Back
[01:45:20] <tjb1> Aero-Tec: He obviously cant handle production :)
[01:54:50] <Valen> http://www.diyfaq.org.uk/humour.html
[01:57:17] <DJ9DJ> moin
[02:01:53] <tjb1> Valen: tl;dr?
[02:02:53] <Valen> funny, read it
[02:03:14] <tjb1-mobile> Its wayyy tooo long
[02:03:28] <Valen> you are american aren't you
[02:03:41] <Valen> and there are 3 stories in there
[02:12:22] <tjb1-mobile> Im lazy
[03:02:30] <Loetmichel> mornin'
[06:04:07] <r00t4rd3d> http://www.cnczone.com/forums/cnc_wood_router_project_log/148018-mistress_build-2.html#post1208201
[06:34:52] <archivist> I bet the string will fail, real cable chains I have seen have a hinge of some sort
[06:41:09] <jthornton> if he used Kevlar or similar it would last a while, but it looks like string to me
[06:41:35] <archivist> depends on flex radius
[06:42:06] <jthornton> and how much CA wicked into the joint
[06:42:48] <archivist> we learned rapidly when we were modifying a black and white printer to colour
[06:43:07] <Tom_itx> shoe string to me
[06:43:22] <archivist> added a solenoid apparatus to the head to move a ribbon
[06:45:42] <Tom_itx> they're not bending far. i suspect it will live a while
[06:46:45] <archivist> it has a shock at the end of the bend as well
[06:48:11] <jthornton> probably a Mack user as they tend to prefer commercial web sites with flashy ads
[07:00:01] <mrsun> oo i rather like those cablechains =)
[07:38:29] <r00t4rd3d> simple design too
[07:40:04] <r00t4rd3d> cable chain is pretty cheap though if you order it from china
[07:41:01] <Loetmichel> r00t4rd3d: or you can make it yourself ;-)
[07:42:15] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_itemId=412
[07:42:26] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_itemId=4420
[07:43:41] <jthornton> what kind of material is that?
[07:44:24] <alex_joni> jello
[07:45:29] <Loetmichel> jthornton: mine? PCM material 1 side 35µm copper
[07:45:33] <Loetmichel> PCB
[07:46:04] <Loetmichel> and some copper studs6mm diameter, 5mm long
[07:46:24] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_itemId=4384
[07:46:30] <Loetmichel> works like a charm ;-)
[07:46:58] <jthornton> nice work
[07:53:04] <Loetmichel> absolutely insane frm material and work cost
[07:53:16] <Loetmichel> but material was spare and work is hobby ;-)
[08:35:36] <skunkworks> Can anyone give me a recap of genes pid spindle thread?
[08:35:52] <skunkworks> 10 words or less?
[08:36:05] <archivist> does anyone know
[08:37:10] <skunkworks> I was hoping...
[08:37:14] <skunkworks> :)
[08:37:31] <mrsun> hmm i wonder how robust 20x5mm flatstock with braces every say 10cm would be in the middle of the braces ... its angled at 45 degrees
[08:39:13] <archivist> flat is not good in compression
[08:45:18] <mrsun> ffs
[08:45:32] <mrsun> tried to cad it and cant get an image out of the cad program even with print screen ...
[08:50:22] <mrsun> http://imagebin.org/239174 like that ... with braces like that every like 100mm
[08:50:26] <mrsun> say 20mm thick
[08:50:33] <mrsun> *10mm
[08:53:41] <mrsun> i do not want to weld in it tho as welding always warps stuff
[09:21:21] <hdokes|werkin> lmao... that last one requires their sdk to be intregrated into the app you want to use it with... they give a list of apps that have it... they are all ipad apps
[09:21:59] <hdokes> ummmmmmmmm... ooops.... sorry guys... wrong window
[09:22:04] <hdokes> :/
[09:30:03] <mrsun> i will never forgive you!
[09:35:23] <archivist> just get some angle iron
[09:44:43] <mrsun> i need the 45 degree V on top and bottom for bearings ...
[09:44:52] <mrsun> 90 degree V
[10:37:29] <Aero-Tec> I am trying to get the new config files to run and they will not
[10:38:02] <Aero-Tec> I copied the dir of a working config
[10:38:25] <Aero-Tec> renamed them each for the new config
[10:39:24] <Aero-Tec> took the starting icon and copied that a few times and renamed and retargeted them
[10:39:31] <Aero-Tec> so this should work
[10:39:34] <Aero-Tec> not
[10:40:03] <Aero-Tec> so why will my new configs not work and the old working config still is working fine?
[10:41:03] <JT-Shop> typpo
[10:41:10] <Aero-Tec> oh and I edited the config files for testing new ways to get the job done
[10:41:25] <Aero-Tec> I looked, looked hard for typos
[10:41:32] <archivist> one small change at a time
[10:42:32] <Aero-Tec> as you can see by my postings typos are one of my specialties, I am what you would call a expert at typos
[10:43:17] <Aero-Tec> so it should work
[10:43:51] <Aero-Tec> I thought it should work, was shocked when it did not
[10:44:10] <Aero-Tec> now the ini has the name of the config in it
[10:44:21] <Aero-Tec> is that used in any way by EMC?
[10:44:58] <Aero-Tec> does it have to match file names or dir names or anything like that?
[10:45:18] <Aero-Tec> there is no error or nothing
[10:45:19] <Loetmichel> mrsun: this way? ->
http://www.cyrom.org/palbum/main.php?g2_itemId=344
[10:45:20] <JT-Shop> what was the error message?
[10:45:28] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_itemId=335
[10:45:28] <Aero-Tec> it just will not do anything
[10:45:58] <Aero-Tec> like you never clicked the icon
[10:46:02] <JT-Shop> what is your line you use to launch with?
