Shapeoko 4 X-axis Linear Rail Build

I originally was trying to order the Shapeoko Pro around the time when C3D was having issues sourcing linear rails, and then the 4 came out and I ordered an XXL the day it hit the store. I figured I was going to be in the occasional use category, but I’ve found I use the machine all the time. As I learned more about CNCs and how they are built I started becoming quite jealous of Pro owners and their shiny rails! After learning how to mill aluminum I did notice quite visible movement caused by the x-axis v-wheels flexing under the load. This dancing around caused inaccuracies in my parts and broken bits from fast dulling. Obviously the solution is to run conservative feeds and speeds, but sometimes the flex was unpredictable.

Since the 4 is effectively a v-wheeled Pro, I’ve designed my upgrade to use the Pro as a template, and also using the the existing holes in the Z-Plus that are used for the Pro since it doesn’t use a carrier plate. On the 4 the carrier plate mounts the drag chain tabs, stepper motor, pulleys etc, and where as on the pro all of these things mount directly to the Z-Plus. C3D did some smart optimizations like this when they designed the 4. This way there is only one part for the Z-Plus or HDZ that work on both machines. Of course I didn’t know all this until I watched videos again on repeat for shots of specific elements, and pouring through the docs for both machines.

Additionally I took some hints from the design the PrintNC, and have used 3d printing extensively to design alignment jigs and adapters. I selected 15mm linear rails from Amazon that were about $50 each. I made several different versions of the jigs for different stages of the rail installation. Here is the first type:

It snaps very securely into the v-rails with no play, and also roughly aligns the holes of the rails together.

The second don’t snap on as securely so they can be moved all over the rail for alignment of each screw.

The third type were a burly version of the first without the alignment pins. I found they were not really necessary and the second type would work.

Here you can see how nicely the jigs align the rail:

This is one of the burly ones, it appears that there are gaps around the rails but that is just a shadow. The rails don’t move.

The process of installing the rails is roughly as follows:

  1. Clip one rail onto the extrusion and align it horizontally. I spent some time with the dial indicator to get an idea of the variation as well.

  2. Move the portable jigs around each hole and use a 4.5mm transfer punch to mark the outer holes for drilling.

  3. Remove the rail and carefully drill the two holes. After drilling some not so straight holes I quickly designed a drilling guide and took 30 min to print it. I also tapped the holes.
    Screen Shot 2021-10-27 at 12.33.50 AM

  4. Reinstall the rail and use m4 screws to fasten it, then use the guide jigs to align.

  5. Use the transfer punch to mark the rest of the holes, drill them, and tap them. Easy! Actually this is the most tedious and critical part. Here you can see the mark left by the transfer punch in the hole:

  6. Since 3d printing is perfect I didn’t bother doing any alignment. Just kidding. I put the dial indicator to work and carefully aligned the rail to the top of the extrusion. This is using the assumption that the extrusion is straight to begin with but since the machine ran on it for alignment already I figured that was good enough. Sometimes the jigs were purely magic and zeroed out the rail in the center (it was .35 before dropping them in):

  7. Once the top rail was good I used all of the jigs to align the second rail and repeated the process of drilling and tapping. I installed all the m4 screws, with blue loctite as well.

  8. Then I spent about an hour using the dial and the jigs aligning the second rail to the first rail. I actually found the jigs were easier to get general alignment than using the dial and messing with each screw with hand pressure. I was pretty surprised just how much accuracy offered. Once I was within 0.02mm across I used a bike torque wrench to tighten them all to 3nm.

  9. I reinstalled the extrusion on the machine. I actually left the endplates on the machine and just pulled the extrusion originally, this worked out pretty well but I was dreading it at first. Then I installed my plate to test it, pulled one block off and tested the alignment again. They were within .01mm across!

Next steps:
After getting the plate on the machine I discovered that my pulley m8 holes are in the wrong location. I made a bet by using the Pro location and it did not pay off. I’ll design an adapter to 3d print for now until I can mill another plate. I’ve designed some of the adapters for the Z-Plus already, but I need to do some for the drag chain mounts. I’ve also noticed the rail blocks are pretty tight and grindy, so I might remove them to clean them out as suggested by youtube.

Going forward I’ll update the thread with my progress. My plan is to release all of my files for the jigs, the plate, and adapters along with a bill of materials so others can use them. But really, don’t buy a 4 with the intention to do this, just get a Pro :joy: .

