Aciera wrote:

Sounds correct. Famous last words

Yea, that was backwards

Makes sense in hindsight: I want the V-DRO to move at twice the speed of the joint, not half.

But everything's working now: switchkins, subspindle synch to main with CSS and the parting slide programmed as "X".

I couldn't compress a GIF small enough to embed, but I've attached a short video as a ZIP file.

parting_slide.comp and the four M-codes are all commented pretty heavily for future reference.

I think this is nearly the last hurdle on the DIY lathe.  Only big challenge I can think of is mounting tools for subspindle work.

Thanks again @Aciera, couldn't have done it without you.

Hakan wrote:

Synchronization of slave clocks can sometimes take a long time at start.
There is a warning in syslog if it takes more than 5 seconds, or is it 30?
Don't know how to fix it, more than hope the slaves' clocks run more equal
when not being nudged by the master.
Add pll-err to the halscope plot. 
 

I saw at my own eyes synchronization of 120 seconds with beckhoff SoE AX5125 driver with older firmware ... its not odd behaviour 

beckhoff IPCs have sometime very similar behaviour with their own hardware from my close experiences...

hi there,

if you do not want to change timing in the TC3 to 1ms (easiest way and no play to errors) then change -

and in ethercat conf xml change -

and in .ini file change - 

check cabling CAT5e with shielding ... its pretty sensitive stuff

check if you have non generic driver ... it has to be native NIC driver at network card with EoE turned off => very important

set time of task in TC3 to 1ms and export it ... 

connect just and only AX5201 to ethercat master... export config and translate it (problem isolation)

change 0x0740 to 0x0730 ... timing have to be correct because it is direct timing extraction from beckhoff master and exported to our shape ... 

Well that's a clever idea.

Just so Im clear, the pos->v = 2*j[2];function results in the V-axis moving twice the distance of JOINT_2?

And the reverse forj[2] =0.5* pos->v;in the Inverse section?

Which means in case 1 I'd want to adjust the X-axis:pos->tran.x = 2*j[2];So the X-axis moves the correct distance when connected to the parting slide joint?


Which means I can adjust the INI stepgen position scale so it's not cut in half, and then adjust both case 0 and case 1 so the V-axis moves half the distance (diameter mode).

• Don't adjust the case 1 X-axis/joint_2 relationship, because X-axis will move at half-distance when in diameter mode.
• DO adjust the case 1 V-axis/joint_0 relationship, because LCNC doesn't have a diameter mode for any axis other than X.

Make sense, or have I got something wrong?

grandixximo wrote:

@PCW
Agreed that an ideal TP should always obey constraints, and the full stop fallback is the correct mechanism to achieve that. No argument there.
The question is whether strict enforcement should be mandatory or an integrator choice. A 20% jerk overshoot on a 1ms grid at a G1 corner is not an unknown uncontrolled violation, it is small, bounded, and predictable, and on drives with internal interpolation the mechanical system never sees it. That is a different situation from the uncontrolled external factors you are rightly warning about.
The practical concern is cycle time. On a complex CAM toolpath with many short segments, mandatory stops at every corner that fails the jerk test could multiply execution time significantly. That is a real regression for integrators who have matched their drive capabilities to their application.
So the proposal is: implement the jerk-aware full stop as an option, off by default to preserve existing behavior, and let integrators who need provably clean command streams or are running drives without internal interpolation opt in. That gives you the mathematically correct behavior when it is needed, without penalizing everyone else.
 

non budget drives always have more step filtering to suppress these behaviour... I think from differential calculus it is exactly that we need count with some limits of movememt caused by curving, jerk implementation fundamentally ...

it should be switchable from Gcode directly by user, who knows whats is doing... if the trajecotry tolerance is too high and velocity in that moment is non sensing plus path is small segmenting... some overlimit during small amount of time are acceptable(timing tolerance and spiking toleracne should be controlled by the user)

bending(trajectory) the tolerance - to handle this situation is non sense 

So here's an interesting problem for anyone following along...

I cut the parting slide stepgen scale by half (JOINT_3) so the V-axis can be programmed in diameter mode.  When set like this jogging/MDI/g-code is just like X in diameter mode - very intuitive.

Getting ready to test G96 CSS and double-checking everything - set G54 offsets, etc.  I enable switchkins and everything's fine...

But now the virtual X parting slide is only moving half the distance it should.  DRO is following the movement - it's moving at half the distance it should as well.

Ok, got it - I've got the scale cut in half, and LCNC is cutting that in half again when I switch the kinematics and am commanding an X move.

Easy solution: change the JOINT_3 scale back to full-distance value when switchkins is active.  But...

LCNC instantly throws a JOINT_3 following error when the scale switches.  Bugger.

I can double the V-axis/JOINT_3 scale and fiddle with the jog values, but that will still leave the V-axis G54 offset wonky.  I write the V-axis offset to the X-axis when switchkins is enabled (along with max/min limits) so the parting slide tool-tip at centerline can be set to "X0" for CSS.

How can I have my cake and eat it, too?

• V-axis/JOINT_3 to be in diameter mode when switchkins is not enabled, and
• X-axis/JOINT_3 to be in diameter mode when switchkins is enabled
• No following error when enabling switchkins
• All programming in G7 mode - switching back and forth between G7/G8 is going to cause my brain to hurt and a crash.

@PCW
Agreed that an ideal TP should always obey constraints, and the full stop fallback is the correct mechanism to achieve that. No argument there.
The question is whether strict enforcement should be mandatory or an integrator choice. A 20% jerk overshoot on a 1ms grid at a G1 corner is not an unknown uncontrolled violation, it is small, bounded, and predictable, and on drives with internal interpolation the mechanical system never sees it. That is a different situation from the uncontrolled external factors you are rightly warning about.
The practical concern is cycle time. On a complex CAM toolpath with many short segments, mandatory stops at every corner that fails the jerk test could multiply execution time significantly. That is a real regression for integrators who have matched their drive capabilities to their application.
So the proposal is: implement the jerk-aware full stop as an option, off by default to preserve existing behavior, and let integrators who need provably clean command streams or are running drives without internal interpolation opt in. That gives you the mathematically correct behavior when it is needed, without penalizing everyone else.

there was a surface map project a while back, not sure if that would be something that was useful or not. but i never really used it much, it seems like it was for larger routers that may have uneven beds, so im sure it could be configured for your use case. maybe the creator of it can chime in. I will ping them on the chat and ask if they can step in and reply.