WARNING: Mach 3.43.59 Feedhold crash bug report

[DaveCVI]

Hi,
I'm sorry to say that there seems to be another serious bug in Mach turn mode. This bug can cause a crash of the machine whenever you resume from feedhold with Mach 3.43.59 (in turn mode).

YOU ARE ADVISED TO NOT USE FEEDHOLD IN Mach 3.43.59 in Lathe mode until this is fixed by ArtSoft.

This bug has been reported to ArtSoft as of 3-29-2012

FYI - here are the links to the bug report videos that have been supplied to ArtSoft:

youtu.be/ZyAxrJBxnNE (part 1)
youtu.be/Ow9p_dGAKMo (part 2)

The bug report was done with stock mach 1024 screens (I really hate those ), this prevents any accusation that it is MSM that is causing the problem. However this bug will crash MSM Turn or Mill-tun operations as well (I mangled metal to prove it ).
Dave

[woffler]

Thanks Dave,
I was just getting ready to update from MSM 1.1.18 and Mach 3.043.037 .
I have a lathe set up on my mill and i use the A axis as a spindle with a chuck. for threading with Mach Turn .
I believe i will wait and see what shakes out of Mach before creating a mess for my self!

[DaveCVI]

FYI - The wait should not be too long - I have a pre-release mach version from Brian that has the feedhold bug and some others fixed. I've sent feed back re my test results and I hope to see a new mach rev in a day or so.

The I'll have to make a new MSM beta rev as one of the changes he is making is not backwards compatible with prior mach versions.

Dave

[thosj]

I see 3.43.60 is posted. Nothing in the changelog, of course. I'm holding out for a new MSM beta and perhaps news from here.

[DaveCVI]

mach 3.43.60 is a no-go from my viewpoint.
It only partially fixed some of the issues - so skip that one.
in fairness to biran, he did say it was an interim snap shot for testing.

He shipped me a .62 for testing that is looking ok wrt to what I've tried so far (feedhold bug and X diam handling). But I don't have a list of all that is in it except for what Brian posted on the yahoo list earlier today.

I think he'll get a better mach version posted once he gets more tester feedback.

Dave

[woffler]

That sounds good, i am looking forward to a good version to play with and see what happen's.

[thosj]

Well, I see Dev. 062 posted. Is this one to be avoided, or should I give it a try?

[DaveCVI]

Tom,
How brave are you?

From the mach change log:

4/3/12
3.043.062
-- Lathe - Program extents worked on. The fixture offset was getting added in

This change is probably benign and I'm not worried about it.

-- Lathe - worked some more on the Dia vs Rad. There was a problem with setting an axis to a position

This entry does not tell the whole story. There have been major changes in the way Diam mode works as of mach 3.43.62. X for WC values is still a diam, but all others are now setup units/radius (G53, G52). This is a good thing in the long haul, but it is NOT backwards compatible with prior versions of mach.

I am working on a rev of MSM that checks the mach version and gives parameters to mach using either old or new semantics depending on the mach version running. I have the code done, but have not yet finished the testing required prior to releasing a new beta.

So if you use MSM 1.3.1 (assumes old semantics) with mach 3.43.62 (new semantics) undesired things may happen. One trouble area is the storage and handling of the TCP. MSM 1.3.1 compensates for the old mach semantics when moving in MC in diam mode - but mach 3.43.62 is using new semantics... so the result may be 1/2 or 2x of where you would have gone in mach< 3.43.62.....

If you use radius mode you should be fine, but diam mode may present surprises until I can get a revised MSM beta out.

-- Lathe - Fixed the vars that get sent to the m1083 macro. The tTaper() was not getting set

Hum, I have a 3.43.62 that I tested - and this was added after I got that, but the posted mach release is also called 3.43.62.....
so I have not yet explored this and have no idea what the impact of this is.

...and what is not stated in the mach change log (which is really important IMHO) is that (at least in the 3.43.62 I have, which we know is different from the 3.43.62 that is posted on machsupport), is that the resume from feed hold causes crash bug is fixed.

So, to not answer your question, how intrepid an explorer are you?

Dave

[thosj]

Probably not that much of an explorer! I'll at least wait for your updated version, and hopefully a SS fix, too.

All these versions and cross compatibility is pretty out of control. For example, I discovered my VistaCNC iMachIII P3 pendant doesn't work right now. I don't know what Mach version broke that, but ALL 4 axes only jog one direction. I've emailed Lee Neuman, but no reply as of yet. So I, at least, can add one more software in the mix.

[DaveCVI]

MSM 1.3.2 beta has been posted.

Highlight for this release:
MSM now has logic required to handle the incompatible changes made to Mach diameter mode (introduced in mach 3.43.62). Internally MSM now adjusts diam mode handling based on the mach version running).

Other details in the release notes.

NOTE: The feed hold crash bug has been fixed in mach 3.43.62. Therefore, mach 3.43.63 and MSM 1.3.2 are now the recommended combination for MSM turn and Mill-Turn operation.

Dave