[momus_cnc]

As has been mentioned in a few other threads, some people have had problems with the carriage shifting diagonally during use. So far the solution to this has been using 5/16” grade 8 bolts to replace the four 1/4-20 studs that clamp the carriage together. There are a few threads that discuss this change. This greatly increases the available clamping force and I have not heard of anyone having any problems after upgrading to the 5/16 bolts. However, I've been working on an upgrade for the next release of the plans that will provide a positive connection between the parts, and not rely entirely on frictional forces. I've attached new drawings of these parts for use on existing machines (plan version 1.2 and older.) I'm looking for some comments and feedback before going ahead and integrating this change into the next release.

The upgrade involves fabricating five new parts: new left, right, and lower carriage blocks, and two new lower spacers. The lower block is slightly taller, to allow for the insertion of four roll pins, which positively lock it to the left and right blocks, as well as the spacers. The additional height (1/2”) may reduce your Z axis travel slightly. On the forthcoming version 2.0 the Z axis will be redesigned slightly to compensate for this, so it will actually have more usable Z travel than version 1.2. Version 2.0 will also have slightly modified bearing blocks, so that the spacers will be .75” thick instead of the odd size.

The new lower block as it is shown here is slightly more complex to fabricate than the old version (I'm contemplating simplifying it, see the end of this description), but it makes carriage assembly much, much easier. No more trying to hold 8 parts in perfect alignment with clamps while tightening the nuts. This design also makes these 5 parts essentially permanently attached to each other, but again, makes overall assembly much easier.

Photo 1- The new parts. The 3 parts at the top center are the same as the old ones. The other 5 are new.



Photo 2- I began carriage assembly with the 3 parts shown.



Photo 3- With an accurate square, scribe guide lines that correspond to where the edges of the spacers will be when the parts are assembled.



Photo 4- Clamp the 3 parts together with 5/16” bolts. Use the scribed lines as guides to keep the left and lower blocks square to each other. Double check that they are square to each other and fully tighten the bolts. Triple check that left and lower blocks are square to each other.




Photo 5- Flip the assembly over and clamp it into a vise on the drill press table. Check again that the parts are square to each other. Insert a 3/16” drill bit into the chuck. Lower the quill next to the parts to make sure that the drill press spindle is parallel to the edges of the parts. You'll be drilling through the thickness of all 3 parts, so the drill needs to go as straight through as possible. Carefully align the drill bit with the existing holes of the lower carriage block and drill down through the other two parts. Align the bit carefully so that you don't ream out the existing holes.



Photo 6- Take the parts out of the vise and disassemble them. Mark them first so that they can be put back together with the spacer in the exact same orientation. Carefully clean the parts of any oils, cutting fluids, etc.. With a hammer, drive two 3/16” x 2” roll pins down through the 3 parts to assemble.



Photo 7- Repeat the procedure with the right side parts. For accurate alignment, use the square as well as taking multiple measurements between the left and right side carriage blocks to make sure they are exactly parallel.



Photo 8- With no bolts installed, the assembly should be very rigid.



Photo 9- Install the remaining carriage parts. Install 5/16” grade 8 bolts and fully tighten.


The entire assembly that is seen in the last photo only took me a couple of hours to fabricate. That includes fabricating all of the parts shown, as well as the final drilling and installing the roll pins. I did use the Momus CNC machine to mill the corner notches and 3/8” holes out of the lower block, but otherwise all was done by hand. I haven't installed the new carriage on the machine yet, as I'm constructing an entirely new machine to test out the version 2.0 plans.

I was a little worried that this design would require lots of precision to assemble successfully, but it all went together quite easily. I actually built two of them to test the procedure. Getting the parts aligned before drilling was actually very easy and quick, since you are only dealing with a few parts at a time. Much easier than the old method of carriage assembly. And once it is assembled, those parts are aligned forever.

I've been debating on whether the new lower block should have the notches at the corner as shown here, or if it should be a simple rectangular bar that is 1.25” over its entire length. Cutting the notches makes it look a little more elegant, and it keeps the bearing bolts a consistent size. Not putting the notches would make the part easier to fabricate, and would only require the use of 2 longer bolts.

Thoughts, opinions, comments?

-Bob

[momus_cnc]

Here are the drawings of the new parts as shown in the post above.

version 1..3 page 49.5 part 16.5.pdf

version 1..3 page 55 part 22.pdf

version 1..3 page 56 part 23.pdf

version 1..3 page 57.5 part 24.5.pdf

[OCNC]

I'm at a loss to understand why you would want to perpetuate the use of the 0.678 dimension. Using multiple components that have to be hand dressed to accuracy in dimension as well as flatness seems quite incongruous with the idea of a DIY machine, particularly an entry level one. I could never support something like that and I hope you'll work that out of your design.

