#emc-devel | Logs for 2010-07-14

Back
[13:26:44] <skunkworks> http://cnczone.com/forums/showthread.php?t=108507
[13:38:32] <roberth_> skunkworks, i noticed this on M61 Qx it does not display the tool is loaded, EMC status shows the tool number loaded but axis display is blank, might be related i did not get time to investigate it more, was on 2.4.2
[13:48:40] <cradek> skunkworks: did you reproduce that?
[13:50:48] <jepler> in sim/axis if I issue m61q1 after power on and home, it still says "no tool"
[13:51:03] <jepler> if the intent is for emc to think in all ways that tool 1 is loaded, that seems wrong..
[13:51:41] <cradek> what's the actual use case for m61q? what problem does it solve?
[13:51:43] <jepler> I didn't reproduce the cnczone problem (reload tool table, then display of tool doesn't work)
[13:52:57] <skunkworks> I have not played with it. I don't have 2.4.2 installed on anything yet
[13:53:10] <jepler> alex added m61 in Jan 2009
[13:53:28] <jepler> * for now it's not allowed to call M61 from a program, only from MDI (or while in manual mode)
[13:53:32] <jepler> that's a bizarre restriction
[13:54:03] <jepler> aabe136 add M61 Qxx
[13:58:22] <cradek> "allows changing the current tool number"
[13:58:44] <cradek> does this mean it's supposed to "rename" the current tool?
[13:59:41] <jepler> emcioStatus.tool.toolInSpindle = number;
[13:59:41] <jepler> *(iocontrol_data->tool_number) = emcioStatus.tool.toolInSpindle; //likewise in HAL
[14:00:00] <jepler> I think it's supposed to mean "contrary to whatever you thought, this tool number is the one actually loaded"
[14:01:01] <cradek> I wish I knew what problem that's meant to solve
[14:01:16] <jepler> me too
[14:01:20] <cradek> I wonder what a random-tc machine should do with it, for instance
[14:01:22] <jepler> if alex_joni drops by, maybe he can remember
[14:01:33] <jepler> yeah
[14:01:50] <cradek> but on a manual-tc machine, why not just issue TnM6
[14:02:17] <cradek> the only difference would be whether it moves to the tc position
[14:02:18] <skunkworks> is it just so you can tell emc what tool is in the spindle?
[14:02:51] <cradek> seems like you could even tell it an invalid tool (one not in the tool table) and it'd be none the wiser
[14:03:06] <cradek> so maybe it's meant to let you add tools to the tool table, I dunno
[14:04:14] <jepler> before random tc, we had the problem of emc not knowing what tool was in the spindle at power-up or after an aborted tool change.
[14:04:15] <cradek> skunkworks: yes I think so - but on what kind of machine is that useful? and what is that supposed to do to the tool table?
[14:04:24] <jepler> was this to solve that problem?
[14:04:39] <jepler> roberth_: what was your task or what problem were you trying to solve when you decided to use M61?
[14:12:07] <cradek> actually I think it could be useful on random. if I load a tool (say T1) from a pocket (say P10) and then pull it out of the spindle and insert T99 instead, currently I have to just edit the tool table (first line, change T1 P0 to T99 P0). Next time I load another tool (say from P3) it will put the new T99 in P3 and T1 will be gone.
[14:13:16] <jepler> so if T1 is in the spindle and you issue M61 Q99, what pocket is T1 in afterwords?
[14:13:34] <cradek> in no pocket
[14:13:35] <jepler> no pocket (i.e., its tool table entry is deleted?)
[14:13:38] <cradek> yes
[14:14:02] <jepler> too bad you lose its offsets
[14:14:14] <cradek> yeah.
[14:15:06] <cradek> it's almost like you need two tables, one with all your tools, one representing what's in the carousel
[14:15:27] <jepler> and then you have to set all the offets for T99 anyway
[14:15:40] <cradek> or, support P[none, you can't load it] in the current tool table
[14:26:03] <skunkworks> what happens when you bring the machine up with the last tool in the spindle? does emc remember that? (have not played with it)
[14:26:17] <cradek> with random-tc, yes
[14:26:42] <cradek> if you have a 16 pocket carousel you can have 17 tools in the machine and it keeps them straight
[14:28:07] <skunkworks> neat
[14:28:49] <skunkworks> did you figure out your tach issue?
