MIcroMill CNC: X andY Axis

A project in process

(aren't they all?)
Having more or less done the Z axis (at this instant, the bracket needs to be done and the connectors wired), I decided to make the X and Y axis couplers and mounts.  They will be substantially the same, with the possible difference of where the connectors will be located, and perhaps some minor things here and there.
This is part two of a 3 or so part series, covering the Y axis.  I'm not sure what part three will be about, possibly the differences between the X and Y axis. 
I decided to use the existing mounts, (after buying an extra set) and then put in thrust bearings in the mounts.  I'd need some couplers, and the next thing is to do them.

Naturally, the coupler design depends on how the motor is to be coupled to the shaft.  I decided to use a spider coupler.  Now there are numerous ways to handle this, and you could argue that the best way is to belt drive the screws.

However, I decided to use the same scheme (however good or bad) that I used on the Z axis. 

The coupler screws onto the existing shaft, which preloads the bearings.  A setscrew locks it in place, so the setting should stay.  Unlike the Z axis, which is belt driven, this is directly coupled.  I will turn down the coupler body to accommodate the spider coupler.  The ones I got were about 1/8 to about 5 mm, and were bored out to 8 mm or 5/16 for the motor shaft.  On the coupler side, I will pick 8 mm (in case I want to do another scheme for attaching the coupler.  On the stepper side, 5/16 to match the stepper shaft.

 
Similar to the z axis, the coupler is bored out to clear the shoulder of the leadscrew, and then threaded to accommodate the threaded part of the screw.  I've turned down the end to fit the spider.  The material is 3/4 inch aluminum.  The nub is 8 mm in diameter and about 0.200 or so long.
I milled flats on it so that I could hold the coupler with a wrench.  It's about 0.60 across the flats, so it is not intended to be a particular size, just kinda decorative.  I wanted the extra 0.050 thickness because this overlaps the bore for the threaded part. 

Dimension wise, the shoulder was about 0.6 inches, the threaded part roughly the same, and I wanted about 1/4 inch solid material between the end of the threaded hole and the nub.  You can see that the flats overlap this.  The setscrew is 8-32.

 
The coupler as it is intended to be mounted.  The thrust bearings are not yet available (backorder at McMaster-Carr), so I will use the pieces as delivered.  I had one very low load thrust bearing, so that's in there right now. 

I found that the amount of preload on the screw is critical, since it's a bare metal to metal contact, you get lots of friction.  The thrust bearings will take care of this.

As before, tighten the coupler to preload the bearings, (you can probably lock the gibs to do it), then tighten the setscrew.
The stepper will be way the heck out here.  I'm not happy about the distance, but I think that with heavy enough brackets and so on, it will work. 

Right now, I'm looking at two ears that go over the exiting holes (which are already M6-1.0), and then two horizontal bars, with a motor plate between them.  Depending on the size of the plates, I might put a cross plate for stiffening.  It should be above the coupler and protect it from any sort of debris.

 
Just a closeup of the flats being milled on the coupler.  I'm using a spin indexer with a 5C collet, and cutting a flat at 0, 60, 120, 180, 240 and 300 degrees.
Couplers 1 and 2.  More or less identical, complete with little machining  "oopses" here and there.  Nobody will ever notice, I hope.

Next, on to the brackets.
I spent a bit of time trying to center the casting in the 4 jaw chuck.  Then I decided that there's a trick to it.

Trick is this.  For an inside hole, use a spud or a rod with a center drilled end.  Using a spud or the center on the tailstock, align visually.  This gets you close.

At that point, use the centering indicator in the tailstock.  Pick an axis.  Find the maximum minus deviation on that axis.  Turn the dial so that is zero.  All the deviation will be plus at that point. 

Turn to 180 degrees from that point.  Measure the deviation, and adjust (with two chuck keys!) to 1/2 the deviation at that point. 

Do the same for the other axis.

Repeat as needed.  With this type of centering indicator, it's quite quick.
Here's the casting centered.  I'll use the lathe to trim the ears flat.  Since I don't have the bearings, I have to wait a bit.  I don't trust their figures quite yet.

 
Here's the result.  I was thinking of putting ears on this, to make a partial box.  Call it a U shape with inner ears that would bolt using the same bolts that fasten the casting to the table.

 
I decided to do something different.  I would not have enough space for the ears, so I milled the ears flat.  This ought to get it close enough, I hope.  IT does mill the whole square surface.  In retrospect, I may not have needed all that.

