CV mode jitter when G1/G2 or G1/G3 alternates

[Robertspark]

Robert, with all due respect, I think you are missing the point..

Sorry you obviously have the issue in hand

[cnv]

You also mentioned acceleration settings.
If CV mode worked properly, then this toolpath (and even much tighter turns) can be followed with very low accelerations.
(as is the case in my second example: test2 - no jitter.tap)

Actually, a high-acceleration setup could mask this problem, as the frequent accelerations/decelerations would be hard to notice.
But the point is, in CV mode it should not accelerate/decelerate between these segments, as there is very very little change in direction.
This is what I'm trying to say.

[cncdrive]

Yes, you do not understand how the CV control works in the UCCNC. Please read the manual.

[cnv]

Robert, with all due respect, I think you are missing the point..

Sorry you obviously have the issue in hand

I sense that I might have offended you.
If so, sorry, that was not my intent.
I wanted to point that you did not respond to my question, but a completely unrelated part of the message.

Obviously I do not know the solution to the jitter problem, otherwise I wouldn't be here.

[cnv]

Yes, you do not understand how the CV control works in the UCCNC. Please read the manual.

I already read it several times in hungarian and english as well.
Can you explain why it behaves different with pure G1 segments and different with alternating G1 G2 segments?
Considering the same smooth toolpath, with very low side accelerations.

Actually, in the toolpath in the example, there is nothing to smooth out, in CV mode it should go through it with full feed, as it does in the other part of my example.

I understand that if there is a corner, CV mode will "round it" so to speak. And if the parameters are very limiting and do not allow rounding, then it will slow down as necessary.
But here there is no corner, no need to slow down.
Where am I wrong?

[cnv]

suggestion to try.....

remove the backlash settings from the axis (or not enable backlash).

and change the linear and corner error max to 0.001

then re-run the code and have a look how they compare.

Thanks, I tried these too.
There is no change in behavior.

[cnv]

To better illustrate my point, I have removed much of the sample G-code and left a simple rounded rectangle.

In the second turn (lines 74 to 134), I have replaced all G2 segments to G1, so now this arc is entirely made of linear segments.
All other turns are still made of alternating G1 and G2 segments.
In each turn, there is the same amount of segments, same length, same feed, same (low) side accelerations.

The second turn does not jitter, all others do.
In my understanding, CV should behave the same in all of these turns.

[cncdrive]

what "alternating means"? I see the feedrate moving between 5.9 and 6.0 which is not much difference and can be caused due the the feedrate is calculated on a small time scale and so the visual output is not that precise in the DRO, but is optimised to update fast instead.
Maybe post your profile file too then I'll test with that too tomorrow.

[cnv]

By alternating G1/G2 I mean code like this:

Code: Select all

G1 X13.212 Y5.876
G2 X13.233 Y5.928 I1.538 J-0.597
G1 X13.277 Y6.03
G2 X13.325 Y6.133 I1.517 J-0.649

So, G1 and G2 segments follow each other in short intervals.

On my setup, feed rate drops from 600 to around 200.
Except the modified second turn, which goes almost full 600, as it should.

I attach my profile.

[Robertspark]

The profile is attached with the earlier tap files