Back
[00:20:36] <AitalMAC_2> http://linuxcnc.org/docs/2.5/html/examples/spindle.html
[00:22:24] <AitalMAC_2> i need help with the spindle soft start
[00:26:45] <AitalMAC> I have connect spindle to PWMgen and the scale is working but it's on the side
[00:26:51] <AitalMAC> i would like to connect them
[00:27:19] <AitalMAC> I'm using 2 PC sorry for the confusion
[00:27:29] <aitek> hello
[00:28:24] <AitalMAC> Right now when i set a speed i see the scale doing his acceleration and decelleration job
[00:28:29] <AitalMAC> in the hal meter
[00:29:01] <AitalMAC> but my spindle is still jumping like hell because the pwm are still outputted like before
[00:29:11] <AitalMAC> Any help?
[00:39:37] <AitalMAC> sorry i got it
[01:50:34] <DJ9DJ> moin
[05:14:40] <Rogue> hi
[05:35:03] <DJ9DJ> hi
[06:05:10] <yoyo_> hi
[06:10:09] <DJ9DJ> hi
[06:11:09] <yoyo_> is there a gcode to touch a over a circuit of object ?
[06:11:56] <archivist> touch a over a circuit? do you mean probe
[06:12:16] <yoyo_> I have object flat object ~h2mm
[06:12:44] <yoyo_> and I would like to probe it over a circuit
[06:13:18] <yoyo_> something like 3d touch scanning but not 3d only 2d over circuit
[06:14:12] <archivist> to do what with the information?
[06:14:54] <yoyo_> have circuit probe coordinates
[06:14:58] <awallin_> http://www.youtube.com/watch?v=_Qr8tZXGXZU
[06:15:39] <archivist> not sure what you mean by circuit in this context
[06:16:19] <yoyo_> points on the circumference of the object
[06:19:44] <yoyo_> scan edges of object not sufrace
[06:20:43] <archivist> I know what you want now, just not aware of a simple method at the moment
[06:21:42] <yoyo_> I can write it by my self but first I wanted to ask
[06:21:50] <cncbasher> could gridprobe.ngc be of use , and turn off the z axis probing perhaps
[06:24:14] <archivist> I think it needs a different algorithm thing of an M or an O part to be scanned
[06:24:19] <archivist> think
[06:26:13] <archivist> or better still a maze
[06:26:18] <cncbasher> simple until you think how the hell do i do it
[06:26:36] <yoyo_> yyy no drama :P
[06:27:02] <yoyo_> if there is no ready gcode I will write it and share
[06:27:38] <archivist> use the standard 3d and slice maybe
[06:27:41] <cncbasher> i'm not aware of one
[06:27:52] <yoyo_> in the meantime new release of mgcodegenerator on wiki :)
[06:29:06] <cncbasher> i have always wanted to scan /probe items , i'll have to give it a go now
[06:39:58] <Loetmichel> *buuurrrrp*... my wife goes on a doc-controlled diet tomorrow... so she bought a "last meal" at McD... and supplied me with the same ;-)
[06:40:08] * Loetmichel is a bit short of bursting ;-)
[06:40:28] <Loetmichel> good monrning/ good day btw.
[06:49:00] <cncbasher> it will be your last meal too , did she not tell you ..
[06:55:26] <DJ9DJ> hi Loetmichel :)
[06:55:35] <DJ9DJ> hehe
[06:58:32] <Loetmichel> cncbasher: she did ;-)
[06:58:49] <DJ9DJ> hmm, until easter?
[06:59:16] <DJ9DJ> lenten season? (says dict.leo.org *G*)
[06:59:26] <Loetmichel> i.e. she told me: "as i have to follow a strict eating code you will have to cook for yourseld the next half year" ;-)
[07:00:08] <DJ9DJ> haha
[07:43:59] <awallin_> more cam experiments:
http://www.youtube.com/watch?v=lfIU_gv0iB8
[07:46:21] <syyl> that is cool :)
[08:01:54] <Tom_itx> cool but alot of air cuts
[08:02:16] <pfred1> cuts air like hot butter!
[08:02:33] <awallin_> yeah..
[08:03:19] <awallin_> now it only keeps track of what has been cut "behind" the advancing cut-arc
[08:03:22] <archivist> are the outer corners supposed to be the same?
[08:04:00] <awallin_> when we come to the end of a cycle in the MA there is cleared material "in front" of the advancing cut-arc, which the algorithm isn't aware of. thus a lot of air-cuts
[08:04:11] <pfred1> archivist define "the same?"
[08:04:24] <awallin_> yes the corners should be the same, but the resolution of the simulation isn't that great. or there's a bug - I don't know
[08:05:17] <pfred1> attributes are not bugs when they can be defined as features
[08:42:51] <joe9> I want to make 8cm x 8cm piece with drilled holes. Is it wise to cut it to size and then drill it or cut it to 8.5cm x 8.5cm, drill and then cut to size?
[08:42:57] <joe9> which is the recommende way?
[08:45:43] <pfred1> usually you size the work then work on it but of i need a piece to fit with another piece I go oversized then trim to fit
[08:47:34] <pfred1> makes stuff come almost looking like I know what I'm doing sometimes when I do that
[08:48:11] <pfred1> the trim to fit technique
[08:48:29] <joe9> pfred1: thanks a lot. To cut the piece to size I am using the heavy duty paper trimmer. Is that a good idea?
[08:48:36] <joe9> or should I be filing it to get it to size?
[08:48:44] <pfred1> sheet metal?
[08:48:46] <joe9> it is 1/32" inch thick
[08:48:49] <joe9> no, pcb
[08:49:09] <pfred1> I came up with a way to cut PCB with a knife and a straight edge I prefer
[08:49:28] <pfred1> what you do is score the board over and over and angle the blade different digging a channel
[08:49:36] <pfred1> then it just snaps
[08:49:50] <pfred1> comes out really straight and you can sand the edge a little but it is clean
[08:50:21] <pfred1> if you're doing it right little wisps and curls of the board should come up out of the channel you're making
[08:50:54] <pfred1> 10 or so passes is about all it takes
[08:51:18] <joe9> ok, thanks.
[08:51:43] <pfred1> I used to use all kinds of cockamamy ways to cut PCB I like the knife and straight edge way better than them now
[08:52:05] <joe9> why? precision?
[08:52:21] <pfred1> yeah it doesn't beat on tools and I get precise clean cuts
[08:52:30] <pfred1> PCB that fiberglass really tears up blades
[08:52:46] <pfred1> but a knife blade I don't care i can resharpen it easily enough
[08:52:51] <joe9> oh, ok.
[08:53:19] <pfred1> I've used circular saws bandsaws hacksaws
[08:53:28] <pfred1> but knife and straight edge way for me to go now
[08:53:48] <pfred1> cut on the copper side
[08:54:03] <pfred1> unless it is double soded then you're going to have to transfer your score t othe other side
[08:54:29] <pfred1> which with a knife isn't hard to do
[08:55:19] <pfred1> but every time you rip the knife down the straight edge angle it a little different side to side perpendicular to the straight edge kind of
[08:55:32] <pfred1> so you dig a channel
[08:56:08] <pfred1> like I said if you're doing it right slivers of the board should emerge from your channel as you go along
[08:56:43] <joe9> ok, thanks a lot.
[08:56:47] <pfred1> I use a utility knife
[08:57:24] <pfred1> steel straight edge of course
[08:58:03] <pfred1> sanding the edge of the board does clean them up
[08:58:46] <pfred1> yesterday I polished the edge of a piece of clear plastic that I cut on a circular saw
[08:58:51] <pfred1> man it came out amazing
[08:59:13] <pfred1> thing is like optical quality
[09:00:02] <Tom_itx> i got a little kit from the local plastic company for stuff like that
[09:00:19] <pfred1> I used turtle wax polishing compound for my final
[09:00:19] <Tom_itx> starts with uber fine sandpaper then goes to paste then wax
[09:00:38] <pfred1> yeah I started out 100 grit 150 220 320 400 600
[09:00:48] <pfred1> then hit it with the polishing compound
[09:00:51] <Tom_itx> this is like 3600 grit or more
[09:01:01] <pfred1> nope it came up after wet sand with 600
[09:01:06] <pfred1> came right up
[09:02:08] <pfred1> now i have some other hunks of plastic i want to do it to
[09:02:39] <pfred1> it is freaky to be able to see through the edge
[09:03:05] <pfred1> I mean I can see clear through it
[09:03:38] <pfred1> stuff I did is about a quarter of an inch thick I'd say
[09:04:34] <pfred1> but i have a bolt in the middle of it I can see the bolt through the side
[09:06:01] <pfred1> it looks like someone flamed the other "good" side I can see the ripples all the way through the piece
[09:13:18] <mikegg> is it possible to use more than one fpga with hostmot2? I.e. a 5i20 and 7i43
[09:13:58] <pfred1> you know what they say anything is possible the real questions are is is practical or is it likely
[09:14:46] <mikegg> I'll give it a shot I guess
[09:15:37] <pfred1> you ran out of IO with one?
[09:15:55] <mikegg> well not really - kinda
[09:16:11] <pfred1> or you just want to split a time task?
[09:16:52] <pfred1> like spread a workload
[09:17:06] * jthornton sees bubbles
[09:17:26] <mikegg> right now I have two servos on a 7i39, a stepper on the z-axis and +/-10 spindle on a 7i33
[09:17:31] <pfred1> jthornton some say our entire Universe is a bubble
[09:18:00] <mikegg> i'd like to replace the stepper with a servo
[09:18:28] <pfred1> I thought mesa made multi servo boards at least that was the impression I had
[09:18:54] <pfred1> they seem to offer complete kits
[09:19:16] <mikegg> yeah, they do. two servos per board. So if you have a 3-axis machine 2/3 of the 50 pin connectors get used up for servos
[09:19:21] <mikegg> on a 5i20
[09:19:25] <pfred1> ah ha
[09:19:37] <pfred1> that is the best you can do is 2 servos on a board?
[09:19:57] <pfred1> so anyone with more than that many axis is running multiple boards?
