my first cnc router, peer review wanted

[tgrandahl]

Hello all,

I am in the beginning design stages of a midsized bench top hobby cnc router with plans for construction during my upcoming winter break in ~4 weeks. I would like to be able to work with a range of materials: plastics, woods, light non-ferrous metals, but mainly pcb etching.

I am also trying to make good use of the plethora of components i already have lying around: all electronics and steppers, gearing, frame materials, z-axis lead screw and bearing. I am by no means restricting myself to these components but would like to make use of if possible.

So on to machine specs;
-Working area of at least 24"x6"x3"
-Resolution of 0.001"
-Cutting speeds between 50-150 ipm

I have already worked up a rough base and gantry design around materials i have and some linear rails i am closely watching. I attached a picture of the frame and a dimensioned pdf.

The rails and bearings are rexroth pn#1622-194-10. The frame would be constructed from 1.5"x3" aluminum extrusions with 1/8" wall, and .5" thick mill finish aluminum.

I am hoping to be able to drive the machine with a single center mounted ACME screw on each axis. The machine will be lightly used and well taken care of.

My current thoughts are to use a 1/2" - 8 ACME 4 start screws, with dumpster cnc anti- backlash nuts.
I have 250oz/in, 200 steps/rev steppers.
I can gear them down 3:1 with the pulleys i already have.

so at 3000 steps/sec i should be able to obtain 150ipm, while staying within the torque curve of my steppers, and have decent efficiency from the lead screw. Does this all sound right?

I was under the impression that racking was typically not an issue when rails are at least 1.5 times their width in length. My x rails are 39" long, and 18" apart on center. The y rails are 18" long, 2" apart on center. Each axis will also be using 4 bearings.

what do you think?

[ger21]

What kind of drives are you using? You're talking about 3000 full steps per second. Most drives would be 1/8 or 1/10 microstepped, which gives you 24,000 or 30,000 steps (micro) per second. You probably would be fine geared 2:1, with more speed.

[tgrandahl]

Originally Posted by ger21 What kind of drives are you using? You're talking about 3000 full steps per second. Most drives would be 1/8 or 1/10 microstepped, which gives you 24,000 or 30,000 steps (micro) per second. You probably would be fine geared 2:1, with more speed.

ger21,
I have three Keling KL-8078 stepper drivers, from what i have read there good drives. micro-step options range from 2-256, the computer that i will be using has posted stable with a 100Khz mach3 kernel. Does this limit my step output to 100Khz as well i assume?

I am not sure if there are speed limitations for the driver though

The reason i wanted to gear it down originally besides already having the pulleys was for resolution. The smallest components i will be working with are 0.4mm or ~0.016", so i thought a resolution 1/10th that or so would be appropriate. With 3:1 gearing i would have a resolution of ~ 0.0008", perhaps i don't need to gear it down at all but i would rather have resolution than speed.

Does my design as of now look acceptable for using 1 lead screw per axis without trouble?

[jalessi]

tgrandahl,

Your design looks very functional.

Make sure you have enough space to fit your screw and ballnut for the X axis, in your PDF it looks shallow.

The stepper mounting may also be a issue because of the two inch wide upright, the screw will offset the stepper.

Looks good otherwise.


Jeff...

[tgrandahl]

Jeff,
Believe it or not i have this crazy idea of putting the lead screw inside the y-axis supporting tube. I was thinking that an ACME screw would be more vulnerable to contamination from cutting debris etc. than a ball screw, so enclosing it would be nice.

Using dumpster cnc's cut down flange: http://www.dumpstercnc.com/images/fl...98_cutdown.jpg
This should fit inside the tube with at least 1/8" of clearance on each side, im not sure if this is cutting it to close. My one worry with the design is the distance the nut will be from the connecting y plane, the y bearings offset it 1" plus at least .5" to the nut inside the support. Im not sure if this connection will be to prone to flex under stress even with a .5" thick flange.

I am with my family for thanks giving until this saturday, when i am back i will be able to draw up a model of what i am talking about. This is still a very open part of the design however so suggestions are welcome and appreciated as always.

