#emc | Logs for 2009-01-21

Back
[01:07:20] <eric_unterhausen> how does a new .ini file improve latency?
[01:12:57] <jmkasunich> I was wondering the same thing
[01:13:18] <eric_unterhausen> face it, jepler is magic
[01:15:26] <tomp> isnt latency tested without emc ( without .ini) ?
[01:15:30] <jmkasunich> yes
[01:16:34] <tomp> thx
[03:01:37] <steve_stallings> steve_stallings is now known as steves_logging
[04:58:29] <Benjamin_Mullin> Its not controlled by EMC yet, but I thought I could share this here anyhow
[04:58:31] <Benjamin_Mullin> http://www.flickr.com/photos/26107619@N07/tags/minicnc/
[05:09:48] <dgarr> what kind of slides?
[05:12:47] <Benjamin_Mullin> Guide actuators from THK
[05:12:49] <Benjamin_Mullin> http://www.thk.com/us/products/class/lmguideact/index.html
[05:13:00] <Benjamin_Mullin> scrap bin at work
[05:13:09] <dgarr> yes -- i've used those (from ebay) verynice
[05:13:27] <dgarr> i notice a pittman motor but not in the final design? is it servo?
[05:13:40] <eric_unterhausen> I have some of those slides, the hole circle for the motors is a very strange size
[05:14:42] <Benjamin_Mullin> The "final design" is a set of surplus stepper motors
[05:15:47] <eric_unterhausen> I like the servo version, that's what I should have done
[05:15:57] <Benjamin_Mullin> The pittman motor and encoder are still around but I chose to start with the steppers
[05:16:59] <Benjamin_Mullin> maybe final design 2
[05:18:10] <eric_unterhausen> I have a small one that's still very long, I'd like to cut it in half
[05:18:13] <dgarr> nice job -- not emc yet? what have you used for testing so far?
[05:18:43] <Benjamin_Mullin> an atmel micro controller and custom software
[05:19:14] <eric_unterhausen> do you have any plans for your z?
[05:19:45] <Benjamin_Mullin> I have another guide actuator that is shorter
[05:19:51] <eric_unterhausen> I see that in the last pic
[05:19:54] <Benjamin_Mullin> no plans on the spindle yet
[05:20:22] <eric_unterhausen> keep wishing I had a better trash bin at work :)
[05:21:12] <Benjamin_Mullin> I've asked several times if its ok to take from the scrap room, they keep saying yes
[05:21:22] <Benjamin_Mullin> apparently they PAY to dispose of this stuff...
[05:21:41] <dgarr> nice fringe benefit
[05:21:43] <eric_unterhausen> I've heard of that
[05:22:03] <eric_unterhausen> I'm pretty sure Penn State makes money on most of their discards
[05:22:04] <Benjamin_Mullin> most of the scrap is electronics so I understand having to pay for most of it
[05:23:17] <Benjamin_Mullin> crappy video of it in action here: http://www.youtube.com/watch?v=51_OKkY9SDk
[05:32:52] <JustinXJS2> those are nice looking slides
[05:35:03] <Benjamin_Mullin> I haven't measured backlash or any other slop, but there doesn't appear to be much
[05:35:46] <eric_unterhausen> it looks like the wear material is aluminum
[05:36:08] <JustinXJS2> how much travel do they have?
[05:37:58] <Benjamin_Mullin> 141mm I think
[05:38:04] <killar> hmm
[05:38:44] <JustinXJS2> oh, a little on the small side for me
[05:39:22] <killar> thats what she said
[05:40:19] <Benjamin_Mullin> it's a cute size
[05:41:14] <killar> hehe
[05:41:21] <killar> sorry, i couldn't resist
[05:41:23] <eric_unterhausen> it's good for that kind of machine. I like the structure you've made. Trying to think of how I will do something like that
[05:42:27] <Benjamin_Mullin> I got tired of trying to hold the pen by hand
[05:42:48] <Benjamin_Mullin> the 2x4 and plywood frame took an hour tops to construct
[05:47:42] <JustinXJS2> heh, thanks killar
[05:47:59] <JustinXJS2> what are you going to use the machine for Benjamin_Mullin?