[09:20:14] <pcw_home> AFAIK you can have any mix of cards and multiples of each
[09:20:50] <pfred1> pcw_home that strikes me as the most realistic either that or their hardware is artificially limited
[09:21:10] <pfred1> something I'm sure their engineers strive to not do
[09:21:13] <mikegg> I thought that was the case
[09:21:20] <mikegg> heh
[09:21:50] <pcw_home> we will have a SSERIAL equivalent of the 7I39 sometime this year (with fewer FPGA pins used)
[09:21:51] <pfred1> well we'd love to sell you more of our product but unfortunetately it all doesn't work together ...
[09:21:53] <morfic-> pcw_home: if pin 49 cable power makes it to the 7i44, should i expect the 7i70 to have any status LEDs lid even w/o firmware loaded?
[09:22:07] <mikegg> ah hah, I was just going to ask about that
[09:22:39] <pcw_home> morfic: yes the red fault LED should be lit
[09:22:46] <mikegg> pcw_home: the 7i40 as well?
[09:22:54] <pcw_home> Yes
[09:24:17] <mikegg> sweet
[09:24:53] <pcw_home> morfic: forgot to say only if you have field power (12-24v)
[09:25:41] <pcw_home> without field power you got nothing :-)
[09:25:41] <morfic-> thanks, in the 5i22 manual i see cable power is 5V, pull up voltage is 3.3V and so is the bus switch mode, this installed 5i22 has all three jumpers for all 4 connectors in the up/5V position
[09:26:31] <morfic-> pcw_home: "w/o field power it'll never show anything in dmesg either" kind of nothing?
[09:26:38] <ctjctj> I am having a problem where gladevcp never comes ready. What is the standard method of debugging this issue?
[09:27:02] <pcw_home> the 5V from the 5I22/7I44 runs the 422 interface on both the 7I44 and the 7I70, but the 7I70 needs field power for its logic and processor
[09:27:27] <pcw_home> so it wont even bee seen without field power
[09:27:37] <pcw_home> (it being the 7I70)
[09:29:11] <joe9> # Pad[-1.6250mm 1.025mm
[09:29:11] <joe9> # 1.6250mm 1.025mm
[09:29:12] <joe9> # 3.900mm 30.000 mil 3.950mm "12" "12" "square"]
[09:29:21] <joe9> sorry, wrong paste.
[09:30:01] <pfred1> now i have to hear Iron Man
[09:30:11] <joe9> Has anyone tried a "Harry Watts Square Drill Bit"? I am trying to buy one. Cannot seem to find a place selling it in the US.
[09:30:32] <joe9> http://technabob.com/blog/2011/11/15/drill-bit-makes-square-holes/
[09:30:37] <pfred1> pcw_home born in the great magnetic field!
[09:32:02] <mikegg> has he lost his mind?
[09:32:26] <pfred1> ah he was turned to steel in the great magnetic field
[09:33:17] <pfred1> I guess being a legendary rock god isn't as easy as it looks
[09:34:09] <Tom_itx> joe9, use a broach
[09:34:32] <pfred1> Tom_itx I ran a huge one of those once
[09:34:50] <pfred1> they had to tear the roof off the building and place it in with a crane
[09:35:05] <pfred1> it wouldn't fit through any doors
[09:35:23] <Tom_itx> they didn't think that out very well
[09:35:31] <pfred1> how do you figure?
[09:35:37] <joe9> Tom_itx: I just have a drill press.
[09:35:58] <Tom_itx> pfred1, if they had to take down the building to get it in
[09:35:59] <pfred1> this thing had 6 foot long plates
[09:36:14] <pfred1> well the building was already there before they got the machine
[09:37:02] <pfred1> it'd rip a dovetail into a chunk of steel an inch deep without even batting an eyelash in one shot
[09:37:20] <pfred1> just come down da da da da blade after blade
[09:37:29] <joe9> and I do not want a through hole. more like a slot.
[09:38:23] <pfred1> Tom_itx the dopes that owned it when they trained me to operate it they taught me to feel for metal chips in the cam vise before loading a piece
[09:38:39] <pfred1> because if there was a chip in there it was bad news
[09:39:10] <pfred1> meanwhile this machine had one inch oil pipe spraying oil all over the place filling the vise fixture etc
[09:39:29] <pfred1> so missing a chip now and again while it was cycling I mean it happened
[09:39:41] <pfred1> I missed a couple
[09:39:55] <pfred1> piece would go shooting clear across the shop!
[09:40:02] <pfred1> boss would come up shaking his head
[09:40:24] <pfred1> after the second time i started waving a megnet in the fixture before I loaded in a piece
[09:40:35] <pfred1> magnet never missed a metal chip
[09:40:44] <pfred1> 40 years they never thought of it
[09:40:52] <Tom_itx> provided it was ferous metal
[09:40:58] <pfred1> all we did was steel
[09:41:31] <pfred1> was no other kind of stock in the whole place
[09:41:36] <morfic-> pcw_home: thanks, i keep asking for power supplies, even if in the end it would work w/o, it's nice to have one for testing
[09:41:52] <pfred1> all the broach ever did was vise noses
[09:42:03] <pfred1> well noses bodies and the base serrations
[09:42:41] <pfred1> oh yeah that is right the noses were serrated too
[09:42:50] <pfred1> these things
http://www.jstool.com/images/clamp.jpg
[09:43:07] <pfred1> that dovetail is broached
[09:43:32] <pfred1> and on the bigger clamps it is pretty deep
[09:44:13] <pfred1> I should have stolen 100 of those clamps when I was working there
[09:44:27] <pcw_home> morfic: well the final machine needs control power (probably 24VDC)
[09:44:29] <pcw_home> you can make due with a 12VDC wall wart, or an old 19V laptop charger
[09:45:07] <morfic-> pcw_home: the 5i22 W7, W8, W9 jumpers are all up, instead of up/down/down as manual says, is this a new shipping configuration?
[09:45:08] <pfred1> morfic- power supplies are easy to make
[09:45:31] <pfred1> for 5V get a 7805
[09:45:40] <pfred1> they cost about a quarter
[09:45:59] <pcw_home> Yes I think we changed to the safest working config
[09:46:47] <morfic-> pcw_home: yes i have a 19V cut up laptop psu at home i used to use to power my furnace control board, just blamed lack of knowledge to load firmware more than power, so don't have anything close by, i'll have him wire it up Monday then
[09:46:47] <pcw_home> the 7I70/7I71 aren't picky they will work with 10-32VDC
[09:47:19] <pfred1> pcw_home they have an on board regulator?
[09:47:22] <morfic-> i mean, i could cut up a walwart, but my luck means it would turn out to be job altering for me to do so o_O
[09:47:41] <pcw_home> Heck 2 9v batteries would run it for a while
[09:48:24] <pfred1> I hate batteries with electronics
[09:48:31] <morfic-> pcw_home: and there i was thinking only i have such "unprofessional ideas" :P
[09:48:39] <pfred1> everything works good when they're fresh but that never lasts
[09:49:08] <pfred1> what is great about batteries is they're the cleanest power known to man though
[09:49:14] <pcw_home> servomotor and a diode (helper turns motor)
[09:49:22] <pfred1> a battery is real DC
[09:50:07] <pcw_home> what scared me is high voltage/high power batteries, no way to turn them off
[09:50:14] <morfic-> pcw_home: now you make me tell everyone how my brother in law's father used to have to pull the transmission belts if the steam engines in basement went out, so the journeyman could run their lathes...."before the big war"
[09:50:17] <pfred1> once i had about 150 9V batteries and we connected them all up in series
[09:50:43] <pcw_home> Yow!
[09:50:43] <pfred1> you can hook 9V batteries up almost like they're legos
[09:51:04] <pfred1> well when you have that many of them they really don't make the best contacts we found
[09:51:19] <pfred1> maybe it wasn't that many but i had a box load of the suckers
[09:51:29] <morfic-> i never had enough 9V blocks to be a pfred1, but u tried the concept...
[09:51:35] <pfred1> I think the best we managed out of it was like 150 volts?
[09:51:56] <pfred1> because them connectors would wiggle loose from each other they made pretty crap contact
[09:52:09] <morfic-> 15? for 135, 17 for 153?
[09:52:17] <pfred1> they were the cheapy ones not the folded under but kind of the open
[09:52:24] <pcw_home> I m old enough to remember 90V batteries (I used for neon light blinkers)
[09:52:58] <pfred1> still we must have been nuts to have done it
[09:53:17] <pfred1> thinking back I don't think I'd do it today
[09:53:18] <morfic-> pcw_home: i am so ready to finally see everything in dmesg though, you have no idea how much the "they are supported in the development version" "ok" changed to "why did we buy incompatible boards"
[09:54:02] <pfred1> morfic- tail -f /var/log/syslog is helpful too while debugging
[09:54:22] <pfred1> it kind of keeps going automatically
[09:54:44] <pcw_home> Well 2.5 should be released soon (and would have been I suspect absent the name change nonsense)
[09:54:53] <pfred1> you're going to have to be root to do that though
[09:55:12] <pfred1> users don't have access to syslog
[09:56:15] <pcw_home> morfic are you running 2.5 from the buildbot?
[09:56:31] <pfred1> pcw_home someone couldn't run sed on the package to change EMC2 to LinuxCNC?
[09:57:39] <pcw_home> well I think thats kinda what happened but ther are so many little corners for stuff to hide
[09:57:50] <pfred1> sed doesn't care
[09:57:58] <pfred1> it'll find every occurance
[09:58:26] <pfred1> sed s/EMC2/LinuxCNC/g;
[10:00:09] <pfred1> once i had to change forward slashes with sed then you just use different delimiters
[10:00:50] <pfred1> sed s#
http://old#http://new#g;
[10:01:13] <pcw_home> I know a number of things were broken for a while
[10:02:30] <pfred1> grep -ir emc2 *
[10:02:56] <pfred1> that just shows them to you though
[10:03:50] <pfred1> grep -ir emc2 * | wc -l
[10:04:00] <pfred1> tells you how many there are
[10:07:16] <pfred1> UNIXalike OSes were built from the ground up to deal with source trees
[10:07:35] <pfred1> that is their reason to be
[10:08:11] <pfred1> all because one of my neighbors wanted to play SpaceWar in a basement up the street
[10:09:44] <pcw_home> I do know this, as do the developers, but I also know it was not that simple...
[10:10:16] <pfred1> what wasn't that simple?