[14:29:07] <cradek> after I hooked up the test equipment, it worked perfectly
[14:29:11] <cradek> so it's healed again (for a while)
[14:29:12] <skunkworks> yeck
[14:29:21] <skunkworks> mill or lathe?
[14:29:26] <cradek> mill
[14:32:37] <jepler> in 2.3, M61 Q1 does change axis's idea of the current tool
[14:32:46] <jepler> so I bet random tc changed/broke m61's behavior
[14:33:05] <cradek> * cradek cries quietly
[20:19:59] <robh__> jepler, it was at the weekend when i was setting a little boxford up for someone with a turret on, and they have no way inside it to tell what tool you are on, so i went to do a M61 Q to set where i was in EMC untill he fits a true encoder to the turret
[20:20:42] <robh__> it seems M61 Q changes the status of current tool # etc, but does not update axis display
[20:22:08] <cradek> robh__: how does it change tools?
[20:22:34] <robh__> how do you mean? sorry
[20:22:58] <robh__> how do you know where you are or such on the turret
[20:23:21] <cradek> does it have automatic tool change controlled by emc?
[20:23:34] <robh__> there is a switch inside, that index each station, so they must count from known station untill you it the called for one
[20:23:40] <robh__> yes
[20:23:59] <cradek> ok I see
[20:24:05] <cradek> it has relative positioning with no provision for homing
[20:24:32] <robh__> no home or tool#1 positon no
[20:24:40] <cradek> any plan for adding a homing sensor?
[20:24:41] <robh__> and you cant get anything inside we looked
[20:25:00] <robh__> only thing todo is take back plate off, bore a hole into there shaft and fix a encoder
[20:25:01] <cradek> sounds frustrating :-P
[20:25:35] <robh__> but you can not take the shaft out easy, so need to make jig to hold it all on mill etc, yes very frustrating
[20:25:50] <robh__> but then its only little stepper lathe ;)
[20:26:25] <cradek> you could always just remember to change to tool 1 before shutting it down... (yuck)
[20:28:33] <robh__> that could be intresting concept that
[20:28:57] <robh__> right now i am bussy retrofitting my sencond superslant again twin turret
[20:29:12] <cradek> neat
[20:29:55] <robh__> i will try and use W on 2nd Z on this as i have 2.4.2 installed
[20:30:25] <robh__> will save using G10 all the time
[20:30:46] <cradek> ah definitely W
[20:30:52] <cradek> 2.4 is finally great for that kind of thing
[20:31:13] <cradek> did you know you can even do a drill cycle in W with G17.1
[20:31:45] <robh__> yea i rember seeing you mention it awhile back, cant wait to try it, also try rigid tapping on it
[20:32:09] <cradek> heh, I bet tapping in W doesn't work
[20:32:45] <cradek> we did not have tapping yet when I added UVW
[20:32:47] <robh__> o well looks like ill still be using the old tapping unit
[20:33:08] <cradek> could be added if someone was interested - just isn't there now
[20:33:22] <cradek> not like there's any fundamental problem in getting it to work
[20:33:39] <robh__> 99% time its big threads so screw cutting it quickest way so its not something i badly need
[20:35:12] <robh__> something that would be nice for someone to work on it to draw UW as seperate displays on the back plot due to how the two turrets work on a twin turret lathe
[20:35:20] <cradek> pick your posion - single point makes nice concentric threads at the wrong pitch diameter, taps make crooked and badly centered threads at the perfect pitch diameter :-)
[20:35:29] <cradek> poison
[20:37:00] <cradek> it's hard to know how to display uvw, since the can mean different things on different machines
[20:37:03] <cradek> they
[20:37:05] <cradek> argh, can't type
[20:39:20] <robh__> i hear you, i think for lathe you can somewhat say XYZ will also be top turret, UVW always be bottom if there, C is always main spindle i think A or B is sub spindle (cant rember of top head dont have twin spindle lathe)
[20:41:10] <robh__> i have to pop off now chat soon