Those 15mm rails look really good on the 4.

18 Likes

Very cool take on the classic linear rail mod!
This is both useful info for tinkerers, and a reason for me to be happy I got a Pro since I realize I would most probably not have been able to do it in as precise a manner as you did, with my clumsy fingers and poor drilling skills :slight_smile:

5 Likes

Excellent project!

Have you done any bench marking tests to quantify any changes once you’re done?

Any plans on putting linear rails on the Y-axes?

Good stuff.

3 Likes

I didn’t end up making a benchmark, but when I mill aluminum again I should be able to tell. I have some ideas about the Y-axis but I’ve never noted any flex on it, so v-wheels might be good enough. I like the design of how the Pro does it with two plates, but I think with the clearances for the drag chains and board enclosure I might be limited to mgn12 rails like the Pro. I have some 12’s on 3d printers and the size difference between both the 15mm rails and carriages is pretty noticeable. If I’m doing new Y plates then I would probably look at adding a ball-screw to the X as well in the same location as the HDM (if there’s space).

Any thoughts on whether 20 mm rails would fit on the X axis? Just going by the specs sheets that say the 15 mm rails typically use M4 fasteners whereas the 20 mm use M5 fasteners.

20mm rails are not that much larger and definitely would fit, but honestly the 15s are overkill from a load and strength standpoint as it is. I think your plate would be more crowed though so I’d model it before committing. The block mounting holes are 10mm further apart, so depending on how it works out you might need a wider plate which would reduce your machine travel. You could probably engineer around the problems, but the 15s worked so I could position the plate in roughly the same location as on the Pro, reducing my risk for other fitment issues.

I set the rails 50mm apart (c-to-c) to align with the black studs on the sides. With 20s you may need to bring them closer together to avoid the inside corners of the extrusion with the m5s which would not look as nice either.

shapeoko plate3.stl (183.1 KB)

Here is the STL of my plate for reference.

One man’s overkill is another man’s room for a heftier spindle…

The platform itself has limits with aluminum extrusions, v-wheels, belts etc., so if one’s plan is to go above a 2.2kw spindle then it’s probably time for a new chassis. Otherwise you are in that awkward scenario like on car forums where someone puts an expensive engine/turbo/whatever in a low cost car… someone will come along and tell them all the components they need to upgrade to make the car handle the power without destroying itself. It’s never cost effective.

Thanks for your notes. I don’t have any disagreement with the rails themselves; I am just adverse to itty bitty screws and itty bitty taps. But maybe that speaks more to the quality of the cheap hardware store taps as opposed to proper ones from tool and machinery shops. I would hope my experience is not unique: one would not hesitate to buy a proper plug tap and bottoming tap in M6 and M8, for instance, but for the smaller less used sizes just use what comes in a cheap set from the hardware store, curse when they break, and then is unable to get out of this loop because buying a single tap is just about as much as buying another cheap set, and maybe the next one won’t break so easily…

A little preview of where I’m at.

5 Likes

Not sure where you are but there are generally trade tool sellers who will sell you a good quality set of 3 taps for a given thread size for sensible money, here’s a local supplier I use for example.

I also found that I chewed way fewer threads and broke fewer taps once I started to use a sprung tap guide in the drill press for my smaller tapping sizes (such as threading the holes on my X extrusion for rails)

https://www.arceurotrade.co.uk/Catalogue/Cutting-Tools/Threading/Spring-Tapping-Guides

HTH

1 Like

Yeah, that is a better price point by about a factor of 2 to 3 (compared to for instance in Canada KBC,3.0X0.5 METRIC TAP SET,1-373-014K,KBC Tools & Machinery, about $24 USSD).

The go-to place in terms of price over here would be Canadian Tire, where you can get a set like this for about $18-20 USD, but in my experience can snap the small ones in short order if not started perfectly straight in a mill / drill press or such. For my home shop I have neither and am still mulling on whether to spend $500 USD on a drill press or $2500 USD on a sizeable desktop mill.

I have a cheap set which I bought years ago, it’s useful for when I don’t have a good quality starter & bottoming set for that size.

Since purchasing it however I’ve decided that the tap wrenches in it were horrible and bought some good quality old used ones from eBay and bought M4, 5, 6 & 8 tap sets from a supplier like EW. I find the good taps take less force and are also less brittle, leaving me spending less time trying to extract them from a damaged workpiece.

