Calypso CNC User Forums

Posted: **Sun Mar 09, 2014 11:23 am**

Hi Dave,
I am experiencing an issue with my MSM Pro when running a tool change routine using M6 Txx, I am using master tool mode with my probe set up as master tool (99).
I call T99 M6 manually and probe the tool change table plate (calibrate) then probe the feature on my workpiece to give my G54 all good so far.
I run my first op which includes a M6 T8, it runs the M6 goes to the tool change position ( tool is changed) then it feeds to the tool change plate and does the double touch and eventually returns to the run page and continues the operation using the tool offset measured, all still good so far, I may not finish the programme correctly with the M30.
I then run the next operation (a separate programme is loaded) and if I use the same tool number that is already the active tool during the M6 tool call the machine goes to the tool change position, the tool is changed or not in this case, cycle start is pressed then I get an M6 end error, the programme then continues without measuring the tool but the TC LEDs are still lit whilst running the programme, further tool changes then do the same thing until Mach3 is reset ( reboot)
Any Thourghts?

Posted: **Mon Mar 10, 2014 2:52 pm**

Hi,
let me start with a couple of questions so I'll know if my assumptions are warrented or not:
1) MSM version?
2) mach version?
3) motion device?

Treb63 wrote:Hi Dave,
I am experiencing an issue with my MSM Pro when running a tool change routine using M6 Txx, I am using master tool mode with my probe set up as master tool (99).
I call T99 M6 manually and probe the tool change table plate (calibrate) then probe the feature on my workpiece to give my G54 all good so far.
I run my first op which includes a M6 T8, it runs the M6 goes to the tool change position ( tool is changed) then it feeds to the tool change plate and does the double touch and eventually returns to the run page and continues the operation using the tool offset measured, all still good so far, I may not finish the programme correctly with the M30.

OK that all sound normal.
Would you elaborate a bit re the "... may not finish the programme correctly with the M30." ?
Is this just a simple situation of you git feed hold or stop instead running thru an M30? Or did something else happen that may be a clue I should be thinking about (Crash? error messages?)
Executing an M30 to end a program is not required - M30 is end program and then rewind program to start. It's ok to end a program with out rewinding it (rewinding is actually a hold over from the days of paper tape programs).
M0 and M2 are other ways to end a program- but even they are not strictly required as far as I remember.

I'd like to know [precisely the way the progra was ended - as this sets up the state for mach just before the 2nd part of your question.

Treb63 wrote: I then run the next operation (a separate programme is loaded) and if I use the same tool number that is already the active tool during the M6 tool call the machine goes to the tool change position, the tool is changed or not in this case, cycle start is pressed then I get an M6 end error,

What is the exact error message? Can you get this to happen and send me the mach history log (Hist button in MSM)? Maybe take a screen shot of the error message?

I've tried loading a t# then loading a program where the 1st T3 called for is the same as the # already loaded - seems to work fine for me. So there must be more to the sequence of steps that I am not doing.

Treb63 wrote: the programme then continues ...

really? all by itself? I'm surprised that it could continue after a M6 error.
Did you have to do anything to make it continue (Press cycle start or ?)?

Treb63 wrote: ...without measuring the tool

That make sense to me - if M6End faulted early only the tool measurement stuff would not get run.

Treb63 wrote: ...but the TC LEDs are still lit whilst running the programme,

Also makes seom sense - the MSM Tool Change led is turne don by M6Start and off again by M6 End. So if M6 End did not run, it would get stuck in the ON state.

Treb63 wrote: further tool changes then do the same thing

By "same thing", do you mean that all M6 sequences give the same error message?

Treb63 wrote: ...until Mach3 is reset ( reboot)
Any Thourghts?

well at least that gets things fixed up.

Is a reboot necessary. or will a click of the reset button to put mach into estop and then another cliock of reset to get out of estop enough to clear the error condition?

Dave

Posted: **Mon Mar 10, 2014 4:46 pm**

Hi Dave,
Guess I should have been posting this whilst at home so I could have been a bit more accurate with the description but will be back on Sat so can fill in all the gaps then.

1. MSM v 2.0.10
2.R 3.043.066
3. UC 300

To finish off the programme my post processor sends all axis to machine co-ords zero then finishes with an M30, sometimes I don't wait for the axis to get to zero so press stop/feed hold, then close the programme in the load page.
Loading the next programme and pressing cycle start is usually the next step.

Will get a screen shot and detailed error log on Saturday. For memory it's something like 'error in M6 xxxx.end'

The fault occurs when the tool is about to enter the auto tool length measure routine after pressing cycle start following the tool change, I expect the tool to head down to the touch plate but the fault occurs and the tool ignores the measure part and proceeds back to the run page and the rest of the programme with the fault displayed at the top of the screen and the TC LEDs still alight. The next M6 in programme will then do the same thing despite calling a different tool.

I have found a reboot of mach will clear it ( not surprising) however I think it also clears by closing the programme being run calling for any tool in MDI which strangely does complete with a tool measure sequence, I can then load the programme and continue normally. I think the MDI tool call sort of re- latches the system somehow. The reset button although putting mach into estop doesn't fix the issue.

