FERROR on Z with float switch

More
22 Apr 2011 16:41 #9260 by jcizek
John,
Well, I am stumped again. I talked with Jarl and he says there are no separate configs for velocity during the probing routine. It should be using the .ini settings. I followed the entire torch fire procedure though, looking up each AND, MUX, etc... it eventually winds up right back to the standard axis motion hal commands that should be using the .ini. Nothing weird or special in there once you actually get to the motion part!! I have sent those videos and explanation to Jarl, he said he's really busy right now but will look when he has a minute to see if he notices anything missing.

In the meantime, I think i'll work on converting your comp routine you sent me and see how that goes. More to come! -James

Please Log in or Create an account to join the conversation.

More
22 Apr 2011 17:29 - 22 Apr 2011 17:30 #9261 by BigJohnT
This is the thcud comp I did yesterday, but have not tested beyond compiling, installing and loading it up in Axis.

File Attachment:

File Name: thcud.comp
File Size:4 KB


John
Attachments:
Last edit: 22 Apr 2011 17:30 by BigJohnT.

Please Log in or Create an account to join the conversation.

More
22 Apr 2011 17:42 - 22 Apr 2011 17:47 #9262 by BigJohnT
Other than up and down are reversed it appears to work fine.

It should be like this:
        if(torch_on && arc_ok && vel_status){ // allow correction
            if(torch_down){
                offset -= correction_vel;
            }
            if(torch_up){
                offset += correction_vel;
            }    
            last_z_in = 0;
        }

I can up load the config I used to test it but it used a Mesa 5i20 so you would have to change that part to parallel port if that is what your using.

John
Last edit: 22 Apr 2011 17:47 by BigJohnT.

Please Log in or Create an account to join the conversation.

More
22 Apr 2011 19:29 #9265 by BigJohnT
I found one more error:
                }
            last_z_in = z_pos_in;
            }
        }
        z_pos_out = z_pos_in + offset;

should be
                }
            }
            last_z_in = z_pos_in;
        }
        z_pos_out = z_pos_in + offset;

Please Log in or Create an account to join the conversation.

More
26 Apr 2011 12:40 #9332 by jcizek
Hi John, just wanted to let you know I'm still working on getting your code to work with my config... I haven't given up, just haven't gotten everything going quite right yet! I built a new bare config to start with and insert your code into. Some of it's working but I still have a few problems. Working through it!! I am back at work now so don't have nearly as much time as I did. I'll let you know how it;s going soon!! -James

Please Log in or Create an account to join the conversation.

More
07 May 2011 01:43 #9631 by jcizek
John,

I wanted to let you (and anyone else that might search for this in the future) know that this is solved!! The stepper.hal file referenced the decoupled Z axis back to the max_acceleration section of the thc300.ini file, but there was NO section in that file tying any of the axis's back to the max_velocity setting in the ini. So the Z axis would obey the max velocity any other time besides the probing, but when probing, it was running limitless.
I solved this by adding this line to stepper.hal:

setp stepgen.2.maxvel [AXIS_3]MAX_ACCELERATION

Many thanks to Chris Radek for getting me on the right track to finding what was missing!
and many thanks to you John for all the assistance here.

-James

Please Log in or Create an account to join the conversation.

More
07 May 2011 11:52 #9632 by BigJohnT
James,

I'm glad you finally solved the puzzle.

John

Please Log in or Create an account to join the conversation.

More
29 May 2011 04:03 #10145 by jcizek
Hey John,

So I have been playing with my table some since I solved that velocity problem... and I think I'm ready to ditch this massively complex HAL setup. I have some very strange things happening... I'm having good luck with some gcode, but others are doing random weird things (like firing the torch while the Z is 4" above the metal and just going to town...even though the first 2 pierces worked fine and use identical code, the 3rd one just fails to lower the Z at all before firing the torch. )

So I was looking at the the thcud.comp file you built, and the touchoff.ngc. One question.. how do you call the touchoff file? Also, do you only touchoff once per gcode file? Or do you run that script at every M03?

Thanks for your help! I am bound and determined to make this work... seems like I am always "SO" close to having a working machine but keep running into these little things that keep me from making parts!!

-James

Please Log in or Create an account to join the conversation.

More
29 May 2011 12:28 #10151 by BigJohnT
I call touchoff.ngc at start of each cut, it goes to pierce height starts the arc then goes to cut height.

I call it using:
(#1 pierce height, #2 pierce delay, #3 cut height)
o<touchoff>call [n][n][n]

www.linuxcnc.org/docview/html/gcode_main.html#cha:O-Codes

You will have to adjust this line to suit your switch hysteresis

G1 Z0.070 (Raise up to Z0)

On the mailing list there was a suggestion to connect the arc ok input to the spindle at speed and forget the M66. I'll try that this weekend as it seems a better solution to me.

John

Please Log in or Create an account to join the conversation.

More
29 May 2011 16:50 #10153 by jcizek
OK John, I am heading out to the shop right now to try and compile in the thcud. see how it goes..

one question... Are you editting your gcode files manually to include the O subroutine calls? Or do you have a post that's doing it for you? Or somehow else?

-James

Please Log in or Create an account to join the conversation.

Time to create page: 0.100 seconds
Powered by Kunena Forum