[10:11:55] <pfred1> "UNIX is basically a simple operating system, but you have to be a genius to
[10:11:57] <pfred1> understand the simplicity
[10:12:10] <pfred1> that is what my neighbor said about it
[10:12:33] * pfred1 grew up in Summit NJ where Dennis Ritchie did
[10:13:41] <pfred1> he moved to Berkley Heights but I always still saw him shopping in the New Providence A&P
[10:15:10] <morfic-> pcw_home: hm2-pyvcp, noone ever made anything like it, or did they? would be nice to wire it up and have a set of buttons and leds and toggle stuff before the actual machine setup is ever written
[10:16:17] <joe9> pfred1: did you know him personally?
[10:16:37] <joe9> Dennis Ritchie, I mean.
[10:18:13] <morfic-> pcw_home: oh, didn't notice the question, not from buildbot, built from git on the actual machine
[10:18:19] <pcw_home> morfic I dont think so but you can always use the show hal configuration for inputs and a halcmd script for outputs
[10:18:56] <morfic-> that's what i did earlier, show hal configuration that is
[10:19:05] <pcw_home> you c
[10:19:27] <morfic-> i think having a UI helps the other guy a lot, he looks lost when i work on terminal
[10:19:29] <pcw_home> an set output with setp
[10:20:28] <pcw_home> setting outputs is easier with halcmd because you can write little scripts
[10:21:56] <pcw_home> also getting a pin/parameter list with show pin, show parameter redirected to a file is helpful when editing a HAL file (so you can copy/paste the names)
[10:22:24] <morfic-> i need to take my time with this, i am trying to shoehorn the SVSS8_16 into an existing ini, i may try to see how far pure halcmd gets me, you refered to the 7i33 manual with out PWM vs PWMgen question, so i'll be headed that way since in the hm2-servo/5i22-small.ini i worked the firmware in, but everything is PWMgen in show hal cmds
[10:22:50] <morfic-> with our*
[10:24:01] <pcw_home> the HM2-servo config should be fine to start with its all correct for what you have, you just need to add the I/O and change the config string
[10:24:35] <pcw_home> are you using Run-In-Place?
[10:25:05] <morfic-> pcw_home: i think once i have actual power and can work around dmesg i'll start to feel more comfy, not seeing the cards immediately makes it easy for the "ok, what of what i don't know is preventing me from showing the pins"
[10:25:11] <morfic-> pcw_home: yes, rip
[10:26:26] <morfic-> i think i have all the docs i need open in either a terminal or browser or acrobat, just needs some paging around to follow steps and verify functionality
[10:26:59] <morfic-> pcw_home: is there a run in place pitfall that makes you ask that? :)
[10:29:49] <morfic-> knowing you agree editing HM2-Servo into a test config alone is great
[10:30:02] <morfic-> btw, did we sufficiently thank you for the quick formware turn around?
[10:30:35] <pcw_home> just make sure that if you do a man command its in a window where you have run the RIP script
[10:31:37] <pcw_home> (so the man matches the RIP version)
[10:33:22] <mazafaka> Do you know of such video cameras which are of a small size and can be fixed to helmet or whatever?
[10:33:36] <pcw_home> firmware that just has common sets of things on different connectors is really easy to make with copy/paste from old to new pinout file
[10:34:01] <pcw_home> GIYF?
[10:35:21] <morfic-> pcw_home: yeah, i may make it so rip-environment gets run for me every time i open a terminal
[10:36:14] <pcw_home> google helmet mount camera (there are some in sunglasses!)
[10:36:57] <morfic-> go to target, they have a ton right between videos and ipads, facing the main aisle "go pro" titled
[10:37:27] <mazafaka> OK
[10:39:37] <morfic-> pcw_home: well, based on how you said is supported in 2.5 vs "will be" or "shouldn't be hard to get working" i saw the 7i70/7i71 as the best solution, i just should have thought to ask you for firmware sooner and things would have been calmer when boss heard about "the untested hardware" i bought
[10:41:24] <morfic-> learned a lot today, one of them will allow me to recognize why linuxcnc doesn't load, if halrun is still open in another screen.....
[10:44:18] <pcw_home> morfic so on the 7I70/7I71 there's LED a red fault LED and a green activity LED
[10:44:20] <pcw_home> when you apply power the red fault LED should be lit, when LinuxCNC establishes communication
[10:44:22] <pcw_home> the green activity LED shoud blink at about 1/1000 the servo thread rate (1Hz at 1 KHz)
[10:44:54] <morfic-> any thoughts on what's easier to do mess with between pyvcp and gladevcp to do quick and dirty test panels, considering my python knowledge is near zero?
[10:45:10] <pcw_home> (and the red fault LED shoud extinguish)
[10:45:40] <morfic-> do the ports on 7I44 light up? or are they just there cause that's just std for how most are sold?
[10:46:21] <pcw_home> I dont know python either so I am not in a position to give advice
[10:46:23] <pcw_home> (the 7I44 LEDS are unconnected)
[10:47:58] <pcw_home> The most common RJ-45 blocks have the LEDs so we use them (the no LED option is custom)
[10:48:00] <morfic-> CR35/CR36?
[10:48:17] <morfic-> for fault and activity?
[10:48:45] <pcw_home> Not sure which is which
[10:49:16] <morfic-> i just mean area, looks like cr8 on side is also led
[10:49:52] <pcw_home> but if you just apply field power you should get the red fault LED
[10:50:43] <morfic-> doesn't matter much, i miss power, after that things should go towards self explanatory again i think
[10:52:35] <pcw_home> without field power there's no communication so no hal pins...
[10:55:05] <morfic-> "field power" is how you call the power source used on the signal side
[10:56:27] <morfic-> too bad he read power on pin49 and stopped there, when i asked him, if we needed power, he only refered to that, seeing the 7I33 light up i guess helped us to run with that misconception
[10:59:24] <morfic-> i'll just refer him to page2 and tell him "that's your 24V" and have him find me field power around my desk
[10:59:58] <morfic-> double touchscreen setup is still on my desk next to my normal monitor and not anywhere close to the lathe :)
[11:04:15] <morfic-> do not buy optical touchscreens for a machine, just saying
[11:08:27] <pcw_home> Yes "Field power" is the isolated power for the "Field I/O"
[11:08:29] <pcw_home> it is the common power for the 7I71s outputs, the threshold reference for the 7I70s inputs
[11:08:31] <pcw_home> and is also used for the isolated side logic and processor
[11:13:23] <morfic-> thanks for the clarification
[11:15:54] <morfic-> looking at page 7 of 7i70 manual it has a pin out again refering to field power and range of 8-32 V with PIN 5 for logic power also having 8-32V, at that point he should have questioned the 5V over pin 49 idea i think
[11:17:43] <morfic-> and repeating the range on page 8 :/
[11:17:49] <morfic-> RTFM goes a long way
[11:18:39] <mazafaka> new type of websites:
http://www.paulrhayes.com/experiments/cube-3d/touch.html
[11:29:52] <pcw_home> Normally pin 5 is connected to field power as well so you dont need to worry about the difference
[11:38:15] <DJ9DJ> mazafaka, i suggest to use a hypercube, because 6 sides might be to few pages ;)
[11:41:29] <tom3p> halgui progress using nodebox
http://edmsolutionsservice.com/pix/halgui-editor-24feb2012.ogv
[11:41:51] <tom3p> ^shift right click brinks up editor for that comp, and that pin
[11:42:00] <tom3p> brings
[11:52:01] <morfic-> pcw_home: and guess who now has a 9V battery
[12:10:47] <tom3p> format of dictionaries for halgui (the palette you can select from, or the template that .hal indexes )
http://edmsolutionsservice.com/pix/haldict03.py
[12:13:12] <mrsun> pcw_home, works nicley =)
[12:13:27] <mrsun> tho with the voltage drop over the thingie i get a maximum of 4.2V out
[12:13:31] <mrsun> at 99% duty
[12:13:42] <mrsun> (using 5V test voltage source)
[12:16:26] <morfic-> pcw_home: lol
[12:17:55] <morfic-> pcw_home: so close
[12:21:48] <mrsun> tho, dont know the pwm frequeny we used but had about 0.1V ripple
[12:22:04] <mrsun> dont know how sensetive these vfds are to ripple in the frequency ref voltage
[12:34:55] <morfic-> pcw_home: 9V not being high enough or did i get the VFIELD == VIN with jumper W1 in "left position" wrong?
[12:35:34] <morfic-> 09:38 < pcw_home> Heck 2 9v batteries would run it for a while <--- because you said 2
[12:37:02] <morfic-> pcw_home: you want 2 in series to get us to the middle of the range?
[12:59:26] <morfic-> alright, got 18V
[13:08:53] <morfic-> never get past red led on, at least i have power to test with monday
[13:10:32] <pcw_home> Yeah the VIN jumper should be in in the left position
[13:16:06] <morfic-> it is, i have 18V between bottom pin VFIELD and top pin GND, green CR8 led on side is on and red fault CR36 is on, firmware loads, should the pins show up as available but unused?
[13:19:12] <morfic-> pcw_home: CONFIG="firmware=hm2/5i22/svss8_16.bit num_encoders=-1 num_pwmgens=-1 num_leds=-1 num_stepgens=0 sserial_port_0=00000000"
[13:23:47] <pcw_home> That should work ( you should get the red out. green blinking)
[13:23:49] <pcw_home> where is the 7I44 connected?
[13:25:39] <morfic-> pcw_home: i need to check layout, to see which P is closest to shield, left if pci connector is down and shield is left
[13:26:28] <pcw_home> 7I44 should connect to P4
[13:27:02] <morfic-> p2 and p3 show in dmesg as encoders
[13:27:42] <morfic-> is that the lvds one, just recall coming across one that was different from the others earlier today
[13:29:13] <morfic-> P3 was the lvds one
[13:29:23] <pcw_home> dont mess with the LVDS option
[13:30:10] <pcw_home> in any case P4 is the first (port 0) SSERIAL interface
[13:30:19] <morfic-> .pin tells me which is where
[13:31:02] <morfic-> sorry i was mostly babbling while finding this info in writing (even if it seems silly over irc, i consider your info verbal)
[13:31:37] <pcw_home> dmesg also gives you the pinout info
[13:32:29] <morfic-> what i looked at is not what it did, but what it expected........
