Starting work on a 4' x 8' going to try out a rotating nut for the long axis due to whip and inertia concerns. Hopefully can get some feedback on the design. Wondering if anyone else has implemented this successfully. I do have access to a cnc bridgeport and decent lathe as well which i plan to use.

Rotating nuts are a good idea if you don't mind the mass of the motor involved becoming part of the gantry's moving mass..

(ie I'd have done a rotating nut maybe if I could have for my axis if I could have had that extra weight out on the gantry how ever I could not see supportnig another 20+ lbs with out seeing to much sag. So I got a rotating shaft (which does limit speed) and the motor out on the end.

It is generally easier to drive the shaft because, as wcarrothers1 pointed out, stationary motors reduce the driven mass and also the need for complex umbilicals.

Performance wise, you should see no difference between driving the nut or the shaft. A driven nut puts the motor right at the driven carriage and thus consumes valuable tool space so it is usually used only for single-end supported systems (like extension arms) and compact coaxial drives where the drive nut is the center of a planetary gear motor system which rides on a shaft.

Either method is perfectly acceptable. I just recommend you include the wiring issues in your decision.

Gerber uses it on there 4x8 Saber router and it seems to work great.

On a 4' X 8' router why not just go with rack and pinion like most of the other machines?

Either nut or shaft rotation will move the gantry. Seems easier to use rotating shaft from my limited perspective and gears on the end with some give on gantry threads-this may be a no no. You can then change end grear ratio-but guess you could use a variable speed drive too..... Seems each drive is fine. jmho,pw

This is the 1st rotating nut design i have seen that is a good visual view, and can see this should work well. It is very similar on the lines of what i have been thinking about doing.

Joe

I don't know how often it would come into play, but changing a belt wouldn't be very easy. You could perhaps tape a spare to the inside of the left bearing plate. Then if you did need it, well there it is. ;)

Here is what my line of thought was... If i go rack and pinion i still have the mass of the motor as part of the driven mass, however; i have reduced the driven mass by taking the lead screw inertia out of the equation. But the rack and pinion is i think more difficult to attain with the same level of precision and repeatability as the ball screw. If i go with the rotating ball screw then i have serious whip concerns especially at this length. So with the rotating nut i have a few more components to make but in turn eliminate these other issues.

How about timing belt then?

Andy

Im not sure what you mean, if your refering to how the timing belt will be changed i think its reall just a matter of a few screws. You can see from this view what i was going for a bit better. So I think all that is needed is to unfix one end of the lead screw, and take out a few screws to remove the assembly shown exploded. The exploded assembly on the right can be kept together for the most part, that outer flange is just to apply preload to the bearings so it may need to be loosened and reloaded after belt replacement.

Sorry, I didn't make myself clear.

Instead of a rack, you can use an open timing belt stretched and held at both ends.

The timing belt is looped over a pulley wheel that is driven by a motor.

The motor turns the pulley and the motor moves because the belt is held at both ends.

Very easy to build and minimal precision required for manufacturing.


Andy

You may never even need to replace the belt, but it looks like it won't take that long if you do. Nice drawing too.

I don't think I would go for the open ended timing belt on a large fast machine. Inertia would seem to wreak havoc on it considering that the belt is responsible for acceleration and braking. Lots of G's on a big machine with a 4 foot gantry running at any decent speed. It has limitations and it is just about as complicated to pull off as a rotating nut.

I am using a jack shaft and racks on my plasma cutter and I think it is good for that application. It would likely work okay on a larger router too, but not quite as good as the rotating nut.

I used fixed nuts on my router. 4 foot long 5/8" ballscrews and anything over about 300 IMP has some whip when at one end or the other. That would be compounded pretty well on a 9 foot screw. I think the rotating nut is a great solution especially if you want to run a some speed.

Re comments on timing belts........




Inertia would seem to wreak havoc on it considering that the belt is responsible for acceleration and braking. Lots of G's on a big machine with a 4 foot gantry running at any decent speed.


Is this assumption or based on fact?


It has limitations and it is just about as complicated to pull off as a rotating nut.



Is this speculation or are your conclusions based on direct experience?