I've posted some images showing the dimensions that I've used for the anti-racking plate Z-axis I'm currently using. The assembly procedure is to pre-drill the 3/4" square stock and the 1/2" x 3/4" flat stock as shown using a drill press. The pieces are then assembled and accurately squared using the 5/16" corner bolts to hold the assembly in place. Then using a hand electric drill and the #21 tap drill (10-32) continue to thru drill the holes in the spacer bars and the anti-racking plate using the holes in the 3/4" square stock as drill guides. Next tap the holes without disassembling. Then disassemble and countersink the holes for flat head screws (10-32) on the exposed side of the 1/4" plate using the drill press again. When disassembling be sure to mark all parts so they can be reassembled in the same orientation in which they were drilled. Then reassemble and check for squareness. Using screws and tapping through all pieces allows me to remove the z-carriage from the y-axis without loss of squareness. This gives me direct access to the back bearings on the y-axis which are otherwise trapped by the z-axis assembly. In the stock design the only way you can do work on the z carriage is to remove it as a unit by disassembling the y-axis or disassemble it in place and lose the squareness.

You'll also notice that I increased the length of the 3/4" bars to give more clearance between bearings. Again I don't understand why you would want these dimensions to be so close that washer tolerances become an issue.

Chris

[OCNC]

Here are the image files.

[momus_cnc]

I'm at a loss to understand why you would want to perpetuate the use of the 0.678 dimension. Using multiple components that have to be hand dressed to accuracy in dimension as well as flatness seems quite incongruous with the idea of a DIY machine, particularly an entry level one. I could never support something like that and I hope you'll work that out of your design.

Chris, you are absolutely correct, hence this:

Version 2.0 will also have slightly modified bearing blocks, so that the spacers will be .75” thick instead of the odd size.

These drawings are for pieces that will directly fit the 1.2 and earlier design. The odd spacer dimension originally came from using a fractional dimension for the bearing hole locations in the bearing blocks. Due to the bearings being metric, the odd spacer size resulted. Yes, a far better solution would have been to use unaltered stock thickness for the spacer, and relocate the holes, or rely on tolerance between hole size and bearing bolt. A combination of those is what will now be happening.

I think your solution using the plate is well done, and I especially like your procedure of tapping through the parts while they are bolted together. There would seem to be some advantages to doing this rather than using roll pins, namely ease of disassembly. I am curious how easy this is to reassemble. The parts need to be tight together to thread the 10-32 machine screws through, yet you can't fully tighten the 5/16" bolts without already having the machine screws in place for proper alignment. Is there enough tolerance in the tapped holes that the machine screws thread through with just hand pressure holding the parts face to face?

I guess the bigger question is how often the carriage would ever need to be removed? On the rare occasions that I've had it off, I simply removed the bearing rails on the left end of the gantry, rotated it up to disengage it from the right rail, and slid the whole carriage assembly right off the end. Took all of about 3 minutes, and didn't require any bearing readjustment upon reassembly.

This gives me direct access to the back bearings on the y-axis which are otherwise trapped by the z-axis assembly.

Which bearings are trapped? Trapped by what?

Again I don't understand why you would want these dimensions to be so close that washer tolerances become an issue.

Where are you finding washer interference? Dimensions were kept as close as possible to maximize axis travel.

-Bob

[OCNC]

I am curious how easy this is to reassemble. The parts need to be tight together to thread the 10-32 machine screws through, yet you can't fully tighten the 5/16" bolts without already having the machine screws in place for proper alignment. Is there enough tolerance in the tapped holes that the machine screws thread through with just hand pressure holding the parts face to face?

The parts would have to be one full thread away from each other for there to be a successful misengagement of the screw, hence it can be reassembled with only light contact pressure between parts.

I guess the bigger question is how often the carriage would ever need to be removed? On the rare occasions that I've had it off, I simply removed the bearing rails on the left end of the gantry, rotated it up to disengage it from the right rail, and slid the whole carriage assembly right off the end. Took all of about 3 minutes, and didn't require any bearing readjustment upon reassembly.

To be perfectly honest I hadn't considered removing the left gantry rail as a method for removing the z-carriage, but 'all of 3 minutes' sounds like a bit of a euphemism with regard to a precision machine and consider that to remove just the carriage you're removing the entire gantry. I also would not be happy lifting the gantry at one end without giving the bearings some play at the other end. As to how often it would have to be moved, well once would be enough for me.

This gives me direct access to the back bearings on the y-axis which are otherwise trapped by the z-axis assembly.

Which bearings are trapped? Trapped by what?

