#emc | Logs for 2009-07-02

[01:20:26] <skunkworks> jmkasunich: any time to work on the gantry?
[01:21:43] <skunkworks> * skunkworks thinks he waited a descent amout of time to ask ;)
[01:21:54] <skunkworks> *amount
[01:23:28] <sed_> do make emc2 start I changed linksp Xminlim <= stg.in-02 to linksp Xminlim <= stg.in-02-not
[01:23:48] <sed_> because the limit switch true/false was backward
[01:24:23] <sed_> did I need to change <= to => or somthing as well?
[01:24:44] <sed_> cus it turns on but as soon as I power up the machine it complains of a limit error
[01:25:30] <cradek> if it's still doing it, maybe you didn't get the right one
[01:25:47] <cradek> you can see on the screen what it thinks is on a limit - it's the |<- icon
[01:26:06] <sed_> well I fixed it in that it now tuns from red to green when the sofware starts
[01:26:11] <cradek> poke each of the switches while watching that and see if they work and/or are reversed
[01:26:22] <sed_> ok
[01:26:24] <sed_> thanks
[01:26:25] <cradek> oh, you're using tkemc, I'm not sure what the colors are
[01:26:37] <jmkasunich> skunkworks: not yet, but I'm getting close
[01:26:38] <cradek> red is on limit, I think
[01:26:38] <sed_> red bad green good
[01:26:51] <jmkasunich> I had another machining project to finish up
[01:26:56] <cradek> ok, you can do the same test then
[01:27:07] <sed_> k will do
[01:27:10] <sed_> brb
[01:33:52] <skunkworks> jmkasunich: great!
[01:35:12] <sed_> yea they work, goes red when I hit the switch back to green when off
[01:35:22] <cradek> excellent
[01:35:32] <sed_> and I get an error in emc not red line
[01:35:40] <cradek> more details please
[01:35:56] <sed_> when i turn the machine on i get limit swith error
[01:36:05] <cradek> what do you mean turn the machine on?
[01:36:29] <cradek> and what is the exact error?
[01:36:50] <sed_> joint 2 on limit switch error
[01:36:54] <sed_> and joint 0
[01:37:29] <cradek> ok, 0 and 2, but never 1? but when you tested them they all work correctly?
[01:37:54] <cradek> and again what do you mean turn the machine on? do you mean switch on the machine, or "machine on" in tkemc, or something else?
[01:37:54] <sed_> didnt test 1... its kinda hard
[01:38:23] <sed_> turn machine on, when I it on in emc no error
[01:39:06] <cradek> ?
[01:39:16] <cradek> sorry I'm having trouble understanding what you're doing and seeing
[01:39:54] <sed_> from estop reset I hit f2 and get the error on either joint 0 or 2
[01:40:03] <cradek> ok
[01:40:09] <sed_> with the machine phisicly off no error
[01:40:13] <cradek> before you hit f2, all the numbers are green? yellow?
[01:40:19] <sed_> yea
[01:40:23] <cradek> which?
[01:40:34] <sed_> yellow
[01:40:41] <cradek> ok
[01:41:00] <cradek> when you turn machine on, that should enable the servo amps. does this cause your hardware to do something that would affect the limit switch signals?
[01:41:19] <sed_> didnt in emc1 and shouldnt anyway
[01:41:23] <cradek> well I take that back - it will do whatever you wired it up to do - the axis.N.amp-enable-out pins turn on
[01:42:10] <sed_> is there another lmit switch setting anywhere?
[01:43:10] <cradek> so when you hit f2, something happens that causes the limit signals to become wrong? what are the axis.N.amp-enable-out pins hooked to in HAL?
[01:44:13] <sed_> looking...
[01:47:16] <roh> hm.. sometimes i wonder.. why is axis uses tk as toolkit and not something.. eeeh.. 'recent'
[01:47:26] <sed_> axis.N.amp-enable-out does not exist in stg_io.hal, we use stg.out-00 for x axis
[01:48:55] <sed_> it is in core_servo...
[01:49:05] <cradek> roh: I guess the AXIS authors didn't have anyone around to tell them what they were doing wrong... they're probably just morons.
[01:49:39] <cradek> sed_: doesn't matter which file it's in. what's it hooked to?
[01:50:02] <roh> cradek no.. thats not what i am saying or asking.. just wondering a bit since tk.. well.. i actually do not know anybody who would choose it in person.. thats why i'm wondering
[01:50:02] <cradek> you can examine it easily with halshow or halcmd (multiple hal files gets confusing)
[01:51:12] <cradek> AXIS is several years old now - programmers choose toolkits based on their own whims - usually they choose whatever they know best. I doubt there's a better answer than that, really.
[01:51:40] <roh> ah.., ok. didnt know when axis was done. thought it was the 'new gui'
[01:51:51] <jmkasunich> I was just trying to recall exactly how old axis is, but too lazy to go digging it up
[01:51:57] <jmkasunich> roh: axis is the new GUI
[01:52:06] <cradek> also, no matter what a programmer picks, someone will think it's wrong, and tell you so - and actually, this does nobody much good.
[01:52:19] <cradek> it's the newest
[01:52:21] <cradek> :-)
[01:52:22] <jmkasunich> this is industrial control software we are talking about, not MP3 players - 3 years old is new
[01:52:34] <jmkasunich> emc(1) was originally developed in 1995
[01:52:44] <jmkasunich> emc2 was developed starting around 2003
[01:53:13] <roh> where does axis take its 'ui' from ? i thougt i'll find some tk-file with that.. or is it all hardcoded?
[01:53:51] <jmkasunich> I don't understand that question
[01:54:49] <roh> i dunno tk good enough how it works internally.. has it some 'layout/design' in a extra file like glade which then gets used by axis.py or are all widgets and their placement hardcoded?
[01:55:05] <jmkasunich> I believe the layout is determined by the code
[01:55:12] <cradek> AXIS builds up its screen in various ways - for instance it adds/removes some controls according to the machine capabilities
[01:55:14] <jmkasunich> it isn't some screen builder tool
[01:55:58] <cradek> right, you won't find a single file that represents the whole gui, but if there was one, axis.tcl would be it
[01:56:48] <sed_> cradek can I pm you?
[01:57:05] <cradek> yes
[01:57:31] <roh> cradek at.. thanks.. will take a look in there
[01:58:23] <jmkasunich> roh: what are you actually trying to do? just learn how it works? or is there some specific feature you want to add? or ?