Andy

As someone who has practical experience with ballscrew driven, ACME driven, belt driven, and linear motor based machines, I can make the following statements:

1) Belt drive machines are fairly straightforward to implement, but not as rigid as any of the other options. Belts do have a little bit of stretch, even with steel reinforcement. The pick and place robots we built were approximately +/-0.025 mm in terms of repeatability (even after years of design effort) due to this fact, and this was without any sort of cutting loads on them. Look at most professional belt drive solutions, and most of them won't claim better than 0.050 mm.

2) The inertia of a large ball screw can definitely outrank the carried weight of a motor and cables. With a good e-chain for cable management, the fact that the motor is moving is not that big a deal. For our ball screw pick and place machines, we used a 30mm ball screw on a machine with only 5 feet of travel. For 8 feet, you'll need a big (read expensive) ball screw to achieve decent speeds if you turn the screw, which means you'll need big motors. There's a reason a lot of professional machines with large travels spin the nut.

3) I've wanted to see someone do a DIY rotating nut machine, and have thought about doing it myself. I say go for it, for all the reasons you have listed (whip, space, etc.) Be careful with the bearings surrounding the nut to make sure you account for thrust loads. Otherwise, you look like you have the CAD skills and machinery to get it done. I'm rooting for you!

Ahren

Is this assumption or based on fact?



Originally Posted by Leeway
It has limitations and it is just about as complicated to pull off as a rotating nut.


Is this speculation or are your conclusions based on direct experience?

To the first question, it depends on which part of the two statements you are refering to. The second sentence is fact. The first sentence is a logical opinion based on the fact in the second half of that sentence.

For the second question, it too is fact. The open belt system like every type system has limitations or drawbacks. They all have pluses as well. They are nearly as complicated to build as a rotating nut design. It isn't rocket science, but there are still a lot of parts to an open belt system.
I have tried some belt drives and i just don't think I would like to build one that way. I live where there is a lot of heat and humidity and I envision to have to adjust tension often. I know there are different types and sizes of belts and perhaps I didn't use the best, but the reason I said what I did was they were my conclusions based on my experience with belt drives and from reading and hearing about other machines guys have built. I am a hobbiest, so read my comments with that in mind.
It wasn't my intention to step on any toes.

Ahren

What you says makes sense and I concur on your observations of timing belt solutions being easy to build but also less rigid and therefore incurring potential loss of accuracy.

I also echo your sentiment of the floating nut having "build" appeal.

However, there is an underlying question. Is high rigidity (and hence repeatable high accuracy) important for a 4' x 8' machine where the material is likely to be wood or plasma head cut metal?

I think horses for courses is important.

I have built using timing belt for x and y as my machine will be cutting material that does not have to work to vertical mill tolerance.

However, I have put very high effort into creating a highly rigid Z axis as I plan to be able to interchangeably trans-mount the same Z axis assembly onto a smaller x, y frame which will be running ball screws.

This way I get two machines for the cost / price of 1 1/2.

The large machine for large wood with (relatively) low accuracy using acme / ballscrew (demountable Z axis) and timing belt (X and Y).

The small machine for smaller, higher precision harder materials using acme / ballscrew (demountable Z axis) and acme / ballscrew (X and Y).

So, the reason I describe this is that I have come to the conclusion that a rotating nut is likely overkill as a) high accuracy is probably not needed for large machine and b) small machine does not require floating nut.

If the appeal is the challenge of floating nut.... go for it.

If the appeal is for a large machine with vertical mill tolerance.... overkill and unnecessary.

Andy

It wasn't my intention to step on any toes.

Wearing steel capped boots in the workshop allows direct questions and direct answers.

Not a problem......all opinions are valid and open to scrutiny .....including mine :D

Andy

Starting work on a 4' x 8' going to try out a rotating nut for the long axis due to whip and inertia concerns. Hopefully can get some feedback on the design. Wondering if anyone else has implemented this successfully. I do have access to a cnc bridgeport and decent lathe as well which i plan to use.