Thanks guys and happy holidays,
~Tyler

[acondit]

Originally Posted by tgrandahl Jeff, Believe it or not i have this crazy idea of putting the lead screw inside the y-axis supporting tube. I was thinking that an ACME screw would be more vulnerable to contamination from cutting debris etc. than a ball screw, so enclosing it would be nice. ~Tyler

The problem is that then you have to have a slot running the length of the Y tube to allow the nut to drive the axis (unless you use some kind of cable setup), and this will weaken the tube and make it more prone to flex.

Alan

[tgrandahl]

Acondit,
I was planning on using a slot so you have a good point about it impacting the structural integrity of the beam, although those 0.5" end flanges will be welded on.

There are so many different orientations that this axis could be configured in... I think it would be quite easy to just mount the lead screw on the back of the tube, would a setup like this be prone to chattering? It seems very common for a lead screw to be centered between the bearings if only a single one is used.

Perhaps i will draw up some quick examples of the other configurations i had in mind when i am back home, i can think of at least 3.

Thanks,
Tyler

[acondit]

Originally Posted by tgrandahl Acondit, I was planning on using a slot so you have a good point about it impacting the structural integrity of the beam, although those 0.5" end flanges will be welded on. There are so many different orientations that this axis could be configured in... I think it would be quite easy to just mount the lead screw on the back of the tube, would a setup like this be prone to chattering? It seems very common for a lead screw to be centered between the bearings if only a single one is used. Perhaps i will draw up some quick examples of the other configurations i had in mind when i am back home, i can think of at least 3. Thanks, Tyler

Tyler,

Take a look at the Joe's CNC Model 2006 forum. You'll see a number of machines where the screw is not directly between the "linear bearings".

Here is a picture from rdhharm's post about a modification that he made that shows very nicely the relative positions of the screw and the pipes used for bearing surfaces.

Alan

[tgrandahl]

Alan,
I looked through the joes cnc 2006 forum and saw a few interesting orientations. I like the design joe used in his own build shown in that picture you posted. I am wondering now how my y bearings should be oriented, It would seem stronger directly under the z plane like i had originally planned. but i could mount them on the top and bottom too similar to the shafts in joes design.

On this note should i be using bearings with a pre-load? and if there is no pre-load how should this affect the bearings placement?

~Tyler

[acondit]

Another option would be two Y beams instead of one with enough space to fit the screw between them. Generally speaking the wider the spacing between your bearings, the smaller the effect of slack or looseness in the bearings.

To exaggerate if you have two bearings 1 ft apart and 0.012" slop in the bearings you would see much more movement at the spindle tip than if the bearings were 12 feet apart with the same slop. That being said the kind of linear bearings you are showing tend to be very precise.

My personal preference is to have the bearings as close to the Z-axis as possible and as far apart vertically as reasonable. These things are always constrained by the specific design and the wallet.

Alan

[tgrandahl]

Ok so getting back in the groove of things I have some screen shots of my two current models.
First the one with the lead screw mounted on the back similar to joe's design, this is still very rough and I am just trying to get an idea of where to locate things. I dont think there will be much of a difference between mounting the beatings vertically or horizontally.

The pictures show mounting holes for a nema 23 stepper, but because the screw will be belt driven i should be able to shave that down quite a bit.

The second shows the cut out that would be needed to run the lead screw inside the y axis. I am not skilled enough to do a stress analysis of this model so I suppose it will be best to leave this idea for now.

I want to get the x and y situated before I start on the z axis. I may have some more time to work on this tonight and if so will post updated images.

[acondit]

Originally Posted by tgrandahl Ok so getting back in the groove of things I have some screen shots of my two current models. First the one with the lead screw mounted on the back similar to joe's design, this is still very rough and I am just trying to get an idea of where to locate things. I dont think there will be much of a difference between mounting the beatings vertically or horizontally. The pictures show mounting holes for a nema 23 stepper, but because the screw will be belt driven i should be able to shave that down quite a bit. The second shows the cut out that would be needed to run the lead screw inside the y axis. I am not skilled enough to do a stress analysis of this model so I suppose it will be best to leave this idea for now. I want to get the x and y situated before I start on the z axis. I may have some more time to work on this tonight and if so will post updated images.

Another thing to think about is, think how much more you would have to tear down in order to fix the nut inside the beam, if anything ever goes wrong (and it always does).

So even in mounting the screw, think about mountings that make it easier to mount and remove (like a bearing plate) that bolts to the outside of the beam flange with a hole big enough in the flange to pull the whole screw through.

Alan