[05:49:35] <Benjamin_Mullin> small pattern making for casting, maybe a few one-off PCBs, misc small parts, ???
[05:49:59] <killar> what CAM software is everyone using?
[05:50:04] <killar> anything directly for linux?
[05:50:50] <JustinXJS2> I am a newbie, nothing yet here
[05:51:29] <Benjamin_Mullin> CAM is what I don't really have direct access to at the moment
[05:51:50] <LawrenceG> vcarve by vectrix runs under wine... kind of a combo cad/cam, but seems very nice, but a little pricey for hobby use
[05:52:00] <killar> i used QCad with a few custom programs i've written, nothign special, standard 2.5D cam stuff...
[05:52:22] <killar> yea, i really don't want wine stuff, i'd liek native
[05:52:57] <killar> sheetcam has a new linux version, but their package i've yet to get to run under ubuntu, or 3 other distributions, no idea what they packaged theirs in
[05:53:16] <killar> synergy's latest version has 3d visuallization issues in gentoo
[05:53:56] <killar> just wish there was some decent solution
[05:54:59] <JustinXJS2> what about pycam?
[05:58:35] <killar> never used it
[05:58:38] <killar> looking into it now
[05:59:18] <JustinXJS2> I have not used it either, I think it is fairly new
[07:46:58] <alex_joni> killar: latest sheetcam has a deb iirc
[08:13:09] <toastydeath> anyone with a mech e background, what methods do i have to calculate the deflection/beam stress of a beam with a heterogenous cross section down the length
[08:13:47] <toastydeath> with holes in it or notches cut out
[09:08:46] <archivist> I dive at one particular book for formulas, Roark's formulas for stress and strains, W. Young
[09:09:30] <archivist> but with notches and holes use finite element analysis if you can
[09:10:17] <archivist> isbn 0-07-072541-1
[10:26:55] <piasdom> g'mornin all
[10:28:17] <piasdom> my terminal syas i have an existing realtime session. but i closed the terminal already, how do i access the latency test results ?
[10:33:17] <piasdom> i reread what i wrote(and ya'll replies) and i thought jeplers' new ini file he gave me yesterday helped..guess i was wrong thinking it helped my latency.....
[11:40:56] <The_Ball> got my 5i23 board and 7i33t board today, man that was quick postage
[12:01:55] <piasdom> can someone tell me if my latency test results from last night are in this command http://paste.org/index.php?id=4941
[12:02:55] <The_Ball> I believe they are yes
[12:03:13] <The_Ball> I think it's the Max-Time values
[12:07:33] <piasdom> but i don't know ANY of what i'm looking at...can you tell me my max ?
[12:12:49] <piasdom> The_Ball: thanks for your help....i'll ask again later
[12:13:46] <The_Ball> piasdom, I think you need to have a safety margin over thouse max values, but someone will be around to tell you for sure.
[12:14:24] <piasdom> The_Ball: appreciate your help
[12:32:13] <JymmmEMC> heh heh
[12:33:13] <archivist> sysops rule
[12:34:00] <JymmmEMC> More like CHANSERV rules in this case =)
[12:34:25] <archivist> they are supposed to get permission
[12:34:43] <JymmmEMC> He joined/and was klined in the same second
[12:34:52] <archivist> we have our own logger so dont need that one
[12:34:53] <JymmmEMC> who is?
[12:35:13] <archivist> IRseekbot
[12:35:43] <JymmmEMC> It's IRC, lots of logging everywhere
[12:36:03] <archivist> they got thrown off completely unless they got perms from channels
[12:36:51] <JymmmEMC> That's unusual. probably was more than just that.
[12:37:37] <archivist> I think freenode changed policy as well
[12:37:52] <JymmmEMC> nfc, if that's the case.
[13:32:39] <The_Ball> arg, bad video card memory stopping me from trying out the 5i23
[13:41:19] <jepler_> jepler_ has changed the topic to: Welcome! EMC (Enhanced Machine Controller) is a linux-based opensource CNC control. | Latest release: EMC 2.2.8 | http://www.linuxcnc.org | http://wiki.linuxcnc.org | Channel logged by logger_emc
[13:42:02] <jepler_> jepler_ is now known as jepler
[14:20:47] <piasdom> can someone tell me if my latency test results from last night are in this command http://paste.org/index.php?id=4941
[14:25:24] <BigJohnT> didn't you just ask that?