[10:47:03] <cncbasher_> did you rename the new name or folder in the INI
[10:47:07] <Aero-Tec> same as the working one with a small change to the path for the new name of the dir of the new config
[10:47:19] <JT-Shop> can you paste it here?
[10:47:29] <Aero-Tec> now I have spaces in the dir names
[10:47:35] <Aero-Tec> does that matter?
[10:47:41] <cncbasher_> yes
[10:47:45] <archivist> spaces!!!
[10:47:45] <cncbasher_> dont use spaces
[10:47:46] <Aero-Tec> also spaces in the file names
[10:47:56] <archivist> dont do that
[10:47:59] <JT-Shop> spaces are not allowed
[10:48:17] <Aero-Tec> bummer
[10:48:22] <cncbasher_> just use an underscore or dash instead
[10:48:27] <archivist> well they are if you learn to quote
[10:48:29] <Aero-Tec> one thing I like better about windows
[10:49:04] <Aero-Tec> ok
[10:49:07] <Aero-Tec> will fix
[10:49:08] * JT-Shop smacks Aero-Tec with a frozen mackerl for even saying that
[10:49:17] <Aero-Tec> lol
[10:49:42] <Aero-Tec> I hate windows for the most part
[10:50:01] <cncbasher_> JT : did you manage to grab all that was in drop box
[10:50:19] <JT-Shop> I suspect that the file naming is not a Linux thing
[10:50:19] <Aero-Tec> for me it is a necessary evil
[10:50:24] <cncbasher_> just want to make sure you got it all
[10:50:37] <JT-Shop> if you didn't add anything I got it! Thanks!
[10:50:51] <cncbasher_> ok i'll delete it this end
[10:51:11] <JT-Shop> should I copy it to a place outside of dropbox folder?
[10:51:14] <Aero-Tec> so linux is fine with dir and name spaces?
[10:51:39] <JT-Shop> in some ways but not in a terminal
[10:52:07] <cncbasher_> yea i would
[10:52:26] <archivist> you have to quote names with spaces SO DONT DO OT
[10:52:29] <archivist> IT
[10:52:41] <JT-Shop> done on my end
[10:52:52] <Aero-Tec> I just looked at some file names and they have spaces
[10:52:58] <Aero-Tec> ok
[10:53:00] <Aero-Tec> got it
[10:53:28] <archivist> you cannot tell if all the stuff you are using is also fine with spaces
[10:54:39] <archivist> spaces have always been a bad idea because of the confusion between options and the name
[10:56:21] <gene77> Morning all
[10:59:06] <JT-Shop> hi gene
[11:16:08] <gene77> Hi John
[11:16:38] <gene77> Trying to start from scratch on this encoder setup
[11:18:04] <gene77> when I set position-scale to the # of slots, its off, 1 rps is about 1.2 at the encoder out, thats 39
[11:18:42] <gene77> when I setp it to about 50, its almost dead on
[11:18:56] <gene77> so what is correct?
[11:19:25] <archivist> what is reading the pulses, parallel port or hardware
[11:19:32] <gene77> pport
[11:19:57] <archivist> slow down and try again
[11:20:40] <gene77> 1 rps is pretty slow
[11:21:12] <gene77> and its convenient to measure with a scope
[11:21:28] <gene77> 1 index pulse per second that is
[11:21:34] <archivist> does halscope see the same as a real scope
[11:22:31] <gene77> no
[11:23:18] <gene77> well, lemme qualify that with a test right now, this wiil take about 4 or 5 mins though
[11:24:39] <gene77> this morning it look like it is, yesterday it was off about 3/1
[11:25:49] <gene77> even with a 1 multipier its dead on
[11:27:28] <gene77> Ok next is scale for motion-speed-out, is .95 atm for 1 rps, not that far off
[11:28:21] <archivist> there should be no approximate numbers, the system is digital
[11:28:29] <gene77> Bur the next + step will go to 101.95, too much of a change, needs to be about half that
[11:29:40] <gene77> where is that step size changed when I click the +- buttons on the axis screen?
[11:30:25] <archivist> your sampling period will limit the resolution
[11:30:46] <gene77> of course, 1 millisec for this
[11:31:18] <gene77> actually 995 hertz
[11:32:17] * archivist remembers nyquist
[11:32:53] <gene77> so my acuracy then is .5%
[11:33:16] <gene77> I can't read the scope that close :)
[11:48:32] <IchGuckLive> hi all B)
[11:48:55] <gene77> But scaling is way off. If I enter, in MDI mode, s60 then motion-spindle-speed-out is 6
[11:49:24] <gene77> ah screw it, override was set at 10%
[11:53:57] <gene77> so, motion-spindle-speed-out is in RPS?
[11:55:18] <skunkworks> probably...
[11:56:32] <skunkworks> well - according to this it is rpm...
http://www.linuxcnc.org/docs/2.5/html/man/man9/motion.9.html
[11:57:18] <skunkworks> there is a separate pin for rps motion.spindle-speed-out-rps
[11:59:35] <gene77> And thats much lower, and gets me down into the digital noise
[12:01:03] <gene77> I need to get the scaling under control, a speed-out=10 s/b in rpms, not almost rps
[12:03:50] <gene77> If I run pid.FF0 up to about 57, then its correct for RPS, not RPMS.
[12:04:13] <gene77> and the pid.error fades to near zero too.
[12:04:46] <gene77> and it is turning 10 rps=600 rpm
[12:05:53] <gene77> that means MDI ot gcode issued speed commands are in RPS?
[12:06:05] <gene77> that means MDI or gcode issued speed commands are in RPS?