[01:59:08] <roh> jmkasunich understand more how it works, gets started, what stuff execs what etc.. and why pyvcp stuff ends up on the right ;)
[02:07:54] <roh> jmkasunich some stuff would really be practical to put some pyvcp things into the manual control tab (i have free screenspace there) and not neccessarily on an extra bar on the right
[02:09:50] <roh> i think i found what i searched in line 3908 of axis.py
[02:11:43] <cradek> where are the rest of the hardware driver docs? This http://www.linuxcnc.org/docview/html//hal_drivers.html is very incomplete
[02:12:01] <cradek> sed_ is asking me about how the stg ports are numbered, and I'm not seeing it
[02:12:30] <jmkasunich> I dunno if anyone ever wrote a chapter for the stg driver
[02:12:47] <cradek> oh they're split out separately
[02:12:51] <jepler> http://www.linuxcnc.org/docview/html//drivers_servo_to_go.html
[02:13:11] <cradek> sed_: go to linuxcnc.org, click documentation, then html, then find the servo-to-go docs
[02:13:16] <cradek> or ... jepler's url
[02:14:18] <sed_> thanks brb
[02:14:19] <cradek> looks cryptic - more flexible than emc1 with regard to what's in/out
[02:14:49] <roh> some other thing: how can i limit the 'jog speed' for manual stepping via axis.n.jog-counts ?
[02:15:25] <roh> it seems to 'step' at max_velocity always
[02:15:36] <cradek> you're talking about using a jog wheel?
[02:16:00] <cradek> it's true it moves at maxvel for the jogwheel
[02:16:18] <cradek> you can smooth it out in a few ways - I like to use the ilowpass component for that
[02:16:24] <roh> i have 6 encoders on a parport which are connected to 6 encoder-hal-components of which 3 are connected from encoder.n.counts to axis.n.jog-counts
[02:17:00] <cradek> neat. try ilowpass in the counts. http://www.linuxcnc.org/docview/html//man/man9/ilowpass.9.html
[02:17:08] <roh> one goes through knob2float and then to all 3 axis.x.jog-scale for 'stepsize' and the other 2 are feed and spindle override
[02:17:49] <cradek> that doesn't answer your question (the answer is there's no way to limit that velocity other than the usual machine limits)
[02:18:27] <cradek> with jogwheels you really want the machine to keep up - that's the whole point
[02:19:04] <cradek> but on a fast machine it's not very smooth (way too responsive!) - you can fix that with ilowpass.
[02:19:19] <roh> i see. well.. it would have been nice to be able to 'switch it' from 'behave like manual G0' to 'same with G1 speed every tick'
[02:20:38] <roh> at feedrate 600 (on a metric machine) even 0.010mm makes a hard 'clock' sound and not a nice precise move i would like in material (which basically rules out manual stepping that way while in material)
[02:21:32] <cradek> ilowpass ilowpass ilowpass
[02:21:43] <jmkasunich> heh
[02:21:51] <roh> doesnt that just make my steps smaller then?
[02:22:20] <jmkasunich> it turns a one-count step from the wheel into 10 (or 100, or whatever) smaller steps, and spreads them out into a ramp
[02:22:21] <cradek> it gives you a continuous smooth position update, instead of a big step
[02:22:49] <sed_> how do we change the limit switch pings to read normaly open versus normaly closed and vise versa
[02:23:16] <cradek> if you want the inverse of an input, use the corresponding -in-not pin
[02:26:58] <roh> hm.. are there any docs about the 'timing behaviour' of hal components? i.e. how they scale (i(1) vs i(n) etc
[02:27:18] <sed_> for example: Xmaxlim <= stg.in-01-not would be NC and leaving the '-not' off would beNO
[02:28:13] <cradek> if your wiring presents a high signal to the STG when on the limit, you'd use in-01. if it presents a low signal to the STG when on the limit, you'd need -not.
[02:28:27] <cradek> NC and NO are not really the issue
[02:29:00] <roh> somehow i don't think hal_input, axis and the aduino reciever script should all eat 15-20% cpu each while axis is running.. something i do must be wrong
[02:29:04] <cradek> but this is easy - pick one - poke the switch - if it's backward, use the other.
[02:29:47] <sed_> we did that with 1 limit while the power to the servos was off and it works fine. as soon as power
[02:30:06] <sed_> to the servos is turned on and enabled we get the limit error
[02:30:42] <cradek> that could be an issue in your hal virtual wiring OR your real wiring
[02:30:50] <sed_> there is nothing wired to the STG card to tell it if the power is on or not and the amps are not wired for faults. they are only wired with the analog signals
[02:31:26] <cradek> that's scary - emc should be able to disable the amps
[02:31:52] <sed_> emc can disable the amps but not kill the power to them
[02:32:05] <cradek> otherwise when you get a following error and emc thinks it disables them, the machine won't necessarily stop (due to dac offset)
[02:32:20] <roh> build some simple chargepump. its just 2 transistors and a latch
[02:32:36] <sed_> i have the Estop wired directly to the power and emc is jumpered and does not know the state of the estop
[02:33:14] <sed_> no the amps will disable and stop motion
[02:33:56] <sed_> when i say kill power i mean the Ac power supply
[02:41:09] <sed_> bbl
[02:44:53] <roh> hum.. i don't get it.
[02:45:37] <roh> i added "loadusr -W hal_input -KRAL Jess Tech Dual Analog Rumble Pad" to custom.hal and i can see all the inputs then, but also it eats 20% cpu just being there, not even one pin of it connected
[03:20:55] <Jymmm> SWPadnos: http://newspaperads.mercurynews.com/ROP/ads.aspx?advid=32664&adid=8021897&subid=26171844&type=
[03:22:06] <Jymmm> SWPadnos: Spec near bottom... http://www.frys.com/product/5801953?site=sr:SEARCH:MAIN_RSLT_PG
[03:26:06] <SWPadnos> Jymmm, you can build a better machine for less from NewEgg
[03:26:23] <Jymmm> ok
[03:26:25] <SWPadnos> without Windows, of course
[03:39:50] <SWPadnos> Jymmm, check your email
[03:40:29] <Jymmm> looking...
