Good idea. From my investigations, there will be lateral flex in the Y axes dependent on how much force the OP uses on the X axis ballscrew. I recommend the testing be done with the X axis gantry at the very front of the machine, as close to the Y axis mounting blocks as possible, to remove this flex as a variable.
1 Like
I couldn’t agree more.
Thanks, Pat
All good suggestions. I’ll report back when I can.
Random thought of dubious value: how are you securing the stock? Is it possible there is more play in one direction for the stock itself?
While it is unlikely that you secured it exactly the same way every time thru all of your experiments trying to isolate the chatter, but I thought I couldn’t hurt to ask since it continues undiagnosed.
Different ways. Hi-Lo clamps, blue tape super glue or even boring through holes and bolting it straight into t nuts in the spoilboard. Nothing made a difference. I’ve tried different places on the spoilboard. No difference.
So the nut at the end of the ball screw wasn’t terribly tight. I did tighten it down a bit, but no help. I don’t feel any play in the ball screw if I try to move it back-and-forth. If I pull the way to the front and try to move side to side it’s somewhere between 0.002-0.004”. if I really push on it I can get it to go a little further but that’s pushing hard.
I uploaded a couple of videos to YouTube demonstrate.
Climb cut:
Conventional cut:
Thank you for taking the time to share the videos - that is always helpful.
If we are willing to consider all possibilities, understanding that the following ideas may not seem to explain any X vs Y direction machining…
-
Have you checked backlash on all axes with a dial test indicator and simple gcode program? Maybe there is excessive ‘slop’ in some of the ball nuts or linear bearings.
-
WRT #1 above, when I replaced my original Z axis assembly, there was ~0.5 mm axial play due to the way the ball nut is held by design. With a heavy spindle gravity may make this indiscernible, but certain toolpaths/endmills may work to add upward/downward forces.
-
Can you try lowering your spindle in the 80 mm mount to reduce the moment force. I know that different setups require certain spindle ‘stick-out’ relative to the mount or gantry assembly, but this would just be as a temporary test.
-
In my rigidity tests, I found that there was a fair amount of flex in the Z axis relative to the other axes (I posted the results in the forum so ‘fair amount’ has actual N/um that can be compared. The 3rd rail did improve things BTW.). Perhaps feeds and speeds need to be further adjusted to take this into account, and certain MRR may just not be realized. I machine mostly aluminum, so will need to rely on others who do similar work as you to add their experiences - perhaps your machine is an anomaly.
-
Have you measured spindle runout and tool TIR to see if they are within a reasonable tolerance/range?
I don’t really have a proper indicator and holder for some of these tests. I’ll order one on Amazon.
When you say a simple code to help test, can you explain more? I’m not sure how I would do that
And yes, I was thinking about the spindle maybe sticking down too far. I remember looking for information on that and nobody seems to have a definitive answer on the proper amount of spindle should stick down.
Basically the backlash tests move the axis back and forth varying distances (+/-) with the distance dependent on the travel range of your indicator. After the initial calibration, the dial should always be returning to the zero set.
Attached are some files I received from Masso support a long time ago as examples…
Z axis test.nc (635 Bytes)
How to use Backlash Test.txt (1.5 KB)
X axis test.nc (615 Bytes)
Y axis test.nc (601 Bytes)
Here is a video - although a different controller…
https://www.youtube.com/watch?v=CS2rosB_ce0
There is also a great deal of information online on climb vs conventional, what causes chatter, how to reduce it, … worth looking into IMO even though the usual disagreements are there.
1 Like
Awesome. Thanks again for all this help. Indicator should be delivered tomorrow. I’ll adjust my spindle and give it another shot.
1 Like
Happy to help where I can. I am still new to CNC machining and learning as I go.
From my experience in the last few years, I believe these are great machines that provide great value. I feel that as long as I continue to explore and learn their limitations, and work within those, I will be able to get the best - realistic - results possible.
Well I choked up on the spindle and it didn’t help.
Is there an easy way to check if my table is the problem? Seems to me that the only way would be to mount the cnc somewhere else and run it again.
I think if your CNC is securely mounted, and you did a reasonable job surfacing, the table may not be responsible for what you are experiencing.
One other thing to try might be to machine your test piece at the very front left side, where the cumulative machine flex should be minimized.
I wish I had data to share on reasonable feeds and speeds for machining wood - hopefully others will joint the discussion with real-life insights.
I am fairly detail oriented, so can understand how frustrating it would be to have this ‘perceived or otherwise’ performance issue. I am optimistic that this can be resolved - more time, testing, and input from others will surely help.
1 Like
I’m willing to put in the e time. I really want to start getting into metal working as well. Thinking adding a small lathe and/or mill to the shop sometime in the next year. So I’d love to get this dialed in so I can use it for light milling/cnc projects.
This chatter was an issue when I tried brass aluminum before.
Ok lots of chatter if I move the workpiece to the left front.
I know it’s not the safe thing to do but if I grab the spindle above where the z slider grasps around, I can basically make the chatter stop with just my hand. Not sure if that’s helpful at all. Maybe pointing to something in the z slider? I don’t feel anything loose in the slider itself. I’ll check for backlash and run out when I get the indicator and stand.
That would make me suspect the Z axis assembly.
It is one reason I made a new Z assembly, and added a second 80 mm upper mount to my heavy and tall ATC spindle.
It may just mean machining within the rigidity limitations of the CNC. There are many variables that can be modified going forward that can help, but you may not always be machining with the speed or DOC/WOC that you expect.
As before, looking forward to other users to add their insights.
1 Like
If it really just the limitation of my z axis then that’s disappointing.
I did notice if I push on the end of the spindle near the clear nut I can get my dial indicator to move 0.010 or more both laterally and front to back.
Even with my new z axis, I still need to be smart about tool paths when machining aluminum.
Although I have much greater confidence now, I still constantly fine tune - higher RPM, lighter cuts (DOC/WOC), flute number, chip load, climb vs conventional, etc… and after that some days are still a motivating(?) combination of successes and failures. 
Well unless anyone else has some suggestions I guess I’ll keep trying different combinations. Maybe I can figure something to stabilize the z. I’ll still check run out, backlash etc.
1 Like
Hey Doconefall,
when I watch your videos, I think this is exactly why I want to replace the entire Z assembly with something with less leverage force between end of the bit and the spindle mount. With its Z travel, the spindle mount is unable to protrude downward beyond the end of the Z rails, this is how it differs strongly from all other CNCs I know. I’ve planned to replace the Z assembly even before I bought the machine, the Z assembly is the weak spot of these machines. It’s the design issue I described here.