Hello Guest it is September 26, 2020, 01:14:22 AM

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - smurph

Mach4 General Discussion / Re: LOVING the PMC!! Thanks Steve!!!
« on: August 31, 2016, 08:12:55 PM »

I added the encoder registers in the list of input registers for the PMC.  So the next build will have that in there.  However, I really don't think the PMC is the proper place for it.  That sounds like something the motion controller plugin should be doing.


Mach4 General Discussion / Re: LOVING the PMC!! Thanks Steve!!!
« on: August 30, 2016, 11:39:18 PM »

At this time, encoder registers are not included in the list of input registers.  But are you needing an encoder value in the PMC?  To drive an MPG?  If so, you can associate an encoder to a MPG object in the Mach config.


Mach4 General Discussion / Re: Re: Mach 4 and Galil
« on: August 23, 2016, 12:04:19 PM »
We are in the final stages of releasing it.  Be aware that it will be a licensed product for Mach 4, meaning there will be a cost to it.  As to what the cost will be, I don't know at this point.  

The "supported" controllers will be the 40x0 and 41x3 controllers.  However, the older 21x3 and 21x0 controllers may work but no testing has been done at this point.  Rigid tapping and lathe threading are not implemented yet.  ISA cards will not be supported at all.  

Galil is working on a special Mach 4 firmware that will enable advanced features such as movement coordinated outputs, single input probe latches, backlash compensation, and screw mapping.  This firmware will be available to users that purchase a 41x3 or 40x0 from Galil.  This firmware will not be available for the older controllers.


Mach4 General Discussion / Re: CNC movements not smooth
« on: August 19, 2016, 04:16:53 PM »
The only M codes that don't break the CV chain are M62 and M63.  M03 and M05 will.  If you think about it, you would never want the look ahead to turn your spindle off (M05) when the machine is actually still 100 lines back in the G code (assuming you have a 100 line look ahead set).  So when the interpreter look ahead runs up to a M05, it simply waits until all of the prior motion is complete (and it will be an exact stop) before turning the spindle off. 

The reason Pedio's code works is because the S word is given just once before a large set of G01 blocks.  CV is in effect for all of those G01 blocks. 


Mach4 General Discussion / Re: Mach4 5-axis
« on: August 19, 2016, 04:07:00 PM »
We do not do kinematics (yet), so a gimble head is not possible.  But 5 axes with a trunion table setup is.


Mach4 General Discussion / Re: CNC movements not smooth
« on: August 19, 2016, 01:37:15 PM »
Just to be clear, the S word separating the G01 moves IS the problem here.  If you take out all of the S words, you will see it run very gracefully in CV mode. 


Mach4 General Discussion / Re: CNC movements not smooth
« on: August 18, 2016, 02:54:18 PM »
It is M62/M63 in Mach 4.  And the motion controller you are using has to support it.  They are special M codes that are an exception to the rule of M codes breaking the CV chain.  I'm not sure if ESS supports them yet or not.  :(  And they also only operate outputs, not some analog value.  So you would have to vary the PWM output based on which output was selected.  maybe 10 outputs to span the range of the PWM?

There is a LUA instance that runs the PLC script.  It basically gets executed at a predefined interval.  That is where I would do any code like what you are talking about. 


Mach4 General Discussion / Re: Modbus and Sim
« on: August 18, 2016, 11:12:10 AM »
No, Brett is not normal at all.  But then again... neither am I.  :)  Anyone that plays with this stuff can't be normal.  Anyway, have fun with the modbus stuff. 


Mach4 General Discussion / Re: CNC movements not smooth
« on: August 18, 2016, 12:12:48 AM »
The issue is all of the G01 lines are separated with S lines.  S breaks the CV chain.  In fact, M, S, and T will all break the CV chain.  In a Fanuc control, they refer to it as MST.

If you think about it, the look ahead would basically set S to the last value it read even when the machine is actually running the FIRST line of G01 code.  So we MUST break the CV chain and WAIT on that G01 line to run before advancing to the next S line.  Then, you will have spindle speed ramp times in there too.  You can probably set them to zero and alleviate that.  But nothing will keep it from going into exact stop mode for the S word.


BL is implemented in the motion controller.  And not all of them do it.  :(