The bearings on the rear carriage blocks, the ones with the vertical axles that have the hex heads cut down to provide clearance off of the top and bottom surfaces of the gantry extrusion, cannot be removed without taking the carriage off of the gantry.

Where are you finding washer interference? Dimensions were kept as close as possible to maximize axis travel.

I'll have to look at the old gantry tomorrow and my assortment of washers to report on this.

These drawings are for pieces that will directly fit the 1.2 and earlier design. The odd spacer dimension originally came from using a fractional dimension for the bearing hole locations in the bearing blocks. Due to the bearings being metric, the odd spacer size resulted. Yes, a far better solution would have been to use unaltered stock thickness for the spacer, and relocate the holes, or rely on tolerance between hole size and bearing bolt. A combination of those is what will now be happening.

I believe it's still a better solution to use the anti-racking plate for a mod to an existing machine. It only requires the 1/2" x 3/4" spacer blocks and the 1/4" plate along with two 3/4" x 1-3/4" bearing blocks (note that for a retrofit the slightly larger 1-3/4" material is necessary to catch the holes for the original offset studs rather than the 1-1/2" material that catches the holes for the centerline studs that I've shown in my as-built drawings above). I believe it just changes the distance between the centerline of the bearing axle and the centerline of the stud holes from 7/8" to 1". Check that though before drilling. For me drilling is always preferred over filing.

Chris

[OCNC]

Here are the image files.

I noticed in the drawings above I neglected to indicate the counterbores for the 8- #10-32 holes used to assemble the plate and spacers. They should be machined into the faces of the square stock components that won't be in contact with the 1/2" x 3/4" spacer bars. I used a 1/4" drill and went 3/8" deep. This reduces the length of the thread that has to be tapped into the assembled pieces.

Chris

[OCNC]

Where are you finding washer interference? Dimensions were kept as close as possible to maximize axis travel.


I had a problem with the washers partially covering the set screw holes where the 0.8125 distance was being used. This could have been tolerances and it's even possible I picked up the wrong washer. I also found it difficult to remove (it might even have been impossible) the y-axis bearing axles (horizontal) that are near the set screw axles (vertical) of the bearing blocks. Too much time has passed for me to remember the exact details but both problems were solved by lengthening the bars as illustrated above. I also switched from the cap screws to button head socket screws. All washers were finally chosen to be SAE #12. These were drilled out to 5/16" for the axle screws. As-is they work beautifully for the 1/4" threaded rod. This made it possible to do the whole machine with one box of one washer size. By using button head socket screws throughout clearances improved and access with a T-handle allen key made adjustments easier. It was even possible to drill an access hole in the right side of the wooden frame so that the axles on the underside of the right x rail could be adjusted using the T-handle.

Chris

[jerrywi]

Hay Bob,
I just finished parts 1 through 15 and came across your design change just in time. I've printed your revisions and will start parts 16 thru ?? this week end. Thanks for the update and I'll let you know how it went.

jerrywi

[momus_cnc]

Chris,

Nice washer diagram. I guess I tend to use the tighter SAE washers in general rather than the sloppier USS, so that .8125" dimension was most likely based on that washer size. I should note that in the plans, that the USS are large enough to both obstruct the setscrews, as well as hang out beyond the width of the 3/4" stock by a noticeable amount.

I like your use of the button head screws, and being able to access the lower bearing screws through the wood frame is a nice benefit.

The bearings on the rear carriage blocks, the ones with the vertical axles that have the hex heads cut down to provide clearance off of the top and bottom surfaces of the gantry extrusion, cannot be removed without taking the carriage off of the gantry.

Ah, gotcha. I guess I had assumed it would be easiest to slide the carriage off the end of the gantry, rather than rip the whole carriage asembly down to remove it. I wasn't trying to be hyperbolical with the 3 minute comment. Remove the two lower left nuts, and that bearing rail drops off. If the bearings on the right end don't have too much preload, the whole gantry will swing up and off the machine. First remove the two nuts holding the belt to the right end of the gantry, and the cable arm mount. With the gantry free, loosen the 4 nuts inside the end of the gantry tube, and slide the top bearing rail/idler pulley off. Unhook the Y belt from the stepper pulley (it might have already dropped off with gravity at this point.) The carriage will now slide right off the end.

You are right to be concerned about doing damage to the bearings or rails by removing it this way, so I probably shouldn't have even mentioned that I've used that method. The gantry becomes a lever arm of considerable length acting against those bearings when lifting it at one end. They could be easily damaged. The best method would probably be to remove the X belt idler pulley, the front X limit switch if you have one, and the cable arm mount. The whole gantry will now slide right off the front of the machine. Now the parts on the left end can easily be removed to slide the carriage off.

-Bob

[Cheeto]

Has this design changed for v2.0 plans ? I would like to start drilling holes on these parts.