Stopping every turn to back off 1/2 turn to break chips and always using a cutting oil has also helped me break fewer taps.

As for the drill press, I use an old $80 cheap wolfcraft nasty which has a 43mm collar for an old Bosch hand held drill. It doesn’t need to be a good drill press, just better than me at keeping the tap upright :wink: It’s worth taking a look on fleaBay or similar marketplaces to see what’s going cheep. :baby_chick:

2 Likes

I never knew about a tap guide before and will defiantly try it out because I murdered my drill tap the other day and don’t want a repeat of that headache.

1 Like

The next step after getting the rails installed was to design and print adapters for the motor mount and drag chain mounts. The dragchain mounts I printed in one shot and they worked well. The dip was intended for cable management, but the only cable that was long enough to fit under it was the z-limit.

After that I set to making the motor mount. Originally I was going to attach the aluminum mount to the Z-Plus and then design the adapter to go on top of it, but after printing a few tries I realized that I was trying to create something that wasn’t really possible once I discovered I needed to raise the motor to clear my larger drag chain (aftermarket).

Winston’s video for the HDZ installation was helpful in understanding how to disassemble the Z-Plus, but then I did it again, and again, and again. Winston gingerly holds the assembly at an angle preventing the rail blocks from falling off, but eventually I learned you could pull the whole thing without a problem if you just do it upside down. Look at that mess:

Further iteration lead me here:

It raises the motor a bit, has mounts for the pulleys, and clears the new X-plate. I was worried this would be the weakest link of my machine given it’s printed, but with aggressive settings like 4mm walls, top/bottom, and 30% infill it is a tank and I can’t feel any flex in the motor. I designed the inside circle after I got tired of printing bricks, the circle inside a rectangle design seems to hold up.

Here is everything together, with the motor mount and drag chains installed in the holes used on the Shapeoko Pro:

Finally I designed a ‘bumper’ to ensure the rail blocks will never lose a bearing and also brings the X-limit switch out so it can detect the plates in their new location. This is the only 3d printed part visible from the front of the machine. I suppose I could have used gray filament or something, but the grape adds a nice pop.

So that’s it, three printed parts that live on the machine. The machine is operational. I’ve ordered a few more rails to start thinking about the Y-axes but that will involve machining new X endplates so that upgrade will be a ways out.

If anyone wants the files for their own conversion post in the thread or contact me via DM and I’ll send them out. Unless there is significant interest I won’t bother packaging them for public release.

5 Likes

Here’s a 1/4" depth 350ipm cut in pine writing a check my workholding couldn’t cash. No skipping or chatter as far as i can tell. Imgur: The magic of the Internet

Carbide Create wouldn’t let me do 400ipm :joy:

Probably because it observed that your stock was loose😂

3 Likes

Next up in my ill-advised SO4 modifications is… Pro end-plates! I found these on ebay, but after I got them I feel a little iffy about their origin. They are boxed like they were supposed to be included with a machine, but looking around on ebay didn’t see any other random Pro parts on offer. It seems like they were sold to a liquidator or something, but I’m not really sure how that would happen.

I knew this could be a total bust and the plates might not fit at all, but they do. The first challenge is that the 4 has the belts moved inward 5mm or so because of the V-wheels. So I designed new clips with an offset.

The second challenge is that the Pro nicely has the power switch and bitsetter run through the right channel. This was designed out of the 4 probably for a mix of cost savings and how the 4 X plates go quite low. This won’t be a problem for me once I figure out my Y-axis linear rail mod, but I won’t be using the right plate until that happens. At that point I will be drilling a lot of holes in the extrusion, so one more for the exit cables won’t be a big deal.

Seems like Carbide 3D could easily make a ‘Pro Plate’ upgrade accessory for the S04. These things are so nice.

3 Likes

Those ebay listings are very interesting. Why would someone have just the end plates for some shapeoko pro’s? That seems very strange. I found 3 different sellers of these who all are located in a suburb of Kansas City. Typically stuff like this is bought at auctions and then sold for a profit on ebay. Maybe someone bought a few machines and made some modifications that included changing out the end plates and then for some reason sold them through an auction.

Yeah it is strange. Mine included the business card signed by the packer too. I wonder if this is what happens to stuff lost in the mail? Or maybe a supplier liquidated some seconds? Mine included the bitsetter and it looks like all of them do. The other weird thing is none of the sellers know how to price what they have either.