Back
[00:56:48] <KGB-linuxcnc> 03Sebastian Kuzminsky 052.7 d867473 06linuxcnc 10src/Makefile pncconf: install the new python modules * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=d867473
[01:27:05] <KGB-linuxcnc> 03Sebastian Kuzminsky 052.7 f7c7957 06linuxcnc 10src/emc/usr_intf/gscreen/gscreen.py 10src/emc/usr_intf/pncconf/pncconf.py Merge remote-tracking branch 'origin/2.6' into 2.7 * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=f7c7957
[01:27:36] <seb_kuzminsky> holy crap, jepler, that "git log --oneline @{u}.." command is super handy, i aliased it to "git pushwhat"
[01:27:39] <seb_kuzminsky> thanks!
[02:39:40] <KGB-linuxcnc> 03Chris Morley 052.7 ef0b29b 06linuxcnc 10src/emc/usr_intf/pncconf/private_data.py pncconf -fix icon/image path error * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=ef0b29b
[03:18:37] <KGB-linuxcnc> 03Chris Morley 052.7 6f35231 06linuxcnc 10src/emc/usr_intf/pncconf/tests.py 10src/emc/usr_intf/stepconf/stepconf.py stepconf pncconf -remove probe_parport command * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=6f35231
[08:47:13] <seb_kuzminsky> yay, thanks cmorley
[09:54:12] <seb_kuzminsky> pcw_home: have you changed fpga parts on some of your boards?
[09:54:53] <pcw_home> Not sure what you mean
[09:55:13] <seb_kuzminsky> i'm looking at commits like b4402ccf78e9
[09:55:27] <seb_kuzminsky> there have been one or two more like that in the past year or so
[09:56:11] <seb_kuzminsky> http://git.linuxcnc.org/gitweb?p=linuxcnc.git;a=commitdiff;h=b4402ccf78e9a2547303321f044fb7c0409667fe
[09:56:59] <pcw_home> those fixed incorrect chip names
[09:57:36] <seb_kuzminsky> so those drivers never loaded, until now? that's surprising
[09:57:43] <seb_kuzminsky> but ok
[09:58:59] <seb_kuzminsky> i thought i had a 5i25 at the hackspace that worked before that commit
[09:59:01] <seb_kuzminsky> hmm
[09:59:02] <pcw_home> Im sure they loaded but the name is not checked
[09:59:27] <seb_kuzminsky> ooohh...
[09:59:43] <seb_kuzminsky> a dim lightbulb goes on over seb_kuzminsky's head
[09:59:49] <pcw_home> (since the name is only on the bit file header and those are cards with on board flash)
[10:01:04] <seb_kuzminsky> yeah, i get it now
[10:01:05] <pcw_home> I think micges noticed when doing mesaflash and using the hm2 driver as a reference (since it deals with the bitfile, it checks)
[10:01:20] <seb_kuzminsky> it's only when we flash the fpga that we check the part number against the bitfile
[10:01:30] <pcw_home> Yep
[10:01:33] <seb_kuzminsky> i'm ok now :-)
[10:03:54] <pcw_home> it _might_ be possible to determine the FPGA name from ICAP but the PCI cards have write only ICAP for the moment (due to lazyness)
[10:06:47] <pcw_home> I can try on a Ethernet card since they have read/write ICAP
[10:23:53] <skunkworks> yes - you do seem like a slacker...
[10:25:36] <pcw_home> well it crossed the "too hard to do today" threshold
[10:28:28] <skunkworks> :)
[10:29:04] <skunkworks> anyone have a source for 36" high resolution inexpensive linear scales?
[10:29:13] <skunkworks> (I don't ask for much_
[11:30:05] <jepler> skunkworks: maybe you should try to 3d print something
[11:38:44] <jepler> skunkworks: more seriously, I don't know what "inexpensive" is.
[11:38:46] <jepler> http://www.aliexpress.com/item/5v-resolution-5um-1000mm-ttl-linear-encoders-linear-sensor-linear-glass-scales-optical-ruler/1385839710.html
[11:39:39] <cradek> I wonder how you guess what those output
[11:40:13] <cradek> some are meant for human reading, so they could generate serial (slowly) or something
[11:41:27] <pcw_home> 60M/min so fast
[11:42:15] <pcw_home> looks like diff quadrature and 1v PP analog (might be an option)
[11:43:15] <jepler> this may be the manual shown in the picture
http://www.t2cnc.hu/downloads/Utmero/uveg-merolec.pdf
[11:43:46] <pcw_home> I wonder is thats what Dabit on the forum referred to as "Cheap Chinese scales"
[11:43:46] <jepler> but .. no information on the outputs
[11:45:18] <jepler> and otherwise it generally reads like the turbo encabulator product presentation
[11:45:33] <jepler> "the indicating grating adhibits the score surface and the top of the grating all the time"
[11:46:31] <pcw_home> Says RS-422 so surely must be differential quadrature
[11:47:24] <pcw_home> My all time favorite is "regal sticked rubbles" for stick-on rubber feet
[11:49:51] <jepler> some sites have shown me TTL single-ended and EIA-422 differential
[11:50:57] <cradek> you can find out for $60 and about a month of your life
[11:51:17] <jepler> cradek: plus surprisingly high shipping
[11:51:19] <pcw_home> Yeah a surprise pachage
[11:51:47] <cradek> oh that's shipping in a few days - $45 is no surprise
[11:51:57] <jepler> Pin 1 2 3 4 5 6 7 8 9
[11:51:57] <jepler> Signal A- 0V B- Shield R- A +5V B R
[11:51:58] <jepler> Color Orange Black Blue Shielding White Yellow Red Green Brown
[11:52:08] <cradek> aha!
