[shorton]

A year or so ago I got my first CNC going. I'm a mentor ofr a high school robotics team and the kids really needed a CNC machine to make their parts. I had always wanted to get one myself so I have the perfect rationalization (excuse) So off I went on my quest for info and parts. I'm an engineer and pretty crafty, too so it wasn't too hard to figure it out, for the basics anyway.

As luck would have it, my search led me to a small machine for sale that ticked almost all my boxes. It was built be a respected member here (louieatienza). What he had for sale was just the machine, no controller or electronics. His design was from 7/8" phenolic with a fixed gantry. I really liked the design and bought it (thanks Louie!).

One of my goals with this plan was to get the kids to do some of their designs in CAD before hacking the parts. The CNC machine was fascinating to some of them and my plan worked. This past season I got several of the kids up to speed in CAD, CAM and operating the machine itself. Our team won the championship and a lot of it was due to the CNC machine.

The machine worked flawlessly, I really enjoyed working with it and teaching the kids. But I wanted just a little more for table size. The current machine's envelope was about 14x8x3. I originally wanted one that could do 12x12, so I thought I'd upgrade this one to 14x12. It involves a rebuild of a good bit of the machine, and since my machine is sort of beginner oriented, I thought I'd document it here, if you are interested..

I didn't take a picture of every step, but I think I've got the good stuff.

[shorton]

This is the original build thread I believe FYI: You can see some of the original unit and the design choices there.

http://www.cnczone.com/forums/cnc-wo...371-forum.html

[shorton]

Machine Specs:

Leadscrews 1/2-8 2 start

Danaher motors wired bipolar parallel
Soigeneris stepper covers
MicJack Connectors

TDK SHS15 bearings all axis. Upgraded X axis bearings NOS from eBay

Limit switches for homing on all 3 axis

Dewalt DWP611 router

PreciseBits Precision Collets

G540 Driver

Electronics case cobbled from spare parts

SuperPID speed controller, fully integrated with G540

Probe pad for Z-setting

Emergency stop switch box

Rebuild, T-slot system for new table, designed to be used with flat and a spoilboard. Can access T-slots with spoilboard in place.

M5 inserts for MDF spoilboard attachment

Dumpster CNC leadnuts, Leadscrew collers

Leadscrew from McMaster CArr

Igus energy chain X and Z wire management

Misumi high rigidity extrusions, machined flat

Misumi gussets, high perpindiculariry version, milled the channel "groove" off one side to allow them to sit flat against phenolic frame.

Frame: 1" brown Phenolic

Stainless hardware on the new stuff, just for looks.

Original envelope ~14x8x3.

Rebuild envelope 14.5 x 14.5 x 4

[louieatienza]

Cool! Glad to see you got this upgrade off the ground, and congrats to your team! You should post of pics of their project, or a link; it was pretty impressive...

[shorton]

Thanks Louie, will do.

[shorton]

When it was born:





You can see a few more details on the design and build phase, especially of the X and Z axis in the original build thread. 1st 3 pages of that thread have all of the machine build

[shorton]

After I got the machine, I added a few personal touches:

Added the energy chain for cable management. Kids and wires don't mix:





Then I added slot covers to the extrusions, make cleanup a little easier, and they look nice



Added limit switches to all 3 axis

[shorton]

So the original goal was to get the table Y axis to 12" I replicated the original phenolic parts in CAD and started adjusting. My first attempt was to do the set goal, increase a little so I simply took the existing sides and made them a little longer. I started looking for quotes on getting the parts recut but unfortunately the quotes were all too high for my limited (rationalized) budget or the kids CNC machine. The original builder offered to help me out cutting the new sides but we don't live anywhere near each other so the overall costs including material and shipping on this heavy stuff pushed the $ envelope, too.

A stroke of luck. While picking up some plastic for something unrelated, I asked the local guy if he dealt with phenolic. He didn't but knew someone who did. Turned out that company was local and was familiar with the robotics team. So I sent him the drawings. We had a couple of discussions and it since his material was 48" wide and he was going to end up with some scrap anyway, I revised my plan to go to the full 14.5" on Y so I'd have a square work envelope.

The original design is very rigid. That's what I love about the machine and fixed gantry. Since I had material to spare in this new design, instead of cutting it out, I optimized the "L" shape so it was bigger on the back side and allowed room for another extrusion and I added gusset supports. The original machine has a nice aluminum bracket made from MIC6 cast aluminum, but it's attached with threaded holes down into the sides. I didn't have a good way to add those vertical holes into the edge profile of the sides, so I designed gussets instead.

I also increased the height of long part of the L so give it more beam strength for the longer travel. Finally, the local guy only had 1" so the sides got a hair thicker, too.

Attached are the 3 side profiles, original in black, plan A in Blue, Plan B/Final in Orange.

Due to the fixed gantry design, in order to add the 6" I needed to go to 14, I had to add 6" to each end. Not so much of a desktop machine now.

