#linuxcnc-devel | Logs for 2016-02-11

[07:07:49] <skunkworks> heh - I am glad steve blackmore is an equal opportunity grump. Every time the lathe subject comes up - he asks if threading is working yet.
[07:08:00] <skunkworks> (mach4)
[07:08:06] <archivist> hehe
[07:09:19] <skunkworks> And this seems to always be the answer.
[07:09:20] <skunkworks> https://groups.yahoo.com/neo/groups/mach1mach2cnc/conversations/messages/150025
[07:09:35] <skunkworks> (I think he makes pokeys)
[07:11:46] <archivist> it seems the problems with some of the external solutions are deep seated from talking to JS the other week
[07:13:01] <skunkworks> how is the hobber doing?
[07:13:39] <skunkworks> he made a comment that he was going to post something about it on homeshopmachinist when it was up again.
[07:14:25] <skunkworks> Looks like there isn't a spindle encoder of any kind on the matsuura
[07:15:18] <Tom_itx> inductive feedback on the armature?
[07:15:27] <archivist> he rang me afterwards after he tidied up and made a batch of gears, seems a happy bunny
[07:15:38] <skunkworks> archivist, great!
[07:16:06] <skunkworks> Tom_itx, vfd - probably doesn't need any feeback
[07:16:10] <skunkworks> feedback
[07:16:14] <archivist> I did have some fun electrical noise problems setting it up :)
[07:16:22] <skunkworks> unless there is something hiding in there
[07:17:22] <skunkworks> I think cradek had a multi pole resolver on the spindle motor - but that was a jr with the same control
[07:29:07] <skunkworks> archivist, encoder noise? how did you fix it?
[07:29:55] <archivist> skunkworks, there were a few sources, so improved grounding/screening and a power supply replacement
[07:30:12] <skunkworks> neat
[07:30:27] <archivist> just have to know how to use a scope and diagnose faults
[07:30:36] <skunkworks> :)
[07:30:46] <archivist> almost impossible to describe in emails or over irc
[07:31:23] <archivist> and I think I left one thing a bit close to the bone
[07:31:33] <skunkworks> mach4 was threading 1.5 years ago.. https://www.youtube.com/watch?v=YvQeXDsVZ0A
[07:32:25] <archivist> skunkworks, those breakouts we have been using ar a bit iffy is you want to use the input optos at 5v
[07:32:58] <skunkworks> not enough voltage?
[07:33:14] <archivist> designed for 12v and switch contacts rather than encoder input
[07:33:21] <skunkworks> ah
[07:34:02] <archivist> was cheating and using the pc usb 5v on the input optos :)
[07:38:03] <skunkworks> I don't think I have used any inputs yet
[07:39:01] <archivist> best to use an open collector buffer (74series) and 12v to the opto
[07:47:22] <skunkworks> can you get to the other side of the opto's easy?
[07:48:02] <archivist> I wanted to keep isolation
[07:48:12] <skunkworks> ah
[07:50:28] <jthornton> does a 6i25 work in any PCI Express slot? 2.0 x 16, 3.0 x 16 etc
[07:54:16] <skunkworks> I don't think it says in the manual
[07:56:17] <malcom2073> It says it's a one lane card
[07:56:51] <malcom2073> So 1x or greater I would assume
[07:57:06] <malcom2073> Or 2.0 or 3.0, misunderstood sorry
[07:57:46] <jthornton> thanks that's why I was asking I didn't see it in the manual
[07:58:16] <malcom2073> I would assume 2.0, but yeah you should verify with PCW heh
[08:36:21] <pcw_home> The 6I25 should work in x16 slots ( the PCIE spec requires this )
[08:36:22] <pcw_home> but it may not work on older MBs due to BIOS issues
[08:40:58] <skunkworks> Sure... Blame the bios.. ;)
[08:50:59] <pcw_home> We've seen so pretty weird stuff with BIO's, though they tend to get fixed over time
[08:52:39] <JT-Shop> thanks
[08:52:44] <pcw_home> I especially like my Jetway Atom MB that goes crazy every 256 power cycles (needs PS shutdown to fix)
[08:53:46] <skunkworks> yikes
[08:54:51] <pcw_home> maybe thats not BIOS but whatever controls the soft power switching
[08:55:56] <pcw_home> the X16 slots not working with narrower cards is mostly old motherboards
[11:18:19] <jepler> LIGO! (I know, I'm hours late)
[13:23:02] <jthornton> when you flash a 6i25 do you just use --device 5i25?