[11:52:09] <jepler> I'm finding a number of sources that give information comparable to this
[11:52:12] <cradek> perfect
[11:52:18] <jepler> http://cnditron.en.made-in-china.com/product/ybBJjMNOdFWc/China-Ditron-Linear-Scale-Galss-Scale-DC10-DC11-DC20.html
[11:54:53] <jepler> linear scale linear encoder tv new listing of multifunctional mosquito rat repellent pest reject drive cockroaches tool tv055
[12:31:29] <skunkworks> I think we decided that 5um isn't high enough..
[12:32:33] <pcw_home> even if its multifunctional mosquito rat repellent
[12:40:31] <kwallace2> It seems, other than the glass scale and sensor mask, the rest, housing and sensor, should be relatively easy to make.
[12:44:04] <kwallace2> ... I guess like writing a novel is just typing a bunch of words.
[13:29:49] <jepler> skunkworks: .0002in? luxury!
[13:30:57] <skunkworks> well - it is a tuning thing..
[13:31:26] <skunkworks> to position within .001 - .0002 is a bit of a stretch.
[13:32:11] <pcw_home> if its just used for slow I term feedback for temperature comp etc .0002 may be fine
[13:32:57] <jepler> all you need to double the resolution is to get 2 of the scales and place them 90 degrees out of phase
[13:33:13] <jepler> or do I mean 45?
[13:34:08] <pcw_home> the likelyhood they would stay at 45 degrees over length and temperture is not high
[13:34:51] <jepler> awww party pooper (the idea was not serious anyway)
[13:35:24] <pcw_home> Are they really glass scales?
[13:35:30] <jepler> I have no idea
[13:36:34] <pcw_home> Dabit's cheap scales use a steel tape (that vibrated, adding so tuning fun)
[13:36:47] <pcw_home> some
[15:29:38] <seb_kuzminsky> cmorley: i'm looking at the 2.6 commits since 2.6.3, for writing the changelog for 2.6.4, and i came across commit 18911987b
[15:30:05] <seb_kuzminsky> which looks like it has a bug in it, where you call tooledit.set_filename() with a hardcoded path in your home directory
[17:08:08] <andypugh> Errr, I got to the bottom of my feed-per-rev problem
[17:08:45] <andypugh> It was a combination of rpm/rps _and_ a feed-over-ride of 14,000
[17:09:03] <andypugh> So, nothing to do with JA or lInuxCNC at all.
[17:09:37] <andypugh> The FO is a mystery. It took me a solid 5 minutes of handle-cranking to get it bck to 100.
[17:10:05] <seb_kuzminsky> heh
[17:10:23] <seb_kuzminsky> well it's good to hear it's not some crazy bug
[17:10:54] <andypugh> There has to be something crazy, I certainly didn’t spend 5 minutes getting the FO up there...
[17:11:35] <andypugh> (5 mins is not an exaggeration, it felt longer, my arm was getting tired!)
[17:11:35] <seb_kuzminsky> does the FO live in a variable in interp? and if so, did your .var file get corrupted?
[17:12:17] <andypugh> Always possible, but I couldn’t find a .var file. and it wasn’t in .touchy_preferences either.
[17:12:32] <andypugh> It is certainly stored somewhere.
[17:15:47] <andypugh> Right, now to make the parts.
[17:18:49] <jepler> andypugh: that's a heck of a story
[18:20:23] <andypugh> I don’t know how I failed to notice the FO number…
[19:49:30] <cmorley> seb_kuzminsky: technically I should comment that out, but that bit of code is for testing only, so not a deal breaker.
[20:52:29] <cradek> FO is not stored in the var file. it's stored in .touchy_something I bet
[20:52:47] <cradek> but yes 14,000 is impressive
[20:54:16] <andypugh> It certainly wasn’t done deliberately
[23:26:11] <seb_kuzminsky> cmorley: ok cool