[14:25:37] <BigJohnT> or is there an echo in here :)
[14:26:06] <piasdom> about 2 hours ago...but no one was here to answer
[14:26:43] <BigJohnT> ok I thought someone answered... so why don't you run "latency-test" from the command promp?
[14:27:12] <piasdom> i did...but this morning i closed the terminal before i saw the results
[14:27:30] <piasdom> and it says halrun is still running
[14:27:30] <BigJohnT> LOL ok
[14:27:50] <jepler> #
[14:27:52] <jepler> 4 s32 OUT 60979 timedelta.0.jitter ==> bj
[14:28:03] <piasdom> i don't want to lose the results from last night
[14:28:07] <jepler> the base-thread jitter is 60000
[14:28:12] <piasdom> jepler: thanks
[14:28:12] <jepler> more or less
[14:28:20] <BigJohnT> cool
[14:28:42] <BigJohnT> piasdom: did you do a dmesg to get that?
[14:28:57] <BigJohnT> ok nevermind I see
[14:29:19] <piasdom> no, did halcdm shop
[14:29:28] <piasdom> show
[14:29:54] <piasdom> don't know about dmesg
[14:30:04] <BigJohnT> me neither
[14:32:13] <cradek_> cradek_ is now known as cradek
[14:49:30] <alex_joni> piasdom: maybe try "killall halrun"
[14:49:36] <alex_joni> that should stop it
[14:49:48] <alex_joni> or "halcmd stop" "halcmd unload all"
[14:52:07] <piasdom> alex_joni: i stopped it, but before i did i wanted to know the results
[15:30:11] <piasdom> have a good latency test results for emc2 ? is there a range ?
[15:32:31] <jepler> http://linuxcnc.org/docs/html/motion_tweaking_steppers.html
[15:32:40] <jepler> "So, what do the results mean? If your "ovl max" number is less than about 15-20 microseconds (15000-20000 nanoseconds), the computer should give very nice results with software stepping. If the max latency is more like 30-50 microseconds, you can still get good results, but your maximum step rate might be a little dissapointing, especially if you use microstepping or have very fine pitch leadscrews. If the numbers are 100uS or more (100,000 nanoseconds)
[15:32:54] <jepler> "ovl max" means approximately the same as "base-thread jitter"
[15:33:10] <jepler> so 60000 is in the "disappointing, but maybe usable depending on your step rate" range
[15:35:53] <piasdom> jepler: thanks..i'm installing a 512 video card today/tomorrow and want to see what i get then
[15:36:50] <piasdom> i just can't do much with the computer while running emc right now
[17:10:23] <eric_unterhausen> jepler: is the requirement for latency lower with servos? Has anyone ever explored using a machine with bad numbers with servos?
[17:10:58] <jepler> eric_unterhausen: yes, if you have no "base-thread" then the latency can be higher
[17:11:42] <eric_unterhausen> I wasn't thinking that I didn't need the base thread
[17:12:50] <eric_unterhausen> my system has pretty consistent latency, so I should probably just use it as-is
[17:13:44] <SWPadnos> what do you expect to do in the base thread?
[17:14:12] <jepler> unless you e.g., run out of encoder counters on a mesa board, there's often no reason for a machine with a smart I/O board to have a base-thread
[17:14:47] <eric_unterhausen> SWPadnos: wasn't thinking, the little message box told me something so I was like "ooh noez"
[17:15:09] <SWPadnos> ok
[17:16:05] <jepler> in 2.2, all or almost all of the sample configurations have a base-thread, even though it's unused in almost everything besides the software stepgen setups
[17:16:22] <eric_unterhausen> so I should delete?
[17:16:23] <SWPadnos> you removed those from TRUNK, right?