[03:42:32] <Jymmm> SWPadnos: It was for you actually =)
[03:42:42] <Jymmm> less the hp part
[03:43:16] <SWPadnos> heh
[03:43:24] <SWPadnos> I don't need one, thanks :)
[03:43:30] <Jymmm> k
[03:43:42] <SWPadnos> I just got that list for my sister, and it's nice and quick
[03:43:52] <SWPadnos> Ubuntu 9.04 installed without a hitch and detected everything
[03:44:06] <Jymmm> cool
[03:46:35] <Jymmm> SWPadnos: I just have way too many computers here to buy another right now.
[03:46:46] <SWPadnos> funny, I have the same problem
[03:47:27] <roh> * roh just built the worlds most complicated and overengineered halogen-dimmer
[03:47:35] <SWPadnos> I thikn I have 9 complete computers in my office right now (not including the ones I have here that aren't mine)
[03:47:50] <roh> but somehow one needs to play with pid loops :)
[03:47:54] <Jymmm> SWPadnos: What I need is storage, lots and lots of storage
[03:48:06] <SWPadnos> I've only got about 6 or 7 TB in here
[03:48:17] <Jymmm> total?
[03:48:19] <SWPadnos> but I don't download movies, so it's a lot
[03:48:22] <SWPadnos> yes
[03:48:53] <Jymmm> well i have that if i add em aall up, but I need it al in one spot
[03:49:30] <SWPadnos> one machine has 4x1TB drives in it, in a RAID5+hot spare configuration (so only 2TB usable)
[03:50:11] <roh> sounds like a usecase for cheap nas appliances with hotplug sata which either do use iscsi to some raid/volume managing box or get used directly as raids in 'chunks'
[03:50:53] <Jymmm> SWPadnos: Sounds like roh is offering us free $700 RAID cards
[03:51:32] <roh> Jymmm they do sw-raid in there. hw raid isnt faster than a dedicated generic purpose cpu. small intel cpus
[03:52:00] <roh> also means the disk layout is default linux.. one can recover easily
[03:52:01] <SWPadnos> I have a RAID card, thank you :)
[03:52:15] <Jymmm> SWPadnos: which?
[03:52:20] <SWPadnos> hmm
[03:52:24] <SWPadnos> 3Ware I think
[03:52:29] <SWPadnos> it's been a while
[03:52:34] <roh> SWPadnos do you have 2 of the same hw revision and or specs about the on-disk layout?
[03:52:44] <Jymmm> SWPadnos: pata or sata?
[03:52:46] <SWPadnos> no
[03:52:48] <SWPadnos> SATA
[03:52:50] <SWPadnos> 2
[03:53:59] <roh> ive seen really nasty recovery scenarios where some controller died and one couldnt just replace it anymore since the ondisk layout changed in new products and the old ones vanished
[03:54:17] <SWPadnos> ah yes, 3Ware 9650SE
[03:54:24] <SWPadnos> rigth
[03:54:28] <SWPadnos> right
[03:54:42] <roh> with 3ware there is a chance.. they are pretty conservative and 'well hung'
[03:54:46] <SWPadnos> there have been a lot of problems, and I'm not sure if this one is much better
[03:54:50] <SWPadnos> heh
[03:55:06] <SWPadnos> there are Linux management tools for this one also, I believe
[03:55:08] <Jymmm> SWPadnos: lot of problems with what?
[03:55:12] <SWPadnos> that was a factor in choosing it
[03:55:26] <SWPadnos> RAID cards and actual disasters
[03:55:45] <Jymmm> SWPadnos: Ah, that's why I said $700 raid card and not $49 one
[03:55:53] <SWPadnos> no matter really
[03:56:12] <SWPadnos> on some HP server hardware, you couldn't even move a disk from one controller port to another
[03:56:14] <Jymmm> SWPadnos: well no mobo has that many ports
[03:56:19] <SWPadnos> and those were $1k cards
[03:56:31] <SWPadnos> I have a mobo with 10 or 12 ports
[03:56:33] <roh> SWPadnos jap. but if one needs to decide if to buy some 500$pc and a 300$ controller + disks or some 600$box and some disks, i'd vote for the second. simpler, cheaper, easier to hack and modify. KISS principle applies aslong as performance survices (home useage <30mbyte/sec)
[03:56:39] <Jymmm> SWPadnos: ext box?
[03:56:40] <SWPadnos> they can be used with software RAID just fine
[03:56:45] <SWPadnos> no, a motherboard
[03:57:05] <Jymmm> SWPadnos: 12 SATA connectors?
[03:57:11] <SWPadnos> sure, a second PC is great
[03:57:21] <SWPadnos> Jymmm, yes, 10 or 12, some large number
[03:57:41] <roh> SWPadnos if you can find some board which has all conns in _hot_pluggable ;)
[03:58:01] <SWPadnos> nope, I'm betting they aren't ;)
[03:58:36] <roh> some could also use the marvell stuff like the raid cards and the appliances. its cheap chips which are quite fast.
[03:58:44] <roh> and do hotplug
[03:59:01] <SWPadnos> ah, 10 ports: http://www.newegg.com/Product/Product.aspx?Item=N82E16813128377
[03:59:07] <SWPadnos> I thought it had a bunch of them
[04:00:38] <Jymmm> what's "GSTAT"?
[04:00:47] <Jymmm> GSATA
[04:01:10] <SWPadnos> no idea
[04:01:22] <Jymmm> it's the white connectors on that board
[04:01:32] <roh> seems to be 6x sata chipset and 4x sata from some JM322 chips (i guess pci-express sata)
[04:01:33] <Jymmm> the blue ones are SATA
[04:01:51] <Jymmm> http://www.newegg.com/Product/ShowImage.aspx?ISList=13-128-377-S01%2c13-128-377-S02%2c13-128-377-S03%2c13-128-377-S04%2c13-128-377-S05&S7ImageFlag=1&Item=N82E16813128377&Depa=0&WaterMark=1&Description=GIGABYTE%20GA-MA790FXT-UD5P%20ATX%20DDR3%20AMD%20Motherboard%20-%20Retail
[04:01:52] <SWPadnos> that's probably their name for the RAID ports
[04:03:28] <roh> hm.. does only halgui use psyco?
[04:05:41] <SWPadnos> no idea
[04:05:57] <roh> rgrep doesnt find any other reference
[04:06:22] <roh> somehow my hal_input and some other python userspace hal modules eat a lot of cputime
[04:12:45] <Jymmm> Hmmm, my dvd drive is hot enough to almost fry an egg
[04:17:59] <Jymmm> !!! MAJIDE !!!!
