I’ve been using the CNC module a lot over the last 9 months, with generally good success. I’ve noticed recently that the bottom of the bit doesn’t seem to be quite level. When doing a facing operation it systematically leaves little ridges on every pass. I’ve attached pictures of my results using a 1/4" flat end mill with 2mm stepover and a very shallow depth of cut (0.2mm). I see this pattern with multiple endmills.
Is there an adjustment I can make to correct this?
Strictly speaking, no, there’s no adjustment. You can, however, do something about it. Shimming either the upper pair of mounting bolts or the lower pair of mounting bolts will tilt the motor axis. This suggestion assumes that it’s a simple geometry problem and not a problem with stiffness or vibration in the X-axis linear module.
Looks like you probably have play in your linear module. Is it tight when you try to tilt it in either direction?
If so, check this thread:
If not, how sharp is your bit? How fast are you trying to cut? At .2mm it really shouldn’t matter, but just wondering. Is there chatter as it’s running or is it smooth?
-S
Do you mind posting a picture of the cutter youre using? Is is worn or chipped?
Also, are you using a zip zag pattern? It could be due to climb vs conventional cutting pattern.
Lastly, if the tool head is canted slightly or flexes it could lead to some poor surface finishes like this. What is your stickout of the tool from the holder?
Climb vs. conventional would only affect outside contours.
At .2mm cutting depth and a 1/4" bit it’s not going to be flex, and even if it was, it wouldn’t be consistently in one direction.
-S
I imagine a long moment arm from excessive stickout could cause the linear rails to flex. Its true the 1/4" end mill itself won’t flex but the system as a whole might.
I think cutting direction is probably more of a contributing factor but it’s always good to have more information.
But then again I’ve been wrong before lol. 
The linear rails are actually pretty sturdy. If your brackets are adjusted properly most of the deflection from cnc comes from the bed deflecting and lowering the spoil board, especially when machining near the front and back edges. (which is why additional rails under the bed is a popular mod) If you’re cutting deep enough and fast enough to cause the rails to flex the spindle is going to stop turning. As long as you keep your depth of cut on hardwoods to 1mm max on roughing and .5mm on finishing passes at 400-600mm/m it does fine.
It’s obvious his passes are against the grain. Passes with the grain will cut easier and smoother and generally give a better finish, but depending on the wood a lot of times the difference is barely noticeable. I do a lot of oak (which has one of the larger grain structures) and I generally try to mill with the grain (especially when doing facing) but sometimes due to the shape of the object I’m milling I can’t. While I can feel the difference it usually isn’t a visible difference. I usually end up sanding the faces either way.
If I was milling this object I’d either rotate the object 90º or change the pass direction 90º
-S
I feel like the linear rails are not very rigid when discussing cutting forces induced by milling. They are acceptable when taking light forces. That’s why they work well for the laser and AM but struggle with milling when people like us push it.
If youre putting enough force on your cutter to completely stop the spindle you need to rethink your strategy.
I agree that milling with the grain of a fibrous material would probably result in an improved surface finish, this was not my point. You can clearly see his step over where there is a constant edge that is raised or fizzy and one the is mostly clean with some small raised/fuzzy parts.
I’m just saying it might be interesting to see a sharp cutter running purely climb or purely conventional with this grain orientation. It might make a difference.
Climb vs. conventional only makes a difference when the cutter is using the side of the bit along a contour or outer/inner edge. For surfacing the top of the workpiece the primary engagement is on the bottom cutting edge and pass direction vs. grain direction is what you want to pay attention to.
In theory climb cutting gives smoother results but that’s with a normal router or a more professional cnc machine that has plenty of power. With the SM the choice of climb vs. conventional is mostly determined because of the difference in the forces encountered by the bit against the wood. With conventional the bit pulls itself towards the wood, with climb it pushes away from it. I personally have found with any contours greater than 10mm in depth that conventional is almost mandatory as it runs much smoother and causes less chatter and torque.
-S
This is a nice illustration to show what climb and conventional cutting is but I think it greatly simplifies the dynamics at play. Respctfully, I think your statement is misleading. He is not using a large face mill that covers the entire part, he has a relatively small endmill in a zip zag pattern.
There is a lot that goes into chip formation. Think of how the chip is formed when a tool first impacts the work and then is evacuated. Upcut/downcut, climb/conventional, tool geometry, number of flutes, chipload etc… all of these play a factor in the finish.
But all of this is just speculation without knowing more specifics.
Also, With deep cuts you might consider running a dynamic clearing tool paths that keeps a constant amount of tool engagement in the work if they are available to you.
Thanks for all the great discussion everyone! It is really helpful to hear more experienced points of view. I’ll run some experiments going with the grain and with different settings. I was doing a zig-zag pattern, so it is definitely possible that it was only happening in one direction.
A couple things I’m not sure about:
I’ll update here when I’ve had some time to experiement.
Single flutes are really meant for clearing passes only.
For wood and the smoothest surfaces you want a downcut bit and 2 flutes.
At least for the final pass. Generally I’ll run clearing passes using an upcut because it will throw the shavings out of the way better and less problems with it getting compacted if I don’t regularly vacuum. I’ll set “stock to leave” at 1mm for clearing and then do the final passes with a downcut bit.
As far as speeds and feeds with a 1/4" bit I usually do .5mm step-down and somewhere between 400mm to 600mm for work speed - depends on the wood (some oak is really tough). I err on the side of slower usually. For clearing you can get away with 1mm step-down but again I generally do .5mm. For me it’s not worth pushing it and I just have less issues.
In theory there’s an optimal engagement of how deep you want to be cutting but in practice I haven’t found much difference on my SM. Try a pass with step-down at .2mm and one at .5 and another at 1mm and compare and see if you can tell the difference.
-S
In my experience on larger-scale CNC machines, that cut pattern almost always indicates a head that need tramming; however, because the Z on this machine has two control points I would guess that is where the problem lies. There isn’t enough slop in the head connection to indicate that would be the site of failure.
It could be that there is some sort of debris under either the bed or between the head and the bracket that’s causing a tilt. Use a square and verify that they’re perpendicular to each other.
-S
I believe the issue is on the 2 “Z” columns, make sure they are tight and perfectly squared to the base, if that doesn’t fix the issue, then it might be the CNC head, try adjusting the screws holding the CNC head and moving it, even placing spacers (very thin washers) in between the head and the bracket, use 2 at the time, bottom first.
Let me know if it works.