[17:16:32] <jepler> I recently changed that in TRUNK
[17:16:35] <SWPadnos> (I saw about 100 commit messages :) )
[17:16:48] <jepler> yeah it creates a lot of commit messages since they're all in different dirs
[17:16:52] <SWPadnos> yep
[17:17:14] <eric_unterhausen> so I just have to upgrade
[17:17:56] <jepler> you can get rid of base-thread by setting BASE_PERIOD=0 in your inifile
[17:18:31] <jepler> upgrade alone won't change anything, because upgrading doesn't edit inifiles or halfiles in your home directory (just upgrades the sample configurations)
[17:20:26] <eric_unterhausen> I need to wipe some of that anyway
[17:21:05] <eric_unterhausen> so I can go from the 2.2 live cd install to 8.04 live cd install without reformat/reinstall?
[17:21:19] <jepler> er, what?
[17:21:32] <SWPadnos> yeah
[17:21:35] <SWPadnos> er, what?
[17:21:39] <jepler> if you have a working 6.06 machine don't screw with it
[17:22:01] <eric_unterhausen> my motivation is the x server
[17:22:48] <eric_unterhausen> it doesn't like my monitor and I have to futz with it sometimes when I lose power
[17:23:04] <jepler> I'm not a big fan of automated OS upgrades, and specifically in the case of a system with emc2 installed I don't think you're left with a working system without additional fiddling around
[17:23:19] <eric_unterhausen> so I should wipe/format/reinstall
[17:23:38] <jepler> you should install to a new partition / new hard drive as possible
[17:23:55] <jepler> to avoid throwing away a working system that has a few annoyances in case you run into trouble
[17:24:01] <jepler> s/as possible/if possible/
[17:24:26] <eric_unterhausen> ok, you just said something that made me wonder if I could just click a button and be updated
[17:24:42] <eric_unterhausen> I'm a long time fedora user, no possibility of any such thing
[17:24:53] <eric_unterhausen> at least there is pain
[17:25:22] <jepler> 2.2.x -> 2.2.y upgrades are easy with net connection, nothing else is terribly easy
[17:28:06] <jepler> (if you run an unmodified ubuntu, they make it pretty painless to upgrade OS versions; but when it comes to systems with third-party software that includes a custom kernel, it doesn't work quite as nicely ..)
[17:28:23] <eric_unterhausen> I see
[17:31:03] <eric_unterhausen> I tried updating fedora on my old computer, and I had two sound systems running sometimes
[17:31:16] <eric_unterhausen> that was strange
[17:31:53] <eric_unterhausen> I was too lazy to learn how to fix it so I got a new computer
[17:32:03] <eric_unterhausen> I also buy new cars when I need an oil change
[17:32:43] <jepler> I tend to stick to one OS version for the lifetime of a machine, which is fine as long as security updates last as long as the machine does
[17:33:15] <eric_unterhausen> fedora has a problem with that
[17:33:30] <jepler> yeah
[17:34:12] <jepler> that's one reason I like LTS so much, I was still smarting from the Fedora Legacy project closing up shop
[17:34:35] <eric_unterhausen> I probably should have installed the lts on this machine
[17:35:42] <jepler> if ubuntu lts hadn't come along, I dunno whether I'd have gone to debian or to one of the repackaged RHELs like centos
[17:35:46] <eric_unterhausen> I guess I shouldn't eat these peanut butter crackers
[17:36:06] <jepler> are you young and healthy?
[17:36:14] <jepler> if so, I say "go for it"
[17:36:17] <eric_unterhausen> 6 known dead so far, I don't even like peanut butter enough to die over it
[17:37:02] <eric_unterhausen> oh, no, not Lil' Debbie
[18:23:07] <eric_unterhausen> too bad there isn't a setting where you can only see the quit/entered messages that are interesting
[18:24:23] <archivist> there is one quit that has a time to the end of the world and he does not have a constant in it
[18:24:45] <eric_unterhausen> that's interesting
[18:26:22] <archivist> some a really funny, others break channel rules, heard of them getting banned
[18:27:35] <eric_unterhausen> i only quit when my router dies
[18:27:50] <eric_unterhausen> does the quit message get transmitted when you join?
[18:28:35] <archivist> no
[18:28:53] <eric_unterhausen> I should have realized that
[18:30:12] <eric_unterhausen> I fried my left index finger last night taking the top off of something I heated in the microwave
[18:30:37] <eric_unterhausen> got a blister
[18:32:09] <eric_unterhausen> how do archivists treat an injury like that
[18:32:10] <eric_unterhausen> ?