[06:23:51] <sed_> (in-00 is PORTA.0, out-15 is PORTD.7).
[06:24:04] <sed_> does that make sence to anyone?
[06:24:19] <sed_> came from this sentence
[06:24:53] <sed_> For each pin, <channel> is the axis number, and <pinnum> is the logic pin number of the STGif IIOO is defined, there are 16 input pins (in-00 .. in-15) and 16 output pins (out-00 .. out-15), and they correspond to PORTs ABCD (in-00 is PORTA.0, out-15 is PORTD.7).
[06:26:11] <SWPadnos> sed_, the way I read that, you have control over the split between input and output, probably in byte-sized chunks
[06:26:14] <sed_> I dont understand how out-15 is PORTD.7
[06:26:40] <SWPadnos> the inputs are numbered sequentially and the outputs are numbered sequentially, but they're separate sequences
[06:27:05] <sed_> but why is one on port A and the other D each port is 50 pins
[06:27:05] <SWPadnos> ok, three 4 characters in the control string (IIOO in the example)
[06:27:07] <SWPadnos> there are
[06:27:14] <SWPadnos> what?
[06:27:29] <SWPadnos> on an STG?
[06:27:31] <sed_> thats how I read it.. there are ports A-D
[06:27:32] <sed_> yea
[06:28:13] <SWPadnos> which do you have, STG1 or STG2?
[06:28:21] <sed_> stg2
[06:29:17] <sed_> http://www.linuxcnc.org/docview/html//drivers_servo_to_go.html#r1_1
[06:29:20] <SWPadnos> the connectors aren't called PORTxx, they're called P1-P4
[06:29:22] <sed_> that is what Iam reading
[06:29:30] <SWPadnos> don't confuse P1 with PORTA
[06:29:43] <SWPadnos> P1 has all the digital I/O on it
[06:29:58] <SWPadnos> err, it has the first 24 bits of I/O on it
[06:30:20] <SWPadnos> P2 has 8 analog inputs, 8 I/O pins, and 8 motor direction outputs
[06:30:37] <SWPadnos> I'm looking at this hardware manual: http://www.servotogo.com/dwnld/hardman2.pdf
[06:31:10] <sed_> hmm
[06:31:32] <SWPadnos> note that on page 10, they describe the I/O ports
[06:31:44] <SWPadnos> (well, starting on page 10)
[06:32:23] <SWPadnos> ah, page 12 shows the pin assignments
[06:32:28] <SWPadnos> for I/O
[06:33:35] <sed_> reading....
[06:33:46] <SWPadnos> have fun. I'm headed to bed :)
[06:35:09] <sed_> thanks for the tip
[07:38:09] <pjm__> pjm__ is now known as pjm
[07:38:39] <pjm> pjm is now known as Guest79472
[07:46:56] <sed_> anyone up?
[07:47:43] <Jymmm> no
[07:47:54] <archivist> we are asleep at work
[07:49:21] <sed_> heh
[07:49:47] <archivist> sed_, the general idea on irc is put your real question people do lurk and answer when/if they can
[07:49:59] <sed_> sorry
[07:50:21] <sed_> can anyone explain how the stg ports are numberd in hal? I just dont get it
[07:51:59] <sed_> For each pin, <channel> is the axis number, and <pinnum> is the logic pin number of the STGif IIOO is defined, there are 16 input pins (in-00 .. in-15) and 16 output pins (out-00 .. out-15), and they correspond to PORTs ABCD (in-00 is PORTA.0, out-15 is PORTD.7).
[07:52:11] <sed_> I cant make sence of that sentance
[07:53:26] <archivist> which part,
[07:54:11] <archivist> think in small steps it should become clear
[07:54:13] <sed_> how hal define the pin on the port
[07:54:41] <archivist> http://www.linuxcnc.org/docview/devel/html//drivers_servo_to_go.html
[07:54:50] <archivist> that page?
[07:54:59] <sed_> yea
[07:55:36] <sed_> http://www.linuxcnc.org/docview/html//drivers_servo_to_go.html
[07:55:46] <sed_> yea same link
[07:57:04] <archivist> fill in the blanks as needed for your board
[07:57:28] <sed_> I have a stg 2
[07:57:53] <archivist> that means nothing to me, I only know what I see on that page
[08:00:19] <sed_> well for instance how do you tell pin1 on port 1 from pin 1 on port 4?
[08:02:40] <archivist> because A has 0-7 B 8-15 in and C 0-7 and D 8-15 out
[08:03:07] <archivist> A=port 1 etc
[08:03:22] <archivist> * archivist assumes
[08:03:32] <archivist> and docs say
[08:06:03] <sed_> each port or what I assume is aport is a 50 pin connecter on the board P1-P4
[08:10:50] <sed_> seems to me they are talking about pins 1-16 on a single port
[08:11:29] <sed_> so how do I define the second third or fourth 50 pin connecter?
[08:11:34] <archivist> I assume from that line in the docs that a port is 8 bits
[08:12:01] <sed_> so each port is only 8 bits and has 50 pins?
[08:12:24] <archivist> you have the board and docs, you tell me
[08:13:29] <sed_> P1Furthest right, has a 5024 bits of digital I/O (Ports A, B, and C)
[08:14:08] <sed_> 5024 should be 50pin 24 pit
[08:14:10] <sed_> bit
[08:14:41] <archivist> 8 *3 =24 A B C
[08:14:56] <archivist> so that is correct
[08:19:09] <sed_> ok I think I get it
[08:45:59] <sed_> one thing I dont get is ever other pin is a ground so only 1,3,5,7,9.. do anything
[08:47:17] <sed_> so when it refers to 24 bits is each ping a pair of pins bit0=1,2 bit1=3,4 ect?
[08:48:36] <archivist> the pdf has each and every pin defined
[08:49:24] <archivist> there are multiple ground pins
[08:51:35] <sed_> yea on connecter 1 ever even pin is a ground, but how do the ground relate to bits?
[08:55:54] <archivist> a voltage is in relation to another, so ground is the reference for a bit/input
[09:29:30] <sed_> I mapped out the pins in a chart according to my understanding.