[12:07:11] <IchGuckLive> gene did you set g-code preloads into your ini
[12:08:24] <gene77> example, or do you mean hal, yes, I'm playing with those from the config screen in this instance
[12:08:50] <IchGuckLive> in ini R7... section
[12:09:12] <IchGuckLive> G97
[12:09:21] <gene77> from mthe MDI screen s1 is a tad slow, but s10 is 600 revs with near zero error now
[12:09:22] <IchGuckLive> needs to be theire
[12:10:01] <gene77> No, not familiar with that, man page?
[12:10:10] <IchGuckLive> http://linuxcnc.org/docs/html/config/ini_config.html#_rs274ngc_section_a_id_sub_rs274ngc_section_a
[12:10:22] <IchGuckLive> RS274NGC_STARTUP_CODE = G01 G17 G20 G40 G49 G64 P0.001 G80 G90 G92 G94 G97 G98
[12:10:31] <gene77> G97 is set however
[12:10:35] <IchGuckLive> put there a sequence of your need
[12:10:47] <IchGuckLive> it needs to be there
[12:10:56] <IchGuckLive> start with Plane
[12:11:01] <IchGuckLive> G17
[12:13:05] <gene77> G17? This is a lathe?
[12:13:23] <gene77> G18 is set in the ini
[12:13:58] <gene77> But G97 wasn't set, no is explicitely
[12:14:07] <IchGuckLive> can you post the line in the ini
[12:14:13] <gene77> correction But G97 wasn't set, now is explicitely
[12:14:43] <gene77> G7 G18 G20 G90 G97 now
[12:15:13] <IchGuckLive> put the G7 beond the G20
[12:15:37] <IchGuckLive> and a G54 into
[12:15:55] <IchGuckLive> also beond G20
[12:15:57] <gene77> ok, done
[12:16:11] <IchGuckLive> is this 2.5.1
[12:16:35] <IchGuckLive> incorrect spindle speed bug
[12:16:43] <gene77> 2.5.1 yes
[12:17:06] <IchGuckLive> it is fixed after 31.6.2012
[12:17:13] <gene77> Cute, fix?
[12:17:52] <IchGuckLive> i only see a bugfix for this
[12:18:10] <gene77> this is 2.5.1-138-g02e7afa
[12:18:24] <IchGuckLive> if you are online with the mashine pc please update to master
[12:18:43] <IchGuckLive> http://buildbot.linuxcnc.org/
[12:18:57] <gene77> ok, apt sources line?
[12:19:11] <IchGuckLive> "deb
http://buildbot.linuxcnc.org/ lucid master-rt"
[12:20:39] <IchGuckLive> if there is a SMI fix to your latency you need to redo it
[12:20:51] <IchGuckLive> as the master will renew the file
[12:22:22] <gene77> SMI? I thought that was in the bios
[12:24:09] <IchGuckLive> no
[12:24:21] <IchGuckLive> if you got latency chumps
[12:24:27] <IchGuckLive> jumps
[12:24:41] <IchGuckLive> http://wiki.linuxcnc.org/cgi-bin/wiki.pl?FixingSMIIssues
[12:25:09] <gene77> I don't seem to have those
[12:25:30] <gene77> and can't recall if I fixed it, been a while
[12:25:44] <gene77> upgrade in progress
[12:25:57] <IchGuckLive> look into the file bevor updating if it is in there smi
[12:26:07] <IchGuckLive> BNP
[12:26:28] <IchGuckLive> NP if please check your latency after update
[12:26:58] <gene77> while thats doing, I have a coffee cup thats so empty its dry... brb
[12:30:24] <gene77> And, I just got a request for sustenance from the better half, and if we want to eat at our usual greasy spoon, I'd better take an hour off & go that, back later
[12:31:00] <IchGuckLive> NP
[12:34:04] <andypugh> I hadn't seen this pattern of V-block before. They have the rather nice property that you can hold them in a machine vice.
http://www.ebay.co.uk/itm/Bilton-Son-No-2-Vee-Blocks-As-Photo-3-9-16-x-2-3-16-x-2-3-16-/150960522280
[12:35:20] <archivist> nice but pricey
[12:36:30] <andypugh> Indeed. Not always, though.
http://www.ebay.co.uk/itm/321036230270
[12:37:00] <andypugh> (Looks shinier in RL :-)
[13:04:43] <mrsun> Loetmichel, nah got V grooved bearings
[13:10:00] <mrsun> Loetmichel, your router? :)0
[13:12:22] <Loetmichel> mrsun: a non-finished project of mine
[13:12:55] <mrsun> ahh looked quite neat but what is the material ? :) like 1mm pcb? :P
[13:12:59] <Loetmichel> state at the moment:
http://www.cyrom.org/palbum/main.php?g2_itemId=359
[13:13:00] <IchGuckLive> the prismas are mutch cheeper in germany
http://www.paulimot.de/epages/paulimot.sf/de_DE/?ObjectPath=/Shops/paulimot/Categories/Prismen
[13:13:18] <Loetmichel> 3mm FR4 without copper
[13:13:42] <Loetmichel> -> about 2 pcbs stacked
[13:13:44] <IchGuckLive> Loetmichel: Ice on the road ?
[13:13:48] <mrsun> ahh =)
[13:14:27] <Loetmichel> IchGuckLive: no. easy going... but only 2Mbit at the company, yubuntu install with nvidia drivers is a PITA....