[18:32:56] <skunkworks> honey
[18:34:45] <SWPadnos> thanks, sweetie
[18:34:51] <skunkworks> ;)
[18:35:45] <SWPadnos> now where is FedEx with those cameras
[18:36:16] <eric_unterhausen> I'm getting a high speed camera at work
[18:36:22] <skunkworks> heh - Just got my digikey order.. I like being close to one of their distributiers.
[18:36:32] <eric_unterhausen> next I have to get some bows and arrows and some water balloons
[18:36:32] <skunkworks> usually 2 days for ups ground
[18:36:35] <SWPadnos> high res also?
[18:36:37] <SWPadnos> heh
[18:36:46] <eric_unterhausen> yeah, fairly high res
[18:36:52] <SWPadnos> you need a WaveSensor. too bad they aren't made any more
[18:37:14] <SWPadnos> (it's a sound/beam trigger with programmable delay time)
[18:37:24] <eric_unterhausen> this has a trigger input
[18:37:36] <eric_unterhausen> it probably has programmable delay time too
[18:37:42] <SWPadnos> do you know the lag time?
[18:37:54] <eric_unterhausen> it's always taking images
[18:38:03] <eric_unterhausen> so the trigger is just for saving
[18:38:07] <SWPadnos> video mode, then a high-res capture
[18:38:08] <SWPadnos> oh,
[18:38:21] <eric_unterhausen> it does triggering off of the images too
[18:38:28] <SWPadnos> so it's kind of like "save that last frame"
[18:38:31] <eric_unterhausen> I think it's always high res
[18:38:40] <SWPadnos> what speed and resolution?
[18:38:42] <eric_unterhausen> like a decent oscope
[18:39:28] <eric_unterhausen> it is up there in speed, thousands of frames per second
[18:39:33] <eric_unterhausen> light becomes an issue
[18:39:59] <SWPadnos> yep
[18:40:02] <SWPadnos> as does readout speed
[18:40:19] <eric_unterhausen> right, it could use more memory
[18:40:26] <eric_unterhausen> 4gb
[18:40:45] <SWPadnos> unless it's a high end CCD with many outputs (and a bunch of high speed A/Ds attached), it's unlikely to be able to capture high res frames at that speed
[18:40:48] <eric_unterhausen> http://www.olympus-ims.com/en/ispeed-3/
[18:40:53] <SWPadnos> no, I'm talking about getting the data off the sensor
[18:41:10] <eric_unterhausen> 1280 x 1024 sensor
[18:41:18] <eric_unterhausen> depends on how much of the sensor you are using
[18:41:22] <SWPadnos> ah. that's not high res in my world :)
[18:41:36] <eric_unterhausen> you have high speed that is bigger than that?
[18:41:50] <SWPadnos> right, that was my point - you can't get high res and high speed readout at that rate
[18:41:52] <eric_unterhausen> pretty sure there is
[18:41:56] <SWPadnos> no, nothing that fast
[18:42:09] <eric_unterhausen> I mean I'm pretty sure there are larger sensors
[18:42:24] <SWPadnos> 2000 FPS at full res (that site says)
[18:42:31] <SWPadnos> up to 150k FPS max
[18:42:38] <eric_unterhausen> for one pixel
[18:42:42] <SWPadnos> heh
[18:42:54] <eric_unterhausen> nice thing about it is you can cut a rectangle out and capture that
[18:42:55] <SWPadnos> you might get QVGA at that rate
[18:42:58] <SWPadnos> yep
[18:43:00] <SWPadnos> ROI
[18:43:05] <SWPadnos> and presumably binning also
[18:43:09] <eric_unterhausen> lots of ROI are not flexible
[18:43:18] <SWPadnos> true, in CCds
[18:43:31] <eric_unterhausen> I have a camera that has a totally worthless ROI
[18:43:50] <SWPadnos> so the ROI isn't very good either :)
[18:43:53] <eric_unterhausen> you can only make a wide rectangle of any height
[18:44:03] <SWPadnos> the ROI of the ROI is bad
[18:44:08] <SWPadnos> oh, that's pretty useless
[18:44:42] <SWPadnos> what's the cost of that camera? (roughly)
[18:44:43] <eric_unterhausen> I can see some uses for it, but I was disappointed
[18:44:47] <eric_unterhausen> 60k
[18:45:01] <SWPadnos> ok, I figured ity would be in the mid-10k's
[18:45:15] <eric_unterhausen> you can get a very useful one under 30k
[18:45:30] <SWPadnos> sure, Redlake has some nice ones
[18:45:33] <eric_unterhausen> not sure about 10k
[18:45:36] <SWPadnos> and others I don't remember the name of
[18:45:47] <SWPadnos> I meant middle-range units multiplied by $10k :)
[18:46:08] <eric_unterhausen> like 50k+/-?
