Thanks Rich,

I have re-read it and the reason for asking was that it goes against a couple of videos on the 'tube - nothing like a bit of confusion

Their idea was to go 100% speed and something 0.001seconds

I will have a play with it and see what makes a difference if anything

You can see taking the backlash out so i guess the controller is ok with it.

Dave,
Can you post a link to one or two of the videos so I can see what they are doing?

RICH

Here's one link...

https://youtu.be/A27nOkTFYDY

A bit long but skip to around 5:30


Dave

Thanks for the link. Watched the video.

My 50-50-50 rule is a "suggested starting point" for settings and one needs to adjust based on the machine and program settings.

Note that BC is implemented at change in axis direction thus the Gcoding comes into play. Motor tuning values define the rate of possible axis motion and are set by the user....and hopefully the settings allow for reliable / repeatable operation. So BC values
could be at 100% and some small time value.
 
Note also that in the video he adjusts BC to get a more perfect circle. What would happen if another circle with a larger diameter ( ie; 10x) was done in at different axis position? There will be a change in cut diameter tolerance. Just a matter of degree....


A few comments on backlash.
 
The first step is to know how much you have if your going to use BC to remove it. Backlash is directly related to axis / controlled point repeatabilty.

Each component of the axis can have backlash and each adds to the total. The total backlash may not apply over the complete travel of the axis. So the backlash value can vary depending on where and how it was measured.It's all a matter of degree of non-movement and what is causing it and additionaly where it is happening.

Any component of the axis is not perfect in it's motion.
An example is the thrust bearing you just replaced. You can only preload so much and must leave some play or have a motor capable of turning the screw.

RICH

Thanks

I'll try it out on some code, do a test piece.

I've measured BL as good as i can, at three points on each axis, all axes show <0.1mm backlash at all points.

Hmm....see what happens with repetitive short moves back and forth.
Would think it's the equivalent  of beating the poop out of the machine!


Similar having a  car stuck and your constantly going from forward to reverse with the pedal to the floor to rock it.
RICH

Yes, this was caused my recent ePid fault query, i was trying some V-Carve engraving for the first time and i had the following error set a bit too tight.

So, whats best to do - I can't really reduce backlash anymore physically without spending stupid money on surgical grade ball-screws which is just not worth it on this machine. Is it best to leave BC off and live with it or leave it on with the risk that it will hammer the backlash worse than it is?

Or set a slow comp time maybe?

I have an interesting experiment for you to try. I'll bet your Z axis is no problem with backlash, because it is gravity preloaded againt the screw and bearings in one direction. So try the same thing on the X&Y. Get a gas spring with a force greater than the cutter thrust and preload the axis in one direction. An initial experiment could be done with enough bungee cords to provide the force required and tie the ends to the wall, another piece of immovable equipment.
I bring this up because I once owned a Spindle Wizard knee mill conversion that had double nuts with bellville spring washers loading a second nut against the first so it stated against one side of the screw. That machine was not fast, 20 ipm cutting speeds, but would hold tenths with no problem, and round holes were like they were were bored!

Yes that would be the only way to guarantee zero backlash, but at speed its not so easy to do. The thing with gas springs or springs in general is that the force is not linear, only a portion in the middle of travel is reasonably linear.

Tried a few different settings, I think it will be ok on the standard mill setup with 50% speed and 0.1s but on the high-speed setup its far too violent when doing V-Carve stuff as its hundreds of very short lines and it just sounds and looks rough on the old girl.

Will do an engraving test tomorrow and see how it is with the new thrust bearings in and the rebuilt Z drive.