4th axis and Feed Hold bug

[bryansnyder]

I recently got a 4th axis setup and so also decided to upgrade MSM. I installed mach .66 as reccomended. I didn't run for very long before doing the upgrade, so not sure if this is a version issue or not, but now when I press FeedHold and then press play to continue I get a junk target position that mach starts trying to move to. So now my only option is to hit pause, then stop, the restart program from previous line (and go thru all the tool changes, intermediate moves, etc). Kind of painful. Searching the forums I see some others have had this problem but don't see any resolutions other than 'maybe it is related to brains', so is it maybe an MSM conflict? The only add in I have is a Shuttle Pro ubs 'pendant'.

A few other minor issues:

1) Spindle speed DRO displays incorrectly now. The one box is the base command, but the actuals box (cmd * SRO) now looks like its displaying acceleration -- it sticks at zero and the surge up and back to zero anytime there is a commanded speed change.
2) Feedhold seems even slower to respond. Takes a solid 1.5 seconds before anything happens after pressing feedhold.
3) Because of the issues with the bad moves after a pause I've been performing axis verification more often. If TLO is on, I get a verification routine, but at the end the Z axis moves up and re-trips the limit switch and faults. Not sure if the verification routine is mach or MSM but looks like its trying to include TLO when commanding a move to machine zero.

Thanks for the help.

Bryan

[DaveCVI]

Hi,

I recently got a 4th axis setup and so also decided to upgrade MSM. I installed mach .66 as reccomended. I didn't run for very long before doing the upgrade, so not sure if this is a version issue or not, but now when I press FeedHold and then press play to continue I get a junk target position that mach starts trying to move to. ...
Bryan

What version of mach did you have before 3.43.66?
What was the prior version of MSM and the current version?
Is this a PP or a SS or a ? for the motion device?
I'm not sure what you mean by "junk target position" - I don't get any type of dialog box after I feedhold then resume.
Have you done anything else during the feedhold that I should also be doing for a test (jogged an axis maybe)?

... So now my only option is to hit pause, then stop, the restart program from previous line (and go thru all the tool changes, intermediate moves, etc). Kind of painful. Searching the forums I see some others have had this problem but don't see any resolutions other than 'maybe it is related to brains', so is it maybe an MSM conflict? The only add in I have is a Shuttle Pro ubs 'pendant'.

For what it's worth MS does not handle feed hold - that's all mach. So one option may be to go back to the prior version of mach to see if the problem goes away. MSM will run with any mach version later than 3.43.22 - but there are various fixes for thing is each mach version so it's sort of what works best for you. We recommend 3.43.66 as it has few complaints as far as we can tell - but it's known not to be perfect (no mach version is that I know of).

Slightly related: When I was being taught to run gcode etc. I always went back to the most recent tool change point when I had to restart a program (those control had no "run from here" button - which often does not work in mach anyhow from what I read). A tool change tends to be a clean reentry point as TLOS get set for the new tool etc. If you still have WC 0 set, then it's less time than running from the beginning.

A few other minor issues:

1) Spindle speed DRO displays incorrectly now. The one box is the base command, but the actuals box (cmd * SRO) now looks like its displaying acceleration -- it sticks at zero and the surge up and back to zero anytime there is a commanded speed change.

The spindle speed stuff has not changed in a long time. If I knew what the prior mach and MSM version were, I could look thru the release notes (well the ones for MSM anyhow, there generally aren't any for mach) and see if something comes to mind.

2) Feedhold seems even slower to respond. Takes a solid 1.5 seconds before anything happens after pressing feedhold.

Not sure why that would be - it's all handled by Mach.

3) Because of the issues with the bad moves after a pause I've been performing axis verification more often. If TLO is on, I get a verification routine, but at the end the Z axis moves up and re-trips the limit switch and faults. Not sure if the verification routine is mach or MSM but looks like its trying to include TLO when commanding a move to machine zero.

I;m not sure I"m following on this one... I hardly ever do a verification (essentially except for trying it for testing, I haven't tended to need to use it). The actions of the verification button are mach code rather than MSM code.
If you'd lead me a bit more step by step as to what you are see4ing, I'll try to duplicate your steps to see what I get to happen.

Thanks for the help.

Bryan

Not sure how much help I was re this post - but "any time"!
Dave