#emc-devel | Logs for 2009-01-02

[04:36:57] <cradek> haha (chris morley's message)
[11:55:15] <micges> jepler: after applying your sane period patch joint velocity are precise close to 0.00001 mm/min, before there was about 0.5 mm/min
[11:56:35] <micges> before: F500, fb_vel = 500,5
[11:57:06] <micges> now F500 fb_vel = 500.0000
[12:10:22] <CIA-1> EMC: 03cmorley 07TRUNK * 10emc2/src/emc/usr_intf/stepconf/ (stepconf.py stepconf.glade): (log message trimmed)
[12:10:22] <CIA-1> EMC: work to improve loading of selected ladder program, copying of ladder program to
[12:10:22] <CIA-1> EMC: user directory. added widgets to select base address and direction type of two
[12:10:22] <CIA-1> EMC: more parallel ports. Stepconf only connects signal names to the pins. for now
[12:10:24] <CIA-1> EMC: it's up to the user to connect to the signals in the custom hal file. added text
[12:10:26] <CIA-1> EMC: to README so user can can lookup examples of these names. Hope nobody is upset
[12:10:28] <CIA-1> EMC: about the parport mods...still need to add dialog to warn about overwiting of
[14:41:37] <BigJohnT> hmmm gs2_vfd still causes a Communications Warning: Illegal data value - The data value received in the command message is not available for the AC drive.
[14:48:22] <BigJohnT> gs2_vfd.spindle-speed and spindle-on do not do anything but spindle-fwd works fine...
[14:49:19] <BigJohnT> but it might only be spindle-on when testing the hz is not displayed until your running
[14:51:01] <BigJohnT> all the items from the drive seem to show up in HAL configuration watch window fine
[15:07:34] <CIA-1> EMC: 03bigjohnt 07TRUNK * 10emc2/src/hal/user_comps/gs2_vfd.c:
[15:07:34] <CIA-1> EMC: fixed parameter and pin having the same name which caused hal configuration show window to show the
[15:07:34] <CIA-1> EMC: value for the parameter instead of the pin
[15:29:11] <BigJohnT> ok, I got spindle-on working but speed-command does not get to the drive... perhaps that is the root of the error
[15:32:35] <BigJohnT> ok got speed-command to work but a value of 30 is 2.0 on the drive...
[15:34:03] <BigJohnT> ok 30 is rpm not hz...
[15:34:27] <skunkworks> heh - sounds like you are having fun :)
[15:35:25] <BigJohnT> yep, about to get it figured out so I can write the manual :)
[15:35:41] <BigJohnT> ok, now the error message is gone
[15:36:06] <skunkworks> again - you are doing great work!! Thanks again.
[15:36:21] <BigJohnT> thanks
[15:36:55] <BigJohnT> weeee it works... now to make a working config :)
[17:33:43] <jepler> what does anything think of adding the ability to specify [HAL]ERRORS=continue in an emc2 inifile? this would arrange to pass the "-k" option to halcmd, and might ease testing
[17:34:48] <SWPadnos> I'm not sure it should be apermanent thing like an ini option
[17:35:01] <SWPadnos> but it sounds like a good idea otherwise (maybe an emc script option)
[17:35:59] <jepler> yeah, that makes a bit more sense
[22:42:26] <jmkasunich> jepler: are you going to apply the period patch?
[22:43:14] <jepler> jmkasunich: no, I still haven't tried to do it "the right way"
[22:43:57] <jmkasunich> ok
[22:44:33] <jmkasunich> I wish "de-legacying" wasn't so hard
[22:51:46] <CIA-1> EMC: 03jepler 07TRUNK * 10emc2/scripts/emc.in:
[22:51:46] <CIA-1> EMC: fix(?) problem reported by Les Newell about realtime not unloading
[22:51:46] <CIA-1> EMC: new flag 'emc -k' to help users incrementally develop their hal files
[23:06:01] <CIA-1> EMC: 03jepler 07TRUNK * 10emc2/src/emc/motion/motion.c: it's not useful anymore to run motion not as a hal realtime component
[23:23:05] <BigJohnT> why is motion.spindle-speed-out a negative number when I issue a S200 M4 from the MDI?
[23:23:44] <BigJohnT> in trunk...
[23:23:59] <SWPadnos> M4 means "go backwards"
[23:24:07] <SWPadnos> so -200 = 200 backwards
[23:24:26] <BigJohnT> I thought spindle-reverse means "go backwards"
[23:24:48] <SWPadnos> yes, that's true too
[23:24:58] <BigJohnT> but you can't have a negative speed...
[23:25:09] <BigJohnT> only a speed...