(Actually, considering what I did do for a motor mount, I did need the thing milled flat.)
A bit of thought showed me that there is a design that ought to be rigid enough.  Extend the casting mounting screws through two long rods, and use that to support a plate at the end of the rods.  The assembly can be made a bit more rigid if needed, I think

Then the stepper can be mounted on small standoffs that allow the mounting surface to be parallel to the plane of the table and perpendicular to the leadscrew. 

To do this, I will need 4 spacers, each 3.6 inches long.  The mounting plate is .25 inches, and the motor will be 0.25 inches above that, which ought to give me enough room for the nut for the M6-1.0 threaded rod. 

I actually had some.

 
Four spacers, each bored to 0.25 inches (will clear M6 hardware), and pretty much the same length.  If needed, I can turn down the ends to fit the counterbore in the casting.  If I do that, then the spacers that mount the motor will have to be longer than 0.25 inches.

 
This is the idea behind the mount.  Imagine a plate across another spacer, then the motor mounted on that.  The main problem will be the torsion on the mounts, but I'm thinking that it will not be too much if the clamping is tight enough.  If not, then it's possible that I can put a cross bar across the two braces.  That ought to keep the torque down.  Main problem would be the amount of aluminum that I have available for the threads.  Ought to be enough.

 
Here's the motor mounted.  The adjustments are a bit crude, but will work.  This gives you a way of managing to adjust for some axial misalignment.  However, we're not done yet.
Note that the motor shaft is cleverly not yet coupled, but it aligns well enough.  I need to ream out the other part of the coupler to 0.375.  The mill end is cut to 8mm.
I cut the studs a little shorter.
The other end of the coupler. 
There's a small bearing between the coupler and the mill support.  It's not going to be there in the final build.  More on that later, after the bearings get here.
Add the spacers.
Mount the motor plate, which has a hole in it large enough to pass the coupler and the setscrew.  That's a 7/8 inch hole, by the way. 

To adjust the motor, screw the capscrews in (they're metric), then tighten both nuts.

 
A closeup of the small bearing.  I could probably leave it in there, but I'll counterbore for the thrust bearings on both ends of the leadscrew.
There's a little clearance between the end of the screw and the motor.  If I wanted, I could put in spacers.  This arrangement also allows you to adjust the spacing on the coupler.
See?  Glad you waited?  This is the whole setup so far.  Even though the motor is out at the end of a long standoff, it's rigidly mounted. 
A view of the Y axis looking towards the X axis.
A view of the spacer and the clearance on the rightmost gib.  There's enough clearance so that I don't have to worry.  This was actually a design goal.
Clearance on the left side.  I think that the gibs are going to be lapped a bit.
This is the maximum forward extension, giving me some idea of the maximum milling area. 
About a good idea of how forward the table will go.  More than this, and I think that I'm moving too far forward.
Testing the X axis.  The Y axis works.  The axis is a bit tight, so it will need some work.  I rather thought it would.

The stepper tester is the little blue box, as always.  Since the steppers are unipolar (run as bipolar), they can be run by this box. 

I got a step rate of about 400 steps/second which is 2 RPS or 120 RPM.  With a 20 TPI screw, that's about 6 IPM at the best. 

Working with the ball bearing says that the right to left movement (with the bearing) is a maximum of 400 or so steps/second.  The other way is about 250 steps/second, which is a metal to metal contact. 

Opinion, you really really need those thrust bearings.
The whole setup running the X axis.  The table is at the maximum rightmost movement.  The  stepper is a 4.5 volt stepper that I'm running with about a 5.6 volt supply (remember that this is unipolar, so no chopper drive at all, you feed the motor the voltage on the coils plus the drop in the output transistors (about a volt or so).

The little box is the frequency counter. 

I managed to clean up the bench before I started full scale testing.  Next to do is to make the brackets and mount the motor connections.  I also need to think about where the home and limit sensors need to be, and I'll have to wire them up.

Maximum rearwards movement, maximum left and right are also indicated.

 
Wonderful news (so to speak).  I got the bearings early, so I now have four thrust bearings, each that will handle over 400 pounds or so. 

First, we put the bearing block in the 4 jaw chuck and center it.  Remember to use the center indicator.  I got it tuned almost exactly.