[13:34:57] <mrsun> hmm, i want to go test my spindle card with the actual vfd, i wonder how much i can destroy it if it all goes wrong ;P
[13:36:42] <pcw_home> if you only supply power from its 10.5V to its analaog speed input I doubt you can do much harm...
[13:36:44] <pcw_home> (famous last words)
[13:44:59] <morfic-> sorry, forgot i wasnt on a screen at home
[13:45:23] <morfic-> green blinking led though as reward for the quick shutdown
[13:45:45] <morfic-> and pins to see in show hal configuration
[13:47:19] <morfic-> woot, and a 9V block is enough to get even that far
[13:47:44] <morfic-> couldn't resist to try it, now where i know what to expect in the config tree under pins
[13:48:52] <morfic-> thank you VERY much pcw_home
[13:52:46] <morfic-> http://www.linuxcnc.org/docs/html/drivers_hostmot2.html everything under 1.7 now makes utter sense, flying over several sections, the connector# didn't sink in
[14:02:17] <pcw_home> Glad you got it going. halcmd show pins is a good way to see what you've got
[14:03:47] <morfic-> sorry this was such a difficult birth
[14:04:34] <pcw_home> currently you have to be careful not to swap serial cables around as your I/O signals will move
[14:04:36] <pcw_home> this can be fixed later by checking that the card serial numbers match what you expect (locking down the config)
[14:06:12] <morfic-> that's kind of how my dual touchscreens still work
[14:07:08] <morfic-> <commandline>:0: Unknown 'show' type 'pins'
[14:07:23] <pcw_home> I suppose the driver could enumerate the cards in serial # order per type but thats a way off (and Andy is unavailable for the next 7 weeks or so)
[14:08:43] <pcw_home> I probably have the halcmd token names wrong
[14:08:53] <pcw_home> man halcmd...
[14:09:34] <morfic-> halrun seems like a good choice for me, otherwise i don't get realtime going, just using halcmd from commandline
[14:09:52] <pcw_home> Ah its "show pin"
[14:10:58] <morfic-> yeah found that, and i am better off to loadrt hostmot2 and hm2_pci with my config, just caught on to you giving me tip after the "well obviously" part was done :)
[14:23:32] <morfic-> pcw_home: the bit i'm missing is what i would need to load to get halcmd to work from terminal, /etc/init.d/realtime start is not doing it, so i use halrun to do have it start RT for me
[14:24:02] <morfic-> so i could just write something up to run instead of typing/pasting in halrun
[14:26:38] <pcw_home> I just start LinuxCNC and the run halcmd from a terminal...
[14:26:47] <pcw_home> then run
[14:27:17] <morfic-> so i am not less elegant than you :)
[14:31:21] <pcw_home> well for halrun you need to kind of make a hal file anyway to get the right thread/comps going so why not just use what you have
[14:40:15] <mrsun> pcw_home, hmm, hooked up the vfd and i get a maximum of +6V at the output pin :/
[14:40:27] <mrsun> without load i get the full 10.5V
[14:40:33] <mrsun> with the analog input i get +6V
[14:41:38] <mrsun> and a buttload of frequency jitter
[14:41:43] <pcw_home> Well that means you have too much series impedance (your 10K)
[14:42:00] <mrsun> pcw_home, hmm ok
[14:42:22] <mrsun> pcw_home, any value you think would fit ? :)
[14:42:32] <pcw_home> best thing might be to add a rail-rail op amp
[14:42:43] <pcw_home> as a 1X buffer
[14:42:55] <mrsun> pcw_home, but that would require me to have an external power supply :(
[14:43:18] <pcw_home> No it should run from the 10.5
[14:43:21] <mrsun> or can i draw the power needed from the analog V+ pin ?
[14:43:29] <mrsun> pcw_home, there is no specs on how much current i can draw from it
[14:44:06] <pcw_home> Does it droop at all with your PWM circuit?
[14:44:19] <mrsun> pcw_home, what do you mean ?
[14:44:26] <mrsun> ahh the power supply voltage
[14:44:28] <mrsun> not that i can see
[14:44:37] <pcw_home> does the 10.5 drop
[14:45:34] <pcw_home> OK so you have a nearly 10 mA load at full PWM (into the 1K load) thats plenty for the op amp
[14:46:13] <mrsun> pcw_home, hmm ok =)
[14:47:03] <mrsun> so after the low pass filter, add an opamp at 1x non inverting and feed that into the analog pin insted? :)
[14:47:14] <pcw_home> based on the 6V it looks like the analog input impedance is ~15K
[14:47:40] <pcw_home> Yes and you can now add another pole of filter
[14:47:52] <pcw_home> (if you have too much ripple)
[14:47:54] <mrsun> so filter -> opamp -> filter
[14:48:43] <pcw_home> I would do filter, filter --> opamp but there are many ways to do this
[14:48:45] <pcw_home> whats your current filter capacitor?
[14:48:54] <mrsun> pcw_home, 1uF
[14:48:59] <mrsun> and the fact that the vfd goes the wrong way, how to fix that :/ make the opamp inverting?
[14:49:10] <mrsun> when i turn the spindle on in linuxcnc it starts at max rpm directly
[14:49:14] <mrsun> then lowers when i press +
[14:49:34] <mrsun> or maybe thats something ic an fix from the linuxcnc side
[14:49:41] <pcw_home> I think you want to invert your output polarity on the PWM
[14:49:43] <mrsun> inverting the pwm output
[14:50:31] <pcw_home> it may still need a charge pump /enable to prevent any motion at startup
[14:58:06] <pcw_home> TLV2371 for example is a 16V rail-rail I/O CMOS op amp
[14:58:08] <pcw_home> so currently you have a 10 ms time constant you can do a
[14:58:09] <pcw_home> fair passive cascaded 2 pole filter by making the first stage lower
[14:58:11] <pcw_home> impedance than the second so PWM --> 10K 1 uF to GND --> 100K --> 0.1 uF to GND --> op amp folllower
[15:01:52] <pcw_home> The 100K is not a problem anymore as the CMOS OP amp is almost 0 load
[15:13:53] <mrsun> pcw_home, inverted the signal and now it stops fine and starts fine
[15:14:13] <mrsun> then i found a parameter that i can set what speed it should be at at 10V .. so if i set that value to like 150 i got full speed at about 5V :P
[15:14:34] <mrsun> pcw_home, btw should i set this setup ive done as "potentiometer option" or as analog input option ? :)
[15:16:52] <mrsun> pcw_home, any schematic of what you are talking about so i can understand it a bit better? :)
[15:17:23] <pcw_home> netlist?
[15:17:58] <mrsun> pcw_home, humm, for something like ltspice? =) or what do you mean ?
[15:18:39] <pcw_home> Well since I cant just type in schematics but can type netlists...
[15:18:58] <mrsun> pcw_home, sure go for it and ill see what i can do with it =)
[15:25:35] <pcw_home> GND = C1.2,C2.2, R1.2,OPAMP.4
[15:25:37] <pcw_home> 10.5V = OPTOCol, OPAMP.7
[15:25:39] <pcw_home> PWM = R1.1, R2.1
[15:25:41] <pcw_home> FILT1 = R2.2,C1.1,R3.1
[15:25:42] <pcw_home> FILT2 = R3.2,C2.1, OPAMP.3
[15:25:44] <pcw_home> OUT - OPAMP.6, OPAMP.2
[15:26:18] <pcw_home> OOPS PWM needs OPTO-EM added
[15:26:56] <pcw_home> opamp pin #s are for single 8 pin OPAMP
[15:32:50] <mrsun> hmm, thought ltspice would be able to open something like that but no :(
[15:35:28] <morfic-> as far as latencies are concnerned, looks like you are more relaxed on what us for jitter you said 50-100 us? often i see 20us as top, that aside, does the amount of servos being updated ultimately tie into the max jitter "target" or limit? or is the same jitter as bad fro a single servo as it is for 4? i susppose if we talk RT then it's the same?
[15:36:21] <morfic-> long story short, i am at ~22us right now, and any PM/second core/auto stuff i shut off doesn't really get me any better
[15:36:56] <morfic-> does seem if i run latency test in pncconf i get a better us jitter than when i call latency test from gnome menu
[15:37:15] <skunkKandT> for a servo machine? That will work great
[15:38:41] <skunkKandT> (non step/dir servos)
[15:39:22] <skunkKandT> (software step gen) - it will work well for that also - but my limit your top feed rate)
[15:40:25] <morfic-> analog voltage ref for speed control servos
[15:40:45] <pcw_home> even 50 usec is only 50 u inches at 60 IPM cutting speed
[15:41:43] <morfic-> 60IPM as base for an easy rule of thumb number?
[15:42:05] <pcw_home> I think Seb was looking at factoring this error out as well (since at thread actuation we know the exact time)
[15:42:55] <morfic-> that ought to be a topic people must love to see gone then
[15:43:02] <pcw_home> well 1 IPS
[15:45:10] <pcw_home> I dont think it helps the software step rate limitations but for servo/hardware step systems even say 1/2 ms of jitter is theoretically tolerable if factored out
[15:45:11] <mrsun> pcw_home, so how i understand it its the low pass filter i got + another low pass filter, that is fed into the opamp ?
[15:46:34] <pcw_home> yes but the first must be lower impedance than the second (that's why I used your original 10K and 1 uF followed by 100K and 0.1 uF)
[15:47:07] <mrsun> http://i42.tinypic.com/118itex.png like that?
