpson, Sextupteron, CoreXZ

Quote
Quwat
Why would increasing the length of the x rails reduce vibrations?

I didn't say lengthen the X axis shafts.
I said to increase to the distance/spread between the X axis shafts.

Yes my bad, that was clearly what you were saying.

I could probably increase that a bit. It comes down to the clearance of the X plate and the drill bit, for if I want to mill something deep etc.

I will need to use 2 Z steppers anyway, the lead screw will be a large obstruction. Using 1 would make the choice of this setup redundant, otherwise I might as well have the Z axis nested within the X. But I want to get that full 200mm range, for if I ever get an extruder hocked up.

Quwat,


I'm just a former U.S. Marine and 35yr old mechanical engineering student, but it seems you're getting some good advice. A2 was on point about spreading the X axis farther apart and beefing up the rod diameters. You can compare it to standing with your feet together vs. shoulder width apart...or standing on something like a soda can vs. a cinder block.

When I designed my CNC, I went looking at current commercial and DIY designs (via Google Images). From there I did the research and decided what I wanted my CNC to do (dimensions of cutting area, fixed vs. sliding Y axis, etc). Then I took features from 5-6 different designs and created my own custom version.

For the 3D printer I am currently working on, I did the same thing. The Z- and X-axis greatly resemble the Replicator 2, but for various reasons the Z-axis uses two 12mm shafts spread ~190mm (roughly 7.54") apart; with a 1/2-10 lead screw (this decision is not final). The X-axis uses two 10mm shafts spread 35mm (~1.37") apart. I haven't started the Y-axis yet, but it will most likely use 8mm shafts.

Because of all the vibrations and other factors related to CNC's, they have to be a lot tougher than a 3D printer. My advice would be to forget about RepRap for a minute and jump over to CNCZone.com and other similar forums. Then come back to the Reprap forums for the electronics. I would not use 3D printed bearings from Thingiverse on a CNC unless the designer is a mechanical engineer and designed them specifically for use on a CNC.

Murphy's Law says that if anything can go wrong...it probably will. So if Einstein were to say that according to his calculations you should use 8mm shafts, then use 10mm. 5/8" MDF...then use 3/4" MDF. "Over-engineering" the original design is one reason the B52 is still flying.

It's hard finding information for small to medium sized CNCs, so if you need some links to get you started I know I have a list hidden deep in one of my external hard drives. I'll do some digging if you would like.

----
@nicholas.seward...I live about 70 miles up Highway 7 from you. You can check out my introductory post in the RepRap Arkansas forum. I really like Wally, Gus, and the Sextupteron. I just haven't slowed down long enough to make some posts about using SolidWorks for the graphics and Adobe for the layout on your bounties for free...if someone can write the assembly instructions. Don't want to hijack the OP, so I'll post something soon on CF.

Thanks for the detailed response nitewing76

Yes I am completely aware of the benefits of increasing the spacing on the X axis, once I change the other parts of the design I will maximize that. But I won't take it to the point where I will have to extend my frame work, to avoid shortening the field of movement. At this level the it won't be worth complicating the frame for extra spacing.

Don't worry I have done plenty of research, a lot of compromises had to be made because I want it to be universal. Ie I don't mind that I can only mill the soft stuff if it makes the design suitable enough for 3D printing. This universal factor may seem like a bad idea, but I'll be moving into a dorm room next year and its the only tool I'll have room for.

I'm going to use Nema 17 stepper motors with built in lead screws, so I'm stuck with 8mm. But using this setup will cost cost at least half compared to externally attached screws, so its a really good deal.

I just found out my local hackerspaces 3D printer is down, so I will not have the option of the solid bearings anymore. I have two options to choose between, you might be able to help with your mechanical knowledge? I could use 4 8*24mm bearings or 2 12*57mm bearings for each axis, what do you think would be more suitable? The 8mm I will be able to increase the spacing dynamically, the 12mm is fixed at 57mm but would allow a thicker rail. Maybe a combination of both would be best? I'm thinking 12mm for x&z and 8mm for y.

Haha good old Murphy's Law. This is a casual project, if things go wrong it won't be the end of the world. I try to aim for a nice balance or risk vs initial cost (over engineering), the more mistakes I overcome the more I will learn.

I didn't have any trouble finding other small CNC's online, but thanks for the offer.