#emc | Logs for 2007-06-28

[01:50:26] <CIA-8> 03cradek 07TRUNK * 10emc2/src/hal/components/Submakefile: s32, not u32, is the one that can go negative
[01:51:55] <Ziegler> got some vids of my machine running at about 30%
[01:52:08] <Ziegler> http://images.myonlinesite.com/cnc/pictures/bb_video/videos.html
[01:53:19] <Ziegler> Friend of mine turned me on to a tip today for keeping my all-thread drive screws from squeaking... he recommended I rub em down with bees wax... made a HUGE difference
[01:53:21] <alex__> * alex__ takes a look
[01:53:29] <alex__> alex__ is now known as The_Ball_
[01:53:53] <The_Ball_> cool player, what is that?
[01:54:37] <Ziegler> oh.. flow player
[01:54:42] <Ziegler> it is very nice
[01:54:49] <The_Ball_> is it open?
[01:54:55] <Ziegler> I believe it is
[01:55:03] <Ziegler> can be controled with php and java also
[01:55:26] <The_Ball_> nice. is this a home made router?
[01:55:28] <Ziegler> http://flowplayer.sourceforge.net/
[01:55:33] <Ziegler> yeah
[01:55:45] <The_Ball_> what motors are you using?
[01:55:46] <Ziegler> this is the end result from the vid: http://images.myonlinesite.com/cnc/pictures/bb_uncover.jpg
[01:55:53] <Ziegler> 200 oz-in steppers
[01:56:11] <The_Ball_> hehe, nice
[01:56:38] <Ziegler> First "job"
[01:57:07] <Ziegler> 8 of those.... got the machine made just in time to start cutting them out
[01:57:22] <Ziegler> runnind a number of them "tweaking the machine"
[01:57:26] <Ziegler> ruined
[01:57:37] <The_Ball_> that's the way it works
[02:13:23] <Ziegler> The_Ball_: you downloaded flowplayer yet ;-P
[02:21:09] <The_Ball_> no im at work ;)
[02:34:43] <The_Ball_> Ziegler, but i will though
[03:16:44] <SWPadnos_> SWPadnos_ is now known as SWPadnos
[03:46:43] <Skullworks-PGAB> OK - so no way to directly read HAL pin status in G-code
[03:46:56] <Skullworks-PGAB> how about via CL?
[03:49:15] <toastydeath> isn't there?
[03:49:26] <toastydeath> there isn't a parameter for that stuff
[03:49:32] <toastydeath> that's how commerical controls do it
[03:49:55] <toastydeath> again, i know zip about it obviously in terms of emc
[03:50:20] <Skullworks-PGAB> yeah - there SHOULD be a way to set a system variable from external input
[03:50:26] <toastydeath> but on commercial controls, #5000 and up or some such similar
[03:50:32] <Skullworks-PGAB> yup
[03:51:17] <toastydeath> owned!
[03:51:19] <Skullworks-PGAB> EMC stores a bunch of stuff in the 5xxx range
[03:52:18] <Skullworks-PGAB> such as the values for G54-G59 etc
[03:52:43] <toastydeath> mabye there could be a feature request?
[03:52:52] <toastydeath> like a set of variables you could link to hal pins
[03:53:16] <Skullworks-PGAB> all I need is a single bit value
[03:53:25] <Skullworks-PGAB> 0 | 1
[03:53:39] <Skullworks-PGAB> on or off
[03:53:44] <toastydeath> right
[03:53:51] <toastydeath> that's how commercial controls do it too
[03:54:30] <Skullworks-PGAB> I figure they are checking status with CL to run the ATC on the Mazak, so there must be a way...
[03:54:40] <toastydeath> but like, you'd be able to do it if there was some sort of mechanism for linking a range of variables to values in hal
[03:54:47] <toastydeath> hmm, perhaps!
[03:54:59] <SWPadnos> toolchange has nothing to do with G-code reading HAL information
[03:55:04] <Skullworks-PGAB> CL can read HAL pins
[03:55:17] <SWPadnos> CL has HAL pins, so it's not that it can "read" them ...