[23:25:40] <SWPadnos> well, that's true too ;)
[23:26:20] <BigJohnT> btw, the gs2 is running now except for the speed issue
[23:27:00] <SWPadnos> cool
[23:27:12] <SWPadnos> I saw the pin name change, thanks
[23:27:16] <SWPadnos> or param ;)
[23:27:22] <BigJohnT> :)
[23:27:55] <BigJohnT> so is it a mistake for the speed to be negative for M4 or does some other thing use it that way?
[23:28:19] <SWPadnos> I don't think it's a mistake, but it still could be dumb
[23:28:35] <SWPadnos> I don't recall what eventually fixed your gs2 problem - what was it?
[23:29:36] <BigJohnT> ok, you must configure the gs2 to use the commands from the modbus or you get an error if you try and take control from the machine :)
[23:29:47] <SWPadnos> oh, heh
[23:30:19] <SWPadnos> now I remember having to change modes at the keypad whenever there was an error I had to reset
[23:30:25] <SWPadnos> and then switch back for more testing
[23:30:48] <BigJohnT> now if I load the gs2 driver in the custom.hal I have to do any net pin things in custom postgui.hal or I get an error that the pin does not exist
[23:31:33] <SWPadnos> are you loading with loadusr, or loadusr -Wn?
[23:31:46] <BigJohnT> loadusr
[23:31:55] <SWPadnos> one of the wait options should make it possible to load and connect in the same hal file
[23:32:17] <SWPadnos> oh, have you been using the command-line options for the serial port parameters?
[23:32:33] <BigJohnT> no I just used the default
[23:32:36] <SWPadnos> ok
[23:32:49] <SWPadnos> I'm pretty sure those are working, but it might be good to check
[23:32:56] <SWPadnos> you can also set a different name with -n
[23:32:58] <BigJohnT> is it -Wn or -W and a number?
[23:33:09] <SWPadnos> you can also set a different name with -n, like "spindle"
[23:33:15] <SWPadnos> the letter n
[23:33:23] <SWPadnos> wait for this component name to become ready
[23:33:56] <SWPadnos> so something like loadusr -Wn spindle gs2_vfd -n spindle -d /dev/ttyS1
[23:34:09] <SWPadnos> will give you spindle.motor-RPM ... names
[23:34:15] <SWPadnos> rather than gss2_vfd
[23:34:17] <SWPadnos> -s
[23:34:32] <SWPadnos> you can also leave out the n and -n options: loadusr -W gs2_vfd
[23:36:08] <BigJohnT> ok loadusr -W gs2_vfd works
[23:36:48] <SWPadnos> yep. if you decide to change the name, or add more VFDs or anything, you'll need the -Wn gs2_vfd -n name form
[23:37:28] <BigJohnT> the name form looks like a good idea to keep things clear
[23:37:42] <BigJohnT> is the -Wn for a name form?
[23:37:58] <SWPadnos> yeah, I figured there's a possibility of having a machine with a lathe and mill spindle, so you'd need different names
[23:38:09] <SWPadnos> -Wn means "wait for a component with the following name"
[23:38:17] <SWPadnos> otherwise the component name is the module name
[23:38:28] <SWPadnos> (gs2_vfd in this case)
[23:39:01] <SWPadnos> but when you have things like pyvcp which can have different names depending on what you're doing, you need to tell halcmd that it should wait for something with a different name
[23:39:53] <SWPadnos> so it's loadusr -Wn name (or loadusr, wait for this named component), then gs2_vfd -n name, to make the VFD have that same name
[23:40:35] <BigJohnT> I gotta get this info in the manual... I just did a quick search and came up empy on "-Wn"
[23:40:45] <SWPadnos> hmmm
[23:40:48] <SWPadnos> man halcmd has it
[23:40:57] <SWPadnos> as does halcmd -h loadusr
[23:41:00] <SWPadnos> (I think)
[23:41:07] <BigJohnT> I tried man loadusr :(
[23:41:13] <BigJohnT> no entries
[23:41:14] <SWPadnos> heh
[23:41:16] <SWPadnos> nope
[23:43:49] <BigJohnT> * BigJohnT finally gets logged in to sourceforge
[23:45:00] <BigJohnT> SWPadnos: that driver is sweet!
[23:45:08] <SWPadnos> cool :)
[23:45:26] <SWPadnos> hmmm. did I default the cycle time to 0.1 second?
[23:45:35] <BigJohnT> I think so
[23:46:12] <SWPadnos> ah, yes
[23:51:56] <BigJohnT> * BigJohnT heads upstairs to cook Crawfish Cardinal