5 axis CNC Router/3d printer

[George Butty]

Hi Tom, got 5.3.1 running with the patch you sent. What a pleasure to work with a machine
that does what it's told to do! Got machine lined up as well and your instructions helped at lot.

Thanks
George

[TomKerekes]

Great to hear George.

Waiting for a cool Video

[George Butty]

Hi Tom, your last fix for the stop start of the rotation axis is working fine. For some reason I am having another problem
which I can't work out. I zero out B and C axis and then I rotate B 90. With tool in horizontal position I zero a dial indicator
at the end of the tool (it's really a bar at a length 20" from center of B rotation). I rotate C-180 and then I rotate back to
C0. The end of bar will not come back to the original zero location. It is out by .030" every time I do this. My backlash in the mechanics
is round .01" at that length. This rotation is in the same directions as run by zeroing of the C axis therefore backlash should be next to zero.
Any chance of some round off error in software? I does this in both RTCP and G49 mode.

George

[George Butty]

Hi Tom, hold off with my last post. I think I am scaling the axis incorrectly. Will play
with it some more and get back to you if I can't resolve.
George

[George Butty]

Sorry but I am back. One other thing that happens is sometimes when I run my c axis home.c , goes home and sets
my home position. When I go back to KmotionCNC it runs in G49 mode and can't set it back to RTCP mode unless I reboot
Kflop and Kmotion. It will note switch when I type in G43.4 H1. I have attached my C home.c.

George

[TomKerekes]

Hi George,

The program is restoring from SaveYLimits which has never been defined. I think you are mixing Z and Y. Shouldn't this be B?

Code: Select all

ch5->LimitSwitchOptions = SaveYLimits;  // restore limit settings

Not sure how that would effect RTCP but it might disable the axis. Please fix this and see if it still occurs.

[George Butty]

Hi Tom, fixed homing file. Unfortunately this did not fix problem. The switching problem between RTCP and G49 mode is there but is very random.
What is not random is I home machine and then set B90 and zero out the end of the setup bar ( 20" from rotation center) to a dial indicator.
I set G49 and then rotate C-180 and back to C0. Dial comes back to zero every time. If I rotate C180 and then back to zero the dial is out by .08" every time . I rotate C-180 again and back to C0, the c axis goes back to the proper dial zero every time.

George

[TomKerekes]

Hi George,

Regarding the "backlash": It isn't clear what you are measuring with the dial indicator. The "length"? That is the "tool" is pointing in the +X direction. The dial gauge is measuring the end along X?

With G49 when moving C nothing else is moving correct?

Regardless, record all the Destinations of all the Axes on the Axis Screen before and after the moves. If the Destinations are the same but things are physically different then it must be something mechanical, or loss of steps, or the drives have following errors.

[George Butty]

HI Tom, I attached a video of the problem. This will give you a better idea what is happening. In G49 mode
you can see that I run the C axis only, to C-180 and then back to C0 then C180 and then back to zero. I set up
a round tip probe so you can see the difference of the C0 going one way and then the opposite direction.
It is out far more than any mechanical error I have in the axis. Also I checked B Axis and its doing the same.
I can't see that the Teknic motors are missing any steps because the zeroing will correct it's self if I repeat the
C-180 to C0 direction and also if you look at the Teknic motor the wrong way, it faults out mediately.
The other thing is the error does not change no matter how many times I cycle it.
The DRO values are correct on the screen in both directions.
Is their any chance that the rotation in the the bad direction is reading an incorrect scaling factor.

https://www.youtube.com/shorts/9FvSK_Fxfwg out a lot.

George

[TomKerekes]

Hi George,

So I would expect that to be caused by an issue with the C or Y axis. Record all the Destinations of all the Axes on the Axis Screen before and after the moves. If the destination is the same but the position has physically moved that should pretty much exonerate KFLOP or any software issue. Other than if KFLOP is somehow not reporting correctly the number of actual steps it is actually generating which I think is unlikely. Backlash correction would do such a thing but your code shows that disabled and would normally be in the opposite direction.

What happens if you move only +/-5 degrees?

What happens if at C0 you push on the tip physically in the +/- Y directions?

It seems you are measuring relative to the floor. So anything in the entire structure could cause this.

You might probe other parts of the structure like points on the y Carriage relative to the floor to see if anything is shifting.

Sorry no simple answer.