I’m still going through my journey of CNC learning but noticed something on my first ever face operation (as in fusion 360 face operation).
If I do a both classic and climbing face operation in the Y axis, a very noticeable pattern can be seen (specially close to the edges), where every pass have a different height:
I’ve tried the same operation in the X asys which alleviates the problem but doesn’t make it dissapear, like wise using only tradicional or climb cut as opposed to both helps too but won’t make it dissapear.
My, clumsy and non educated guess of the root cause is the extreme flimsiness of the bed. Just with my finger if I touch any of the ends of the Y axys the bed will move sensibly. The X asys while better is flimsy too.
Do you think this makes sense? Is it a problem I only have? And finally, is there any hack for a more sturdy bed for CNC?
If the cutting parameters are inconsistent, such as high depth of cut/fast machining speed, the machining impossibility will be high and the tool head will escape or the bed will escape.
Thanks. You might be entirely right.
My doubt comes from the fact that the bed can me easily moved with a finger in any of the Y axis ends (perhaps only in my unit?). But certainly, as you said, the load might be too high.
@Blockmodule Sure you can say the load is too high, but you can also say the bed is to weak instead. Its good enough for 3D and laser but for CNCing it is not so good as i have said a couple of times. The baseplate is weak too but that is another theme for itself.
Hi Steffix
When I first mentioned high load machining, it was in consideration of the bed sagging during CNC machining.
I think the bed will loosen if you keep machining at high loads.
The Snap Maker A250 and A350 are structurally affected by the cantilevering moment as you move away from the center of the bed.
Expecting too much from such a structure will not solve the problem, so you will have to deal with it through usage.
As you said, it is one of the themes, but it seems to be a theme of the newer new models, not the current models.
The only way to prevent the bed from deforming up and down is to make a new rail.
I think someone in the past suggested some kind of moving auxiliary rail with bearings to prevent bending, but I don’t remember.
I think you have a problem if it moves easily by just pushing with your finger.
Please contact Snapmaker support with as much detail as possible, including a video of it actually working.
If it is determined to be a problem, they will take care of it, but it may take some time.
It’s not just your unit. It’s a basic design flaw. The mounting point are 40 mm apart on the single carriage mount. That’s about the distance between the bearings inside the linear module. The bed is ~320 mm in length. It’s roughly an 8:1 moment arm, yielding force multiplication on the bearings. By contrast, that number for a traditionally-support work platform on linear slides is less than 1:1, which is to say that it’s not an issue and other things limit accuracy.
The bearings inside the linear module are arranged to directly resist torques (rotations) in the X-Y plane. Pushing on the end of the work platform is a torque in the Y-Z plane. The bearings inside are not particularly great against that kind of torque. Combine that with torque multiplication and you’ve got a design that works poorly for such loads. That always means CNC and can mean for FDM printing if the work piece is positioned off-axis.
The best work-around without rebuilding the machine is as you outline, to greatly limit cutting depth and thus also the forces on the bed that distort it. The downside is greatly increased run times.
I’ve designed and printed a small sliding support that clips on the underside of the base and floats 0.2mm above the actuator (see dark blue piece). When there is downward pressure it will contact the actuator and limit the vertical movement of the bed. If you make 4 of them (2 of each side) and install them in the front and back you will get containment in both directions.
Here is a link to the geometry in thingiverse: Snapmaker A250 Sliding Support by ppmendonca - Thingiverse.
It’s still a work-in-progress, but I hope it helps. Any comments are more than welcome.
This thing is a toy. You have to treat it as such. With an underpowered 12k spindle you really need to follow HSM practices or even a little less. By that I mean SMALL radial cuts with moderate depth of cut as well. I see a lot of people doing this in reverse. They take a .010" depth of cut at 1/2D stepover…and try to go fast. No good. I’d suggest a max of a 10% stepover and up to (but probably less than) a 1xD cut. Once you can get away from the linear tool paths of Luban, the better off you will be as well. There are much better, more efficient tool path options out there. The Adaptive Strategies of Fusion 360, for example, provide constant tool engagement (load) while cutting cycle time by an unbelievable amount compared to something like Luban’s offerings.
Just for my understanding: So if I have a 3mm bit what you suggest is to do a stepover of 0.3mm and up to 3mm step down? At what speed?
I usually do up to 50% stepover (so 1.5mm) and stepdown of 1mm and relatively high speed of 1000mm/min. All depending on material obviously (mostly worked with mdf so far).
But last time I’ve been doing larger step downs it usually ended in broken bits or skipping steps and getting the head stuck in the material.
I mostly work with hardwoods, oak, walnut, maple etc.
For that I’ve found that my step-over doesn’t really matter.
I tend to go for 1/2 to a third because I seem to end up with better finish and less tearing out (even with large clearing passes with 1-2mm of stock left).
But I’ve had no problem running 90% for facing when I was just trying to remove material fast and didn’t care about finish.
For hardwood with a 3mm bit I’ve kind of settled on 400mm/m with a .5 step-down. While I can push it faster this seems to work pretty well and don’t have to worry about it running into parts of the wood that may give it a little more trouble. When I ran 600 it would work great for a lot of it, and then it would hit a knot or a more dense part of the wood, or on cross grain it would bog down and either slow the bit down, deflect the head, or the worst: cause vibrations and chatter. With the way the brackets are designed adjusted this will cause problems and lead to needing to adjust the rollers in the linear module. Not sure what long term effects as far as wear will be.
For smaller bits I’ll lower the speed accordingly. For 1/4" I may push up to 600mm/m, but generally I just stay at 400mm/m. For speed, I always just run at 12000 on wood.
If you listen to what the bit and the head is doing it will tell you a lot. It’s pretty obvious when it’s pushing too hard and straining.
The SM isn’t a professional machine but it is extremely capable, just not fast.
-S
