Setting up a plasma cutter and looking for some help.

More
07 Apr 2020 01:42 #162997 by phillc54

As for paying attention to that part it has been one of the bigger headaches of this ordeal.

I am glad you did, no one (including me) has noticed it before.


Ended up using a not instead cause I was not understanding your code at the time although I am fairly sure I know where I messed up and could make it work now

Rather than using a not component you could have used:
net plasmac:breakaway hm2_7i96.0.gpio.008.in.not => debounce.0.1.in
But, in the end they achieve the same result.

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

More
07 Apr 2020 15:06 - 07 Apr 2020 15:50 #163070 by Birdman3131
EDIT: Its not very clear that it is units per second not minute. Seems to be working now. Now to figure out how to cut something.

So one rewire later I have the ohmic probe working mostly. (Pull low for the relays it seems.)

However now I have an issue with Z. I had set its max accel and velocity when setting it up. Its a stepper so needs more care. However it seems to ignore the values. the jog value in plasmac is at 117 in/min and it will gladly overdrive it. (60 is closer to right.) The ini file has it was lower even but it does not seem to be effecting it.
[AXIS_Z]
# set to double the value in the corresponding joint
MAX_VELOCITY = 3.33333333334
# set to double the value in the corresponding joint
MAX_ACCELERATION = 30.0
# shares the above two equally between the joint and the offset
OFFSET_AV_RATIO = 0.5
MIN_LIMIT = -2.8
MAX_LIMIT = 0.1

[JOINT_2]
TYPE = LINEAR
HOME = 0.0
FERROR = 0.5
MIN_FERROR = 0.05
MAX_VELOCITY = 1.66666666667
MAX_ACCELERATION = 15.0
# The values below should be 25% larger than MAX_VELOCITY and MAX_ACCELERATION
# If using BACKLASH compensation STEPGEN_MAXACCEL should be 100% larger.
STEPGEN_MAXVEL = 2.08
STEPGEN_MAXACCEL = 18.75
P = 980.300854332
I = 0.0
D = 0.0
FF0 = 0.0
FF1 = 1.0
FF2 = 0.0
BIAS = 0.0
DEADBAND = 0.0
MAX_OUTPUT = 0.0
# these are in nanoseconds
DIRSETUP   = 5000
DIRHOLD    = 5000
STEPLEN    = 5000
STEPSPACE  = 5000
STEP_SCALE = -2540.0
MIN_LIMIT = -2.8
MAX_LIMIT = 0.1
HOME_OFFSET = 0.000000
HOME_SEARCH_VEL = 0.166667
HOME_LATCH_VEL = -0.016667
HOME_FINAL_VEL = 0.000000
HOME_USE_INDEX = NO
HOME_SEQUENCE = 0
#******************************************
Last edit: 07 Apr 2020 15:50 by Birdman3131.

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

More
07 Apr 2020 16:42 #163078 by snowgoer540
would need to see the full ini, but check your overall max velocity. It may be capped below your max for the individual axis’

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

More
08 Apr 2020 16:39 #163213 by Birdman3131
Well I finally cut a thing. Got 2 issues.

1. Firing up the plasma trips the limit switches. I ran it with them unplugged and it worked. I can likely fix this but its still odd as they are normally open and should not be that susceptible to noise. But I will try a couple things and see what I can do.

2. The bigger issue was some jerking while cutting. This occurs in dry run as well so i likely have to tweak my motor settings. You can see it on the wrench here. Not sure yet what it is. Im gonna look at backlash and mess with the motor accel settings.

the edge of the wrench

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

More
08 Apr 2020 17:24 #163217 by tommylight
Tripping the switches is a sure sine you have problems with grounding/shielding.
On the phone so i can not do much, but do a search for proper grounding, there is some very useful info somewhere on linuxcnc pages.

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

More
08 Apr 2020 17:47 - 08 Apr 2020 17:48 #163221 by Birdman3131
I am rewiring them to NC rather than NO to see if it helps. (Assuming noise on a nc won't trigger it whereas I can see noise on a NO triggering the input if it induces a current along the wire.)

It does not help that they are unshielded wire. (Everything else is shielded. The wires to the limits are the only ones not and I reused the old wires.) If that does not work ill go through the grounds a bit more. If nothing else I have the shielded wire to re run them. Just would rather not.

I have seen the grounding doc you mention. I should be able to find it again if needed.
Last edit: 08 Apr 2020 17:48 by Birdman3131.

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

More
09 Apr 2020 00:50 #163261 by tommylight
Just be aware that you are risking burning something to a crisp, the amount of induced field when firing the torch is staggering.

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

More
09 Apr 2020 18:32 #163378 by Todd Zuercher
If you feel you've done all that you can to mitigate the noise problem on the limit switches. You could add a debounce to the limit inputs in your hal file. It won't get rid of the noise or induced voltages on the input circuits (or the potential damage they could cause) but it could successfully filter the signal within Linuxcnc.

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

More
09 Apr 2020 21:19 #163393 by rodw
What voltage are you using through your limit switches? There is a big difference with noise immunity with 24 volt v's 5 volt.

I would go back to Clearpath about the drive tuning. Clearly something is not right if you are using shielded cable

Also make sure the shield on any cables is only connected to GND at one end.

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

More
10 Apr 2020 09:39 #163465 by Birdman3131
So been a fun couple days. I keep breaking things. Have not fired the plasma side of things up again.

Zhome/breakaway switch died. Not sure if it got zapped from the plasma or when I knocked a pair of 24v wires loose. it has been replaced by a normal mechanical switch.

I believe the limits were on the 12v rail at the time. I could switch them to 24 but have not atm. I have switched them to NC though. (Also I forgot to mention but i got rid of the paraport board a while back as I ended up having sufficient inputs on the mesa board. Think I have 1 input free now.)


I don't believe it is an issue with the clearpath motors. Fairly sure it is a mechanical issue involving the rack and pinion gear system. I had noticed mild quirks moving it by hand but had ignored them as it is easier to be willing fix those if we know the machine works. The actual mechanicals are from the early 80's I believe. If I get a chance monday ill get a better video of this to show.

I will probably have the oscilloscope recording next time I fire it so as to verify any spikes or lack thereof. Its rated for 400V P-P on the high voltage inputs so should be safe for it unless the voltages go way above what the max torch voltage is.

Most of my time was either on other projects or swapping the axises on it. I had set it up the way it was labeled but according to my coworkers it was a stupid setup and required flipping the sheet in autocad. And if that is gonna be done now is the time to do it. Had it mostly figured out and then managed to find a specific route that broke one of the jack chains that was supporting the plasma lines from the ceiling. Everything is ok but I have to re hang the lines to they don't drag.

So on the grounding side of things I do need to go back over all that but have not as too much other stuff has sidetracked me.

We are off friday so ill get back to it monday

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

Moderators: snowgoer540
Time to create page: 0.128 seconds
Powered by Kunena Forum