Can't return to TCP

[Calum]

Hi Dave,

I have something strange going on with the TCP that I can't get my head around.

I Ref everything then set a TCP now when I move off the TCP then hit the Go TC Pos't button the current position is off by around .0188 mm in X and Y and a smaller amount in Z.  If I reset Mach3 by using the ready button the machine will return to the TCP when I press the Go TC Pos't button.

I'm sure I didn't have this problem when I got MTM going the other day, I have replaced the XML with a copy from my original Mach3 and started over 2 or 3 times with the same result now, must be something I'm missing any ideas?

Cheers
Calum

[DaveCVI]

Calum,
I haven't experienced this.

1) Are you remembering that TCP is always in MCs?
Are you maybe looking at a current position in WCs  and wondering why it is not the same as a MC value?

2) Put the current Position panel (to the left of the TCP panel) in M-Coords (MC) mode - now you can see the current MC position next to the TCP (in MCs) position.
click "Go TC Pos'n"
When it stops short are the MC DROs different from the TCP DROs?

3) The core of the GoTo TCP button is pretty simple and uses G53 to go to the TCP; and the TCP is just the contents of the TCP DROs you see on the tooling page.
If the curr position DROS (In MC mode) are not at the TCP when motion stops, that is very weird.
Something has to be stopping mach's motion... So let's think about what could stop mach from getting to a commanded position -

Soft limits? Are they stopping you from getting to the TCP?

4) Are you running 3.43.10 or 3.43.12?
have you tried a mach reinstall?
(as an experiment, I'd just install mach on top of itself to see if it makes any difference).

Dave

[Calum]

Hi Dave,

1)I had my current position DRO in MC mode.

2)The current position in MC mode is different to the TCP by about .0188mm.

3)Weird indeed.  No soft limits set.  I have a Smooth Stepper so I thought I would up the pulsing frequency but no change there either.

I was running 3.34.10 and have now reinstalled that.  I rebuilt my XML from scratch and still have the same problem.  I may blow it all away and start with a clean sheet just incase there is something hanging around from a previous install.

Calum

[DaveCVI]

0.0188mm is not very large -
What's the distance of a single step on your machine?
Is this an integer multiple of a step?

Dave

[Calum]

Well spotted it is exactly 1 step, weird that it is always out by 1 step.   This is enough to stop MTM working as if reports that the tool isn't @ TCP, can you think of anything I can do to get around this?

Calum

[DaveCVI]

Does this happens when you try to use settings-common page, "Set TCP TP MCz"?

Please confirm that this is where you get the MSM message re not at TCP.

Within that routine I see a bug in the code - two sets of registers are compared for being within a tolerance vlaue - but I forgot to scale the tolerance value for in vs mm  machines 

Once I know that is the context for the issue, I can build you a custom probing lib. That will get you running until the fix shows up in the next release.

Dave

[DaveCVI]

Calum,
Given the time zone difference between us, I tried to go ahead and email a patch to you via direct email.
Alas, the email address your user account name shows, bounces with and unknown host error.

PM me with a good email address and I'll send the patch to you.

Dave

[DaveCVI]

Calum wrote the following:
('ve moved his post to a new thread to separate the topics)
Dave

Thanks Dave, That the 0.70 Probing lib fixed the positioning issue and I can now run the "Set TCP TP MCz “ function every time.

<content snipped / moved - Dave>
Calum