[13:27:31] <jthornton> ok found it in the 6i25 manual
[13:29:59] <skunkworks> wow - sounds familiar... https://groups.yahoo.com/neo/groups/mach1mach2cnc/conversations/messages/150047
[13:41:35] <seb_kuzminsky> jepler: +1
[15:56:04] <PCW> jthornton: the 6i25s almost always just use 5i25 firmware
[15:57:34] <JT-Shop> thanks PCW
[16:22:12] <skunkworks> so - I think we are going to use a 72 tooth gear for rigid tapping on the mattura
[16:43:39] <andypugh> skunkworks: I think a tap would work better.
[16:43:58] <JT-Shop> less exciting for sure
[16:59:04] <KGB-linuxcnc> 03Dewey Garrett 05joints_axes11 7d41ca5 06linuxcnc 10docs/src/getting-started/updating-linuxcnc.txt 10lib/python/rs274/glcanon.py 10src/emc/usr_intf/axis/scripts/axis.py glcanon.py, axis.py work icons for DRO JA#39 * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=7d41ca5
[16:59:04] <KGB-linuxcnc> 03Dewey Garrett 05joints_axes11 e03bb06 06linuxcnc 10share/axis/tcl/axis.tcl axis.tcl set_mode_from_tab no ensure_manual JA#40 * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=e03bb06
[17:02:46] <skunkworks> heh
[17:04:10] <skunkworks> it fits in the spindle housing and we have some gear sensors
[17:04:48] <skunkworks> shoule work up to 20k rpm with 60 teeth
[17:06:26] <skunkworks> http://www.digikey.com/product-search/en?keywords=620-1328-1-nd
[17:07:35] <seb_kuzminsky> dewey ftw
[18:50:26] <jepler> I wonder if he considers those to close the two github issues
[18:50:57] <jepler> there is a syntax for commit messages that is supposed to close the issue, should work for him to write it
[18:52:24] <andypugh> Is Dewey on a one-man crusade to finally get JA mainlined?
[18:52:41] <andypugh> If so, I applaud the effort wholeheartedly.
[18:53:19] <andypugh> I just hope it isn’t interfering with him make beautiful wooden things
[19:23:43] <skunkworks> those sensors say they will work up to 20krpm with a 60 tooth gear
[19:25:17] <skunkworks> sorry 12krpm. that might be cutting it close. (10k spindle)
[19:26:24] <skunkworks> (not that you would be tapping at that speed) but it would be nice for general feedback
[19:30:17] <andypugh> index?
[19:31:03] <skunkworks> figure a notch in the gear for the index sensor
[19:31:19] <skunkworks> somewhere
[19:34:55] <andypugh> That ought to work
[19:35:28] <andypugh> We should work on missing-tooth index. It’s mechanically so simple.
[19:36:27] <skunkworks> where it some sort of masking with a/b?
[19:37:58] <andypugh> You just have one tooth a bit shorter. If a gap is exactly twice as long as expected, you count double and set index.
[19:38:51] <andypugh> In automotive use they use encoders that give speed, direction, and index on one wire. (wires cost money. On a million cars they cost a lot of money)
[19:39:33] <andypugh> It’s a square wave where the duty cycle encodes direction, and a missing pulse indicates index.
[19:40:19] <skunkworks> http://imgur.com/a/UvDnl
[19:40:44] <skunkworks> neat