Plasma component in Gmoccapy

More
15 Jan 2017 03:39 #85805 by rodw
rodw created the topic: Plasma component in Gmoccapy
Guys,

I'm using a mesa 7i76e and a THCAD-10 card
I just moved my config from axis GUI to gmoccapy which was surprisingly painless :)

Well that is until I tried to connect my THC pins to gmoccapy
The plasma sim has this line
net curvolt     plasma.Target-Voltage   => plasma.hal-lbl-cur-volt

so I assumed all I needed to do was change it to
net curvolt  thc.volts => plasma.hal_lbl_cur_volt

The system then complains that plasma.hal_lbl_cur_volt does not exist. The plasma component is loaded when i Look in halshow so I'm not really sure what I've done wrong. Any help would be appreciated.
More
15 Jan 2017 12:43 #85813 by newbynobi
newbynobi replied the topic: Plasma component in Gmoccapy
Hallo rodw,

I am pretty sure, that you do load the mentioned line in a normal hal file.
As the plasma stuff are just different embedded tabs, you have to put that line in hal file, that is loaded with the embedded tab.

The tab is loaded as follows:
EMBED_TAB_NAME = coolant_spindle_panel
EMBED_TAB_LOCATION = box_coolant_and_spindle
EMBED_TAB_COMMAND = gladevcp -x {XID} -u plasma.py -H plasma.hal plasma.glade

In your case the mentioned line need to be in the plasma.hal file.

The reason is, that the embedded tabs are loaded after the GUI is displayed, so the hal pin are not available at earlier time, that after loading the embedded tab.

Norbert
More
15 Jan 2017 17:58 #85832 by islander261
islander261 replied the topic: Plasma component in Gmoccapy
rodw

You have used different names for the pins. Underscores are not the same as dashes. I do this all the time because I type so poorly.

John
More
15 Jan 2017 18:01 #85833 by islander261
islander261 replied the topic: Plasma component in Gmoccapy
Sorry

I will post my working HAL lines for the THCad when I get to my linux machine.

John
More
15 Jan 2017 18:38 #85834 by islander261
islander261 replied the topic: Plasma component in Gmoccapy
rodw

Here are excepts from my HAL files. The arc voltage display and settings are working as far as I can tell from connecting an external voltage source to the THCad input.
From main HAL file:
#*******************
#  THC stuff
#*******************

# save for future THC with PID
setp   pid.thc.Pgain     10     
setp   pid.thc.Igain     1    
setp   pid.thc.bias      0     
setp   pid.thc.Dgain     0     
setp   pid.thc.FF0       0     
setp   pid.thc.FF1       0
setp   pid.thc.FF2       0
setp   pid.thc.deadband  1
setp   pid.thc.maxoutput 100 
#setp   pid.thc.error-previous-target true

#net thc-index-enable  <=> pid.thc.index-enable
#net thc-enable        =>  pid.thc.enable
#net thc-vel-cmd-rpm     => pid.thc.command
#net thc-vel-fb-rpm      => pid.thc.feedback
net scaled_arcvolts    scale.arcvolts.out => pid.thc.feedback
#net thc-output        <=  pid.thc.output

# ---Encoder Setup for THC--

setp    hm2_7i76e.0.encoder.00.counter-mode 1
setp    hm2_7i76e.0.encoder.00.filter 0
setp	hm2_7i76e.0.encoder.00.scale -1
net thc-vel-in	hm2_7i76e.0.encoder.00.velocity => thc.encoder-vel 
# Value for my THCad card
setp thc.scale-offset 124400
#Value for my THCad card and voltage divider (18.94181:1)
setp thc.vel-scale    .0002238191

#this may need changing
setp thc.correction-vel .0001


#this starts arc
net Torch_On      and2.torch-enable.out
   
#starts motion on arc transfer to work this will need to change for my setup
net arcok     thc.arc-ok motion.digital-in-00 

#*********************************
# experimental use of components for testing future THC enhancements
#********************************

#these settings are scaling encoder counts to arcvolts using scale component
#Value for my THCad card and voltage divider (18.94181:1)
setp     scale.arcvolts.gain .0002238191
# Value for my THCad card 124400 * .0002238191*-1
setp     scale.arcvolts.offset -27.843

net thc-vel-in     scale.arcvolts.in

#these settings are to get the arc voltage rate of change

net scaled_arcvolts     ddt.arcvolts_rate.in

This is from one of my auxilary HAL files:
#link arc voltage to both tabs
net curvolt_tab     thc.volts => plasma_tab.hal-lbl-cur-volt-tab 
net scaled_arcvolts     plasma.hal-lbl-cur-volt

#link target voltage values between tabs
net target_volts_tab	plasma_tab.Target-Voltage => plasma.hal_lbl_target_voltage_main

I have two sources of arc volts shown because I was testing the scale component for for scaling the TCHad output. Note that I have two places on my screen that display the arc volts. One is on a tab in the main screen and one is on a tab with other tabs in the preview location.

