Kb58's granite CNC router build

[kb58]

The second Y servo assembly is complete. Unlike the first, reassembling the ball nut went much faster - having learned how the hard way. Like the first assembly though, both finished bearings were stiffer than before disassembly. As mentioned previously, they're anti-backlash units, with two bearing sub-assemblies that push against each other to virtually eliminate backlash. Preload between the bearings is set with shims between the units, which are then locked in place. For some reason, the shims act as if they're slightly too thick - go figure. The worry about too much friction caused me to carefully lap a fraction of a millimeter off each, and now they turn about as easy as before. Yes, someday I may find out what I did wrong, and then thicker shims might be needed, but that's a problem for Future Me...

For fun, I connected one of the Y assemblies to the controller just to see it move. I'd already run one motor on its own so wasn't expecting much... hah! As soon as I commanded a move, it started banging back and forth and making a big fuss. That made it clear that the Teknic (servo manufacturer) advice to tune each servo in the environment where it'll be used isn't just a suggestion. Ran the tuning software and all is well.

Next up? Probably the moving Z axis assembly. I can't start on the X axis components because the final size of the X axis granite beam is still an unknown (file this under "work with what you can get). Speaking of that, tomorrow I'll have words with the boneheads at the granite fabrication business. Either they'll be shamed into getting on with it, or I'll look elsewhere. There's also a dashcam to install in our car. Been watching a bunch of videos on that, weighing doing it myself versus paying $250 to have someone else do so, though it irks me to pay the same for installation as for the camera itself!

 

[kb58]

Spindle mounting plate complete. Cut out as much as possible with the band saw, squared it up with the mill, and used the DRO to accurately place the holes. It was a bit awkward because my mill isn't large enough to handle the panel with one setup and had to be shifted around. That made it harder to ensure critical dimensions stayed correct, but there aren't many of those. The Z carriages, ball nut adapter, and spindle mount all attach to it.

 

[kb58]

This is the first time that I've used tooling plate of this size and it looks quite the business

. The Z axis back panel is complete other than the ball nut adapter mounting holes. They're rather critical and will be double-checked before committing.

The Z axis ball screw was shortened - you guys haven't lived until you've machined a hardened steel* part on a lathe. Hair-thin shavings fly around and when they land on your skin or clothes, they burrow in like a hungry mosquito. They're so small that they're nearly invisible without a serious magnifier, and very hard to get hold of once embedded in your skin. There were so many on my T-shirt that I threw it away because I didn't want that stuff getting into the washer or drier and into other clothes - the stuff is downright evil. I realize that I should wear an apron but I'm not sure even that would stop this stuff.

*I tried annealing it but it had little effect. The X axis ball screw still needs to be modified so I'll give it another go, especially since it's larger diameter. I don't think I got it hot enough last time so will pick up some MAP gas and try again.

 

[kb58]

So it being Sunday and with tropical storm Hilary here (which is turning out to be a non-event), it gives time to think about the router and The Useless Granite Company. I was planning to suck it up and have them drill the holes - having done it myself and knowing how much work it is. With that now off the table, I'm questioning how the use of granite for the gantry. Yes, it's stiff and self-damping, good for reducing resonance and ringing, but it's also proving expensive and laborious to deal will. Then there's the potential of breaking it (again), cracking if it gets tapped wrong, or while drilling. Then there's reconsidering the decision to drill only one third of the mounting holes for the rails, mostly to keep my sanity in drilling so many. I'm starting to think that one in every three holes needs to be more like every other one, to prevent the rail from flexing. Factoring in the mechanical "magnification" due to cutter offset, and it becomes double that. So...