A rotating nut will work great, apply some preload to the hollow shaft bearings to remove end play, and pull the screw on both ends as much as your frame work will allow, and depending on your loads, you might want to use a keyway on one side to lock the screw, I have used rotating ball screw nuts with good success, but one issue with rotating a ball nut is that some ball nut mounting flanges are not 100% centered or balanced and have to be adjusted for when mounting it to your hollow drive shaft, I will post a couple of pics, of one that I have built, when I dig it out from the big pile, IMO the two main advantages are the use of smaller dia. screws with higher speed servo motors,

www.rockcliffmachine.com (http://www.rockcliffmachine.com/)

Hi,
I work at one of the film industries most prominant SFX workshops in New Zealand and we operate a Techno Isel machine which uses a rotating nut assembly, its perfectly happy being operated at anywhere up to 22metres a minute ALL day without a problem cutting polystyrene to alloy.

Now that the winter is over i have started to make some more progress on this machine. Here are some pictures of the rotating nut assembly finished. I attached this assembly to the gantry and i am awaiting some shaft supports to fix the lead screw to the frame.

Looks good Coop, You're off to a good start on a really nice machine.

I can't see much of your frame design from the closeup but I think you need to add bracing to the rail your drive assembly will mount to. That drive is going to be capable of huge accelerations. That force will be transfered to the beam both well way from the end supports and above the center line of the beam.

The longitudinal resonance of the beam is low enough that it should dampen out quickly and therefore not be a big problem. The twist of the beam, however, is much more of a problem. You could no doubt incorporate gentle acceleration or just live with the device shaking the hell out of you carriage after rapid moves but a simple cross brace would allow you to realize the full potential of your machine.

Resonance is probably the most commonly neglected aspect when building a frame structure. Symmetry just looks so good you can't help but want to make everything evenly spaced or placed exactly in the center but in frame structures that can be very bad. More than one bridge has been symmetrical enough to completely destroy themselves through simple oscillation.

Pete

Hi coop2005

Go with your thoughts Have a look at your BridgePort series 1 CNC if it is a Boss 3 through 6 they have a rotating nut on the X axes & the motor does not move If you make it like theres you will have a winner the motor does not have to move to get the job done. With the motor not moving the wiring is easy as well.But you can do it both ways motor moving or not

Your design looks good are both bearings supporting the drive
& the ballnut if it is it should work well with no problems & like you said to change a belt thats no big deal, the belt will last for years anyway

I wasnt aware that there were mills that used this method of driving. I see its benefits for longer axis but didnt see it as nessesary for that type of machine. I will have to look in to what other benefits other than stationary motors and wiring those machines take from a rotating nut, maybe it will make me feel even better about my descision.

Hi Coop

I'm going to use a ball nut on my Z and if I could afford two 3 meter length ball screws for the Y I would have done that too but am opting for the fixed belt drive option for that due to cost.

I like your design to convert a standard ball nut into a rotating one but I'm sure you are aware that they do make ball nuts that already have all the bearings built in for this type of application.

See BLR type on this page:
http://www.thk.com/eng/products/class/ballscrew/rol_ballscrew.html

Cheers

Mark

Hi Mark

Ya i was aware but cost prevailed, my time building is free so i bought the cheapest ballscrews i could find, they were a few hundred bucks i think. If i went with the premade rotating nut it would have cost me a significant amount more for these. And building it is fun, which im sure 99% of the people on this forum would agree.

Hi Coop

I didn't realise the prebuilt jobbies where that much more. I agree with you 100% that the journey is as much fun as the destination and finding ways around the high cost of parts is definetly where a lot of the inovative designs come from. I've got a bit of the reverse problem on my build as I have a few BLR's (free out of decommisioned gear) and need a fixed nut for the X.

Still hoping the Y axis belt drive does the trick otherwise I'll be on the lookout for some very long ball screws and copy your design.

Cheers

Mark

Kudos on the great design.

Those parts you fab'd are pretty nice i might add. Good Luck, very innovative.

Got the ball screw mounted on the y axis and got the gantry moving. Its got plenty of power that is for sure. Just did some quick testing but it seems to do 150 ipm easily with no hope of stopping the gantry with my strength. Going to get the remaining axis running this friday then i can start some real performance testing. Ill add some pictures on the weekend.