[15:47:41] <mrsun> im a real noob when it comes to stuff like this, digital stuff ic an handle but this analog stuff ... yuck =)
[15:47:43] <pcw_home> Yes but with a RRIO op amp
[15:48:56] <mrsun> TLC272CP is rail to rail
[15:49:17] <pcw_home> I dont think its RR in
[15:49:38] <mrsun> what do you mean ? :)
[15:50:23] <pcw_home> input common mode range does not extend to the rails (it does have RR outputs)
[15:50:51] <mrsun> hmm
[15:50:57] <pcw_home> they type you want is RRIO meaning rail to rail inputs and rail to rail outputs
[15:51:01] <mrsun> was one i could find where i usaly buy my electronics =)
[15:51:12] <mrsun> pcw_home, aha =)
[15:52:15] <morfic-> pcw_home: if it wasn't for coming in one day and "dude, you're getting a dell" i might not have to worry about jitter anyway
[15:53:26] <pcw_home> As skunkKandT said, 20 uSec is fine for servos
[15:54:41] <pcw_home> mrsun TLC272 inputs can only go to about 1v less than the positive supply before funny things happen (the do go to ground or V-)
[15:55:06] <mrsun> pcw_home, i hate op amps =)
[15:56:11] <pcw_home> well they will take your 10K output impedance filter and make it ~0 ohms
[15:57:58] <morfic-> pcw_home: right, i meant it in a "i wouldn't have had to worry about it" kind of way
[15:58:08] <mrsun> pcw_home, yeah but to find one that suites me in this very limited list ...
[15:58:32] <mrsun> http://www.electrokit.com/opforstarkare.c455 that is the list that i choose from as ordering from other places costs alot more etc :P
[16:07:51] <mrsun> ahh here found one datasheet says "rail to rail input/output"
[16:08:21] <mrsun> MCP6001
[16:09:08] <mrsun> 5.5V max... sigh
[16:09:12] <pcw_home> only 6v
[16:09:58] <mrsun> mm
[16:14:14] <mrsun> Common-Mode Input Voltage Range Includes VCC+ <-- does that mean the same thing ?
[16:15:49] <pcw_home> Yes as long as it also includes GND (or V-)
[16:16:16] <mrsun> but aparently not single supply
[16:16:21] <mrsun> sigh i hate op amps more and more
[16:21:26] <mrsun> TS922IN might work ?
[16:21:30] <mrsun> runs at 12V atleast
[16:24:30] <pcw_home> should work
[16:28:14] <mrsun> now if only eagle could have had a model for it
[16:36:47] <DJ9DJ> gn8
[16:37:22] <mrsun> pcw_home, do i still need the 1k to gnd ?
[16:37:39] <pcw_home> Yep
[16:37:47] <morfic-> pcw_home: thanks again, time to head home
[16:39:41] <pcw_home> bye
[17:00:24] <mrsun> pcw_home, as the opamps has like infinite input impedance i can connect both the opamps non inverting sides to the low pass filter right? (from what ive been told they can start oscillating if they are not tied down and this one is a dual opamp with one channel unused)
[17:06:49] <pcw_home> I would just connect them both as followers but ground the unused + input
[17:08:18] <pcw_home> (since that opamp has ~100 nA bias current which will cause a ~10 mV offset into the 100K no need to double this)
[17:08:45] <mrsun> ahh ok =)
[17:10:31] <mrsun> thanks again, will try and route this tomorrow somhow =)
[17:10:59] <syyl_> had a good deal today :D
[17:11:07] <syyl_> http://dl.dropbox.com/u/24396704/wohlhaupter.jpg
[17:42:27] <Tom_itx> does anybody here know of linux drivers for the D525 atom board?
[17:46:57] <KimK> Tom_itx: Hi Tom. Linux drivers? You mean the D525 doesn't "just work"? Please keep us posted, I might be looking around for a "standard" CPU board myself.
[17:47:17] <Tom_itx> it works fine for linuxcnc
[17:47:29] <Tom_itx> i'm trying to set it up as a router
[17:47:47] <Tom_itx> and it won't see the onboard lan with a default debian install
[17:48:09] <KimK> Ah, OK.
[18:25:09] <joe9> want to check your thoughts on this, please? I have a drill press with a runout of 0.025 inches. I measured it by putting a harbor freight transfer punch in it and slowly rotating the chuck manually. Is it acceptable?
[18:25:31] <joe9> the quill did not move when I pull it with my bare hands. no movement there.
[18:38:54] <morfic> Tom_itx: realtek 8168b?
[18:44:19] <Tom_itx> looks that way
[18:44:37] <Tom_itx> r1816e
[18:44:51] <Tom_itx> r1816e-2.fw is the missing file
[19:25:19] <cstop> Has there been discussion of the OpenDRO topic? Comments?
[19:43:18] <KimK> joe9: Depends on what you're doing with it, but I'd guess most folks would say no. It would be especially hard on small drill bits. Imagine trying to drill a .050" hole, the bit would be trying to go back and forth half its diameter. If you're just roughing out stuff, maybe OK.
[19:47:36] <jthornton> joe9 if your still around try and take the chuck off and clean the taper and re-mount to see if your runout gets any better
[19:48:40] <jthornton> Tom_itx: my 525 works fine, have you upgraded the bios on yours?
[19:49:20] <Tom_itx> jthornton, it's not the emc install
[19:49:25] <Tom_itx> that works fine
[19:49:36] <Tom_itx> it's a new debian 6.0.4 install
[19:50:17] <Tom_itx> i've found a couple threads saying not everything was supported
[19:52:32] <jthornton> ah, ok
[19:52:42] <jthornton> had me going there for a minute
[19:52:58] * jthornton hears lots of big bubbles in the background
[19:53:09] <Tom_itx> i was doing some testing on it before i committed the changes to my server
[20:09:09] <joe9> KimK: jthornton: I got the chuck and arbor out. I notice a 2 mil (0.002 inches) runout on the inside of the quill (wall of the morse taper socket). Is that acceptable or normal?
[20:14:44] <joe9> i am using a 1/72 inch drill bit to drill some small holes in a pcb.
[20:15:12] <KimK> joe9: You didn't say the brand of drill press. It might be typical for Harbor Freight? Someone with the appropriate precision grinding setup might be able to improve the socket taper for you?
[20:15:25] <KimK> joe9: Ah, that's different...
[20:15:27] <joe9> sorry, the radius of the drill bit is 1/64 inch.
[20:15:47] <joe9> KimK: Central Machinery 38142 is the drill press.
[20:16:02] <joe9> let me find a link. I will post it.
[20:17:28] <joe9> kimK,
http://www.harborfreight.com/heavy-duty-16-speed-bench-drill-press-38142.html?hft_adv=10062&gclid=CPyDmrXHuq4CFQff4AodvlTlPg
[20:18:13] <joe9> KimK: for such holes, do you consider 2 mil acceptable (in the quill)?
[20:18:32] <joe9> I read that if the quill has a big runout, just chuck the drill.
[20:18:50] <joe9> If it is the chuck or arbor, then it can be replaced and things are ok.
[20:20:04] <joe9> I am trying to ascertain if the 2 mil on the quill is considered too much for holes with a drill bit of radius 1/64 inch.
[20:20:09] <KimK> joe9: ...In that case, another way out (cheaper, faster, not as "good") presents itself. You might consider using a "sensitive drill" adapter
http://www.sherline.com/1012inst.htm and shimming it as needed to get the runout better. Yes, it's a lot of fiddling around, but you'd only have to do it once per PCB drilling festival.
[20:21:01] <joe9> KimK: do you consider 2 mil on the quill bad for my purposes?
[20:23:36] <KimK> I'd like to see it closer, yes, but you could try it and see how bad it is for you. If your drills are "long" (relative to their diameter), maybe there will be enough "flex" to make it work? Not the greatest, but you're speccing your own holes, right? So if they "look good", you're done?
[20:23:55] <joe9> ok, thanks.
[20:24:38] <KimK> I'd try it and see what happens.
[20:25:00] <joe9> KimK: "It is difficult to maintain tolerances closer than +.003" -.000 with drills. If greater accuracy is required, a reamer should be employed." -- I read this in the link pasted.
[20:25:21] <joe9> If so, then the quill with a runout of 0.002" is probably ok?
[20:25:49] <joe9> I am more concerned if runout at the quill increases exponentially after I put in the chuck and etc.
[20:26:02] <joe9> or, if it stays the same provided the other stuff is perfect.
[20:27:13] <joe9> 0.002" runout is ok for my purposes. but, if the runout will increase by a high proportion even with a perfect chuck/arbor, then I am wasting my time with this drill.
[20:27:18] <joe9> Does that make sense?
[20:29:15] <KimK> What you say is true. And since the spindle nose is the basis for everything from that point on, in a perfect world your spindle nose runout would be zero. Otherwise you (as a perfectionist) will say, "The runout is not zero, I cannot continue!" But in a world of limited budgets, I'd try it and see if it is "good enough".
[20:31:18] <KimK> You might keep an eye out for 1/8" dowel pins, as 1/8" is usually considered the largest "small" drill bit. You could chuck the 1/8" dowel pin in your setup and check for runout at the drill tip.
[20:31:24] <joe9> kimk, these are the instructions I am following: "The run-out could be in the quill, the chuck, or the arbor. First, remove the chuck (I'm assuming that it's mounted on a morse-taper arbor) from the quill using a drill drift, and run the point of your dial indicator on the inside of the quill - on the wall of the morse-taper socket. If the run-out is there then the drill is junk for sure. If the socket is good then the problem is in ...
[20:31:30] <joe9> ... the arbor (cheap to replace) or the chuck (not cheap to replace, but often worth replacing just to have one that works smoothly). It's also possible that the chuck might not have been assembled properly onto the arbor, or the arbor might not be seated properly in the quill. If you can remove the arbor from the chuck then you can install it in the quill and check the Jacobs-taper surface for runout. The arbor is, honestly, not ...
[20:31:36] <joe9> ... likely to be that far out, even a cheap one. Some chucks have a hole in them so you can bang the arbor out using a pin punch. If your chuck doesn't have that hole then you either need to drill such a hole (often requiring another drill press) or you need wedges made for the purpose to remove it. Once you know where the problem is you can work on it. As to how much run-out is acceptable, I'd say that 2 thou total runout is ...
[20:31:42] <joe9> ... reasonable for a cheap drill press, but 10 thou is actually well within the needs of the average woodworker just drilling holes in wood. However! Just because you don't really NEED the accuracy doesn't mean you should settle for junk."
[20:31:46] <joe9> kimK, oops very sorry. bigger paste than I imagined.
[20:31:51] <Tom_itx> woah
[20:32:00] <joe9> i got them from
http://forum.canadianwoodworking.com/showthread.php?26529-Drill-Press-Runout-what-is-reasonable
[20:32:06] <joe9> Tom_itx: very sorry about that.
[20:33:17] <joe9> kimK,
http://pastebin.com/sriiivzi
[20:34:07] <joe9> KimK: I have the machinist's transfer punch and I am checking the runout using the transfer punches.