This leads (back?) to considering steel or aluminum. Using either, my mill is too small to handle a piece 42.5" (1080 mm) long but fortunately I have a buddy with a larger mill. As for which to use, it's not helped by no knowing quite a few important things such as: What will the router be cutting? Not sure yet - different things. How much force will there be on the cutter? Unsure, it depends. How much deflection is acceptable? Unsure. How much accuracy do I "need." Not sure. So based on these non-answers, aluminum may be just fine, though it comes down to availability and price. Options on how to use aluminum include thick tool plate and deeming it "good enough" as-is, or maybe bolt tooling plate to a section of C-channel, effectively creating a tubular assembly - I kind of like that approach, but regardless, it'll depend on what's available as remnants to contain costs. Steel isn't out completely, but in a home shop, aluminum is just so much easier to deal with. Regardless, if it ends up as a tubular assembly, it leaves open the option of filling it with sand/gravel/epoxy to give back much of granite's advantages.

One thing I sort of ignored is the susceptibility of the ball screws to flying chips. The Y drives are well protected by the stiffening plates - totally by accident. The X ball screw shaft... I'm not sure. I think it shouldn't be too hard to add covers to both it and the Z screw. As an aside, I briefly considered fabricating a beam out of carbon fiber cloth and honeycomb core. I'm kinda "meh" on the idea because of the labor, mess, dealing with threaded inserts, and weight not being a driving requirement. Regardless of the eventual material choice, the above has helped me decide to terminate the deal with the granite company.

 

[kb58]

Bought a section of 3 x 8" (75 x 200mm) rectangular tubing and some bar stock, costing about 20% that of granite... Wasted several hours being humbled by my apparent inability to accurately locate holes in an item longer than the bed of my mill. I'll go old-school next time and mark them out manually... Anyway, that aside, it's taking shape. Here are a few pics of the various subassemblies coming together.

 

[kb58]

The bar stock I used was hot rolled, which isn't the best for this purpose because it varies in size slightly, and the sides are often tapered. For these reasons (excuses, really) I'm telling myself I'm not too disappointed wasting time on them. Bought some cold rolled - which was crazy more expensive. I think prices on cold rolled has gone way up, as the hot rolled was far cheaper. Anyway... I got schooled on accurately placing a row of holes in a straight line - never had to do that on the cars, at least not where it mattered. It seems so simple, as you just:
1. Put the bar stock in a mill large enough to drill all holes in one go... nope, mine is too small.
2. Put the bar stock in the mill vice and shift it along so that the small mill can reach all the holes. Nope, see below.
3. Scribe a line down the middle of the bar stock and place marks along the line. Nope, see below.

The problem is that bar stock is never really straight - always bent some amount in one or both axis. I ended up reverting to the only way I know of to ensure alignment - using the parts as templates. The bar stock was clamped to the linear rail, then the hole positions marked by using a punch placed in the linear rail bolt holes. Sure enough, when it was done, the holes line up, and the bar stock's curve is more noticeable. The ends are perfectly centered below the rail, but at the center, it curves outward by as much as 2mm. I'll probably bolt the rails to the bar stock when tack-welding the bar stock to the gantry to ensure that the bar stock stays put. Thinking about it, it might even be a good idea to bolt the carriages to the Y back plate assembly well to help keep the bars in-plane (even though it'll get skimmed later).

 

[kb58]

Turns out that even though my buddy has a mill with a 45" table, its actual working envelope is around 36". There's that, but also how I have this hardheaded drive to do as much as possible myself, even if it presents extra work and chances of mistakes. It is, however, much more rewarding when it works out. That said, your way is much better at ensuring that everything ends up where it's supposed to be. My way, there's some concern about weld distortion pulling the bar stock out of alignment. Plan to do tack welds, alternating sides. Should probably also bolt the bar stock precisely to the gantry before welding. Like I said, more work...

Regarding the rails themselves, yes, I can confirm that they're very hard on the outside, and less so inside. Discovered that yesterday when shortening the X axis rails with a cut-off wheel. This is the first project that I've used an angle grinder for cutting material as I've always been a bit scared of the blade breaking. In fact, it did scare me once when I visibly saw the spinning disc becoming wavy. Decided then and there that I'd sacrifice one blade per rail. I know many people use them all the time for cutting steel and never have a problem, but still...