[09:29:31] <sed_> Connecter 1
[09:29:32] <sed_> PortA.0 in-00 Opto-23, C7 pin 1
[09:29:32] <sed_> PortA.1 in-01 Opto-22, C6 pin 3
[09:29:32] <sed_> PortA.2 in-02 Opto-21, C5 pin 5
[09:29:34] <sed_> PortA.3 in-03 Opto-20, C4 pin 7
[09:29:37] <sed_> PortA.4 in-04 Opto-19, C3 pin 9
[09:29:39] <sed_> PortA.5 in-05 Opto-18, C2 pin 11
[09:29:42] <sed_> PortA.6 in-06 Opto-17, C1 pin 13
[09:29:44] <sed_> PortA.7 in-07 Opto-16, C0 pin 15
[09:29:47] <sed_> PortB.0 in-08 Opto-15, B7 pin 17
[09:29:49] <sed_> PortB.1 in-09 Opto-14, B6 pin 19
[09:29:52] <sed_> PortB.2 in-11 Opto-13, B5 pin 21
[09:29:54] <sed_> PortB.3 in-12 Opto-12, B4 pin 23
[09:29:57] <sed_> PortB.4 in-13 Opto-11, B3 pin 25
[09:29:59] <sed_> PortB.5 in-14 Opto-10, B2 pin 27
[09:30:02] <sed_> PortB.6 in-15 Opto-9, B1 pin 29
[09:30:04] <sed_> PortB.7 in-16 Opto-8, B0 pin 31
[09:30:07] <sed_> PortC.0 out-01 Opto-7, A7 pin 33
[09:30:09] <sed_> PortC.1 out-02 Opto-6, A6 pin 35
[09:30:11] <sed_> PortC.2 out-03 Opto-5, A5 pin 37
[09:30:14] <sed_> PortC.3 out-04 Opto-4, A4 pin 39
[09:30:17] <sed_> PortC.4 out-05 Opto-3, A3 pin 41
[09:30:19] <sed_> PortC.5 out-06 Opto-2, A2 pin 43
[09:30:22] <sed_> PortC.6 out-07 Opto-1, A1 pin 45
[09:30:24] <sed_> PortC.7 out-08 Opto-0, A0 pin 47
[09:30:27] <sed_> +5V pin 49
[09:30:29] <sed_> Connecter 2
[09:30:32] <sed_> PortA.0 out-09 Opto-15, D7 pin 17
[09:30:34] <sed_> PortA.1 out-10 Opto-14, D6 pin 19
[09:30:37] <sed_> PortA.2 out-11 Opto-13, D5 pin 21
[09:30:39] <sed_> PortA.3 out-12 Opto-12, D4 pin 23
[09:30:42] <sed_> PortA.4 out-13 Opto-11, D3 pin 25
[09:30:42] <archivist> pastebin
[09:30:44] <sed_> PortA.5 out-14 Opto-10, D2 pin 27
[09:30:47] <sed_> PortA.6 out-15 Opto-9, D1 pin 29
[09:30:49] <sed_> PortA.7 out-16 Opto-8, D0 pin 31
[09:30:54] <sed_> sorry
[09:30:57] <sed_> http://pastebin.com/m54e038f2
[09:30:59] <sed_> I ment to paste the link
[09:32:02] <sed_> yea that was the paste to paste bin, the link didnt copy to the clipboard....
[09:32:09] <sed_> very sorry
[09:33:47] <sed_> * sed_ staples note to head, check clipboard befor pasting to channel...
[10:09:31] <sed_> if somone has time to check please pm me if I got it right, again sorry for the paste..
[11:36:34] <Itsmee> Bit of a stupid question im presuming but is there anyway to instal emc2 onto ubuntu 9? keep saying about python needing <2.6 and mine is over that
[11:44:49] <EbiDK> EbiDK is now known as EbiDK|AWAY
[11:53:44] <Itsmee> no one any ideas?
[11:54:12] <archivist> few
[11:55:19] <Itsmee> Anything that could help me?
[11:56:55] <archivist> been some mentions in here, look at the logs, and look at the mailing list
[11:57:27] <archivist> logger_emc: bookmark
[11:57:27] <archivist> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2009-07-02.txt
[13:24:12] <robh_> hi, is it possible to do gear hobbing with EMC? sync my indexer (4th axis) to spindle speed
[13:26:40] <skunkworks_> there is spindle synced motion http://www.linuxcnc.org/docs/2.3/html/gcode_main.html#sec:G33,-G33.1:-Spindle-Synchronized
[13:26:48] <skunkworks_> what you do with it...
[13:28:19] <skunkworks_> it is pretty powerful - (you can do more than just threading) http://www.youtube.com/watch?v=ACvRilmIKDQ
[13:28:43] <robh_> i saw X Y Z listed as axis, so it would take A axis command
[13:34:07] <archivist> robh_, I use real hobbing machines and I can see needs that should be added at a deeper level
[13:35:07] <robh_> i just want to cut a spur gear
[13:35:42] <archivist> I want to do it in EMC to one day, but the cycle needs a stop but the spindles remain in gear
[13:37:00] <robh_> i guess its back to old idea, do each tooth at once with a cutter
[13:37:06] <archivist> there are a few things to take into account 5xais is really needed for the hob angle to be included
[13:38:24] <skunkworks_> it doesn't explicitly say it can't use abc... That would be a question for chris
[13:39:29] <archivist> skunkworks, its the pas, stop measure start again that is a killer
[13:39:40] <robh_> archivist, out of question, do you know any good places to get hold of module involute gear cutters in uk?