[03:55:51] <SWPadnos> the ATC works (relative to G-code) because there's a pretty clean interface between the interpreter and the ATC
[03:56:21] <Skullworks-PGAB> how does CL know certain steps have been completed during the tool change cycle without some form of feedback?
[03:56:28] <SWPadnos> and it's mostly output from the interp - there's only one input bit "tool-changed", which the interpreter uses as a signal to continue
[03:56:38] <SWPadnos> CL is part of HAL, in this implementation
[03:57:03] <SWPadnos> so it's not quite accurate to say "it can read HAL data, so the interp should be able to as well"
[03:57:11] <Skullworks-PGAB> well - all I need is 1 bit
[03:57:15] <SWPadnos> heh
[03:57:30] <SWPadnos> does it need to pause motion, or do you want to change the G-code program based on that bit?
[03:57:52] <Skullworks-PGAB> if/then check in O sub loop
[03:58:04] <SWPadnos> if "bit" then done or similar?
[03:58:55] <Skullworks-PGAB> state would add an inc value - state would determine if value was positive or neg
[03:59:09] <SWPadnos> you may be able to use the optional block delete input (if that works - I don't know how much testing that feature has had)
[04:00:45] <Skullworks-PGAB> this is for Z axis servo control of EDM using the EDM voltage comparitor to tell if it should feed forward or retract.
[04:01:00] <SWPadnos> hmmm
[04:01:15] <SWPadnos> I don't think you can use the interp for that - it's not realtime
[04:01:22] <SWPadnos> but that's an interesting approach
[04:01:36] <Skullworks-PGAB> It is a easy macro - about 20 lines
[04:02:04] <Skullworks-PGAB> 2 loops
[04:02:28] <Skullworks-PGAB> do/while and if/then nest
[04:02:59] <SWPadnos> well, if optional block delete works, you can do an optional call to a backup subroutine if the voltage is too low
[04:03:22] <SWPadnos> (usinga comparator to get a logical bit from the V feedback)
[04:03:48] <Skullworks-PGAB> I could write it up in BASIC to run just the Z axis - but wanted to use full EMC to provide for X/Y positioning also
[04:05:38] <Skullworks-PGAB> cradek mentioned trying a probe cycle also
[04:06:24] <SWPadnos> hmmm. I'm too tired to look through web CVS to find out if the optional block delete input is actually implemented. I may be able to do that in the next day or two though
[04:06:34] <Skullworks-PGAB> no rush
[04:06:57] <Skullworks-PGAB> this is still just theory
[04:07:26] <Skullworks-PGAB> I've got only 75% the parts for the EDM power generator
[04:07:34] <SWPadnos> heh
[04:08:15] <Skullworks-PGAB> its just that I never thought it would be so hard to read an external input
[04:10:23] <Skullworks-PGAB> maybe it is an area that needed exploring - for one thing it opens avenues for load sensing feedback.
[04:10:45] <Skullworks-PGAB> of servos or spindle drives
[04:10:56] <SWPadnos> there are ways to control motion via HAL - feedhold and feed override are two of them
[04:11:11] <SWPadnos> but they don't cahnge the path, only the rate the path is followed
[04:11:41] <SWPadnos> optional stop can abort a program, and optional block delete should allow individual lines of g-code to be skipped
[04:12:08] <Skullworks-PGAB> yes but to change direction feed override would be a neg number - that is likely an illegal situation.
[04:12:16] <SWPadnos> it is illegal
[04:13:06] <SWPadnos> but optional block delete would take care of that. do the backup move in an optional block, and it only happens if the voltage comparator says there's a short
[04:14:03] <SWPadnos> but you would have to write your G-code programs carefully - you couldn't do a 1" move and expect the comparison to happen before the end of the move
[04:14:13] <Skullworks-PGAB> that is a very unusual application of block delete...
