Comments needed for this design...thx!

[andyew]

Hi all,

It's been some 4 months doing research from ground up for both electronics and mechanical parts of my coming 1st CNC machine! After some longgggg time reading some hundreds of eye soaring posts from here and there, i've come out with this design (mixture of many expert designs i've seen from pictures)!

I need yours inputs about the whole structure especially the Delrin plastic homemade slide/guide on round tube (stainless steel) for all the axis. The X-axis has "half" of a cube that stays on the tube (or simply pipe) while the other Y and Z have "full" cube with hole that hug on the pipe.

The leadscrew is 12TPI ACME of 3/8" diameter and the stainless steel pipe is 30mm diameter. I'll use, as seen, MDF of 30mm thickness for the whole structure. Delrin, as said above, will be use for slider/guide on pipe, leadscrew homemade nuts (for tight fit) and also the thrust bearing pillow. Delrin is represented in "orange" color..... (except for the spindle which is B&D RTX for me)

One notable structure i've chosen is that, the linear guide has wider base than the leadsrew nut base. THUS, X and Y axises have longer travel distance WHILE keeping the load stable enough. That is why the stepper (of X and Y axis) is mounted "innerly" where the moving base (of X and Y axis) will clear over the stepper when reaching the ends....

OK, please do give me a hand and wish to do it very very soon after getting yours expert point of views! (worrying that my design is only a crap and it wont work at all!!!)

Thanks!
Andy

[andyew]

As for dimensions, it has working area of (X)770mm (Y)670mm (Z)70mm. The Z is not deep as it will probably cut stuff about that thick (70mm is around 2-3"). It'll be a waste if Z is designed to have "extra" travel distance and it'll be less stable, i think.

Delrin blocks used here are all 50x60x50mm so that it's easier to produce using simply boring tools for the 30mm holes.

Let me know if you need more angle or zooms of the design, thanks!
Andy

[Graham S]

Just sitting you gantry on the rails is asking for trouble. As the router plunges the gantry will want to lift. If you used an over half cube then that would be OK.

You might want a few lengthways stiffeners in the bed to stop it drooping in the middle.

You could also do with some bracing of the Z axis. A back plate and then some gussets would be OK. Four slides on the Z would also be preferable.

Other than that get building.

Graham

p.s. You can mount your steppers outboard if it helps but make the ends of the leadscrews inboard.

[The Wizard]

Hi Andy,

Just looked your design over and one thing I don't see is any means of adjusting the rails for plane and parallelism. Without some means of setting up your rails you will never get the machine to work. Simply drilling the holes for the pipes will not work unless the machine you use to make the frame is accurate to a few tenths of thous (.oox mm). You will find the slides will bind if the rails are tapered or they will 'rock' or cut tapers if they are not in the same plane or square to each other.

I am about to start building my first CNC router/engraver too but I have been around machine building (on and off) for over 30 years. I started out in the early 70's as an engineering apprentice and toolmaker and then in later life I worked as a technical writer/technical illustrator for a number of machinery manufacturers (and even freelance for a while). Getting the machine set up true and square in all planes and axes is probably the most important factor of any machine building project.

It may be that you have already taken this into consideration but not shown it on your sketches, if so great, I just wanted to give you a 'heads up' on possible problems down the road.

HTH,

Larry

[Graham S]

His X axis rails are adjustable in terms of parallelness, the flatness would be dictated by the flatness of his base which hopefully will be flat

The Y rails may be tricky but some folks have managed, making the end plates as a pair can help and doing the final glueing with rails attached should also be a benefit.

Graham

[andyew]

Thanks Graham and Larry for your suggestions and notes.

Graham:
- The X linear pipe is supported by another piece of MDF and both ends fastened with brackets. I thot this will stiffen it from deflecting in the middle? (btw, instead of half cube, i'll use a 3/4 cube for the delrin bearing)
-Let's say if i mount a back plate for the Z, you mentioned using 4 slides for the Z, how does it look like and what is the advantage?
-Lastly, what do u mean by glueing with rails? Where to glue at?

Larry:
-As Graham said, the flatness and parellelism are determined by the flatness of the MDF i use. But you're right, this isnt going to be working as the material is NOT going to be perfect flat. As for adjusting the parallelism you mentioned, could you explain more about how to do it? Is it by using some kind of a bracket with 4 adjusting bolts that hold the pipe? correct me if wrong....

BTW, i've "copied" the Y design of http://www.team-haase.de/images/CUT_2000_gross.jpg where the back linear shaft is put further away from the leadscrew compared to the front linear shaft (the Y). So, it'll balance the weight of the Z-axis while keeping both the linear shafts horizontal and more balancing can be achieved. Just a thot...

I know i might sound idiotic but help me pls..... )
Andy

[Patrick2by4]

It looks to me that you may have 'racking' issues with the gantry. The y-axis, as it pushes into the work piece from the side will cause the rectangular gantry to become a parallelogram. You need to put angle blocks or something to help the gantry to keep it's squareness.

[Graham S]

Andy,

With your Z axis as it stands it will have very little stiffness, the whole thing will tend to twist and is also likely to tip back when you plunge. The machine you link to is ali and the rails on it's z-axis go into deep holes in the Y block. I just don't think you will be able to make it as strurdy as this in MDF. I have attached the modification, the gussets need not be so thick of course as they will be in compression. Using 4 carriages on the Z instead of two will reduce play and spread the load. The machine you link to probably uses 4, either that or two long ones.