[13:40:17] <archivist> yup arc euro trade £25 ish
[13:40:47] <archivist> for hobs dunno about other pricing at the moment
[13:41:05] <archivist> robh_, what do you need in what material
[13:42:03] <robh_> 1.5mod
[13:42:09] <archivist> because off the shelf gears can be cheaper than making
[13:42:10] <robh_> hss would do
[13:45:33] <robh_> i did find, www.precisioncuttingtools.co.uk & www.crtools.co.uk ill drop them call later i think
[13:46:43] <archivist> plain milling gear cutter arc euro trade dont have
[13:46:55] <archivist> rdg only had mod 1
[13:47:48] <robh_> rdg i had emailed, tracytools can get them in
[13:50:41] <archivist> robh_, j&L industrial
[13:51:29] <archivist> hmm not them
[13:51:30] <robh_> to order i think they where, same as cromwelltools
[13:52:03] <robh_> what is it with UK companys, never make them selfs too easy to find on web
[13:52:17] <archivist> heh true
[13:52:57] <archivist> just get ready made from HPC
[13:53:42] <robh_> just not as fun hehe
[13:54:04] <robh_> need make 50 or so realy, then have to do some midding in backof them
[13:54:45] <archivist> hmm best hobbed at that qty
[13:55:31] <robh_> ye, y i think nowi wunder if my new EMC machine could do it,
[13:56:04] <robh_> about 88mm od, 14mm thick
[13:56:06] <archivist> my machine is nearly there, needs spindle encoder and updated interface like mesa card
[13:56:22] <robh_> i got order some more mesa cards soon
[13:59:02] <archivist> ready made from HPC $24 ish
[13:59:07] <archivist> £24
[14:58:22] <piasdom> g'mornin all
[15:13:54] <Vq^> evening
[16:17:52] <geo01005> So I need to sense an analog signal on this machine I'm building, but the signal needs to be isolated from the control system.
[16:18:16] <geo01005> Does anybody know of a high frequency analog isolation module I can use?
[16:18:56] <SWPadnos> define high frequency
[16:19:41] <geo01005> ~100k
[16:19:52] <archivist> heh near dc
[16:19:53] <geo01005> 10k would be alright I suppose.
[16:20:34] <SWPadnos> ok, so you're using this signal in an analog control circuit, not in software
[16:20:36] <SWPadnos> ?
[16:21:36] <geo01005> Well it would make the most sense to convert the signal to digital, and then just send the digital signal over isolated channels...
[16:21:47] <SWPadnos> not at 100 kHz
[16:21:51] <SWPadnos> not necessarily anyway
[16:22:07] <geo01005> Let me restate what I need.
[16:22:10] <SWPadnos> heh
[16:22:13] <geo01005> I'm not sampling at 100k
[16:22:17] <archivist> accuracy needed
[16:22:25] <SWPadnos> ok, that's cricital information :)
[16:22:33] <geo01005> I'm planning to sample at about 1-5k
[16:22:55] <SWPadnos> when you say "high frequency" and "analog" in the same sentence, there's no reason to think of the frequency as a bitrate :)
[16:22:57] <geo01005> but I don't want the signal to be heavily filtered.
[16:23:24] <SWPadnos> this is a THC or similar?
[16:23:31] <geo01005> arc voltage.
[16:23:33] <SWPadnos> ok
[16:23:50] <archivist> * archivist contemplates method used in switchmode supplies
[16:24:09] <geo01005> I have seen these guys:http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=AD202KY-ND
[16:24:25] <geo01005> looks like it would work ok.
[16:25:30] <geo01005> only 5kHz bandwidth...
[16:25:50] <SWPadnos> it says it has 2 kHz GBW
[16:26:38] <SWPadnos> ok, the 204 has 5 kHz BW
[16:28:20] <geo01005> If I'm sampling at 1-5kHz, does it make sense to try to get a bandwidth much higher than 5kHz?
[16:28:36] <SWPadnos> no
[16:28:58] <SWPadnos> in fact, if you sample at 5 kHz, you probably want a filter that's sonewhat lower than 2.5 kHz
[16:29:04] <SWPadnos> somewhat
[16:29:33] <SWPadnos> you have nyquist issues above Fsample/2, and you have aliasing issues at Fsample/2
[16:29:35] <geo01005> I see.
[16:30:05] <SWPadnos> so you want the input bandwidth to be below any frequencies that are likely to alia
[16:30:06] <SWPadnos> s
[16:30:21] <SWPadnos> err, s/be/cut off/
[16:32:13] <geo01005> I have used these for process control stuff before, but I don't think it is probably to heavily filtered and have some delay: http://web1.automationdirect.com/static/specs/fcsignalconditioners.pdf
[16:35:17] <geo01005> I do think it is to heavily filtered... I can't type.
[16:35:43] <SWPadnos> there are opto-22 analog I/O modules as well
[16:36:27] <SWPadnos> gotta run. see you later
[16:36:46] <jepler> some days, I feel like quitting would be a success
[16:37:05] <BJT-Work> :)
[16:39:34] <archivist> hmm opto-22 might be what Im looking for for a job
[16:55:41] <BJT-Work> yea my new egg is here
[16:56:09] <archivist> break it and make omelette
[16:56:55] <BJT-Work> that's why I got a new one the last one broke
[16:58:07] <EbiDK> EbiDK is now known as EbiDK|AWAY
[17:10:57] <Guest342> I have a quick question regarding the emc controller. Can the emc controller simultaneously move 3 or more axises? e.g. G00 X-0.54 Y1.21 Z0.1
[17:12:26] <pjm__> Guest342 yes it can
[17:12:32] <Guest342> The router machine I'm currently using has a controller that can not use 3 axis at the same time. When you issue a 3 aixs move it will move the z-axis then it will move the xy
[17:12:52] <pjm__> u need to replace your controller with EMC2 then!
[17:13:10] <archivist> Guest342, up to 9
[17:13:15] <Guest342> What is the maximum number of axises controlled?
[17:13:15] <archivist> I have 5
[17:13:26] <archivist> 9 axis
[17:13:28] <Guest342> thank you, looks like I need to investigate emc2
[17:13:29] <pjm__> i have 3 currently plus one in production
[17:13:58] <Guest342> Thanks, ill send some photos later
[17:14:17] <archivist> nothing else around at the price can match
[17:15:20] <archivist> pjm__, what are you doing for the 4th?
[17:15:34] <pjm__> mod'ing my vertex rotary table!
[17:15:40] <pjm__> i just need to get round to it
[17:15:53] <pjm__> things keep getting in the way
[17:16:02] <pjm__> like friends of mine wanting me to machine things for them!
[17:16:07] <archivist> ah, nip to the shop for some round tuits
[17:16:17] <pjm__> hah
[17:16:26] <pjm__> i will be able to make them once my 4th axis is done!
[17:17:31] <archivist> play time.. back in a few
[17:17:43] <pjm__> ok
[17:49:31] <SWPadnos_> SWPadnos_ is now known as SWPadnos
[18:24:17] <frallzor> yo
[18:26:16] <frallzor> having an issue when trying to save stepconf from the installed ubuntupackage, gives me an error saying something about lines 1810 and 1071 in the stepconf
[18:26:50] <cradek> please pastebin the full error
[18:26:52] <skunkworks_> can you pastebin.ca the error?