[04:14:23] <SWPadnos> yes it is ;)
[04:14:31] <Skullworks-PGAB> not sure how that would work
[04:14:38] <SWPadnos> and I'm not positive there's a HAL pin for that at the moment
[04:15:19] <SWPadnos> there are supposed to be some aux outputs you can set from G-code. I'm not sure if there are supposed to be corresponding inputs that can be read from G-code
[04:15:39] <SWPadnos> and I'm pretty sure that neither works well at the moment - I think that code hasn't been discussed or looked at in about a year
[04:15:55] <Skullworks-PGAB> I was like feeding forward in a loop say .0001" incramental, do a check , jump/cont, feed forward .0001 etc.
[04:16:37] <SWPadnos> ok - you could have a G1-style loop that takes a total distance to move, and gets there in small increments
[04:16:45] <Skullworks-PGAB> right
[04:17:04] <Skullworks-PGAB> using a do/while loop to check total depth
[04:17:06] <SWPadnos> well, I'd look for the optional block delete pin
[04:17:10] <SWPadnos> yep
[04:17:34] <SWPadnos> move incremental, add value to distance moved, see if you're at the target yet, lather, rinse, repeat
[04:17:50] <Skullworks-PGAB> basically
[04:18:05] <SWPadnos> well, it's bedtime for me. good luck thinking about it
[04:18:20] <Skullworks-PGAB> thanks for the food for thought
[04:18:28] <SWPadnos> sure
[04:25:28] <Jymmm> Hey SWPadnos, how ya doin?
[04:26:11] <SWPadnos> I'm doin' tired - how about you?
[04:26:15] <Jymmm> ditto
[04:26:31] <SWPadnos> so I think I'll actuallly head for bed now ;) good night
[04:26:39] <Skullworks-PGAB> night
[04:26:40] <Jymmm> G;night
[05:14:35] <x3rox> I have problems with converting AutoCAD R14 drawings into G-codes. I tried "realize" and "ACE". ACE seems to work better, it also creates a .nc file _with_ content from the sample file, but it produces an empty file when I try a self generated DXF file. Can somebody, please, help? It seems that all this programs are so old that they do not work properly with my AutoCAD (which is also not very new).
[05:20:34] <toastydeath> mastercam?
[05:21:24] <x3rox> Oh, thanks for your answer. Is mastercam a windows or linux program?
[05:21:40] <x3rox> ... and is it free?
[05:33:57] <ds2> if you know the right people...
[05:34:21] <x3rox> :-) Do I know them...?
[05:35:01] <x3rox> I just found Gcode 2000 which claims to be such a tool: http://home.carolina.rr.com/gcodemcode/gcode.htm Do you know it?
[08:54:52] <dmwaters> {global notice} Good day all, In about 15 minutes to half an hour i'm going to begin maintenence on a few servers. About 5000 total users will be affected by this and this should not take long at all. I will give any further information in wallops, /mode your_nick w if you wish to see any further info.
[13:03:46] <CIA-8> 03jepler 07TRUNK * 10emc2/src/hal/components/at_pid.c: fix warning: implicit declaration of function 'abs' in sim
[13:39:06] <awallin> hi all
[13:39:23] <jepler> hi awallin
[13:39:50] <awallin> I need to figure out how to connect a transformer
[13:40:03] <awallin> http://www.anderswallin.net/sandbox/trafo.jpg
[13:40:07] <skunkworks> wire?
[13:40:16] <skunkworks> normally copper
[13:40:18] <skunkworks> ;)
[13:40:32] <awallin> the secondary is 2*30V and they are the black wires at the top of the pic
[13:40:41] <awallin> but there are lots of wires on the primary side
[13:40:58] <awallin> two green, one red, one blue, one yellow and one brown
[13:41:24] <awallin> I want to connect those to single phase 230VAC
[13:42:15] <skunkworks> have you ohmed them out yet? made a little diagram that way?
[13:42:35] <awallin> nope, tell me more on how to do that!