By glueing I was refering to the assembly of your y axis, when you come to screw/glue or join endplates if you have the rails and the carriages installed already you can ensure that there is at least enough parallelism to get movement of the carriage.

Graham

[mvaughn]

I'm using much the same tactics as your delrin bearing design. I've found that by using hand tools that it is very difficult to get the bearing blocks perfectly aligned. However, make sure you are precise as possible and you should be fine. Having said that, I don't see any means of adjustment for you y-axis rails.

My other suggestion is to carefully consider your horizontally aligned rails of your y-axis. In that configuration any weight is unevenly distributed on the the leading rail. IMHO it would seem that deflection of that rail will be greater because of the torque. It the rails were vertically aligned the torque should be lessened because the zaxis would be closer to the rails.

Here is a link to my machine in progress... your design is similar yet different.

http://cnczone.com/forums/showthread.php?t=4716

Cheers,

[mvaughn]

Also, I forgot to mention that I would add some stiffening to your table. Maybe some square tubing under it for support.

[The Wizard]

Originally Posted by andyew Thanks Graham and Larry for your suggestions and notes. Larry: -As Graham said, the flatness and parellelism are determined by the flatness of the MDF i use. But you're right, this isnt going to be working as the material is NOT going to be perfect flat. As for adjusting the parallelism you mentioned, could you explain more about how to do it? Is it by using some kind of a bracket with 4 adjusting bolts that hold the pipe? correct me if wrong....

The blocks with the 4 adjusting bolts can be used to align the pipes/rails yes (that is the method I will use when I start building my machine in a few days). You will also need a method of checking that the pipes/rails are parallel in both planes.

Assuming the axis (doesn't matter which one as the principle is the same for all of them) is laying flat on a bench the up/down parallelism can be taken care of by using a set of 'blocks' that are all at an identical height. These could be steel (like parallels), plastic or MDF scraps. The blocks need to be high enough to bring the centre of the pipes/rails up to the desired location.

Note: If you are using linear bearings/slides that are full coverage or cover more than 50% of the diameter of the pipe/rail you will have to put these on before setting up your pipes/rails. If you are using skate bearing type carriages these can be added after the pipes/rails are set up.

Place at least 4 blocks (6 on longer runs) under the pipes/rails so that they are supported without sagging. Decide which of the rails is going to be the 'master' rail and using a square make sure it is sitting square to the end face of the frame. Now adjust the up/down screws at each end until they just 'pinch' the pipe/rail. Check that the pipe/rail is correctly positioned in the side to side plane and if correct adjust the side to side screws until they just pinch the pipe rail too. Leave the blocks under the pipes/rails at this time.

To set the remaining pipe/rail you will have to make a 'gauge block' out of a piece of scrap MDF that is the width of the centre line distance between the pipes/rails MINUS the diameter/width of one pipe/rail. Make sure that the ends of the MDF block are perfectly square to the sides and parallel with each other.

Insert the 'gauge block' between the pipes/rails at one end and make sure it is sitting against the 'master' rail. Bring the other pipe/rail into contact with the 'gauge block' and adjust the side to side screws until it is just 'pinched'. Repeat this process at the opposite end of the axis. Now screw in the up/down adjusters until they just 'pinch' the pipe/rail. Check in several places along the length to ensure the 'gauge block' just passes between the pipes/rails. The 'gauge block' should just slide between them with a little drag but no slack. Once you have the pipes/rails set you can double check that all the screws are fully tight and then lock them in place with the nuts. Then once more use the 'gauge block' to check the distance between the pipes/rails and also check that there is the same amount of tension on all the blocks under the pipes/rails. If there is an error then you will need to re-adjust as required.

Once everything is set up correctly your pipes/rails will be parallel in both planes and your slide should not bind or rock. Remove the blocks under the pipes/rails once set up is complete.

PHEW!.......hope you can follow all that and if anyone sees an error in my methodology please shout up!

Larry

[andyew]

Graham, i thot u said 4 shafts and now i got u! 4 carriages instead of 2 for Z will be stronger. Thanks for the "touchup" you made on my pic!

MV, the reason i use horizontal Y axis instead of vertical is, i m totally inspired by the German design (team-haase.de) where the back shaft is pushed further away for balancing purposes. While with Vertical design, from side view, the spindle and its carriage/shaft pose a whole heavy block "hanging" on 1 side. However horizontal makes its weight distributed on both the shafts. Anyway, i m still opened to any design which is easy and good! )

Larry, got u! i assume what you're trying to say is that, using a "gauge" block for both parallelism and flatness adjustments.... will work on that!

Btw, suddenly i have another idea for the linear motion, i might use square steel solid rod (instead of stainless steel round pipe) AND skate bearings mounted perpendicular to each other on L-bracket (********** design). I think it's better than using ROUND rod/pipe because the bearing will roll across over wider "contact surface" and thus less pressure (force/area). This will reduce the wear on the rails (solide steel square rod in this case)....

Orite, thanks and keep posting (although i m always away when u guys r online since the time zone here is almost reverse from US)!
Andy