[13:14:33] <Loetmichel> so i am a little late ;-)
[13:15:47] <Loetmichel> mrsun: should have been a series of small 400 by 300 by 100mm CNC mills
[13:16:02] <Loetmichel> but accumulating the cost the FR4 is just too expensive
[13:16:17] <mrsun> ahh
[13:16:56] <IchGuckLive> Loetmichel: did you search for turret boad
[13:17:09] <IchGuckLive> there are low cost FR4 3mm plates
[13:17:12] <Loetmichel> i now use "siebdruckplatte" for my newest design, still not finished but a bit more sturdy and cheap as fu** ;-)
[13:17:24] <IchGuckLive> around 10euros for a 500mm sheet
[13:17:32] <Loetmichel> thats expensive
[13:17:58] <Loetmichel> mrsun:
http://www.cyrom.org/palbum/main.php?g2_itemId=8872
[13:18:13] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_itemId=8884
[13:18:54] <Loetmichel> same construction, 600mm by 400mm movement, dirt cheap to build... (IF you have a cnc mill handy ;-)
[13:19:49] <IchGuckLive> compare to 2years ago there is so many minmill sellers around
[13:19:57] <Loetmichel> right
[13:19:58] <IchGuckLive> the market is filling up
[13:20:25] <IchGuckLive> plasma is the new goal to be filled
[13:20:37] <IchGuckLive> all this metall fences around
[13:21:00] <Loetmichel> but still the most sellers have weired wobbly constructs.
[13:21:22] <Loetmichel> nothing worth mentoining ;-)
[13:21:22] <IchGuckLive> chees mills are good for pcb
[13:21:38] <IchGuckLive> weigel does sell for 10k
[13:22:00] <Loetmichel> my intention was and is to build mills that are capable of light aluminium milling
[13:22:13] <Loetmichel> and have no measureable backlash
[13:23:03] <Loetmichel> IchGuckLive: i intend to build the wooden ones plug and play for a LinuxCNC-PC for less than 1kEur
[13:23:12] <IchGuckLive> thats the goeal everyone is looking fore
[13:23:31] <Loetmichel> for the 640 by 400mm mill
[13:23:48] <IchGuckLive> you can go al rack for this price
[13:24:26] <Loetmichel> ?
[13:24:47] <IchGuckLive> on cnc-discount rosa rail system its cheep also to go ballbearing
[13:24:50] <Loetmichel> did you mean "aluminium frame"?
[13:24:55] <IchGuckLive> yes
[13:25:09] <IchGuckLive> the 45x45 heavy
[13:25:15] <Loetmichel> aluminium frames are wobbly OR expensive
[13:25:46] <Loetmichel> ans i have never seen a extruded aluminium profile that wasnd like a banana
[13:25:53] <Loetmichel> *and
[13:26:11] <IchGuckLive> with bolds every 100mm its not
[13:26:27] <Loetmichel> ?
[13:26:46] <Loetmichel> the PROFILES are bent
[13:26:49] <Loetmichel> not the rails
[13:27:10] <Loetmichel> and if you mount the rails on the profiles the rails are also bent
[13:27:44] <IchGuckLive> ok did you see cnc-step reptor production
[13:28:17] <IchGuckLive> he is now on 30people manufacturing this tiny things
[13:28:38] <IchGuckLive> for that price they go as "Warme Weck"
[13:28:47] <Loetmichel> if i want to do a extremely STURDY mill, i will go for rectangular steel tube, welded to a fat frame, glown for some hrs in the oven (stress relieve), and then milled over to be precise
[13:29:26] <Loetmichel> welded to BE a fat frame
[13:29:48] <Loetmichel> cnc-step?
[13:30:03] <IchGuckLive> dot de
[13:30:08] <Loetmichel> the ones with the "multitool"?
[13:32:15] <Loetmichel> god in heavens, THAT company... i would NEVER buy from them, even if they HAD good products because of the attitude of their boss... to try to push the competition out of the market by filing complaints about missing "not aus" and stuff has rendered them "not buyable by me
[13:35:00] <IchGuckLive> http://www.youtube.com/watch?v=6VMoPsPH6GY
[13:35:16] <IchGuckLive> but it fits the need of most at 2,3K
[13:39:51] <IchGuckLive> http://gocnc.de/ nice chees
[13:40:37] <IchGuckLive> he uses EMC2 not linuxcnc
[13:42:49] <IchGuckLive> ok by im off
[13:51:32] <gene77> I'm baaaack! Now, what about this SMI fix, how do I check it?
[13:53:36] <gene77> No more latebcytest?
[13:53:45] <gene77> latencytest
[13:55:22] <gene77> After update, latencytest not found
[14:14:45] <Aero-Tec> will got it working, running the new configs
[14:14:55] <Aero-Tec> have tested one of them
[14:16:26] <Aero-Tec> changing B axis from angular to linear makes no difference, also one can use the 360 roll over on linear axis, go figure
[14:17:14] <Aero-Tec> I have to get this job done so I will not play any more with this now
[14:17:40] <Aero-Tec> will have to rewrite the Gcode for the units per rev test
[14:18:57] <Aero-Tec> and redo the HAL and redo the Gcode for the X,Z axis test, that one I think has the best chance of working, I may just try it as it should speed up production
[14:19:58] <Aero-Tec> being ableto have the 360 roll over on a linear axis will be of a big help
[14:20:16] <Aero-Tec> was worried how I would be able to handle that
[14:22:00] <Aero-Tec> still have to see if the 360 roll over will work on a axis that is used for linear, it may just work on axis that are known to be angular
[14:22:43] <gene77> Need one change to latencyplot, it needs to run as a user who can stop it, it leaves the system dirty now
[14:23:14] <gene77> because the user apparently can't run lahrun -U
[14:23:34] <gene77> halrun -U
[14:24:02] <Aero-Tec> the test should be there
[14:24:16] <Aero-Tec> I have done 2 update and the test was there each time
[14:26:32] <gene77> locate cannot find it now since updating to lucid master-rt
[14:27:27] <gene77> latencyplot is there, but really needs to be run by root if a clean quit is desired.
