[louieatienza]

the problem is that I will used round flange housing like this :

Attachment 315906

Is it also possible?

Depending on design, no modification may be needed.

And contrary to what is otherwise said, backlash does matter significantly. Of course a lot depends on machine use. Truing casting surfaces? Maybe not. Jewelry work, PCBs? Maybe. But by far, the biggest problem I see here in this entire forum regarding positioning? The circles are not round. I have minus tolerances on my inside cuts and plus tolerances on my outside cuts. I saw one thread where the guy couldn't grasp this concept and it drove him nuts, to the point of selling his mill.

And no, in a perfect world we'd make all our machines of cast iron and epoxy granite. But we make with what's available to us and what we can work with. And there's absolutely no way to say what is ideal without asking the ultimate use of the machine. For cast iron, stainless, Inconel... maybe not. For aluminium and brass, plastics, PCBs, it's well adequate for the job.

[N.Constans]

my goal being to realize this machine initially to produce aluminum parts (maybe PCB in the futur), and get familiar with this machine.

[AVRnj]

What are you using for your spindle?

[N.Constans]

What are you using for your spindle?

I use HF motor from Teknomotor:
http://www.teknomotor.com/eng/prodot...ries-19-57.php
It have a constant torque (0.37Nm) up to 18000rpm.

[AVRnj]

I use HF motor from Teknomotor:
3140 Series, High frequency motors with ER collet chuck, Teknomotor
It have a constant torque (0.37Nm) up to 18000rpm.

Is this going to be a metal cutting machine, or wood?

[N.Constans]

Is this going to be a metal cutting machine, or wood?

mainly for metal (Aluminium) cutting. But I think it could be also used for wood too. In any case, I will used wood for first tests !

[rusel]

This is how I mounted my two ordinary ball nuts on my x axis. The one on the left is clear of the mount and held tight with about 400kg force from some thick spring dome washers between the mounting nuts and ballnut.
It adds little to the force need to turn the screw and remove all backlash, This mounting did not reduce any travel of the axis. As fact when I finished the conversion I had more on the x axis.
Used double nut( 2 ordinary ballnuts) on all my axis..no problems

Attachment 316066

[G59]

This mounting did not reduce any travel of the axis. As fact when I finished the conversion I had more on the x axis.

N.Constans, model it up. You have SW so drag your table from end to end and measure the distance travelled when it makes contact with your bearing mounts. You'll see you will lose the length of the ballnut. Add another ballnut and now you've lost the length of that ball nut in total travel. The only way to compensate for the loss, is to now increase the length of the ballscrew and move the bearing mounts further out.
Don't believe everything you read here. If adding ballnuts increases your travel, why not just get a ballnut the length of the screw?
Again advice from just anybody is poor advice. If you need an example, I'll be glad to model it up for you. I build and design machines for a living, I think, (I may be wrong according to the replies you got), but I think I know what I'm talking about.

And BTW, a misleading comment was made earlier, your linear rails/guides, don't dictate your total travel.

In anycase, I won't get into a pi***ng match over it and good luck with your project.

[109jb]

Many untruths in this thread.

1 - The fact is that if your ball nuts are under your table and they don't protrude past the ends they will not have any affect on the travel of the table. It is a very simple matter to mount them in this way. Done this way the linear rails will be the limiting factor. Anyone who says otherwise is quite simply wrong.

2 - Dual ball nuts for rolled screws are far cheaper than any ground screw and matching nut available. The dual ball nuts if properly installed will eliminate backlash. They will do nothing for accuracy, but they can and will eliminate backlash

[G59]

Many untruths in this thread.

1 - The fact is that if your ball nuts are under your table and they don't protrude past the ends they will not have any affect on the travel of the table. It is a very simple matter to mount them in this way. Done this way the linear rails will be the limiting factor. Anyone who says otherwise is quite simply wrong.

2 - Dual ball nuts for rolled screws are far cheaper than any ground screw and matching nut available. The dual ball nuts if properly installed will eliminate backlash. They will do nothing for accuracy, but they can and will eliminate backlash

Hmmm. OK.... If all that is truth, then I've been racking my brain and wasting my time on quite a few projects. Kinda makes me want to just tear my degree in pieces.

You all must be using magic ballscrews and nuts that require no bearings at either end. The ball screws you guys are using, grow when you add a second or third nut?

Give me a break. It shows this is a hobbiest's forum.

Attachment 316070

[G59]

Dual ball nuts for rolled screws are far cheaper than any ground screw and matching nut available.

Did I ever say anything about costs???

[pippin88]

Attachment 316070

Your diagram is quite correct. Adding a second ballnut reduces the available travel of that ballscrew. No one is saying it does not.

Axis travel is limited by rail travel, or ballscrew travel, whichever is shorter. In the vast majority of builds (and this one looks the same), it is rail travel that limits things. If ballscrew travel is already longer than available rail travel, adding a ballnut won't reduce axis travel. This is what everyone is saying.



Look at that picture. The X axis ballscrew travel is obviously much longer than the available rail travel (overall rail length minus bearing spacing). Adding a second nut would not reduce axis travel.