[18:46:20] <SWPadnos> yeah, 30k minimum, and upwards
[18:46:34] <SWPadnos> but that was just my ballpark estimate
[18:46:47] <eric_unterhausen> I got broken in to that price by talking to flir
[18:46:48] <SWPadnos> (I'm not thinking of any specific camera that's like that, at that cost)
[18:46:53] <SWPadnos> heh
[18:47:07] <SWPadnos> they make the stabilized helicopter mounts also, right?
[18:47:12] <eric_unterhausen> they sell really nice cameras for under $10k, the one I wanted was over $100k
[18:47:22] <eric_unterhausen> yeah for car chases
[18:47:26] <SWPadnos> yep
[18:47:47] <eric_unterhausen> the ads said they were affordable, so I wasn't ready for "you can't afford it"
[18:47:59] <eric_unterhausen> it was one of those "if you have to ask..."
[18:48:04] <SWPadnos> yeah, affordable for government and poorly managed companies
[18:48:11] <eric_unterhausen> they tried to sell me the cheap 70k version
[18:48:26] <SWPadnos> "cheap"
[18:48:27] <eric_unterhausen> they make a $5k camera for BMW that is excellent
[18:48:52] <eric_unterhausen> that would be fun
[18:48:53] <SWPadnos> I was looking at getting 60-90 cameras when I was looking, so I quickly started to ignore anything about around $3K
[18:48:56] <archivist> hey I work for a poorly managed company, and there aint no cash here
[18:48:59] <SWPadnos> s/about/above/
[18:49:19] <SWPadnos> archivist, bummer. did they at least get some nice toys with the cash?
[18:49:30] <archivist> what cash
[18:49:46] <archivist> useless boss here
[18:49:52] <eric_unterhausen> I think poorly managed companies don't always start with money
[18:50:09] <archivist> this one is nose diving
[18:50:17] <eric_unterhausen> the trick is to have a good liar in management to attract investors
[18:50:33] <archivist> a banker!
[18:50:49] <SWPadnos> or lawyer
[18:51:04] <eric_unterhausen> I shouldn't get into my theory of academic success
[18:51:38] <eric_unterhausen> but to do well in most things, you either have to like the smell of your own farts, or your farts have to smell pretty good
[18:51:55] <SWPadnos> academic and business (and life) success don't have too much to do with each other, I've found
[18:52:12] <eric_unterhausen> depends
[18:52:32] <eric_unterhausen> everything is sales, someone is doing it or you don't eat
[18:53:22] <eric_unterhausen> it's just in academics, everyone is in sales
[18:53:44] <SWPadnos> no, I mean that academic success is largely irrelevant to business (financial) success or to happiness (life success)
[18:54:23] <eric_unterhausen> in case I was confusing anyone, teaching is not involved in academic success
[18:54:32] <eric_unterhausen> except from personal fulfillment
[18:54:39] <archivist> you cant get by the dumb HT recruiters without the right papers
[18:54:40] <SWPadnos> neither is learning
[18:54:46] <archivist> HT HR
[18:54:52] <eric_unterhausen> learning is usually required
[18:55:18] <eric_unterhausen> I've been learning a lot recently, it makes some of the nonsense worthwhile
[18:55:55] <archivist> worst customers Ive ever had were usually uni profs
[18:56:16] <eric_unterhausen> there is a reason why there are lots of businesses around universities, some academics are fairly good at it
[18:56:25] <eric_unterhausen> archivist: I can imagine
[18:56:27] <SWPadnos> to be academically successful, you need (in my experience) to remember some things that were covered in the course, and (more importantly) figure out what the teacher/professor wants to see in your assignments