[14:27:37] <Aero-Tec> also to check to see if you have SMI fix in place by looking at ->"/etc/linuxcnc/rtapi.conf
[14:28:41] <gene77> and an S10, which if rps, should be 600 rpms, is now 1260 rpms
[14:28:44] <Aero-Tec> one way to see if you have a problem with latency is to run the program and see if it complains
[14:29:01] <gene77> in progress
[14:29:04] <Aero-Tec> check INI and HAL
[14:29:43] <Aero-Tec> also check your over rides
[14:29:55] <Aero-Tec> see if they are set to over drive the speed
[14:30:18] <Aero-Tec> to run at set speed they need to be 100
[14:30:55] <Aero-Tec> the over rides are in the EMC GUI
[14:31:11] <Aero-Tec> not the INI or HAL
[14:31:31] <Aero-Tec> the limit set and other options are in the INI
[14:31:57] <Aero-Tec> not sure if spindle sped has a over ride
[14:33:11] <Aero-Tec> http://wiki.linuxcnc.org/cgi-bin/wiki.pl?FixingSMIIssues
[14:33:15] <Aero-Tec> smi info
[14:58:26] <Aero-Tec> that was painless
[14:58:49] <Aero-Tec> changed HAL from B to Z
[14:58:59] <Aero-Tec> and redid the INI and Gcode
[14:59:13] <Aero-Tec> now for the moment of truth
[14:59:31] <Aero-Tec> will 360 roll over still work and will the G64 work
[14:59:45] <Aero-Tec> BRB
[15:19:56] <Aero-Tec> getting limit error
[15:20:13] <Aero-Tec> I set the limit in the INI to 12 places
[15:20:30] <Aero-Tec> yet for some reason the z axis will not go
[15:20:49] <Aero-Tec> says move would be beyond the limits
[15:20:55] <andypugh> Aero-Tec: Did you home?
[15:21:07] <Aero-Tec> so is there a over ride limit?
[15:21:24] <Aero-Tec> I have ignore home set
[15:21:34] <Aero-Tec> no I did not home Z
[15:21:55] <andypugh> You still need to tell it where it is on the slide, even with no-force-homing
[15:22:17] <Aero-Tec> but I do have limit set at +/- 999999999999
[15:22:19] <andypugh> Press # and you will see where the machine thinks it is in machine cordinates.
[15:22:37] <Aero-Tec> ok
[15:25:48] <Aero-Tec> still complains
[15:25:53] <Aero-Tec> I did a home
[15:26:10] <Aero-Tec> the z was at 4800 and change
[15:27:08] <Aero-Tec> is there some other place other then INI to set axis limit?
[15:27:53] <andypugh> No, only the INI
[15:28:14] <andypugh> This is fun: Google as it would have been in punchcard days.
http://www.masswerk.at/google60/
[15:29:51] <gene77> I guess the only way I'll get the spindle speed right is put a scale operation between limit2.0.output and pid.0.commnd.
[15:30:03] <gene77> Is this normal?
[15:30:46] <andypugh> No, and it makes no sense. Any scaling ought to be on the feedback side.
[15:31:00] <Aero-Tec> Z is axis 2 right?
[15:31:09] <andypugh> Aero-Tec: Yes
[15:32:21] <Aero-Tec> so why does it not accept my limits in the INI?
[15:32:47] <Aero-Tec> just changed it to angular, see if that helps
[15:35:44] <gene77> Ok Andy, encoder.x4 is off, mode is 0, position scale is then the number of slots=39, and its a bunch closer
[15:36:39] <Aero-Tec> well that does not work
[15:37:37] <Aero-Tec> even homing Z makes it complain about axis 2 being out of over limit
[15:38:14] <Aero-Tec> I could have been done my run by now if I just let it stop at moves end
[15:38:22] <Aero-Tec> looks like I will have to do that now
[15:38:32] <Aero-Tec> lost time and no real gain
[15:38:43] <Aero-Tec> found out some limits of EMC
[15:38:54] <Aero-Tec> BTW am I doing anything wrong?
[15:39:38] <Aero-Tec> why will Z not take the limits of the INI?
[15:40:18] <Aero-Tec> there must be some internal limit for Z, or I am messing up some how
[15:40:51] <Aero-Tec> any gurus out there that can help?
[15:42:01] <gene77> we need a real time halscope, this 30+ seconds for an update is killing my back
[15:43:06] <Tom_itx> Aero-Tec, what's MIN_LIMIT and MAX_LIMIT set for on that axis?
[15:43:28] <Aero-Tec> 999999999999
[15:43:34] <Aero-Tec> -999999999999
[15:44:41] <Tom_itx> try a smaller value?
[15:45:33] <Aero-Tec> I just checked and made sure the min was the - one and the max was the + one
[15:45:37] <Aero-Tec> can do
[15:45:41] <Aero-Tec> how small?
[15:45:49] <Tom_itx> 6 digit maybe
[15:46:02] <Aero-Tec> 12 worked for B axis
[15:46:36] <Aero-Tec> actually the B was tons higher
[15:46:39] <Aero-Tec> 9999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999.0
[15:46:51] <Aero-Tec> did not count how many that was
[15:47:25] <Aero-Tec> with 6 digits I run out of room
[15:47:58] <Aero-Tec> when doing the same run over and over it adds up
[15:48:13] <Aero-Tec> but for the fun of it will try
[15:49:27] <Aero-Tec> lol
[15:49:30] <Aero-Tec> oops
[15:49:38] <Aero-Tec> looks like that are both -
[15:50:04] <Aero-Tec> funny how one can over look a little- sign
[15:53:46] <Aero-Tec> ok
[15:53:49] <Aero-Tec> it worked
[15:53:56] <Aero-Tec> nice and smooth
[15:54:06] <Aero-Tec> whole thing in one shot
[15:54:49] <Aero-Tec> thing is wrap rotary did not work
[15:55:11] <Aero-Tec> and it is set at angular so wrap axis should be working
[15:55:59] <Aero-Tec> using G64 and G94
[15:56:08] <Aero-Tec> so units per min
[15:56:31] <Aero-Tec> did not need to do G93
[15:58:40] <Aero-Tec> any idea why wrapped rotary did not work?