[13:42:49] <archivist> watch for isolated 110 windings on the primary
[13:44:09] <awallin> the two green wires were connected together when I got the transformer
[13:44:29] <awallin> does it make sense to leave them that way or should I disconnect them before I do the measurements?
[13:45:12] <skunkworks> I would disconnect them to make the measurments.
[13:45:34] <skunkworks> just remember what was hooked up for future reference.
[13:46:02] <archivist> but that may well be the separate windings connection joint or a screen
[13:46:08] <jepler> that many wires on the primary side may be a "quad primary": 100V, 120V, 200V, 240V, similar to the configuration shown at the top left here: http://dkc3.digikey.com/PDF/T071/1916.pdf
[13:47:45] <awallin> I'll do the measurements right now and get back to you guys!
[13:48:10] <archivist> parts of an ohm matter
[13:48:43] <archivist> or can do
[13:59:15] <awallin> archivist: yep, looks like my multimeter shows 0.2 ohm when short circuited, and that's what it shows for two of the coils too!
[14:00:10] <awallin> it looks like there are three primaries, two that on my meter show 0.2 ohms (yellow-green1) and (blue-green2) and then theres a primary with a bit hogher resistance 0.7 ohms between (red-brown)
[14:01:36] <awallin> any tricks or tips for measuring these small resistances?
[14:01:59] <archivist> I use a 5 digit meter
[14:02:52] <archivist> else stuff some dc current in and measure the voltage across the windings
[14:03:31] <awallin> right... what about putting an ac signal on the secondary and looking at the ac signal on the primary?
[14:03:57] <archivist> yes that too
[14:04:35] <archivist> that helps to check phase as well
[14:07:26] <awallin> I guess all primaries need to be in use to achieve the specified power rating?
[14:07:42] <archivist> not allways
[14:07:55] <archivist> but most
[14:11:15] <skunkworks> Just be careful - 30v on the primary will probably give you 230+ volts on some of the secondary windings.
[14:25:55] <awallin> hmm, I'll try a signal generator, but they're only made for driving 50ohm loads...
[14:28:37] <Bo^Dick> does emc work with xubuntu as well as ubuntu?
[14:32:38] <jepler> Bo^Dick: yes, you can install the emc2 packages on xubuntu 6.06 and kubuntu 6.06. whatever additional packages are needed will be installed by apt.
[14:32:45] <jepler> IT"S LIKE MAGICK
[14:38:57] <Bo^Dick> that's awsome
[14:39:15] <Bo^Dick> i need xubuntu since i've got an old computer and want decent performance
[14:39:50] <awallin> more on the transformer: 8.28 VAC input on the secondary side and for the primary coils I measure 22VAC 20VAC and 60VAC
[14:41:07] <awallin> the spec is about 220 to 30 VAC, that's a ratio of 7.33 so my measured 60 to 8.3 is about right
[14:41:43] <awallin> but the other two primary coils are not 30VAC (which would correspond to 110VAC input...)
[14:42:24] <archivist> it wont be exact as there is designed in regulation and other tolerances
[14:42:48] <awallin> there's a noticeable 'humm' , 50Hz presumably, when I connect the siggen. I hope that's normal
[14:43:32] <awallin> archivist: so you think I can connect in series the two smaller 110V? primaries, and then have that in parallell with the 220V primary
[14:43:59] <Bo^Dick> what would happen if i attempted emc on the 'feisty fawn' ubuntu version (7.04)?
[14:44:01] <archivist> and if the signal distorts from the siggen some measurements will be in error
[14:44:17] <archivist> but yes series the 110 primaries
[14:44:20] <jepler> Bo^Dick: you would not be able to use precompiled packages -- either emc or realtime kernel.
[14:44:25] <Bo^Dick> ok
[14:45:43] <jepler> there are instructions for building the simulator (no realtime) version on 7.04: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Installing_EMC2#Simulator_on_Ubuntu_Edgy_Eft_7_04
[14:46:10] <Bo^Dick> i'm gonna attempt xubuntu-6.06.1-desktop-i386.iso
[15:15:53] <awallin> hrm, now I did some more measurements and I'm getting confused....
