Messages

281

Mach4 General Discussion / Re: How to read/write Analog Inputs and Analog Outputs

« on: May 15, 2020, 04:13:31 AM »

I haven't got that implemented yet.   

Steve

282

Mach4 General Discussion / Re: Fusion 360 simultaneous machining on the A axis

« on: May 14, 2020, 04:50:10 PM »

Try using a Fanuc postprocessor and see if they have one of those setup for the 4th axis.  Then you may be able to adapt what they are doing for the Mach 4 post processor. 

Mach 4 is built to emulate a Fanuc 21i control, BTW. 

Steve

283

Mach4 General Discussion / Re: cant understand mcCvDisToleranceV001 behaviour

« on: May 14, 2020, 03:34:12 PM »

I'm talking about the special CV settings.  I run my machine in CV mode all of the time.  Constant stop only on special occasions.  My point is, I do not need to run the CV wizard in order to have good CV. 

The CV wizard is used when the machine is not capable of doing what is asked of it.  Like a machine with real low acceleration parameters but capable of high speeds.  A lot of people want to cut with their routers at 500 inches per minute but are running steppers that don't have much torque at those highers speeds.  They end up reducing the acceleration parameters so the steppers won't lose steps.  Then they program the 500 IPM feed rate and CV will cut a WIDE corner.  So the CV tuning Wizard allows the user to compensate for this by slowing the corner approach automatically based on angle.

CV is a function of the programmed speed and the acceleration parameters in the motor tuning.  CV will sacrifice accuracy in favor of speed.  So natural CV, without using the CV Wizard, will look at the path and calculate the maximum velocity that it can maintain to get around the corner.  If the acceleration parameters are lacking, CV will make a wide turn around the corner.  If you decrease the feed rate, CV will follow the path more closely. 

So what is the for this? 

1. Slow down and increase the acceleration parameter. 
2. Get stronger motors that are capable of having higher acceleration parameters. 
3. Program the part path to slow down before the corners. (Some CAM packages actually have a setting for this).
4. Use the CV Wizard to try and mitigate the issue.

Steve

284

Mach4 General Discussion / Re: cant understand mcCvDisToleranceV001 behaviour

« on: May 13, 2020, 11:03:58 PM »

I haven't looked at yours yet.  But to answer your question, no.  I don't believe anyone else is having these issues. 

MOST of the time, you don't even have to do ANY CV settings.  I have never run the CV wizard on my mill, for example.  The reason is my machine is very rigid and has powerful servo motors on it so I can have good acceleration parameters in the motor tuning section.  The CV tolerance wizard is usually used on a machine that has poor acceleration settings to try and modify the trajectory's profile so that the corner can be cut at a speed that doesn't excessively round the corner. 

Steve

285

Mach4 General Discussion / Re: Tool change Problem.

« on: May 12, 2020, 10:58:56 PM »

I don't know what you mean about "taking the same tool and run"...  Do you have entries defined in your tool table?  All for them for the same tool? 

Also, try putting the G43 Hx on their own line. 

The preload tool changer G code goes something like this.

%
T1 (preload tool 1 into the tool changer, no tool in the spindle)
...
T2 M6 (load tool 1 into the spindle, preload tool 2 into the tool changer)
M02
%

It will NOT change your offset.  It will change which tool Mach thinks is actually in the spindle!  So your H code was different than the tool Mach actually thinks is loaded in the spindle.  The fact that the offset is closer that way tells me that you have something else going wrong somewhere. 

Steve

286

Mach4 General Discussion / Re: Tool change Problem.

« on: May 12, 2020, 07:45:12 PM »

Art, the "Tool on the M6 line is next tool" is for tool changers that preload the next tool.  Unless you have one of these preload style tool changers, this is NOT what you want. 

Steve

287

Mach4 General Discussion / Re: Mach4 feature vs Mach3

« on: May 12, 2020, 07:41:36 PM »

24. Separate interpreters for the different control types.  Meaning the Mill interpreter is different than the Lathe interpreter.  Mach 3's Lathe interpreter was the same as the Mill!  This led to conflicts with canned cycles, among other things.  There is a G76 cycle in a milling application AND a lathe application and they do not do the same thing. 
25. Multiple planners.  This allows stuff like Retract and Cut recovery to work.  Plus there is a jog planner for every axis and all can be run with their own settings.
26. Out of band axes, aka non coordinated axes.  This just didn't exist in Mach 3.
27. A customizable tool table.  Add your own fields if you need them!!! 
28. A proper reset function.
29. All of the motion is chained together better in Mach 4 vs. Mach 3.  It makes it feel more smooth and graceful, and less clunky.  Even if running in exact stop mode. 
30. Real parameters that can be set with G10L50. 
31. SETVN implementation.
32. M code scripts that can parse letter arguments.
33. Global data in the script environment (big on the mach 3 wish list)!
34. DRO update and modify scripts.  You can do some wicked stuff with these.
35. #var expansion in comments for G code program debugging purposes.  e.g.  "#123 = 64 (#123 = #[123])"  This will print "#123 = 64" in the status history. 
36. Fully Fanuc compatible expression evaluation.  e.g.  "#100 = [#124 - #123] * [#124 - #123]"  Try that one in Mach 3 and watch it bomb. 
37. Set outputs and read inputs directly in G code.

Whew...  I feel a bit like Bubba telling Forest what all you can do with shrimp.   

Steve

288

Mach4 General Discussion / Re: pInst and pcall

« on: May 12, 2020, 07:05:47 PM »

I'm not sure about pinst.  I think that is just a variable name.  But pcall is a LUA "Protected Call", hence the "p" in pcall.  It gives a means to trap and handle errors without bombing out the whole LUA chink that is running.  In other words, it is usually used to handle errors gracefully. 

Steve

289

Mach4 General Discussion / Re: Switching from Mach 3 to 4 - How Do You Change Motor Tuning Time

« on: May 12, 2020, 07:03:12 PM »

It is not wrong.
The time shown to accelerate and decelerate is right there.
The time at constant speed is irrelevant.

You took the words right out of my mouth, Bill. 

The slew speed time is irrelevant and simply meant just to give a visual representation of the trajectory.  I think the time IS expanded if the acceleration and deceleration portions of the trajectory exceed the default time of the graph. 

The time it takes for the acceleration and deceleration is the only thing that will ever change on that graph, time wise.  The max velocity will determine the height range of the graph.  But otherwise, the <slew time> =  <default graph time> - <acceleration time> + <deceleration time>.

What should we do?  Have a constant slew time?  1 second?  2 seconds? And always shrink or expand the total graph time to the the sum of <acceleration time> + <deceleration time> + <slew time constant>?

I think it is just semantics as to what portion of the graph is constant.  And I'll be the first to admit it doesn't work like Mach 3.  But I always say Mach 4 is not Mach 3.  It is Mach 4. 

But anyway, we have a time portion on the graph simply because it is part of the little graph widget that we used.  And you can see your accel and decel times, as they are relevant.

Steve

290

Mach4 General Discussion / Re: G68 Angle and Probing

« on: May 12, 2020, 05:03:39 AM »

Yeah, the one in your Docs folder if your build is newer then 4300.  If not, download a recent dev build from the FTP site and install it to a different directory.  Then look in that directory's Docs folder. 

Also, you could probe the sides of the part in two places each and figure their intersection (corner) with determinants. See https://en.wikipedia.org/wiki/Line%E2%80%93line_intersection#Given_two_points_on_each_line  I believe Renishaw probe routines have this capability built in. 

Steve