One more thing. Today was the first time all four X roller carriages were attached to something, and they were surprisingly hard to move as a unit, the drag of each adding up to perhaps 10kg? Time will tell whether the servo is large enough to overcome that in addition to the cutting loads... I admit that price was a driving factor, as the next size up was quite a bit more expensive. I just received the proper grease to lubricate the roller bearings, so hopefully it'll reduce the friction some.

 

[kb58]

Regardless who skim-cuts the bar stock on the gantry beam, I'm very curious how it's done. If the gantry is bolted down in any way, which it has to be, it bends the gantry slightly. The skim cut is performed, the piece is released... and the whole thing is going to spring back to its original shape. Not sure how to avoid that. One way I'm guessing is to support the beam on narrow supports so that any bow doesn't affect the cut and subsequent release. Even that though doesn't cover the case where the beam is twisted. Maybe a three point support?

In other news, I was in the middle of welding when I heard a faint "pop", wondered what it was, but kept going. It wasn't until I shut off the welder did I hear a slight hiss of escaping argon. The source turned out to be the over-pressure release valve inside the flow regulator. It isn't that the tank had too much pressure; it's that I've had this equipment for so long that stuff is wearing out - especially anything with gaskets or rubber diaphragms. I had a choice of buying a new regulator or having the old one rebuilt. I didn't want to wait two weeks for a rebuild so bought a new one from a local welding shop. Only later did it occur to me to check the price of regulators on Amazon. Instead of their prices making me think that I should have bought from them, it backfired. That is, Amazon's prices were way WAY cheaper, making me worry about how well they're built. I don't feel safe running 3000 psi argon through a $14 flow regulator.

 

[kb58]

Slowly things are rising off the table, however it seems like every component has to be modified in some way. I don't have much experience with ball screws and their supports, and assumed wrongly that all this stuff is standardized. For example, I thought that if a bearing for a given size ball screw shaft slides nicely into a motor adapter, the bolt circle would match. Nope; they often have slightly differing bolt circles. One might be 45mm while the other is 50mm - nice.

In other news, yesterday I got fed up with my noisy shop vac, always wishing for ear protection when using it. Fortunately, Dewalt saw an opportunity for those fed up with the "a loud shop vac must be powerful" marketing BS. They have three sizes, all with a part number ending in "-QT", if you're interested. I believe they only vary in height, and since the 12-gallon size seemed most common, I went with that one.

 

[kb58]

Odds and ends:

Yesterday was spent attempting to flatten the gantry assembly after I came up with a brilliant fixture for my small mill. In short, it was a complete failure, with terrible tool chatter and really uneven material removal. It's bad enough I'm not providing photos, with the only saving grace is that there's enough material left that it can be surfaced again - on a proper mill.

We're having one of our typical summer heat waves and it gets very unpleasant in the afternoon. I went back into the garage briefly at 9:30 pm tonight and it's still 90 in there. It reminds me of dealing with the heat during the car builds and now that tradition carries on with the router as well - good times. On to more pleasant things.

Made another motor adapter ring to mate the Z servo motor with its inch measurements, to the metric motor/ball screw shaft adapter.

Cut down the last ball screw shaft (X axis) and suffered again through machining the end. My second attempt at annealing the material went a little better than before but I still don't think that it was heated enough. The surface material machined okay, but just under the surface was hard enough that sparks were coming off the cutter. Worse, the cutter would ride up on the material, then dig in, making it difficult to hit an accurate diameter.

Spent hours drilling four more holes into each granite side plate to epoxy studs in for mounting the Y axis servo assemblies.

In the last shot you can see my garage helper. In his old age, Midi has become very attached to me, never wanting to be too far away, and often that means being right underfoot. He serves as a constant reminder to keep the floor well-vacuumed so that he doesn't get aluminum or steel chips in his paws. He sleeps about 20 hours a day now, like how many people do in rest homes. In dog years he's 106, which is pretty amazing for a dog his size. Little dogs reach that all the time but larger breeds typically only make it to about 10-12. I count each day with him as a wonderful blessing.