[15:59:15] <gene77> Andy: What is a typical pwmgen.scale?
[16:10:50] <gene77> Andy, there is not a balance point between pwmgen,scale and pid.FF0, its a never ending fight
[16:10:54] <gene77> Ideas
[16:11:05] <DJ9DJ> gn8
[16:16:06] <gene77> Do I change the encoder position-scale to arrive at a balance?
[16:18:20] <Aero-Tec> any help with wrapped rotary for z and why it does not work?
[16:18:35] <Aero-Tec> the info talks about G91.2
[16:18:50] <Aero-Tec> but EMC says it is not Gcode
[16:20:14] <Aero-Tec> Z is set to rotary and wrap is turned on
[16:26:23] <Aero-Tec> should I be using G91.1 instead of G91.2?
[16:30:12] <andypugh> gene77: FF0 only works to the extent that your pwm to voltage is linear. After that point you need P and I terms.
[16:31:03] <andypugh> I am convinced that you should have no FF1 and FF2 in your spindle controller
[16:32:13] <andypugh> You need to set encoder position-scale so that one turn of the spindle = an increase of 1 in encoder position. That is the correct scale. There is no other correct scale.
[16:35:14] <andypugh> Aero-Tec: Which LinuxCNC version. I remember wrapped-rotary being introduced, but it wasn't very recent.
[16:36:16] <andypugh> Three ball nut drive designs. Which do you all prefer?
https://picasaweb.google.com/lh/photo/EqgIdXH-kCH8I90oYKfm2dMTjNZETYmyPJy0liipFm0?feat=directlink or
https://picasaweb.google.com/lh/photo/DELJ44ystXfa1HavUrYFmtMTjNZETYmyPJy0liipFm0?feat=directlink or
https://picasaweb.google.com/lh/photo/X9ZXwL-oMXpF9cc6wAIbQNMTjNZETYmyPJy0liipFm0?feat=directlink
[16:38:57] <gene77> Andy, and it seems fairly linear IF the pwmgen.offset is trimmed
[16:39:22] <gene77> I don't have any FF1 or FF2 ATM
[16:42:06] <gene77> Now I think I have a usable range from 1 rps to 10 rps setup and am looking at pid.out to see if its suitable for use as a fuse trip, but to get that scaled to on screen in the halscope, the range has to be set to about 100v/div!
[16:43:57] <gene77> And, it doesn't seem to be affected by spindle load. Odd. The error is, but thats got way too many startt-stop jumps to be really usefull.
[16:46:26] <gene77> I've copied off everything into the hal file, and will restart it using motion-speed-rps as those steps are much smaller when using the axis buttons for speed control, brb
[16:46:38] <joe9> i am looking for something similar to pcb-gcode
http://phk.freebsd.dk/CncPcb/index.html for linuxcnc. but, am not sure if anyone has something similar. Any thoughts, please?
[16:47:18] <joe9> it probes the height dynamically using G38.2 and then uses that to calculate the Z offset
[16:58:38] <MercuryRising> i have linux cnc running (installed, not livecd) on my computer, everything works except for pncconf. when i try to run it, it complains about a line in expatbuilder, "ExpatError: not well-formed (invalid token)". I tried to find the line in the git source, but I don't know how to look at previous versions (as the line numbers are all different now). Supposedly, this error comes from "Double dashes are not valid within comme
[16:58:40] <MercuryRising> nts (other than when ending them)" (via stack overflow). Any tips?
[16:59:23] <JT-Shop> what version are you running?
[17:05:36] <andypugh> gene77: pid.out is the output to the PWM, and that will be proprtional to velocity, pretty much.
[17:05:50] <andypugh> Looking at pid.0.error ought to work better.
[17:13:02] <MercuryRising> the latest one, i got it a couple weeks ago (linuxcnc - 2.5 and ubuntu 10.04 i believe)
[17:17:43] <gene77> Looks like, but I just found a huge discrepancy between motio-speed-out and motion-speedout-rps
[17:17:45] <JT-Shop> but you did not use the liveCD to install?
[17:18:04] <MercuryRising> i used the live cd to install
[17:18:09] <MercuryRising> but it's running from my hdd
[17:18:15] <gene77> what was an s10 command to get 600 rpms, is now an s2550!
[17:19:28] <MercuryRising> oh wow, on git it's tagged for the versions
[17:20:41] <gene77> actually to get 10 rps out, its s2609
[17:20:42] <JT-Shop> so your running it by going to Applications > CNC > Pncconf?
[17:20:45] <MercuryRising> i have linux cnc 1.5.1
[17:20:47] <MercuryRising> yes
[17:21:04] <MercuryRising> it loads a small UI box "pncconf is starting" then errors out
[17:21:26] <JT-Shop> did you check the iso file md5sum?
[17:21:41] <MercuryRising> nope
[17:22:02] <JT-Shop> might need to reinstall and check the iso first
[17:22:24] <andypugh> gene77: That makes no sense at all.
[17:23:22] <andypugh> S600 should set motion.spindle.speed-out to exatly 600, and motion.spindle-speed-rps to exactly 10.