[15:16:10] <awallin> there's clearly one primary which has the correct 220/30 ratio
[15:16:25] <awallin> but then there are primary coils that show less voltage than on the secondary
[15:16:42] <awallin> I put in about 4VAC, and read about 1 and 2 VAC on these extra primary coils ??
[15:17:03] <SWPadnos> awallin, if you have another transformer - like a 24V one, you can attach it to the secondary and measure the voltages on the primary side
[15:17:14] <SWPadnos> or a 12V - it doesn't matter
[15:17:19] <SWPadnos> (as long as it's AC :) )
[15:17:46] <awallin> SWPadnos: I'm driving the primary with a signal generator now. do you think the 4VAC I input is too little ?
[15:18:02] <SWPadnos> probably. try putting that into the secondary and measuring the primary
[15:18:10] <SWPadnos> that'll get you the ratio
[15:18:31] <awallin> that's what I'm doing (sorry I should have written secondary above)
[15:18:32] <SWPadnos> use the ohmmeter to find out which wires are connected togther
[15:18:34] <SWPadnos> ah
[15:18:44] <awallin> and I get the correct ratio for one of the primaries
[15:18:57] <awallin> but strange ratios for the two other ones
[15:18:59] <SWPadnos> you read 1 and 2V on the primary side when there's 4V on the secondary?
[15:19:05] <awallin> yes!
[15:19:07] <SWPadnos> those aren't primaries then :)
[15:19:29] <SWPadnos> hmmm. actually, they could be other taps on the primary
[15:19:36] <awallin> but something like 28-29VAC on the one that seems correct
[15:19:42] <SWPadnos> do you get continuity on all the primary wires?
[15:19:59] <awallin> nope, there are three coild that are disconnected from eachother
[15:20:30] <SWPadnos> ok. those could be used to make minor changes to the input voltage (like 220 vs. 230 vs. 240)
[15:20:49] <awallin> the ones that read 1VAC and 2VAC have very low resistance, lower than I could measure (<0.2ohm)
[15:21:11] <awallin> yes, that could be one explanation
[15:21:22] <awallin> I wonder if it makes any difference for the power rating if I just leave them out
[15:21:25] <SWPadnos> 1 or 2 VAC out of 28-29 VAC seems about right for the kind of thing I just described
[15:21:29] <SWPadnos> no, it doesn't
[15:21:33] <SWPadnos> (I think)
[15:21:43] <awallin> great, I'll just forget about them then
[15:21:56] <awallin> this transformer is 1.8kva btw ;)
[15:22:16] <SWPadnos> I've seen a lot of transformers that are 110, 115, 120 (or 220,230,240), which is important sometimes, but probably not for your purposes
[15:23:08] <SWPadnos> the limiting factor will be the current rating of the secondary - if you connect it so you only get 28V out (instead of 30), then you'll have a bit less power available
[15:23:27] <awallin> it will become DC for jon elsons servo-drives. the 2x30V secondaries are now in parallell, but I'll put them in series to get 60VAC or something like 1.4*60 VDC for the motors
[15:23:51] <SWPadnos> ok, that'll give you 85V or so, at ~12A
[15:23:54] <SWPadnos> DC
[15:23:58] <awallin> the motors are rated at 90VDC I think, but it won't be a catastrophe if I don't reach 3000rpm
[15:24:37] <SWPadnos> note that some people who make transformers derate by a factor of 1.8 when used in a bilk DC supply
[15:25:04] <SWPadnos> (that's how I got 12A - 1.8kVA / 1.8 = 1KW, 1KW/85-ish = 12-ish)
[15:25:18] <skunkworks> ish
[15:25:27] <awallin> I'm hoping this one is over-rated enough so it doesn't matter
[15:25:44] <archivist> plenty of ish in transformer making
[15:25:51] <SWPadnos> bulk supplies can usually provide 10x or so for a few seconds every minute, so it should be fine
[15:26:35] <awallin> how much heat-sinking do you think the diode-bridge will need? is it ok if I just bolt it to the alum.-case ?