Did I ever say anything about costs???

The opening poster is building a 'milling' machine out of aluminium and states use of 1605 ballscrews, steppers and a 0.75kw low cost spindle. Does that sound like a build that would / should use ground screws?

[louieatienza]

Hmmm. OK.... If all that is truth, then I've been racking my brain and wasting my time on quite a few projects. Kinda makes me want to just tear my degree in pieces.

You all must be using magic ballscrews and nuts that require no bearings at either end. The ball screws you guys are using, grow when you add a second or third nut?

Give me a break. It shows this is a hobbiest's forum.

Attachment 316070

HELLO... the ballnut sits in between the linear bearings. Even a double ballnut here would still leave room at both extents of the linear rails. So if you have to put the ballnut carrier to one side slightly, big deal?! This is all pretty obvious from the first set of drawings. The total potential travel of the ballscrew has no bearing on the total potential travel of the table, because that has been pre-determined by the linear bearing spread, which is more than twice the length of the potential double ballnut

And it's hobbyist, a noun, not hobbiest, an adjective (an unreal one at that). The spell checker doesn't catch this?

And yes, I do agree that I don't see how adding a ballnut would increase travel, without also adding a longer ballscrew. And yes, I agree that if one were to go through the expense of all other components, ground ballscrews should be an option. Even a lower cost option from PMI would be better than most all the eBay rolled stuff.

[G59]

The opening poster is building a 'milling' machine out of aluminium and states use of 1605 ballscrews, steppers and a 0.75kw low cost spindle. Does that sound like a build that would / should use ground screws?

Nope but it was mearily a suggestion. Who knows he could of found a good deal somewhere?

[G59]

HELLO... the ballnut sits in between the linear bearings. Even a double ballnut here would still leave room at both extents of the linear rails. So if you have to put the ballnut carrier to one side slightly, big deal?! This is all pretty obvious from the first set of drawings. The total potential travel of the ballscrew has no bearing on the total potential travel of the table, because that has been pre-determined by the linear bearing spread, which is more than twice the length of the potential double ballnut

I'm well aware of where ballscrews sit.

You can word things in your favor all you want, but the fact remains, total travel of your table is dictated and in direct relation to the length of the thread of your ball screw. If the thread is 10 inches you get 10 inches minus the ballnut length.

That's all I will say about this.

Oh, BTW spell checker obviously didn't catch it.

[louieatienza]

I'm well aware of where ballscrews sit.

You can word things in your favor all you want, but the fact remains, total travel of your table is dictated and in direct relation to the length of the thread of your ball screw. If the thread is 10 inches you get 10 inches minus the ballnut length.

That's all I will say about this.

Oh, BTW spell checker obviously didn't catch it.

This is true, but the table travel without ballscrews attached is the total rail length, minus the spread of the linear bearings. Which is far less than the nut travel in this instance. Unless of course you let the bearing block fly off the rail. You can make the screw as long as you want, the table can only practically go to its extents.

Now if the screw and motor were mounted to the moving table, not stationary, then it'd be a different story....

[G59]

but the table travel without ballscrews attached is the total rail length, minus the spread of the linear bearings. Which is far less than the screw travel in this instance. Unless of course you let the bearing block fly off the rail. You can make the screw as long as you want, the table can only practically go to its extents.

I agree.

I think I need to find a different hobby on my days off. I noticed I'm being way too critical and fail to notice certain things that are obvious to the majority here.

It's a different world having to live engineering everyday, and then signing on here on the Zone, watching someone design a machine who is free of any constraints or choice of materials or even parts price/availability.

Kinda has me jealous.

[rusel]

Well there is some back and forth going on here.
As I was referring to a conversion and not a scratch build design, the length of my x axis was increase with the reshaping of the end bearing blocks as seen below which allowed the table to move further.
As for accuracy you can not rely on roll screws, you can use ground screws or as I have done and use glass linear scales as seen below. Both are expensive but i believe that scale are more accurate in the long run.

Attachment 316072

[louieatienza]

I agree.

I think I need to find a different hobby on my days off. I noticed I'm being way too critical and fail to notice certain things that are obvious to the majority here.

It's a different world having to live engineering everyday, and then signing on here on the Zone, watching someone design a machine who is free of any constraints or choice of materials or even parts price/availability.

Kinda has me jealous.

Don't worry about it... happens to all of us at some point. Hey I got a chuckle out of it. And you don't have to tear up your degree

I'm jealous too, and I have a pretty capable mini-mill. Heck this spindle is worth more than my CNC router! It's fun to watch. Of course it's not engineering in its pure sense, but sometimes it's a good exercise to think of how you would build something if you had almost no restraints. Gives good insight.

Also goes to show the level of sophistication in these builds. Far ways from salvaging plotter steppers and using garage door rails...

[rusel]

Yes we all our days
For N.Constans
I would like to say that having converted a chinese mill to cnc and if having my choice over I would built from scratch then convert. I think I would get what I wanted and be cheaper in the long run.
russell