[shorton]

The original machine has the Y axis' step motor mounted on the rear of the machine. On the new machine, since it was getting so much longer, I wanted to move the motor to inside the 2 ends. I still had plenty of lead screw travel in that configuration, and it would keep the machine shorter and a few more wires tucked inside the machine perimeter.

My original plan was to use the original front and back parts, but this change would require moving the motor down relative to the bearings, so new front and back would be required. I new lead nut mount would be necessary, too.

Obviously I had to have longer bearings, and rails to mount the bearings, So at this point I was pretty much changing out the entire base of the machine. Much more than I originally planned.

What the hell, the new front and back pieces were relatively small and couldn't raise the phenolic cost that much. And they would match.




Side CAD work:




Profile ready to cut. Sure wish I'd have had a big machine at my disposal.

[shorton]

Got a lot done today.

Finished getting the new frame together and tightened up. Installed the bearing support rails and bearings.

Started pressing the leadscrew bearing. They got all my holes undersized and I was able to drill them out. Can't drill this counterbore though. I managed to get the bearing in there, but it's probably not coming out LOL. May have to ask him to redo these. For now, I got it in there....



The original machine used standoffs between the motor mount plate and the body. Room for the couplers and screw collers. Since I relocated the motor to the inside I had to make new motor standoffs which woudl be as tall as the motor, reversed from the original standoff direction. I have an old Southbend lathe so I got some 6061 hex stock from McMaster and whipped up a set of 4, drill, tapped and parted to just the right length. Came out nice:

[shorton]

For my new bearings I ordered the hole covers. Got the direct form THK, they aren't very expensive. Figured it would be a nice touch. BUT, those little suckers do not go in well, do not stay in well and are very difficult to work with. No way they were going to sit flush in the holes. I had to make a special tool on the lathe to press them just below the surface to keep them from popping out. The first machine cleaning with the vacuum will probably get them right back out, we'll see. Disappointed with these. Don't recommend them.

[shorton]

Here's some pictures of the frame all together. Like a kid with a new toy, I've gotten ahead of myself a few times and had to undo ome thing to do another, but that's part of the fun I guess.





As I described before, I chose to modify the detail on how the X-Z system was attached to the base. I already had some rated-for-perpendicularity gussets from Misumi. But mine are for the next size up profile. Those have M8 holes and tabs for a smaller slots. I had to machine the tabs off one side anyway so they would fit flush with the phenolic, so I just used what I had. I also had to machine the tabs off of the other side due to the smaller slot width in the 30x60 profile my X axis is made of. Those larger gussets are 30mm wide though to match the rail width. So a little more contact area. If I had to do it again, I'd use the size 8 gussets. It's the identical extrusion, but they cut the narrower. If you did this yourself, you could use the 8 series gussets adn hit the little tab/nibs with a sander or grinder. You don't have to have a mill to use them for this. I also used T nuts that have M8 threads but are for the 8mm slots. Misumi 8mm slots hardware only goes to M6 fasteners. Bosch Rexroth makes 8mm slot, M8 thread nuts, but I dislike the local Bosch supplier. Ordered some from a different company, fit fine.

Here's some shots of the gussets. I haven't machined all the tabs on the gussets off yet. (see, told you I was getting ahead of myself):





These gussets also attach to the rear support rail. I added this rail just because it was easy to add, could increase rigidity, and gave me a convenient place for a rear cover to keep all the swarf contained, and kids protected. I ran out of silver, I had 4 black left so on they went.

[shorton]

.
.
OK, finished milling the gussets.



Rearranged the black ones while I was at it. Very square, nice fit

[shorton]

.
.
Next, prepare the table.

I found the phenolic isn't as flat as I thought. But it seems to lay down on the bearings pretty well.

Mount the Lead nut to the underside. I used a scrap piece of acetal for the block. I can mill it fine, but I didn't have much that will cut it well. Table saw and band saw heat it up too much. I made it nice and long to resist any moment imposed by the lead nut being off it's axis farther than it used to be. Attached with M6 SHCS from the top. Those holes in the sides are to allow thread chips to clear, I didn't have a spiral M6 tap handy.



Next mount the limit switch plunger. I made a mistake on this part. Not sure if it was my measuring when I drilled, or my CAD dimension. It lined up, but not perfectly as it should have. Whipped up a new one and moved the plunger hole a little. On my original base's limit switch, I had mounted the limit switch to the back of the base. It simply got pushed by the table itself as it moved rearward. That worked fine, but it was out in the open, sort of, this is a little neater, tucked under the table. And fine adjustable.




I will be using MDF spoil boards on the table. Plans call for .75" thick. To attach them, I added some inserts to the phenolic. Probably not necessary, but easier to do now than later.

[shorton]

.
.
Table attached.



Assemble drive components. I got a new leadscrew from McMaster cause the old one wasn't long enough of course.