[15:27:26] <SWPadnos> at 12A, it'll be roughly 12W (I'm assuming a 1V drop here), so you'll definitely want a good thermal connection to the heatsink
[15:27:44] <awallin> I wish I had some adjustable power resistor so I could load-test this psu when it's done...
[15:27:51] <SWPadnos> use a disconnected motor
[15:27:55] <SWPadnos> or two
[15:28:16] <awallin> a good idea. thanks.
[15:28:24] <SWPadnos> heh - hook up a paddle to the motor, and stick it in a bucket of water :)
[15:28:30] <SWPadnos> (the paddle, not the motor)
[15:28:34] <skunkworks> ;)
[15:29:21] <archivist> stand well back
[15:47:56] <anonimasu> hm
[15:48:04] <anonimasu> does anyone have a sane feed for machining aluminium?
[15:48:23] <anonimasu> feed per tooth..
[15:55:11] <jepler> anonimasu: here's one table I found just now while looking for that information: http://www.harveytool.com/technical/speeds.php
[15:59:09] <anonimasu> thanks..
[15:59:16] <anonimasu> 6mm endmill at 2krpm..
[15:59:45] <anonimasu> that ends up as 0.05 * 2000 * 2
[16:01:57] <dmwaters> {global notice} Good day all. It appears we had a bit of routing trouble with one of our european hubs, I've rerouted around it for now, so we'll see how well it behaves.:) Thank you for your patience, and thank you for using freenode!
[16:30:12] <cnc_engineer> Hi there..
[16:30:45] <skunkworks> Hello
[16:34:12] <cnc_engineer> How can I compile and install a single module of EMC2?, like I want to modify the parport driver, compile it and install it. The way I know doing this demands compilation/installation of all the package.
[16:38:37] <skunkworks> you will need to get emc from source. then build it. http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Installing_EMC2
[16:38:40] <skunkworks> #3
[16:38:53] <skunkworks> then you can change the source - then make.
[16:39:05] <skunkworks> (the way I understand it ;))
[16:47:25] <cnc_engineer123_> I want to modify only the parport driver, compile and reinstall it, without recompiling/installing all EMC2 Source.
[16:49:58] <cnc_engineer123_> I want to modify only the parport driver, compile and reinstall it, without recompiling/installing all EMC2 Source.
[16:52:36] <skunkworks> That would be a question for the developers then..
[16:53:08] <skunkworks> like cradek: jepler: alex_joni: SWPadnos: jmkasunich: to name a few.
[16:56:48] <cnc_engineer123_> I want to control spindle rpm using PWM but I am unable to get linearity in rpm changes
[17:02:54] <skunkworks> feedback would not be an option? otherwise - I wonder if there is a function block in emc that would alow you to do some form of F(x) or such.
[17:03:55] <jepler> skunkworks: there isn't any generic F(x) type component now.
[17:04:32] <cnc_engineer123_> I want to control spindle rpm using PWM but I am unable to get linearity in rpm changes
[17:04:59] <archivist> you said that already
[17:05:37] <cnc_engineer123_> Please help me
[17:06:08] <jepler> cnc_engineer123_: the program "comp" from emc2-dev is for compiling components (including drivers) from source code. It appears to work for hal_parport.c. however, I don't understand why you think recompiling the parport driver will fix the linearity of your spindle.