[20:35:04] <joe9> the belts are pretty bad. so, I will be replacing them too. I read about the "vibration less belts"
http://www.paragoncode.com/shop/link_belts/ . wondering what you think of them/
[20:35:20] <KimK> OK. A transfer punch is usually kind of long, though. I thought an 1/8 dowel pin might be more "circuit board drill-like".
[20:35:22] <joe9> s/are pretty bad/seem bad/
[20:36:33] <joe9> there are no cracks or anything in the belts, but, I bought the drill "used" and there was a lot of rubber dust near the belts.
[20:36:50] <joe9> hence, am thinking my accuracy could get better with "vibration less belts".
[20:40:05] <KimK> I have no info on their vibration-reducing qualities, but I am presently searching for various technologies of "breakable" v-belts for a machine where the belt must be broken to thread through holes. I didn't know about these, so, thanks!
[20:49:30] <joe9> glad to be of help.
[20:58:04] <KimK> The "less vibration" business might have come from the McMaster catalog description:
http://www.mcmaster.com/#v-belts/=ger646 Not sure how they are defining vibration, but nevermind.
[21:05:05] <joe9> KimK: the chuck says "1 - 16 mm JB". Does it mean that it is a Jacob's chuck? Where do you buy the chuck/arbor? McMasterCarr?
[21:16:13] <KimK> Not sure about the "JB", but I'd expect that means Jacobs. But they don't specify the taper? Usually the Jacobs taper is marked JT something. See
http://www.jacobschuck.com/pdf/Technical-Information.pdf Yes, you can get chucks and arbors at a lot of places, mscdirect.com, many more. The Arbrecht(sp?) hand-tightened chucks are nice, but take more turns to tighten/loosen. Usually pricier too.
[21:24:24] <KimK> Oh, I was thinking about it, and on your .002" spindle shaft error, you (or someone) should be able to fix it with just a lathe and collet setup to hold the shaft (check accurate centering first), and then a die grinder with a small stone on the compound rest at the taper angle. You only have to take off (very carefully) a thousandth or two to fix it.
[21:28:51] <MattyMatt2> I've seen ancient leather belts like those ones
[21:29:23] <KimK> Ha, maybe that's what I need, lol!
[21:29:36] <MattyMatt2> I thought it was just a way of using scraps
[21:30:40] <MattyMatt2> and avoiding stiching the join
[21:33:30] <joe9> ok, this is what I have with the chuck/arbor in the quill, runout at the quill: 0.002 inch, runout at the end of the chuck: 0.010 inches.
[21:33:37] <KimK> The break (or breaks) would be doubly handy on this machine because I need a break not only to install it (to thread it through some holes), but also to select the high or low pulley (no way to bring the pulleys together to de-tension them).
[21:33:50] <joe9> runout at the arbor: 0.002 inches.
[21:34:08] <joe9> so, I guess the chuck is the bad part.
[21:35:31] <joe9> I cannot figure out how to get the arbor out from the chuck. I tried punching a machinist's tap with a rubber mallet at the center between the teeth of the chuck and nothing is budging.
[21:35:41] <KimK> Maybe. It's also possible that the arbor hole is off-center and/or off-axis (at an angle).
[21:35:41] <joe9> Is that how the arbor is removed from the chuck?
[21:37:11] <KimK> Is the spindle female and the chuck male? Some are one way, some are the other.
[21:39:29] <KimK> Oh, some chucks have a securing screw inside the chuck, where the tail of the drill bit goes. Sometimes an allen screw, sometimes a phillips, sometimes a torx. Also sometimes left-handed, be careful.
[21:44:47] <MattyMatt2> chuck may not be the runout, it could be the top bearing in the spindle
[21:46:07] <MattyMatt2> that would make the chuck seem bad just from the length of it
[21:46:41] <MattyMatt2> joe9 did you get the taper out of the spindle but still have the chuck stuck to the arbor?
[21:47:52] <joe9> MattyMatt2: yes, I got the chuck+arbor out of the spindle, but, the chuck is still stuck to the arbor.
[21:48:19] <joe9> MattyMatt2: "chuck may not be the runout, it could be the top bearing in the spindle" that would be scary, isn't it?
[21:48:38] <joe9> would have to replace the drill, I guess, in this instance.
[21:49:14] <joe9> KimK: I do not see a screw there. just a small round slot kind of opening.
[21:50:55] <MattyMatt2> I guess you could ceck the runout of the top of the spindle to check, then if it is bad maybe you could bore & sleeve it or something, if it's a solid bearing
[21:53:03] <MattyMatt2> I guess if it's tight, the line between the spindle bearings is your datum, and so the taper is the issue
[21:53:39] * MattyMatt2 stops guessing
[21:53:43] <KimK> FWIW, the product manual for the drill link you gave does not show a chuck securing screw either. The manual does show two different ball bearings (probably radial), no doubt of a low-cost type.
[21:54:46] <KimK> And it shows the arbor as male, chuck as female.
[21:57:30] <MattyMatt2> I use a metal hammer to unscrew a chuck usually. it's a light sharp tap on the chuck key that does it
[21:57:34] <joe9> The chuck and the arbor are pretty tight. I do not see any movement when I try to press with my hands. There are 2 things that could be wrong, it appears, 1. the spindle bearing may be gone. (I do not know what the "spindle" is, would have to google it. 2. the chuck might be a little bit off when seated in the arbor or the arbor is off when seated in the quill (is this even possible?)
[21:58:10] <MattyMatt2> by spindle I mean quill I guess
[21:59:20] <joe9> oh, the quill. I checked it by pushing/pulling it around. it did not budge. It was lose and moving around initially, but, I tightened the screw and not it is really snug there. no movement that I could feel with my hands.
[21:59:25] <MattyMatt2> a bit of dirt or damage on the arbor will make it sit off-centre in the quill
[21:59:55] <joe9> I put grease on the quill. Is that a bad idea?
[22:00:09] <joe9> MattyMatt2: I cleaned up the arbor pretty well before putting it in.
[22:00:25] <MattyMatt2> and did you check inside for dirt?
[22:00:30] <joe9> damage, probable.
[22:01:04] <joe9> MattyMatt2: I looked in it and did not see anything as far as I could see. but, it is a small opening to really get a lot of details.
[22:01:07] <joe9> Does that make sense?
[22:01:28] <joe9> I could clean it by putting a cloth or something in that hole from the sides.
[22:01:38] <joe9> or, is there a better way of cleaning it?
[22:01:45] <MattyMatt2> yeah it'd be hard to see I guess, unless you had sunlight shining right up it
[22:02:18] <joe9> I removed the chuck twice and put it back in. I think if there was some dirt it would have been atleast displaced, correct?
[22:02:33] <joe9> both times, the runout was still 10 mil (0.010 inches).
[22:02:48] <MattyMatt2> i poke a parrafin soaked rag up on the end of a pencil etc, on an MT1, but I dunno if that's clever
[22:03:27] <MattyMatt2> a metal chip might be squashed into one place
[22:03:50] <MattyMatt2> or a grain of sand etc
[22:04:38] <joe9> possible, I bought it "used". I think my best option is to buy a new chuck+arbor and check the runout.
[22:04:53] <joe9> if the runout is still the same 10mil, then it is the drill or dirt.
[22:05:01] <joe9> MattyMatt2: would you concur?
[22:05:30] <MattyMatt2> ideally you'd want one of the precision gauge bars with a taper and 12" or so of precision straight shaft. they are expensive
[22:06:15] <MattyMatt2> that's what I need if I'm ever going to get my worn out mini lathe in any state of precision
[22:07:01] <MattyMatt2> they are like boring bars but without the tool
[22:07:01] <pfred1> does anyone in here know about CAM software?
[22:07:31] <pfred1> MattyMatt2 I've been in a few boring bars
[22:07:58] <pfred1> in my experience they're usually filled with tools though
[22:08:07] <MattyMatt2> and they are usually full of tools :)
[22:08:14] <pfred1> jynx!
[22:08:47] <pfred1> I guess a boring bar without the tool is an empty bar then?
[22:09:36] <pfred1> anyhow this one tool on DIY-CNC insists it is the CAM software's task to generate G Code
[22:09:57] <pfred1> I always thought CAM was supposed to read G Code
[22:10:03] <MattyMatt2> yeah I guess that's usual definition of CAM
[22:10:20] <MattyMatt2> nope it's the controller that reads gcode
[22:10:24] <MattyMatt2> aka emc2
[22:10:35] <pfred1> so EMC2 isn't CAM software?
[22:11:10] <MattyMatt2> not in the narrow definition that's emerging
[22:11:34] <MattyMatt2> it's a machine controller software
[22:11:35] <pfred1> I'm not concerned with any emerging definition just the classical one
[22:12:30] <MattyMatt2> it has been called CAD/CAM for decades, so the split occured back then between CAS that doesn't produce gcode, and CAM which does
[22:12:42] <MattyMatt2> s/CAS/CAM
[22:12:46] <pfred1> CAD/CAm is different than just CAD and CAm
[22:12:48] <MattyMatt2> s/CAS/CAD
[22:13:09] <pfred1> puttingthe slash in there says it is an integrated system solution
[22:13:13] <MattyMatt2> CAD/CAM is an industry rather than an actual app
[22:13:36] <pfred1> that integration implies some kind of internal translation
[22:13:53] <pfred1> but to imply it with the separate terms is incorrect
[22:14:05] <pfred1> that is what the slash means
[22:14:29] <MattyMatt2> RhinoCAM produces g-code
[22:14:31] <pfred1> either that or it is a waste of electronis ink etc
[22:15:09] <pfred1> oh I'm not saying for an instant that there aren't any number of applications that make loose with the terms
[22:15:30] <pfred1> and make up their own definitions as they go along
[22:15:48] <pfred1> that doesn't automatically make them right either
[22:16:16] <MattyMatt2> I think most users would agree that "The CAM software" is the bit that converts CAD files into g-code
[22:16:59] <pfred1> OK so it is the in between software then that doesn't run the machine?