[18:27:11] <frallzor> its on another comp but ill try get networking working on it
[18:28:03] <frallzor> http://pastebin.com/m65edc90
[18:28:06] <cradek> great, pretty impossible to help otherwise
[18:28:28] <cradek> you have no ~/Desktop directory
[18:28:32] <frallzor> I just installed the live-cd edition and started to conf
[18:28:54] <frallzor> what should I do then?
[18:30:41] <cradek> after you installed the live cd did you do anything else special? anything to gnome? change window managers?
[18:30:50] <frallzor> nope
[18:30:57] <frallzor> installed, and then nothing
[18:31:04] <cradek> in stepconf what did you name your machine?
[18:31:24] <cradek> on the first page "basic machine information"
[18:31:33] <frallzor> didnt change, just left my-mill stay for now
[18:31:46] <frallzor> change it?
[18:31:51] <cradek> huh
[18:31:54] <cradek> no, that should be fine
[18:32:04] <cradek> in your home directory do you have a Desktop directory?
[18:32:37] <cradek> I just tried stepconf with emc 2.3.1 and it worked fine
[18:32:38] <frallzor> oh im really new to linux too btw
[18:33:04] <frallzor> in home all i have is a cnc dir
[18:33:12] <frallzor> hmm
[18:33:13] <frallzor> ahh
[18:33:18] <frallzor> desktop
[18:33:20] <cradek> what do you mean a cnc dir?
[18:33:32] <frallzor> nm, cnc is my compname
[18:33:33] <frallzor> but
[18:33:36] <frallzor> Im using swedish
[18:33:45] <frallzor> desktopfolder is named skrivbord
[18:33:51] <cradek> ohhhhhh
[18:33:58] <cradek> that's a stepconf bug then
[18:34:16] <cradek> on the initial stepconf page, uncheck "create a desktop shortcut"
[18:34:43] <jepler> and probably "create a desktop launcher" as well
[18:34:47] <frallzor> damn
[18:34:49] <frallzor> ah ok
[18:34:59] <frallzor> gave me error with launcher only too
[18:35:16] <frallzor> now it worked
[18:35:28] <cradek> thanks for the bug report
[18:35:43] <frallzor> np :P
[18:36:04] <jepler> open up a terminal, and cut and paste these lines into it:
[18:36:04] <jepler> test -f ${XDG_CONFIG_HOME:-~/.config}/user-dirs.dirs && source ${XDG
[18:36:05] <jepler> _CONFIG_HOME:-~/.config}/user-dirs.dirs
[18:36:05] <jepler> echo ${XDG_DESKTOP_DIR:-$HOME/Desktop
[18:36:24] <jepler> argh, that's a bad paste, hold on
[18:36:24] <cradek> you got a wrap you didn't want
[18:36:32] <frallzor> care to pastebin them for me
[18:36:36] <jepler> vyes jas
[18:36:37] <frallzor> so i can open on the linuxcomp
[18:36:54] <jepler> http://pastebin.ca/1482172
[18:37:13] <jepler> after the second line, does it print the right path for your "Desktop" directory?
[18:37:38] <jepler> argh, it's still a bad paste
[18:38:31] <frallzor> take your time =)
[18:38:45] <jepler> http://pastebin.ca/1482174
[18:39:37] <frallzor> it does
[18:40:08] <jepler> OK
[18:40:20] <jepler> thanks for testing
[18:41:07] <frallzor> should the stepconf work now then?
[18:41:13] <frallzor> with the links and stuff
[18:41:39] <jepler> no, but those lines are part of the bugfix I haven't actually put into the next release yet
[18:41:49] <frallzor> ah ok
[18:41:55] <jepler> you're using the Ubuntu 8.04 + emc2 live cd downloaded from linuxcnc.org?
[18:42:00] <frallzor> yes
[18:42:18] <jepler> I was under the impression that this didn't affect 8.04, so it was low on my list of priorities
[18:42:22] <jepler> now I know otherwise
[18:42:27] <frallzor> =)
[18:42:55] <frallzor> btw, got a nice tip for cam for this ubuntu?=
[18:43:52] <jepler> some people have collected these links about it: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Cam
[18:43:54] <frallzor> brand new to linux and its toys =)
[18:44:05] <frallzor> ah nice
[18:44:27] <jepler> I use eagle for milling circuit boards, so most of those cam type programs are irrelevant for me
[18:44:36] <jepler> I hope you can find something that's right for the kind of projects you have in mind
[18:45:31] <cradek> for 2.5D, you can use qcad + sheetcam (not Free) or qcad + a number of Free cam programs that read dxf
[18:45:42] <cradek> I wish I knew what the good ones are, but I don't
[18:46:01] <frallzor> ill have to try =)
[18:46:25] <cradek> many things to examine listed here: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Cam
[18:46:43] <geo01005> you could try HeeksCAD
[18:46:58] <geo01005> Still in development, but very useful.
[18:49:47] <geo01005> http://code.google.com/p/heekscad/
[18:50:47] <frallzor> looks nice
[18:50:50] <jepler> I need to give it another try
[18:51:04] <jepler> I checked out the svn, and for some reason the display had text scribbled all over it
[18:51:28] <jepler> I guessed it was some debugging/testing thing.. (this was a few weeks ago)
[18:51:33] <geo01005> jepler 64 bit?
[18:51:36] <jepler> yes
[18:52:07] <geo01005> I think that is fixed now. archivist, do you still have that problem?
[18:52:42] <geo01005> cool new constrained sketching in well in the works now.
[18:52:58] <archivist_attic> not checked out for a few days will check later
[19:48:54] <sed_> I revized my stg pin out chart http://pastebin.com/m2ebd0cb6
[19:49:18] <sed_> if anyone has time to take a look
[19:50:16] <archivist_attic> sed_, we have to get the manual to double check else means not too much
[19:51:29] <sed_> http://www.servotogo.com/dwnld/hardman2.pdf
[19:51:33] <sed_> page 10+
[20:07:04] <sed_> it apears that the linksp command in the sample files is obsolete, and replaced by net, is that right?
[20:08:46] <cradek> yes
[20:08:50] <frallzor> btw, a Q about the samples
[20:09:01] <frallzor> the penguin 3d mill
[20:09:14] <frallzor> why wont it run when just playing around?