These are the thrust bearings.  You note that unlike the "normal" ones, these have formed washers.  What you cannot see is that:

1) the two outside races are very slightly different in diameter, say a thousandth or so.  (actually, about 0.008)

2) the inner race is smaller than the outer races.

 
I'm going to bore out the recess for the bearing.  It's going to be about 0.628 wide by 0.195 deep. 

This is important, in terms of a setup, so I'll spend a bit of time on it.  On the right is my dial indicator on a magnetic mount on the tailstock.  The little adaptor is something I made a while back.

The block is centered with the 4 jaw chuck, how is covered earlier.

Take the lathe boring bar and move it until it contacts the work surface.

Then set the dial indicator for the desired depth of cut.  You do this by moving the dial indicator in until it reads the depth of cut.  Remember that the indicator reads as it is pushed in, so you cannot get a reading unless the indicator is pushed in by the carriage movement.

Since we don't have that situation, we adjust for the maximum reading, then go down to zero.
As you bore the hole, the indicator will go down to zero.  Stop when you get there.

This is at the start of the cut.  Try not to move the magnet mount, but it does work. 

The really nice thing is that cast iron is very very nice to machine.

 
This is at the end of the cut.
A look at the boring process.  This is almost done.  The smallest cutter will just pass the boring head. 

Since the setup is almost identical for the other block, I'll cut it next, then flip the block and do the other bearing surface. 

I'll probably bore down the entire width, then add 0.195.

This will preserve the existing amount of shaft out the bearing block, which is about what the coupler is designed to do.

 
This is the block as bored. 

The little step is exactly what I wanted.  It's about 0.100 deep, which is about half the bearing thickness.  It's about 0.030 wider than the whole inside bearing.

Remember that I said that the two bearing races were different sizes.  This is machined to pretty much exactly fit one of them. 

The bearing itself is smaller than the race, so it will not contact the block.

The outer race is the same size, but you want to make sure that the bearing race will turn with the coupler and not drag on the block.
You can see the milled ears of the block, which were needed.  You can also see the bearing itself.
Now you can see the bearing clearance.  It was arbitrary and was about 0.030 wider.  It's just exactly what is needed for this.

The bearing sticks up just a little above the surface of the block.  If it rubs, then all I need to do is to add a small washer to space out the connector.

You can also make a small shoulder on the coupler so that the coupler is slightly smaller than the bearing surface.  That's the best way to go.

When I get the whole bearing assembly done, I'll ream out the middle to 8 mm.  You'll see why I need to do that later.
This is the original bearing block (with label).  You can see just exactly how far out the bearing sticks. 

I also managed to take off the edges on the milling nubs on the key slot.
As done, the Y axis.  I had to place a small washer between the bearings and the end of the coupler.  It has nothing to do with friction, but a lot to do with how deeply the coupler is bored.  I needed to have the coupler bear on the thrust bearing.  Without that washer, too much gap. 
The back of the bearing block, ready for boring.  You want to bore out 0.195 deep past the inner shoulder, and bore out to 0.629 to 0.630. 
You want it to look like the bottom one.
Center the inside bore of the bearing block.  Procedure has been given earlier.  The center bore is the part that needs to be concentric to the shaft.
Rotate the dial so that everything reads positive, then adjust the chuck for half the displacement. 

When you have the thing properly adjusted, you will read that same zero on all the directions. 

Boring the hole for the thrust bearing.  Take care that the end of the camlock on the tool post does not run into the base of the bearing block.

 
Yes, it fits, and it's about 0.002 out a bit.  So it will wobble a trifle, but only about 0.001 or so.  Not bad.

Take care to run the tool back and forth a bit over the full depth.  Not only does it polish the hole a bit, but it also makes sure that the hole will not suffer from tool spring. 

It will also make sure that the hole is consistent through the entire depth.  That allows the bearing to sit on the bottom of the hole.

 
Bore out the center.  I made it 8 mm.  The screw shaft is now riding on the thrust bearing, and does not need the bearing block to align the shaft. 

This could have been a mistake, but it does allow a bit more slop, and the bearings are fairly tight on the shaft, so I won't get much radial slop.

Based on the initial tests, the X axis works wonders, and the Y axis needs to be played with.  However, it is the same need that it had before I put the bearings in.

Right now, my opinion is that this is a very nice modification, and cost me about 35.00 USD for the bearings.  They are 8 mm heavy duty thrust bearings