[17:24:11] <gene77> Where is that multipier? not to me thats for sure. But that is what it takes... Found it, spindle slider was way down
[17:24:12] <cradek> andypugh: remember spindle override %
[17:24:15] <skunkworks> (depending on SO)
[17:24:29] <JT-Shop> MercuryRising: did you run the latency test?
[17:24:45] <andypugh> Ah. Yes, assuming no operator error, that is.
[17:24:46] <gene77> and that is what I'm getting on the meters now
[17:24:47] <MercuryRising> yes
[17:26:09] <JT-Shop> http://www.youtube.com/watch?v=eyCEexG9xjw
[17:27:06] <cpresser> JT-Shop: "Dieses Video ist in Deutschland nicht verfügbar, weil es möglicherweise Musik enthält, für die die erforderlichen Musikrechte von der GEMA nicht eingeräumt wurden." <-- the respective copyright owner banned this in germany
[17:27:11] * cpresser hates the internet
[17:28:23] <JT-Shop> what a turd head
[17:28:47] <andypugh> cpresser: Are you actually in Germany? I get that very frequently, and I am not in Germany.
[17:29:28] <cpresser> andypugh: yes, germans get this message a lot
[17:29:42] <cpresser> andypugh: its based on a geo-ip database.
[17:30:04] <andypugh> It is even more annoying when you are in the UK. And your IP is part of a class 2 (92.x.x.x)
[17:30:58] <MercuryRising> JT-Shop: would a poor latency score make pncconf not work?
[17:31:02] <andypugh> Ford Motor Company have more IP space than Africa. I am in the UK, Ford are American, why does Youtube think my office is in Germany?
[17:31:47] <andypugh> Dodgy banner ads have no trouble at all in offering me unconvincing examples od "Basildon Milfs" so why do Youtube struggle?
[17:32:17] <cpresser> andypugh: try running your IP on a geoip database. those banners use other techniques. like flash-cookies
[17:32:22] <MercuryRising> they probably want to be extra sure they aren't letting any germans see it
[17:32:54] <JT-Shop> MercuryRising: no, but that does indicate you have the realtime kernal
[17:33:08] <JT-Shop> can you check the md5sum of the iso?
[17:33:35] <MercuryRising> trrying, it took the whole cd which gave a different md5sum
[17:34:02] <andypugh> Talking of YouTube, I am now feeling a bit irrationally bad:
http://www.youtube.com/watch?v=vGgow60KKXM
[17:34:47] <JT-Shop> LOL
[17:35:04] <JT-Shop> that the chap that stole your video?
[17:35:12] <andypugh> That's right
[17:36:50] <JT-Shop> anyone know of a profiling insert that would cut a ball type shape with one tool?
[17:37:59] <gene77> And your right, pid.output seems to have no relation to spindle load. I would think it would rise markedly when the spindle is loaded.
[17:38:41] <gene77> So, if I'm to use the pid.error, I'll need a time delay to absorb the on ramps
[17:38:54] <andypugh> gene77: There may be an internal speed controller in the spindle drive.
[17:38:59] <rob_h> JT-Shop, depends how big ball is, but a vbmt tip might get u close, the large V type shape tip
[17:39:34] <andypugh> gene: Mask it with a oneshot triggered from a rising edge of spindle.is-on
[17:40:06] <andypugh> JT-Shop: Internal or external?
[17:40:10] <gene77> I've got a C41 in its place, that is the interface, suitabvle speeded up, Arturo's version has major fractions of a second lag
[17:40:51] <gene77> And they are killers where the relays are concerned.
[17:41:05] <rob_h> vcmt vcgt is the other type but not sure what holder as dont know how much of aball u need to make/turn...
[17:41:05] <JT-Shop> andypugh: external
[17:41:12] <andypugh> http://www.ebay.co.uk/itm/GLANZE-12MM-INDEXABLE-PROFILING-LATHE-TOOL-FROM-CHRONOS-/370675613989?pt=UK_Home_Garden_PowerTools_SM&hash=item564dfd0125
[17:41:32] <JT-Shop> http://www1.mscdirect.com/CGI/NNSRIT2?PMPXNO=16846728&PMTERM=09829284
[17:42:02] <andypugh> Mine is cheaper :-)
[17:42:03] <gene77> I can't use enough I or D gain without Dgain above .01 making it hunt like a starving dynosaur
[17:42:23] <andypugh> I wouldn't expect you to need any D gain
[17:43:01] <JT-Shop> right size almost
[17:43:18] <gene77> so they are essentialy turned off. IGain seems to just put wrinkles in the speed for a second or so after the ramp up.
[17:45:17] <andypugh> You might need to set the FF0 at a level where I is always pulling in the correct direction.
[17:45:43] <andypugh> I is a bit funny, really.
[17:45:54] <gene77> I'm relatively happy with it I think, as 1 to 10 rps covers 99% of what I'd do on this toy anyway. Now to hook up a limit2 between error and comp in2 & set it for about 2x the response time that the first limit2 imposes on the speed of the ramp.
[17:46:53] <gene77> Tommorrows project I suspect, I need to find some sustainunce for the evening now.
[17:47:58] <gene77> Thank you much for the clues and hand holding. Your video now has a copyright claim by you blocking it, kewl!
[17:48:21] <andypugh> Yes, I just wish it didn't state my name quite so clearly.