[18:57:09] <SWPadnos> actually knowing what was taught at a level where you can use it later on was only rarely related to the grades I got
[18:57:16] <eric_unterhausen> I'm talking about someone working in academics, not students
[18:57:22] <SWPadnos> oh, as a job
[18:57:27] <SWPadnos> well in that case it's different :)
[18:57:28] <eric_unterhausen> yeah
[18:57:57] <eric_unterhausen> academic success as a student is overblown
[18:58:09] <archivist> I wish accademics were forced to get a real job for 10-15 years before they did any teaching
[18:58:19] <eric_unterhausen> funny you should mention that
[18:58:28] <SWPadnos> heh
[18:58:34] <eric_unterhausen> in my experience, it's a real problem
[18:58:39] <eric_unterhausen> to do that
[18:58:51] <eric_unterhausen> you will never get anywhere
[18:59:04] <SWPadnos> schools would have to (a) pay enough and (b) be interesting enough for people to leave the "real world"
[18:59:11] <eric_unterhausen> best to have an entire career, retire early, and get a job as a professor
[19:00:11] <eric_unterhausen> even working a couple of years is a problem
[19:00:34] <eric_unterhausen> I'm going off to uni now to be a problem, see ya'll
[19:01:21] <archivist> I remember a teacher moaning when he did a real job for a few years, full days work, no free paper pens etc....
[19:02:31] <dmess> LOL... i had 1 prof come from the real world... he was NO teacher thats for sure.. but he did know his crap boy if you could figure out what questions to ask him.
[19:03:32] <dmess> told stories of launching rockets back in the early '70s stuff he desighned..
[19:07:03] <motioncontrol> good evening.i have one question please.i want read in classicladder the tool number preload.The pin in emc is ok, but in classicladder i thing necessity create one integer variable.i search one example the plc mazak retrofit , but i don't see it.thanks for all
[19:32:31] <jepler> I think the demo_mazak configuration was made before it was possible to connect any signal type but "bit" to classicladder.
[19:33:18] <jepler> 32772 s32 IN 0 classicladder.0.s32in-00
[19:33:21] <jepler> 32772 s32 OUT 0 classicladder.0.s32out-00
[19:33:29] <jepler> however, when you run the demo-sim-cl configuration you will see these HAL pins
[19:33:49] <jepler> they can be used to connect pins of type 's32' to classicladder
[19:34:29] <jepler> hmmmmmm
[19:34:44] <jepler> I have to take that back: it looks like inputs and outputs of type 's32' are not in version 2.2.x, they are only in the development version.
[19:34:53] <motioncontrol> ok thank jepler
[19:36:15] <motioncontrol> i read the classicladder doc and i have abilitation the s32in es32out with loadrt classicladder_rt numS32in=5 numS32out=5
[19:36:28] <motioncontrol> is ok?
[19:36:40] <cradek> I'm sure 2.2 has s32 in CL but it requires those command line arguments
[19:36:44] <jepler> if you read the documentation, then you know as much as me
[19:37:53] <motioncontrol> i don't understand, i have emc2.2.8 is necessity thw add loadrt line ?
[19:40:10] <motioncontrol> ok i understan excuse for my englis , sometime i use the vocabolary for traslate.
[19:42:26] <motioncontrol> excuse for another question.in another cnc when i write in partprogtam t1 m6 the machine stop at the line t1 m6 and i reset m fuction to plc.i start t1 m6 in emc and the fuction reset immediatley, because?
[19:45:19] <dmess> is this to prestage t1 or to put t1 into the spindle??