John
Attachments:
More
15 Jan 2017 20:22 #85847 by rodw
rodw replied the topic: Plasma component in Gmoccapy

newbynobi wrote: Hallo rodw,

I am pretty sure, that you do load the mentioned line in a normal hal file.
As the plasma stuff are just different embedded tabs, you have to put that line in hal file, that is loaded with the embedded tab.

The tab is loaded as follows:

EMBED_TAB_NAME = coolant_spindle_panel
EMBED_TAB_LOCATION = box_coolant_and_spindle
EMBED_TAB_COMMAND = gladevcp -x {XID} -u plasma.py -H plasma.hal plasma.glade


Norbert, thanks, I think this is my problem. I copied my .hal and .ini files over from another config into a fresh plasma sim config and went to town on them. My Ini file is still pointing at plasma.hal, not the .ini file I am editing. Thanks a lot!

islander261 wrote: rodw
You have used different names for the pins. Underscores are not the same as dashes. I do this all the time because I type so poorly.
John

John, the typo is just confined to the forum post above. I had it right in my hal file. Thanks for posting your config. I will review it. We are both on the same journey just mine is slower as I have to build everything, not just retrofit an existing table...
I've got all the THC signals tested and working except for the interface to the GUI which I'm just starting on. Once I get that done, I want to try a few things in the THC component.
More
17 Jan 2017 11:24 #85901 by rodw
rodw replied the topic: Plasma component in Gmoccapy

rodw wrote: Guys,

so I assumed all I needed to do was change it to

net curvolt  thc.volts => plasma.hal_lbl_cur_volt


Well, I'm glad we are all back online :)

I'm still having trouble with this. I've got all the THC component working in my main hal file but this line breaks the plasma display section in gmoccapy.

Can plasma.hal access components in the main hal file?

I've had a couple of distractions but my next step was to move all of the THC stuff into plasma.hal. Is that necessary?
More
17 Jan 2017 16:37 #85949 by islander261
islander261 replied the topic: Plasma component in Gmoccapy
rodw

The short answer is yes you can spread the hal instructions around, just be careful with naming. No you do not have to consolidate all of the THC related stuff in one hal file. All of my THC setup and main connections are in my main HAL file and then the voltage display connections are in another as you are doing. The easiest way I found to sort these problems out (and believe me I have had plenty) is to start Linuxcnc from a terminal and look at the error messages. Post your HAL and .ini files and we'll have a look. It really took me a while to get the hang of multiple tabs, .py files and hal files. Mine still doesn't work perfect but is close, a recurring problem on reloading persistent variables that I don't know how to trouble shoot..

John
The following user(s) said Thank You: rodw
More
19 Jan 2017 11:05 #86139 by rodw
rodw replied the topic: Plasma component in Gmoccapy - BUG REPORT
Well, I've resolved this error. Its a bug in Gmocappy plasma sim and Islander spotted it earlier, so thanks John! Also thanks to the tip to run from the command line

This line was generated when I created a brand new sim from the LInuxcnc chooser.
# plasma.hal_lbl_cur_volt          (float)

A gmocappy plasma sim generates this plasma.hal file that uses underscores (_) where it should be using hyphens. (-)
############### plasma hal pins #################### 

# out pin
# plasma.CHL-Threshold             (float)
# plasma.enable-HeightLock         (bool)
# plasma.enable-HeightLock-not     (bool)
# plasma.hal-btn-torch-not         (bool)
# plasma.hal-btn-torch             (bool)
# plasma.Target-Voltage            (float)
# plasma.THC-Speed                 (float)
# plasma.Piercing-autostart-not    (bool)
# plasma.Piercing-autostart        (bool)
# plasma.Pierce-Gap                (float)
# plasma.Pierce-Delay              (float)
# plasma.G0-Gap                    (float)
# plasma.Cut-Gap                   (float)

# in pin
# plasma.led-pos-limit             (bool)
# plasma.led-neg-limit             (bool)
# plasma.hal-lbl-corr              (float)
# plasma.hal_lbl_cur_volt          (float)


# Hook up plasma pins using the net command

# for simulation mode we connect them directly
net torch       plasma.hal-btn-torch    => signals.led-torch-on => signals.led-arc-ok
net curvolt     plasma.Target-Voltage   => plasma.hal-lbl-cur-volt

Now if I had chosen any other signal, the bug would not have bitten.

So to wrap this up, can somebody confirm that building my config in plasma.hal is best practice?
More
19 Jan 2017 21:04 #86198 by rodw
rodw replied the topic: Plasma component in Gmoccapy
Hmm, there is also a problem in signals.hal. Pin names do not agree with what is used in signals.glade so the tab does not load

I've only had time to attach one signal. Once complete I will post all errors as a bug report.
Moderators: newbynobi
Time to create page: 0.125 seconds
Powered by Kunena Forum