I will endeavour to get a screen dump and error log ASAP.

Regards
Rob

Posted: **Mon Mar 10, 2014 5:11 pm**

DaveCVI wrote:Hi,
let me start with a couple of questions so I'll know if my assumptions are warranted or not:
1) MSM version?
2) mach version?
3) motion device?

Treb63 wrote:Hi Dave,
I am experiencing an issue with my MSM Pro when running a tool change routine using M6 Txx, I am using master tool mode with my probe set up as master tool (99).
I call T99 M6 manually and probe the tool change table plate (calibrate) then probe the feature on my workpiece to give my G54 all good so far.
I run my first op which includes a M6 T8, it runs the M6 goes to the tool change position ( tool is changed) then it feeds to the tool change plate and does the double touch and eventually returns to the run page and continues the operation using the tool offset measured, all still good so far, I may not finish the programme correctly with the M30.
OK that all sound normal.
Would you elaborate a bit re the "... may not finish the programme correctly with the M30." ?
Is this just a simple situation of you git feed hold or stop instead running thru an M30? Or did something else happen that may be a clue I should be thinking about (Crash? error messages?)
Executing an M30 to end a program is not required - M30 is end program and then rewind program to start. It's ok to end a program with out rewinding it (rewinding is actually a hold over from the days of paper tape programs).
M0 and M2 are other ways to end a program- but even they are not strictly required as far as I remember.

I'd like to know more precisely the way the program was ended - as this sets up the state for mach just before the 2nd part of your question.

Treb63 wrote: I then run the next operation (a separate programme is loaded) and if I use the same tool number that is already the active tool during the M6 tool call the machine goes to the tool change position, the tool is changed or not in this case, cycle start is pressed then I get an M6 end error,
What is the exact error message? Can you get this to happen and send me the mach history log (Hist button in MSM)? Maybe take a screen shot of the error message?

I've tried loading a t# then loading a program where the 1st T# called for is the same as the # already loaded - seems to work fine for me. So there must be more to the sequence of steps that I am not doing.

Treb63 wrote: the programme then continues ...
really? all by itself? I'm surprised that it could continue after a M6 error.
Did you have to do anything to make it continue (Press cycle start or ?)?

Treb63 wrote: ...without measuring the tool
That makes sense to me - if M6End faulted early only the tool measurement stuff would not get run.

Treb63 wrote: ...but the TC LEDs are still lit whilst running the programme,
Also makes some sense - the MSM Tool Change led is turned on by M6Start and off again by M6 End. So if M6 End did not run, it would get stuck in the ON state.

Treb63 wrote: further tool changes then do the same thing
By "same thing", do you mean that all M6 sequences give the same error message?

Treb63 wrote: ...until Mach3 is reset ( reboot)
Any Thourghts?
well at least that gets things fixed up.

Is a reboot necessary. or will a click of the reset button to put mach into estop and then another cliock of reset to get out of estop enough to clear the error condition?

Dave

Posted: **Sun Mar 16, 2014 4:49 am**

Hi,
Attached are two files the error log and the programme being run.

Sequence was as follows,
Power up
Reference
Load T99 M6 (spindle probe master tool) in MDI
Probe table touch probe. (g54 part offset set earlier)
load programme
run programme in single block
M6 tool call loads t3 and measures correctly
programme completes with the M30 All OK so far

press cycle start to re run programme not in single block
M6 call loads tool
Press cycle start to complete tool change and measure
fault is displayed and cycle then continues without measuring tool and without needing cycle start again.

fault is 'Scripter Compile error in :M6END.M1S '

On this occasion the fault occurred the second time the programme was run the first run the programme finished correctly with no apparent alarms or faults.
The programme was not closed and re loaded just re run.

Any help appreciated.

Posted: **Tue Mar 18, 2014 12:40 pm**

Hi,
This is a bit troubling... I believe you see the scripter error as shown in the history log.
Yet, I've run this program using your instructions at half dozen times and I can't get a fault to happen.

The scripter error is really strange as the M6End script is not being compiled when it is run- it's already compiled and all mach does is load it and run it. Of course, I have no idea if the mach error message is precise or not.

I'm guessing that there may be a mach internal timing hole that is impacting the loading of the M6 scripts.. but that is only a guess based on gut feel from lots of messing around with mach - I have no good way to prove or disprove that theory at the moment.

I'll put my thinking cap on, but until I get an Idea, I'm not sure what to try next to make some more clues show up .
I
ll try the test program on different machines, but it will be a couple of days before I can get the other machines free for testing.

On idea as I was typing: it's a long shot: turn off the "T Path" button on the run page. It's sort of well known that the tool path plotting stresses mach - perhaps if the tool path is off, the program will behave differently...?
What type of CPU horse power is in your Mach PC?