[17:06:28] <jepler> full command: sudo comp --install hal_parport.c
[17:06:48] <jepler> note that if you later install a new version of the emc2 package, your modified driver will be replaced
[17:10:57] <cnc_engineer123_> Actually I want to get linearity in PWM Output according to spindle rpm command. Practically I faced non-linearity in PWM output with respect to spindle rpm command
[17:14:47] <anonimasu> :)
[17:28:02] <toastydeath> ffff
[17:38:57] <jepler> cnc_engineer123_: the is component, which I've just written, might be helpful. It performs linear interpolation with up to 8 data points: http://emergent.unpy.net/files/sandbox/linear8.comp
[17:43:24] <jepler> you would find several pairs of (pwmgen input):(measured spindle speed); use (pwmgen input) as the yi values and (measured spindle speed) as the xi values
[17:46:24] <jepler> if you use less than 8 value pairs, then set the 'n' parameter to the number of pairs you chose to use
[18:02:51] <jepler> cnc_engineer123_: you're welcome
[18:03:00] <skunkworks> :)
[18:03:28] <anonimasu> http://www.io23.net/files/frame2.jpg
[18:03:28] <skunkworks> I wanted to ask what he ment by non linear.. cradek has had good luck with it,
[18:03:33] <anonimasu> emc-cut part :)
[18:03:54] <jepler> anonimasu: looks nice, what's it for?
[18:04:02] <awallin> see you later guys, psu progressed nicely today, maybe I'll get to testing tomorrow
[18:04:05] <anonimasu> a robotic toy..
[18:04:14] <anonimasu> http://imagebin.org/9127
[18:04:41] <jepler> skunkworks: it was not at all clear to me what he meant -- I think I need a better statement of the problem
[18:07:30] <jepler> skunkworks: however, I know that emc needs some kind of "F(x)" or interpolation component beyond the one it has now ("scale") -- or maybe more than one kind
[18:07:55] <JymmmmEMC> /sbin/iptables -A FORWARD -p ALL -i io23.net -s all --dport 80 -j DROP
[18:32:16] <ds2> EMC can scale parts with a global setting?
[18:38:14] <jepler> there's no a gcode for scale (e.g., to make the distance between X0 and X1 equal 2mm instead of 1mm). you can modify your inifile and say that there are 400 pulses per mm instead of 200 (or whatever), modifying your maxvel and accel by the proper amount as well. you could write gcode that explicitly incorporates a scaling factor (e.g., g0x2 becomes g0[#1*2]).
[18:39:19] <JymmmmEMC> I was gonna say to redefine the inch =)
[18:39:50] <JymmmmEMC> withi the ini
[18:39:52] <JymmmmEMC> within
[18:42:23] <JymmmmEMC> jepler: Actually, wouldn't that work?
[18:46:05] <JymmmmEMC> Hey, have you guys seen something like "thermal tape", sorta like thermal grease, but in a tape form?
[18:46:05] <ds2> oh hmmm
[18:46:16] <ds2> what about mirror then?
[18:46:28] <ds2> JymmmEMC: frys sells stuff like that, no idea if it is any good
[18:46:47] <JymmmmEMC> ds2: You've seen it? on a roll?
[18:47:16] <jepler> ds2: negate SCALE
[18:48:25] <ds2> JymmmmEMC: not in a roll, but cut to length peices
[18:48:46] <ds2> jepler: is that an answer from someone writing the code or an answer from someone who has tried it?
[18:49:33] <JymmmmEMC> ds2: over in the cpu accesory area or the electronic parts area?
[18:50:03] <JymmmmEMC> ds2: was it double-sided (sticky wise)?
[18:51:32] <ds2> JymmmmEMC: whatever they keep the CPU fans thesedays... I used to see it next to/under the CPU glass cases
[18:51:45] <ds2> JymmmEMC: yep, double sided with a carrier film on both sides; sort of blueish
[18:51:46] <JymmmmEMC> ah ok
[18:52:05] <jepler> ds2: I've never personally tried this, but I am pretty confident of the way multiplying by a negative number works
[18:52:42] <jepler> ds2: if I were in the position of wanting to mill something that was in a different size than the gcode file I had in hand, though, I'd be most likely to try to get a different gcode file before trying any of these things
[18:52:47] <JymmmmEMC> ds2: ...but make sure your limit switches are working before trying it out =)
[18:52:58] <ds2> jepler: it works for the program but is it impact jogging and other setup activities?
