Getting bumps in curves

[Gary Campbell]

What do I do about that? I was thinking I could tune it out.

Didn't you just upgrade to bigger motors or servos? If so, then you may have more torque than the frame and/or motion system can handle. You can "detune" to deliver less power and reduce acceleration, but as a rule, never expect software to fix hardware problems.

The weaker the motors, the easier it is to hide machine deflection. Is you machine self built or a purchased machine? If purchased, what is the motor power recommendation from the builder?

[slodat]

Understand and agree on all points. I bought the machine from H**** Tools. I knew I didn't know enough to do it on my own. Turns out they don't know a ton about machinery either. I learned this after the fact. It is a Chinese manufactured machine with literally zero information available from the factory or the company I purchased from. So, I have almost $25k in the machine and I'm hoping to make the best of what I have. There are links in my signature to my machine build threads with photos, etc. The machine weighs about 4000 pounds and seems to be quite rigid. I'm suspect of the assertion the gantry is flexing and I'm leaning toward the drive/motors not controlling the inertia due to gains needing additional adjustment. I cut three rows across the full X axis of these parts last night and the bumps are uniform across the x axis travel. They are occurring after a sharp corner.

<edited to add:>
Machine specs from my purchase docs:

X,Y,Z working area 5ft * 8ft *12in 
Frame 3/8" steel square tubes 
4- YAKO drivers and motors 
X, Y stainless steel helical rack and pinion drive 
Spindle air cooled, (6 hp) 
Z axis German precision ball screw 
25mm Hiwin stainless steel rails 
X, Y, Z traveling positioning accuracy ±0.001" 
X, Y, Z traveling re-positioning accuracy ±0.001" 
Max rapid travel rate 1200in/min 
Max working speed 780in/min

It came with NEMA 34 steppers. I upgraded to NEMA 34 DYN 750w servos. I feel like the issue is what I have loosely called "tuning". I understand I am grouping terms here.

It's not that I'm expecting "software to fix hardware problems". It's that this is what I have, I'm observing these bumps (that's what I'm calling them) and I want to do what I can to get the machine running as well as I can. I'll live with what I have.

I sincerely appreciate everyone's help. I'm getting an idea of what I have. I'm thinking worst case scenario is at some point I fabricate a more rigid gantry.. It was advertised as 3/8 steel. I didn't measure it, but it's in that range.

I'm going to connect an oscilloscope to each drive when it gets here and see what's going on with position error. I am wondering if it's inertia causing this and if that can be compensated for with servo gain.

Again, thanks for the help!

Steven

[Gary Campbell]

Steven...
In most cases a machine that is that heavy has pretty close to "all the right parts". That said, and this is after disassembling a couple dozen of them, there are usually LOTS of loose fasteners. I choose to believe it is due to vibration from a long ocean voyage rather than workmanship????

My guess is your deflection is due to something being loose, rather than bending the gantry while leaving a 1/4" bit intact.

[slodat]

Gary,

I was thinking the same thing. I'm going to check fasteners this evening. Thank you for the recommendation!!

[slodat]

Didn’t find any loose fasteners.

[Dan M]

Didn’t find any loose fasteners.

Is it messing up where the endmill is plunging into the material? Where I'm going with this is I noticed you're not ramping and doing a straight plunge. I always use ramping and was curious if there's a reason you don't use it? If it's happening where it's plunging, try ramping and see if it works.

Dan

[Sword]

Was kind of wondering about that as well, full load on the bit at the start point. That is if the bumps are right after the start point. You could try changing the start point along with a ramp in. Another thought, do you have "Sharp external corners" checked? If so, that can set up an abrupt stop/start at corners, where if unchecked, it will "roll" the corner.

Is the tool designed for straight plunges? Looks like it is happening at start points as well as after a plunge of a tab. Try using 3D tabs so it has more of a flow up and over the tab instead of a straight plunge and go.

Cut the part with a small allowance and bottom skin and then follow up with a full depth no allowance pass running a bit slower to clean up the cut.

Just some thoughts on toolpath settings that can affect edge quality.

[slodat]

The green node by the 3 in the drawing above show the start point. This is happening after a sharp corner (2 in the drawing).

[slodat]

It took some figuring things out, but I have an oscilloscope looking at DYN4 drive DB25 pins 8 & 1 which is motor position error. It's a 0 - 3.3vdc analog. When the motor is on position the analog should be 1.65v.

Oscilloscope setup:

• X: yellow
• Y: cyan (left end of gantry)
• Y2: magenta (right end of gantry)

X axis

DS0007.PNG (10.92 KiB) Viewed 2802 times

spindle on left side of gantry

spindle on right side of gantry

[slodat]

OK.. I've spent close to thirty hours figuring this out. I've had an advisor that's knowledgeable of this stuff providing some much appreciated tech support. Thank you!

Where I started this morning doing rapid moves on X and Y separately:

DS0008.PNG (11.48 KiB) Viewed 2782 times

Where I ended up at the end of all of this:

And, cutting the part in question, in air:

final following error on part

Part in question:

Part