Dave

Posted: **Tue Mar 18, 2014 1:33 pm**

Hi Dave,
I have mach3 running on this pc/nettop; Running with Windows 8.
Acer Revo RL80 Nettop PC, Intel Celeron 1007 1.5GHz, 2GB RAM, 500GB HDD, NOOPT, Intel HD, Wifi,
Also the U300 claims to ease the load on the processor see below, so far it has been excellent and very easy to work with.
I will try the suggestion below tonight and see if it helps any I must admit I do like to see the tool path when running and with a touch screen monitor do occasionally rotate the plot mid programme.
regards
rob T
----------------------------------------------------------------------------
http://www.cncdrive.com/UC300.html
Mach3 is a popular CNC machine tool control software running on desktop computers and
laptops under Windows operating system. Mach3 in default working the computer's
parallel (printer) port and sending all signals via this port. Windows is not a realtime
operating system and therefor it is a difficult and CPU time consuming task for Mach3 to
generate these signals in a precise manner. As the output signals frequency gets higher
with using todays more and more advanced motor control electronics Mach3 needs the
more processor time of the computer. If the computer gets overloaded by Mach3 or by
running other software(s) in the background may leads to a lagging of the G-code code
execution to even a falter in the motion interpolation which may lead to a catastrophic
failure of some applications.
Another problem with the LPT port is that in today's informatic technology this kind of port
is no more used and is a dying out device and non of the new laptops having this kind of
port built in anymore.
The UC300 motion controller overcomes these problems with removing all time critical
tasks from the control computer and Windows and executing all these tasks on inside it's
own high speed DSP control chip outside the PC.
The UC300 controller connects to the PC via the USB port
Because of using a high speed DSP core the UC300 making the interpolations and other
timings much more precise as Mach3 could ever do using it's standard LPT port driver and
because the time critical tasks are no more done on the PC side, but on the UC300
controller the PC's CPU time usage is much lower and the risk of overloading the PC
resources lowers with factors as the UC300 has a long enough data buffer so that it can
handle a 100% CPU usage and overload for some seconds without interrupting the motion
and the G-code execution.
--------------------------------------------------------------------------------

Posted: **Wed Mar 19, 2014 3:55 pm**

OK, very bad Day today lots and lots of annoying errors which are hard to pinpoint a few Cyprus errors which only stay on the screen for a short time so can't copy them but all required a reboot and required set up to continue. Rant over I still much prefer the power of MSM.

Now I have tried lots of methods to be able to reliably recreate this issue and it seems to happen at almost random times however something that may of interest is that when I get the M6 end compile error I can stop the programme immediately then run the same command in MDI and it completes correctly ie M6Txx then does the measure routine. I can then continue the programme correctly, there shouldn't be a lot of difference between an MDI M6 and M6 being called from within a programme?

Something I want to try to simplify my set up is just have one Tool table saved in the PC, what is the best method to clear it all out and start again? Is there any standard name for a tool table that Mach loads by default or should I be doing this manually as part of the start up procedure? Just though it may be a contributing factor to the M6 error?
Regards Rob

Posted: **Fri Mar 21, 2014 5:46 pm**

Hi,
Mach only has one tool table internally. See the MSM manual for the relationship between MSM tables and the internal mach tool table.

If you have the equipment, the best test you could do would be to swap from the UC300 to a known good and supported motion control device: Either a parallel port or a Ethernet smooth stepper.

No other MSM customers have reported this script error with M6end problem and I can run your code test case over and over but can't get any such error (I've tested on a PP driven mach setup). I suspect a timing issue internal to mach - possibly caused by the plugin interface of the UC300.

Dave

Posted: **Sat Mar 22, 2014 6:49 pm**

Hi Dave,

Lots more testing, taking your comment about a timing issue I have tried to adjust the look ahead down to 1 line then 0 lines no improvement now back to 20, deleted all the tool tables and started again, no improvement.

So I thought for safety and to get something to work I would try putting a M1 op stop after the T** M6 line so I could at least run a MDI M6 after the fault in order to change tools manually and carry on machining.

Surprise surprise the M6 works fine now in Auto with the tool measurement routine BUT after the tool measurement the control returns to the run page and ignores the next line (M01) and continues!

I have tried lots of different commands there M0, G4 etc. but it ignores them all, this has been hidden in my programs because the M3 S*** call after the tool change has been ignored and not noticed by me before as I am switching on my spindle manually at the moment anyway.

As soon as I remove the next line after the M6 line the fault returns.

I haven't tried putting the M3 S*** line down a few lines to see what happens yet but will play around with the possibilities tomorrow some more.

I'm guess you haven't seen anything like this before

I'm guessing the fault likes with the UC300? I can see some more emails to the manufacturer in the near future.

Regards
Rob

Calypso CNC User Forums

Auto tool change issue?

Auto tool change issue?

Re: Auto tool change issue?

Re: Auto tool change issue?

Re: Auto tool change issue?

Re: Auto tool change issue?

Re: Auto tool change issue?

Re: Auto tool change issue?

Re: Auto tool change issue?

Re: Auto tool change issue?

Re: Auto tool change issue?