[22:17:20] <MattyMatt2> yeah
[22:17:24] <pfred1> then Mach3 and EMC2 aren't CAM if I am following you
[22:17:38] <MattyMatt2> yep they are NC Controllers
[22:17:45] <MattyMatt2> CNC machines
[22:17:55] <pfred1> Numerical Control Controllers
[22:17:59] <pfred1> like NCC?
[22:18:02] <MattyMatt2> when integreated into the machine like a Fanuc
[22:18:09] <pfred1> isn't that what was on the Enterprise?
[22:18:18] <pfred1> NCC 1707?
[22:18:44] <pfred1> Fanuc makes machines wiht built in controllers don't they?
[22:19:17] <pfred1> so it is safe to say that Fanuc isn't in the CAM business
[22:19:39] <pfred1> they are in the CNC business
[22:19:43] <MattyMatt2> not by current usage of the term
[22:20:01] <MattyMatt2> so yeah Fanuc make controllers. Rhino do CAM
[22:20:01] <pfred1> see that is what I am trying to avoid is semantic slippery slopes
[22:20:15] <FinboySlick> Rhino does CAM?
[22:20:21] <pfred1> not what every knucklehead believes today but what actually is
[22:20:31] <MattyMatt2> speak French then. their language is controlled by a central commitee :)
[22:20:32] <FinboySlick> We talking the makers of Rhino 3D here?
[22:20:38] <pfred1> because i understand there is some confusion surrounding all of this
[22:20:57] <FinboySlick> MattyMatt2: Oui, mais pas en France ;)
[22:21:28] <pfred1> MattyMatt2 so this is turning into a gay marrige debate is it?
[22:22:08] <MattyMatt2> Rhino 3D is what I'd call CAD, but afaik they do a version/plugin that outputs gcode and that's what I'd call CAM
[22:22:30] <pfred1> well sure a CAD package can have an export filter
[22:22:36] <FinboySlick> MattyMatt2: It was a genuine question. I didn't know they had G-Code generation.
[22:22:39] <pfred1> I'm perfectly fine with that
[22:22:55] <MattyMatt2> which turns it into CAD/CAM
[22:23:02] <pfred1> a CAM controller could even have an import filter
[22:23:07] <FinboySlick> I only used Rhino pre 2000.
[22:23:18] <FinboySlick> Back then, it was super-new and exciting.
[22:23:19] <MattyMatt2> I call blender a CAM package, simply because it has a gcode output script :)
[22:23:32] <pfred1> but what I'm trying to determine is if an NC controller without a G Code generator isn't still a complete package
[22:24:06] <MattyMatt2> pfred1 yeah if an engraving machine could work direct from jpeg or dxf, I'd agree that was CAM
[22:24:08] <pfred1> because I'll be honest I consider the NC controller the CAM
[22:24:41] <pfred1> if chips ain't flying you ain't machining
[22:24:49] <pfred1> simple as that
[22:24:55] <MattyMatt2> it's CAM in the greater general term
[22:25:38] <MattyMatt2> like compiling is a general term for preprocessing, compiling and linking
[22:26:06] <pfred1> you can look at an object file and it ain't plain text no more
[22:26:16] <pfred1> something has indeed been operated on
[22:26:45] <pfred1> but I can look at G Code and it isn't putting a finished part in my hand
[22:27:20] <pfred1> no that seals it the M in CAM
[22:27:26] <pfred1> I know what machining is
[22:27:58] <pfred1> it seems even what CAM stands for has slid a little
[22:28:05] <MattyMatt2> I hoover with my VAx dyson
[22:28:30] <pfred1> it used to be CAD Computer Aided Design and CAM Computer Assisted Machining
[22:28:37] <pfred1> now i see Aided in both contexts
[22:29:59] <FinboySlick> http://www.youtube.com/watch?v=SQIYZ-iKuG4 Here, distraction for you two.
[22:30:06] <MattyMatt2> yeah but now CAD refers to draghting, and CAM refers to manufacturing design
[22:30:29] <pfred1> actually design comes before drawing
[22:30:58] <pfred1> although during the drawing process designs are usually refined
[22:31:12] <Tom_itx> cam takes the cad to the machine
[22:31:37] <pfred1> Tom_itx so I can load a DXF file and I'm good to go?
[22:32:04] <MattyMatt2> I blame the unions, they were the ones doing the demarcation when it was humans not software packages :)
[22:32:05] <pfred1> I know in some cases this is true but I'm saying semantically is that absolutely correct?
[22:32:26] <Tom_itx> mine is a combined package
[22:32:38] <Tom_itx> not all are that way
[22:32:40] <pfred1> but is that to say that htey all have to be?
[22:32:43] <pfred1> exactly!
[22:32:56] <MattyMatt2> FinboySlick, already drooled over that one :)
[22:33:07] <pfred1> I understand it is becomming the rule today but all I'm saying is it really isn't THE RULE
[22:33:09] <Tom_itx> which is crap really since the end product requres both usually
[22:33:35] <Tom_itx> my bud spent $75K on a catia seat and still had to get a windows post for it
[22:33:42] <MattyMatt2> pfred1 the rule today is the rule. language is communication
[22:34:34] <pfred1> MattyMatt2 well in that case there are going to be an awful lot of dissappointed people out there when they find their setups break arbitrary rules made up by idiots
[22:34:46] <pfred1> as they stumble along
[22:34:47] <MattyMatt2> 7 billion and rising
[22:35:08] <FinboySlick> Catia is $75k?
[22:35:56] <pfred1> FinboySlick insane what some have the gall to charge for strings of ones and zeros today isn't it?
[22:36:04] <MattyMatt2> 7 milliard and rising :)
[22:36:06] <Tom_itx> i don't know what sort of agreement he made but yes it cost that to get set up with it
[22:37:25] <FinboySlick> I was formally trained on Catia v4 on Sun hardware a while back. It was sure impressive for its time but for playing with the windows version a few years back I'm not sure what justifies the cost over SolidWorks.
[22:37:43] <pfred1> FinboySlick greed
[22:37:52] <Tom_itx> yeah i have a windows demo disk here i've never opened
[22:37:57] <FinboySlick> I know it can manage every bolt and rivet of an Airbus... And was for a while the largest software project in all of Europe.
[22:38:06] <MattyMatt2> when I was young a billion meant 10¹² but then in the 70s we started hearing "billion dollar deficits" on the news and all started using the american meaning
[22:38:09] <Tom_itx> his was on a sun iirc
[22:38:33] * pfred1 took apart a Sun pizza box with nothing but a crowbar once
[22:38:37] <Tom_itx> i never used it myself as i had my cad cam on windows already
[22:38:49] <pfred1> they really built them suckers
[22:38:54] <Tom_itx> i did see it though and it looked nice
[22:38:56] <FinboySlick> pfred1: You need a crowbar to open these, they're built with 1/8 steel.
[22:38:59] <pfred1> we didn't have a scrwdriver :)
[22:39:08] <Tom_itx> except for not having a post
[22:39:17] <pfred1> took us a good minute to pry it open
[22:39:40] <FinboySlick> I remember having to rack E250s.
[22:39:42] <pfred1> it looked a bit like a large potato chip when we were done
[22:40:14] <pfred1> then we were pissed because the RAM modules in it were weird
[22:40:31] <FinboySlick> All Sun hardware was weird.
[22:40:32] <pfred1> so we tossed it into a dumpster
[22:40:41] <FinboySlick> They blamed their CPU bugs on sunspots too.
[22:40:48] <FinboySlick> Ah the good old days.
[22:40:56] <pfred1> yeah I was demoing out a pherma company
[22:41:02] <pfred1> pharma even
[22:41:14] <pfred1> we got truckloads of computer gear out of that place
[22:42:25] <pfred1> the people who invented LSD
[22:42:41] <FinboySlick> Pff, that's the CIA.
[22:42:43] <pfred1> Sandoz
[22:42:51] <pfred1> nope was Sandoz
[22:43:07] <pfred1> they sucked to work for
[22:43:16] <MattyMatt2> CIA just invented novel uses for it
[22:43:39] <pfred1> they'd hold you up at their guard shack for ah alf a day sometimes
[22:43:56] <pfred1> we didn't care we still got paid
[22:44:06] <FinboySlick> See, CIA front. My conspiracy self prevails.
[22:44:25] <pfred1> oh they had tractor trailers of cocaine going into that joint sometimes
[22:44:55] <pfred1> they had big numbers painted on the tops of the trailers and they'd folow them with helicopters they weren't playing no games
[22:45:10] <MattyMatt2> mm novacaine. I'll be getting some of that soon I hope. mainlined into my gums
[22:46:06] <pfred1> I had a doc shoot me up wiht that stuff once 4 times I could still feel I was like just pull the tooth the shots are killing me!
[22:46:07] <MattyMatt2> or maybe I'll just pull my remaining stumps out unaided
[22:46:53] <MattyMatt2> the shots are nothing compared to toothache, in my experience
[22:47:21] <pfred1> oh when I had my wisdom teeth pulled i told them to gas me
[22:47:34] <pfred1> a decision I don't regret
[22:48:01] <MattyMatt2> I fought them going under and coming out, the one time I was gassed
[22:48:05] <pfred1> then I drank responsibly on Darvon for like a week
[22:49:25] <pfred1> now count backwards from 100 by threes ... 97
[22:50:03] <pfred1> that is al lI remember
[22:52:05] <Tom_itx> the last one i had out the guy left a chunk of it so i pulled it out when i got home
[22:52:41] <Tom_itx> it just didn't feel quite right for a couple days so i poked around a bit and jerked it out
[22:53:53] <Thetawaves> my install of the ubuntu 10.4 linuxcnc edition keeps freezing the display
[22:54:06] <pfred1> Thetawaves turn the heat up
[22:54:09] <Thetawaves> emc2 continues to run the gcode though
[22:54:24] <Tom_itx> bad graphic driver?
[22:54:31] <Tom_itx> someone here i think had issues with that
[22:54:35] <pfred1> sounds like you're getting hit by GL issues like happened to me
[22:55:02] <pfred1> change your conf file to tkemc and see if it works then
[22:55:11] <pfred1> where it says display = axis
[22:55:30] <Thetawaves> what?
[22:55:54] <pfred1> are you getting kicked to the console from X?
[22:56:02] <pfred1> like black screen with big white text?