The front bearing, the one I had to press in, seems fine, It still turns with finger pressure, so no sign it's binding. But the new leadscrew just barely went through it. I checked a different flanged bearing that's part of the assy, it slid fine. I guess the tolerances on the leadscrew must be a little wide. Fortunately the first 6" or so would go through it. It only got tight on down the rod. I had planned to insert screw, measure and cut to length. But I had had to do it the hard way. Then I had to remove the motor and insert lead screw from rear.

Lead nut threaded.



Motor coupler attached




Now with the old design the "push-pull" of the motor was handled by bearings on each end of the machine. One bearing handles push, the other pull. But in the new design, the motor is inside so the rear can no longer provide support for the leadscrew axial motion. The motor coupler is a slip fit so it can't be pulled on, and you wouldn't want the motor bearings to take that load anyway.

My solution was to add a second flange bearing on the back side of the front of the base. One screw collar on the front handles when the screw is pulling table toward the the front, the "back" flanged bearing and a second collar handle pushing back when the screw is pulling the table toward the rear. Motor floats as it should. Motor end does have to "hold up" the end of the screw when the table is all the way forward, but there is no radial load other then the weight of the screw, so I figure that will be OK.

All the parts, spaced out
From the left, single screw collar, brass spacer, pressed in 1/2" bearing (can't see), front body, flanged bearing, brass collar, double screw collar


The dumpster CNC screw collars have a shoulder that is designed to go up against an inner bearing race. So I don't think the brass shims are needed at all, but I had them and they couldn't hurt, so on they went.

The flanged bearing isn't attached to anything. The hole through the front is sized so as not to touch the inner race or seals. I think that will work fine. I could always fix the outer race if it needed it. It is 100% a thrust bearing. So I think it's fine as is.

Here they are together.



Drive assy, complete

[shorton]

.
.
Limit switch, after a little rework on the plunger, adjusted and fits perfect.

[shorton]

.
.
My old base's table has these super nice, deluxe T-slot extrusions. They make up the entire table. Very heavy duty. But I didn't use them much, having to mainly cut out sheet goods. I attached a piece of MDF to the T-slots and we just screwed the materials to the MDF for "clamping". That worked OK, but it was time consuming, and I had a couple occasions where the kids didn't use the prescribed screws, going too deep and thus they dinged my T slot. No real damage, just cosmetic but the T-slots weren't helping me with the kids.

So I decided to try a different table design. My plan, which is yet to be seen if it will work, was to add some T slot extrusions. Just some single slots that woudl cover 99% of my clamping situations for sheet goods. One slot across the front, 2 spaced equidistant up the middle, and one across the back, more for looks than anything.

I plan to make some special "clamps" to fit this setup and the few thicknesses the kids work with. My original plans called for putting these extrusions on a shim to raise them up closer to the surface of the spoil board, but not to it. I've changed that plan for now, and am leaving the slots against the table. I'll use dedicated spacer blocks to work with my clamping system to get down to the slots from the table surface. Less chance of digging into a t-slot extrusion.

But, another nice thing is these extrusions are relatively inexpensive. Won't hurt ($) to replace.

The front and rear have "skirts". A L shaped cross section, so they will have some added strength, and they cover the visible edge of the phenolic table, so nice cosmetic touch, too.

My original plan for the T-slot extrusions called for button head cap screws up from the bottom, threading into the T-slot rails. But when I was drilling the rails on my little mill I made a mistake and made the holes for M4 through instead of tap. oops. No biggie, I went right back and counter bored them so I can put the screw in from the top. Nuts on the bottom. Next time I replace them, maybe I'll go back to to tapping them. But not a big deal so no need to scrap and redo now.

T slots in place:


M4 screws are just a hair wider than the slot, so you get the little wide spot.


If you look close you can see the front "skirt" of the front profile. These cross sections are in the CAD show previously, too.


One last detail with this design which is on purpose, is there is no shearing potential where the table overlaps the frame. My original machine, the T-slots do overlap just a little. In the new design the front of the table butts up against the frame. That way if someone has a finger in there when a move control is executed or pushed, said finger may get smashed, but it won't get cut off . May need to stick something in there and see what it takes to stall these motors just so I know.

[shorton]

.
.
Pretty much ready to transfer the X-Z axes to the new larger base.

Hoping to get to that over the next few days. Start the new wiring, and post some info about my electronics.


Hope this helps someone. I love this design. Very scalable for small to medium machines. If it gets much bigger than this it starts getting more practical size wise to have a moving gantry I think. But I really like this one so far.

[shorton]

I have no idea what happened to all of my photos. Hate to start over. I posted a question in the forum problems subforum. Meanwhile, anyone know what's up? Did I hit some photo limit?

Nevermind, I reloaded the photos. Not sue what happened to them. Seem to be OK now...

[louieatienza]

This is looking great... and bullet-proof (literally!) I think you should be just fine with the stepper bearing holding up the screw. In fact these particular steppers are designed with oversized bearings so they can directly drive a load.

I still believe in the couplers though as a vibration and shock buffer for the steppers.