I was examining the play in my MPCNC in preparation for cutting more aluminum and it seems that there’s a non-negligible amount of flex in the plastic spindle mount. It’s not a huge amount, but it would be a good excuse to cut some aluminum … to try to improve my machine to cut aluminum better
However, this is non-trivial, since you need the horizonal thru holes to bolt to the z-conduit, and for the bottom part there’s a significantly vertical extension. I only have 6mm and 12mm thick aluminum to work with. I expect I might have to stack 2-3 cuts and tap and bolt them together.
Anyone else do anything like this? I don’t want to go through all the design work if someone has already done this.
When I push on the router in the +X/-Y diagonal I’m seeing some shear deformation in the clamp in the XY plane. When I apply torque as if the bit was trying to feed in tough material I also see a small about of torsion.
Both are pretty small deflections, possibly negligible under normal loads. But I thought it would be a fun excuse to cut some aluminum.
A beefier printed clamp mount could mitigate this as well.
Look at the system as a whole, I don’t think the mount moves. If you have a dial indicator, it is fun to isolate each component and see what is actually moving and how much.
In reality everything is moving a little, some things more than others but they all add up. I spend hours and hours trying to improve each component on a regular basis.
Not to discourage you in any way from milling an aluminum mount, but there may be some benefits to using a beefier, 3D printed mount. I know many of the 3rd Party Makita mounts for the Burly were substantially bigger than the one Ryan designed. Most were a single piece rather than the two-piece model you are using.
And if you do mill an aluminum one, please post pictures.
Edit: It just occurred to me that another approach would be to purchase a “standard” 65mm aluminum mount and figure out a way to mount it to the Primo. There are a variety of aluminum 65mm mounts on Amazon.
That’s a very good point about buying one on Amazon. I would still have fun trying to mill an aluminum adapter for it for the MPCNC.
For the fun of it, I just designed a rather complicated bracket out of two 6mm cuts and a 3mm spacer to go in between (which might be printed). No idea if it will work but I’ll have fun trying. I’ll post here after I give it a shot.
I do not have a dial indicator. But I spent about 20 minutes with the motors engaged carefully stressing the core in all 6dof and observing the deflections. The torsion of the mount is definitely present, but it does seem to be the least significant of the three main sources. The motors holding torque are the most significant, but I can’t do much about that The third is the interface between the core and the XY cross bars, allowing some rotation of the core in the X and Y axes – I can visibly see the core clamps moving up and down relative to the conduit. I have tried tightening bolts to hug the rail better but the ones I need to tighten are already tight… Still trying to figure that one out. I have another potentially-overengineered solution for that one
That should not be true, they should be the least.
When you do these tests you need to pull from the tip of an endmill, 10lbs of force would be highly exaggerated but a good place to start. 1-5lbs would be far more realistic.
There is almost no force in Z at all.
I have always felt like those are suspect because they act like an insulating jacket. The Tools get very hot and wrapping them in 1/2" of plastic seems like a bad idea for what I assume is little to no added rigidity.
However, I speculate that this holding torque on this random microstep might be exaggerating the effect.
I bet that the average torque over dozens of smoothed microsteps may not actually result in so much movement distortion. Either way, there’s not much we can do about that particular effect without switch to lead screws or something else infeasible.
Try with the forces I talked about earlier, at the endmill and check again. Pushing on things individually with an unknown force is not really helping. You would be surprised how much you can push with your thumb.
Like I said everything moves on all machines ever, it just depends how much.
If the movement is only coming from your steppers that is amazing, that means I built a great machine. You can always turn up your steppers, we run them significantly underpowered. (All steppers are springs more holding current makes them stiffer, but while in motion it is not really the case).
I have a new LR core and beam started. I have been testing for a few months at this point. I am interested to see what you find, so far we are not seeing the same things. That testing from the end of the tool is important, though. It puts the whole machine under tension and things move different and transmit the forces differently.
You are probably right about the heat. One of the more interesting Makita Burly mounts had “fins” facing inward, creating channels for air flow. Don’t know if this fins-and-channels design provides enough openings for good airflow.
Even if kept mostly open, I’ve wondered if connecting the two rings of the mount would provide additional stiffness…if connecting them would spread the forces more evenly between the four attachment points. Musings from a non-mechanical engineer.
A single piece certainly would be stiffer, but it would need to maintain airflow as mentioned. That’s why many of the more expensive CNC spindles use water cooling (so they can have both a rock-solid mounting bracket and sufficient cooling).
I also think that designing a 1pc bracket might require printing with supports which Ryan has done a great job avoiding to this point.
I just posted about my weird aluminum spindle mount clamp, because I thought it deserved its own thread. It seems that even with a stock build, aluminum milling is workable. Just a bit slow…