[20:09:26] <cradek> ?
[20:09:46] <frallzor> the code wont run
[20:09:50] <archivist_attic> you should see an error
[20:09:54] <frallzor> but the other 3d sample works
[20:09:59] <cradek> what is the error?
[20:10:16] <frallzor> ill check but im pretty sure it just didnt run
[20:10:19] <frallzor> nothing more
[20:10:53] <frallzor> booint lappy again
[20:11:00] <frallzor> *booting
[20:12:29] <frallzor> using a preset for the sherline
[20:12:40] <frallzor> 4 axis
[20:13:21] <frallzor> yup, it just doesnt do anything
[20:13:24] <frallzor> no errors
[20:15:11] <frallzor> might be that it just doesnt like the sherline 4axis preset?
[20:16:32] <archivist_attic> unlikely, does it fit the travels
[20:17:26] <frallzor> I dont know that, i tried my own present before then it complained it didnt fit travels
[20:17:38] <frallzor> i assumed it should fit the sherline then due to no errors
[20:18:04] <archivist_attic> it should give an error is too large
[20:19:44] <frallzor> its too large to the sherline ?
[20:26:42] <archivist_attic> I have no idea can look here
[20:26:46] <archivist_attic> cant
[20:27:39] <frallzor> its very odd =)
[21:07:48] <sed_> can you change the state of a .out pin to True or Fales? -not does not work
[21:10:43] <frallzor> ah lovely, first parts tomorrow for my cnc =)
[21:10:51] <frallzor> or rather cnc-mill
[21:14:44] <toastydeath> just a cnc mill?
[21:14:46] <toastydeath> unacceptable
[21:14:50] <toastydeath> go cnc full or nothing
[21:15:07] <toastydeath> no, i kid, congradulions =)
[21:15:09] <frallzor> at least itll be 4-axis :P
[21:15:19] <archivist> 5 rulz
[21:15:24] <toastydeath> 12 axis
[21:16:03] <toastydeath> PEDESTRIANS
[21:17:09] <sed_> i changed net Xminlim <= stg.in-02 to net Xminlim <= stg.in-02-not did I need to change "<=" as well?
[21:17:39] <frallzor> oh i have a request/question for the software :P
[21:18:06] <frallzor> does it support 3dconnexioncontroller or could it be implemented =)
[21:18:20] <geo01005> I wish...
[21:18:33] <frallzor> its the ultimate way to control 4 axis
[21:18:41] <frallzor> manually that is
[21:19:01] <toastydeath> that would be quite frightening
[21:19:13] <toastydeath> i only move rotaries with MDI
[21:19:23] <toastydeath> sometimes an MPG
[21:20:13] <frallzor> it would rock =) a little screen shows the values and you can decide speed by just using the controller too =)
[21:23:22] <toastydeath> just like, from experience here
[21:23:40] <toastydeath> i really suggest getting/using an mpg over other styles of controlling manual machine movement
[21:24:22] <toastydeath> moving multiple axes at the same time is a recipe for disaster 80% of the time
[21:26:10] <archivist> fun programming 3 when one is a rotary at an angle
[21:31:49] <frallzor> how much is a mpg?
[21:31:56] <frallzor> basic one
[21:32:13] <cradek> $40ish?
[21:32:14] <SWPadnos> $5 to $500
[21:32:24] <frallzor> where to get the $5 one? :P
[21:32:29] <toastydeath> uh
[21:32:32] <archivist> you make it
[21:32:34] <toastydeath> you're better off with a keyboard
[21:32:35] <SWPadnos> cardboard and an optical sensor
[21:32:39] <toastydeath> than a cheap mpg
[21:32:57] <frallzor> ah make one, i can do that =)
[21:33:22] <archivist> yup build a cnc , make accessories
[21:35:57] <geo01005> it might be hard to get the detention feel without spending the $60 that it would take to buy a decent one.
[21:36:16] <frallzor> ill just save for one
[21:36:41] <toastydeath> one of our machines has a cheap mpg and not only doesn't it have clear detents, the machine doesn't go 100 pulses when you crank it 100 pulses
[21:36:51] <toastydeath> if you crank it quick around once, it might only go 30
[21:36:55] <SWPadnos> I'd check ebay for good ones. they will be in the $50+ range most likely, but there's nothing like good equipment
[21:38:19] <cradek> $50-100 is common
[21:38:30] <cradek> the new imports are a little cheaper
[21:39:59] <frallzor> I did allmost cheap out on the controller for the stuff =)
[21:40:22] <cradek> or not - this one's $59: http://www.cnc4pc.com/Store/osc/product_info.php?cPath=40&products_id=163
[21:40:24] <frallzor> the site I got motors from hade one of those you can find plenty on ebay
[21:41:41] <frallzor> but my mind came to its sense =)
[21:45:36] <frallzor> what ya all using electronicswise btw?
[21:45:51] <cradek> which part?
[21:46:01] <frallzor> controller
[21:46:56] <frallzor> controller/driver
[21:47:42] <cradek> my lathe retrofit has an old pc motherboard, mesa 5i20 + 7i33 + 7i37 + 7i37, original servo amps etc
[21:48:12] <cradek> oh also two of the pico systems resolver-to-quadrature boards
[21:52:49] <Spida> cradek: FOUR mesa boards? why?
[21:53:16] <cradek> one is a servo/encoder interface, the other two are isolated IO
[21:53:18] <alex_joni> Spida: look them up ;)
[21:54:46] <Spida> ah
[21:54:49] <archivist> old P4 and parallel port to Chinese stepper driver modules
[21:55:17] <cradek> yeah I have a few other machines that are parallel port based only
[21:55:18] <Spida> alex_joni: didn't realize they made stepper/servo drivers to.
[21:55:33] <Spida> alex_joni: I had only seen their anything-ios
[21:55:37] <cradek> mesa makes all sorts of useful stuff
[21:55:52] <cradek> pico systems does too
[22:01:40] <frallzor> hello DaViruz
[22:03:39] <DaViruz> good evening
[22:14:11] <dmess> i'm using some Mexican made boards...
[22:14:30] <dmess> and a few cnc4pc boards
[22:19:10] <EbiDK|AWAY> I lost the game....
[22:46:40] <alex_joni> g'night all