[22:56:13] <Thetawaves> Tom_itx, i had major issues just getting ubuntu 10.4 to run.. compiz would lock up my graphics
[22:56:23] <pfred1> oh man ditch compiz
[22:56:24] <Thetawaves> nope, just lockup
[22:56:31] <pfred1> ah different problem then
[22:56:31] <Tom_itx> why didn't you use the live cd?
[22:56:33] <Thetawaves> unrecoverable... can't install openssh-server either
[22:56:47] <Thetawaves> i did use the live cd?
[22:56:52] <Tom_itx> hmm
[22:57:04] <pfred1> I didn't use the live CD because Ubuntu sucks
[22:58:18] <pfred1> Thetawaves you check syslog to see if there are any clues what is going on?
[22:58:30] <Thetawaves> i did... absolutely nothing
[22:58:36] <pfred1> sweet
[22:59:00] <pfred1> well I'd still change from axis to tkemc
[22:59:17] <pfred1> see if you're not having a GL issue but you're going to have to kill compiz too
[22:59:56] <Thetawaves> in the ubuntu appearance settings you can tell it to not use any funny business
[22:59:59] <pfred1> really if it was me I'd change the entire window manager
[23:00:08] <Thetawaves> that reduced a lot of the freezing
[23:00:12] <pfred1> I recommend fluxbox
[23:00:22] <Thetawaves> lol
[23:00:32] <pfred1> what do you find so funny?
[23:00:37] <Thetawaves> i'd rather just try a newer linux and see if the bug has been fixed
[23:00:38] <pfred1> you want it to work
[23:01:04] <pfred1> because it is far easier for me to sit here and go lol it works here
[23:01:46] <Thetawaves> yeah but if i don't solve the root cause i'll forever be plagued by random hangs
[23:02:04] <pfred1> I narrowed it down to my GL libraries I should use software rendering if i want it to completely work
[23:02:17] <pfred1> nah if I don't use GL at all I'm fine
[23:02:29] <pfred1> soon as RTAI runs and GL does I'm sunk though
[23:02:53] <pfred1> but like I said if I wanted to completely fix it I'd have to change the GL library
[23:02:56] <Thetawaves> yeah but why is that
[23:03:04] <pfred1> I've chosen to simply avoid all GL
[23:03:13] <pfred1> because it is complicated
[23:03:21] <pfred1> and RTAI doesn't like hardware GL
[23:03:32] <Thetawaves> why!
[23:03:43] <pfred1> apparently there are conflicting signals
[23:03:50] <Thetawaves> Still have problems with a newer nvidia card?
[23:03:54] <pfred1> a message gets sent out and hangs forever
[23:04:08] <pfred1> you're not using the nvidia driver I hope
[23:04:23] <Thetawaves> nah gma3100 i thik
[23:04:37] <pfred1> well still make sure you are using software GL
[23:05:12] <pfred1> but it is easier to switch to tkemc and kill compiz to make sure that is even your issue
[23:05:24] <pfred1> or you'll be fixing what isn't evne broken
[23:06:56] <Thetawaves> gma945 actually
[23:07:34] <pfred1> run wihtout GL first to see if that is even what is causing your problems
[23:07:57] <pfred1> until you can isolate the cause it is all speculation
[23:09:07] <pfred1> and by GL i mean all instances of it because even glxgears caused it on my system when I was using tkemc
[23:09:08] <Thetawaves> anything can call opengl
[23:09:23] <pfred1> I know
[23:09:30] <pfred1> that is why I want you to make sure nothing does
[23:09:47] <pfred1> by killing that fool compiz trash
[23:09:57] <pfred1> and not using axis
[23:10:16] <pfred1> then if doing that makes the problem go away then you can change your GL libs
[23:11:03] <Thetawaves> so how do i use tkemc instead of axis
[23:11:14] <Thetawaves> i've already disabled compiz
[23:11:17] <pfred1> edit your machine's conf file and change your display = line
[23:11:38] <pfred1> a line will say display = axis remove axis and type in tkemc
[23:11:47] <pfred1> save and exit
[23:12:23] <pfred1> them when you fire up emc you'll see theit butt ugly Tk/tcl window pop up :)
[23:12:47] <pfred1> but it really isn't that bad reminds me of the BDI days
[23:14:02] <Thetawaves> none of the files have display=
[23:14:36] <Thetawaves> there are a whole bunch of files, which one should i put display=tkemc in
[23:14:42] <pfred1> I have a stand alone install i don't know how the CD image install is structured
[23:15:08] <pfred1> in my system I have an emc dir in my home dir that contains my machine's conf file
[23:15:47] <pfred1> the file should end with the suffix .conf and you should need to select it when you run your machine too
[23:16:11] <pfred1> least that is how it works here
[23:16:16] <Thetawaves> i have a shortcut
[23:16:44] <pfred1> but it should still pop open a dialog box that makes you select the conf associated with the machine you are operating
[23:16:58] <Thetawaves> nope
[23:17:02] <pfred1> you have to navigate to that conf file then edit it
[23:17:05] <Thetawaves> thats the point of the shortcut
[23:17:31] <pfred1> well then check the shortcut's properties it must be a command line parameter that feeds emc your conf file
[23:17:47] <Thetawaves> there is a .conf file generated by stepconf
[23:17:49] <pfred1> I start emc from the CLI
[23:17:58] <pfred1> yes that is your conf file
[23:19:30] <pfred1> someplace in it there should be a setting for what interface you use when you run EMC
[23:19:49] <pfred1> there must ba a half a dozen of them but today the default is axis if you run stepconf
[23:20:05] <pfred1> axis is nice but it is GL
[23:26:20] <Thetawaves> well no lockups this run using tkemc
[23:26:26] <pfred1> there see
[23:26:37] <Thetawaves> btw
[23:26:37] <pfred1> it was just as I suspectes
[23:26:40] <Thetawaves> my machine file was .ini
[23:26:57] <Thetawaves> and my greps for display went unanswered because it was DISPLAY =
[23:26:59] <Thetawaves> :)
[23:27:08] <pfred1> always grep -i
[23:27:31] <pfred1> anywho now you have to try changing your GL libs
[23:27:55] <pfred1> see if software rendering clears your issue up
[23:28:07] <pfred1> it should
[23:28:13] <pfred1> well, it might
[23:28:36] <pfred1> if it doesn't you're hosed let's just put it that way
[23:28:44] <Thetawaves> not being able to install openssh-server because it depends on the 3rd build of openssh-client when i have the 4th installed
[23:28:50] <Thetawaves> pretty much spells doom for this install
[23:28:58] <Thetawaves> i'll probably install debian on it
[23:29:04] <pfred1> I run debian
[23:29:10] <pfred1> hey want my notes how to do it?
[23:29:21] <Thetawaves> sure
[23:29:28] <pfred1> I'd have to say the entire build is rather non-trivial
[23:30:07] <pfred1> http://pastebin.com/cYYqyKPX
[23:30:23] <pfred1> oh junk that's right I did it on Lenny
[23:30:32] <pfred1> can you still even install Lenny today?
[23:30:52] <pfred1> I know Lenny recently passed into EOL
[23:30:58] <Thetawaves> i like how you start off with removing network-manager and avahi
[23:30:59] <Thetawaves> lol
[23:31:08] <pfred1> well
[23:31:13] <pfred1> I'm a purist
[23:31:20] <pfred1> if things don't work I tend to avoid them
[23:31:38] <Thetawaves> i have a different philosophy
[23:31:50] <Thetawaves> i try to work within the normal methods and practices for the distribution
[23:32:13] <Thetawaves> i don't like when applets fight hand entered values into configuration files
[23:32:14] <pfred1> I been running Linux too long to think anything like that
[23:32:47] <pfred1> that is what chattr is for :)
[23:33:14] <pfred1> set the immutable bit on a file
[23:33:57] * pfred1 don't play when it comes to UNIX
[23:34:07] <pfred1> root God what's the difference?
[23:34:27] <Thetawaves> unix is pretty insecure
[23:34:34] <Thetawaves> in that respect
[23:35:16] <pfred1> yeah i like the total control aspect of it all though
[23:36:40] <pfred1> banging EMC into Debian is a nice display of it too ;)
[23:38:04] <Thetawaves> i'm thinking slackware
[23:38:37] <pfred1> I like debian's package manager though I haven't run slack in about 10 years now
[23:38:44] <Thetawaves> i ran slackware for many years, i think it'll work well for a system with such a defined task
[23:38:51] <pfred1> so I really don't know what they're doing today
[23:39:17] <pfred1> I ran slack from 3.0 to the end of 7
[23:39:29] <pfred1> that was enough for me
[23:39:42] <Thetawaves> i ran from slackware 7 - 12.1
[23:39:50] <pfred1> I started on 8 but I was like I've had enough of this
[23:40:12] <pfred1> then you missed out on Redhat 7.3
[23:40:23] <pfred1> which was probably the greatest Linux distro of all time
[23:41:06] <pfred1> they invented up2date
[23:41:31] <pfred1> man that was sweet!
[23:42:01] <Thetawaves> lol
[23:42:06] <Thetawaves> redhat was a disaster
[23:42:09] <pfred1> hey after running slack
[23:42:22] <pfred1> not 7.1 to 7.3
[23:42:37] <pfred1> everyone and their dog was running it back then
[23:43:07] <pfred1> right up until they only let you have one system registered for free
[23:43:35] <pfred1> that was the death knell
[23:43:44] <pfred1> but before that it couldn't be beat!
[23:43:46] <Thetawaves> i was always quite afraid of rpms
[23:44:05] <pfred1> well that was their high water mark then it all just worked
[23:44:42] <pfred1> but redhat themselves trashed it all
[23:45:10] <pfred1> I guess for bilions i might have done the same though
[23:45:46] <pfred1> at one point Linus was worth over 20 million in stock RH gave him
[23:46:02] <pfred1> he should have sold
[23:52:15] <Thetawaves> 20million!
[23:53:30] <pfred1> not a bad pay day if he'd have cashed in
[23:56:10] <Thetawaves> i wonder what the latest kernel rtai runs on
[23:56:56] <pfred1> what'd I build it against 2.6.35?
[23:57:11] <pfred1> should be new enough for whatever hardware you have
[23:57:52] <pfred1> I have a box upstairs with a 945 GPU in it and it runs on the same kernel