[19:47:08] <dmess> if t1 isnt staged and you want IT in the spindle it will have to fulfill all the prestage logic before the mcode will act and complete... as best i know anyway
[19:51:40] <SWPadnos> in EMC2, the T1 and M6 words are separate. You can issue many Txx lines before issuing M6
[19:51:45] <motioncontrol> i dont't understan.example in mdi . in splindle i don't have tool.write t1 m6 the pin emc relative at tool prepare set true and immediatley at false , and the program start at the next line in partprogram
[19:52:14] <SWPadnos> all Txx does is set the tool-number (or whatever it's called) pin to the number you specify
[19:53:28] <SWPadnos> M6 is what actually starts the process, and in EMC2, that process is that motion is stopped, the tool-prepare output is set high, and things only resume after the tool-prepared pin goes high
[19:54:13] <motioncontrol> excuse. the problem is this . i start the tool change t1 m6 , read in classicladder the pin tool prepare and the plc logic change the tool in splindle.normaly i reset the m function and the nc read the next line in partprogram
[19:54:17] <SWPadnos> so whatever you do in HAL to make the tool get changed needs to start when the tool-prepare output goes high and it needs to signal that it's done by setting the tool-prepared input high
[19:54:39] <SWPadnos> I don't understand "reset the M function"
[19:56:12] <motioncontrol> ok in siemens 840 d and fanuc exist 2 type the m fuction.first type is dinamic and the second type is static
[19:56:18] <cradek> reset = report back that it is completed?
[19:56:22] <motioncontrol> one moment i write
[19:56:55] <motioncontrol> yes reset m fuction= m6 is complete
[19:57:59] <motioncontrol> when the nc read m6 the motion is stop , when i reset the m6 the motion start
[19:58:45] <motioncontrol> this in my emc2.2.8 non is. the m6 reset fast and itself
[19:58:54] <cradek> did you read what SWPadnos said?
[20:01:09] <motioncontrol> the static fuction start nc and write in plc the memori x, the nc stop at line the m fuction call and the start only i reset the memory
[20:02:02] <cradek> did you read what SWPadnos said?
[20:02:09] <motioncontrol> the dinamic the nc read m fuction and for one cycle the plc set the memori x , the nc not stop and the nc read the next line
[20:02:22] <cradek> he answered your question
[20:02:38] <SWPadnos> in EMC2, the tool number will be there for the entire tool cahnge, so you don't actually need to save it in the PLC
[20:02:39] <motioncontrol> in emc because the m6 not stop the nc read next line?
[20:06:15] <motioncontrol> excuse : i have g0 x10 y10 , after t6 m6, g1 x10 f10.the nc when read t6 m6 start in plc the tool change, when tool change finish whith plc , the plc reset m6 fuction and nc read g1 x10 f10
[20:07:47] <jepler> in emc during an m6, iocontrol.0.tool-change is asserted. Execution of the part program does not proceed until the toolchanger asserts iocontrol.0.tool-changed.
[20:08:02] <jepler> "asserted": value of pin changes from FALSE to TRUE
[20:08:41] <motioncontrol> ok thanks jepler i understand
[20:08:57] <alex_joni> same for Txx (tool prepare)
[20:09:27] <motioncontrol> excuse for my question , but i wand use emc for retrofit my cincinnati machine and i not want acquire the siemens 840d
[20:09:34] <jepler> in certain sample configurations there is a signal connecting tool-change with tool-changed. This makes the tool change operation finish immediately. You would remove this signal in a machine that actually has tools to change.
[20:13:31] <motioncontrol> where i disable the pin connection
[20:13:45] <jepler> by editing your hal file to remove the "net" or "link" command that creates it
[20:14:21] <motioncontrol> ok i understand i remember this line in hal file .very more thenks jepler
[20:14:59] <motioncontrol> thenks at all
[20:15:14] <motioncontrol> excuse thanks in english
[20:15:57] <motioncontrol> ok shutdown the pc. at the next question CIAO
[20:50:50] <jensor> I need some clarification on the sense of the limit switches. For example, is the switch that gets actuated when the table moves to the right (X negative), defined as the negative limit switch
[20:53:32] <SWPadnos> yes, the switch that gets activated when the cutter moves to the minus limit is the minus limit switch
[20:53:33] <SWPadnos> whether that means that the table moves right or the gantry moves left or whatever
[20:53:39] <jensor> thanks
[20:53:59] <SWPadnos> sure
[21:15:36] <dmess> i fell asleep for a few hrs last nite after supper... woke up about 9:30ish and the kids had built a snowboard park in my front yard ... exiting onto the street of coarse...
[21:17:47] <dmess> has a 2 M take off mound right beside my van. and we have 4-5m verticle to the road level... but only 15m to lose the verticle
[21:18:57] <dmess> sorry that was meant for another channel