[17:52:28] <MercuryRising> JT-Shop: md5sum matches image on CD and ISO on hdd and displayed md5sum on linuxcnc
[17:52:40] <JT-Shop> hmmm
[17:53:59] <JT-Shop> try running pncconf from a terminal and see what the error is
[17:55:18] <JT-Shop> I see 3 warnings when I run pncconf from a termial but it starts up fine with a LiveCD install
[17:55:20] <MercuryRising> the UI displays the error (glade), when I hit "OK" on the error message, pncconf releases from terminal without status
[17:55:44] <JT-Shop> open a terminal and type in pncconf
[17:55:53] <JT-Shop> that way the terminal won't close
[17:56:15] <MercuryRising> i did
[17:56:45] <MercuryRising> it doesn't do anything, it captures the terminal until i hit okay, then it exits without any message
[17:57:06] <MercuryRising> this is the error message (sorry for not cropping):
http://i.imgur.com/8Ls10.png
[17:57:25] <MercuryRising> top left is the UI window that opens until it errors
[17:58:21] <MercuryRising> can i recompile and install just pncconf from the source?
[17:58:22] <JT-Shop> ugh do you have halrun running? they can't run togethere
[17:58:33] <JT-Shop> close the hal window
[17:58:59] <JT-Shop> and make sure it is unloaded with halrun -U I think
[17:59:23] <kb8wmc> halrun is not running
[17:59:42] <JT-Shop> it may not have closed
[18:00:12] <JT-Shop> it does look like the prompt is halcmd:
[18:00:31] <MercuryRising> i did before, it's not now, i tried halrun -U and it seg faulted, i'm restarting the computer and trying again
[18:01:15] <kb8wmc> hmmm....just attempted to close with -U and got an error window
[18:01:24] <JT-Shop> strange I can have halrun going and pncconf works
[18:02:33] <MercuryRising> what a great time for my graphics card to take a dump :)
[18:02:41] <JT-Shop> maybe do a memory test
[18:03:07] <JT-Shop> in any case I'm the chef de jore and must start cooking for the MIL
[18:04:14] <JT-Shop> otherwise she will eat a 0.99 cent TV dinner :(
[18:04:29] <MercuryRising> hahaha, alright, thanks for the help!
[18:04:59] <JT-Shop> post to the pncconf section of the forum the author monitors that
[18:05:08] <MercuryRising> alright awesome, thanks!
[18:05:09] <kb8wmc> I will reboot computer
[18:07:35] <kb8wmc> same after reboot
[18:07:58] <MercuryRising> is there a huge problem with spam on the linuxcnc forums? there's 3 'human' checks to pass to register
[18:13:26] <skunkworks> and #2 runs youtube vids
[18:16:14] <andypugh> MercuryRising: Yes. I personally rejected 800 spammer manually one day when we only had one Captcha. I think JT did 300 that day.
[18:17:11] <MercuryRising> wow, it doesn't seem like a board with 3,000 users would be that big of a target (although I'm guessing they just point to a domain and see how many registrations they can get)
[18:17:16] <MercuryRising> what's the spammer rate now?
[18:18:50] <andypugh> Zero. :-)
[18:19:22] <MercuryRising> good, cause it took me (a human) three tries to get the security code correct
[18:19:47] <andypugh> Maybe you are not as human as you thought? :-)
[18:20:23] <MercuryRising> i feel pretty humanlike, but who knows, only time will tell
[18:22:33] <MercuryRising> what's the correct way to calibrate the machine to match expected movement and real movement? I wanted to use PNCConf to do that (I saw a screen that had something like move an inch), stepperconf has the 'run an inch' option, but it always moves back and forth (even when just + or - is selected)
[18:26:34] <skunkworks> MercuryRising: do the math...
[18:26:54] <cradek> best to count teeth on your pulleys if you're seeing a scale that's the slightest bit weird
[18:27:09] <MercuryRising> i have direct drive on all leadscrews
[18:27:23] <cradek> oh then it should be trivial to calculate it if you know your screw pitch
[18:27:25] <skunkworks> then it should be east
[18:27:29] <skunkworks> easy
[18:27:46] <MercuryRising> don't know leadscrew pitch, just turn them to see how long it takes to get an inch?
[18:28:12] <cradek> thread gage? uh, tape measure?
[18:28:38] <MercuryRising> ahh that would make more sense
[18:30:16] <PCW_> pncconf from linuxcnc 2.5.1 runs on my machine though its really slow
[18:32:07] <PCW_> pncconf beta 1 on splash screen
[18:42:05] <MercuryRising> Alright got them all calibrated! Awesome!
[18:44:42] <MercuryRising> now I just need to wait for my extruder stepper driver to come in the mail, and i'll be ready to spew some plastic
[21:28:27] <MercuryRising> so i attached a marker to my x/y/z, and it started to work for a little bit, but then somehow a G54 got in there, and now whenever I 'home' my z axis, it is offset from where I want it to be (it's ~1 inch through my table), how do i get rid of the g54 offset?
[21:29:39] <jdh> jog z to wherever you want 0 to be, hit z, hit <end>, hit <enter>
[21:30:32] <MercuryRising> awesome! thank you!
[21:41:49] <RyanS> are wabeco mills designed for Europeans with far too much money? (Which is is a little tough these days)
[23:21:00] <jdh> I built a new d525 atom box for one of my machines. It's been sitting inside doing nothing for a week and just got an RTAPI Error, unexpected realtime delay.
[23:25:11] <Jymmm> jdh: CONGRADULATIONS!!!
[23:26:04] <Valen> jdh: when you first start it?
[23:26:19] <Valen> or was it running axis for a week?
[23:26:23] <jdh> no, I started it a day or two ago
[23:26:33] <Valen> thats new
[23:33:56] <L84Supper> maybe it's just a ripple in the time continuum
[23:34:10] <tjb1> Valen I read…not bad
[23:36:22] <Jymmm> jdh: might enable/raise logging and try agian
[23:36:32] <Jymmm> again
[23:37:27] <Jymmm> jdh: But if it was just idle, might just be a bug.
[23:40:31] <jdh> doesn't sound buglike, just something didn't let it run in time.