[18:53:02] <jepler> for instance, tool radius compensation and tool length compensation will not know about this different coordinate system you're trying to impose, etc
[18:53:12] <ds2> JymmmEMC: no limit switches. just one big red power switch on the powerstrip!
[18:53:44] <jepler> ds2: it will affect everything (jogs and gcode) equally
[18:53:59] <jepler> because you're not resizing the gcode, you're lying to emc about the physical details of your machine
[18:54:00] <ds2> jepler: was I wanted is a variant on using cutter comp to tweak sizes
[18:54:08] <ds2> Ohhhh
[18:54:53] <JymmmmEMC> jepler: Actually, wouldn't that work? Redefining an inch within the ini file that is
[18:56:07] <JymmmmEMC> I think you'll loose resolution, but it would scale
[18:56:53] <anonimasu> change to a bigger endmill ;)
[18:57:18] <skunkworks> there is scaling - then offseting. two differnt things ;)
[18:57:58] <ds2> anonimasu: you mean like welding toolsteel on and regrinding to new size? ;)
[18:58:22] <anonimasu> ds2: heh, no I mean changing it to match whatever scale you want ;)
[18:58:49] <ds2> won't scale Z :P
[18:59:10] <anonimasu> heh, if you redefine " in the ini ;)
[18:59:31] <ds2> wouldn't been nice to be able to scale on the fly but I'll stick with using perl to regenerate G code
[18:59:53] <anonimasu> well, I want sacling and rotation..
[18:59:53] <ds2> which is probally less of an issue since I can run it on the controller instead of having to reupload like on a haas
[18:59:54] <anonimasu> :D
[18:59:56] <anonimasu> and mirroring..
[19:00:07] <anonimasu> and arabity planes..
[19:00:08] <ds2> rotation by arbitrary degrees!
[19:00:11] <anonimasu> yeah..
[19:00:16] <anonimasu> ;)
[19:00:44] <ds2> anomimasu: on a less sarcastic note, if someone will do the math for arbitary planes, it could for the basis of a CMM type setup (alas, the machine...)
[19:01:07] <anonimasu> :)
[19:01:29] <ds2> sample 3 points, compute plane, use as datatum for further measurements
[19:01:35] <anonimasu> :)
[19:01:36] <anonimasu> yep
[19:02:09] <ds2> it should all boil down to some ``trivial'' matrix multiplication for the transforms
[19:02:22] <anonimasu> heh..
[19:02:37] <anonimasu> brb
[19:03:25] <ds2> might even be usable for machining on a machine with sufficient number of axis... avoid reclamps =)
[19:03:53] <anonimasu> I hear that e=mc2 is trivial stuff too :D
[19:04:03] <anonimasu> if you know it ;)
[19:05:56] <jepler> ds2: you could write gcode that explicitly incorporates a scaling factor (e.g., g0x2 becomes g0x[#1*2]).
[19:06:24] <ds2> jepler: is there some tutorial on doing math with G code?
[19:07:38] <jepler> ds2: read near here: http://linuxcnc.org/docs/2.1/html/gcode/main/#1001 (also somewhere in the pdf documentation)
[19:08:21] <ds2> is being able to do math EMC specific?
[19:09:11] <jepler> I am sure it's not universal
[19:10:17] <ds2> nice
[19:10:38] <ds2> is there a parameter editor in axis?
[19:10:41] <jepler> no
[19:10:43] <skunkworks> you could do scaling th same way... X[#1+sinwhatever off center point defined somwhere and such]
[19:11:19] <jepler> you can set parameters by issuing MDI commands like '#1=3.141' but there is no facility for the GUI to inquire as to the value of a parameter
[19:11:45] <ds2> is this in a text file somewhere?
[19:13:24] <jepler> http://linuxcnc.org/docs/2.1/html/gcode/main/#1001
[19:13:25] <jepler> er
[19:13:39] <jepler> http://linuxcnc.org/docs/2.1/html/gcode/main/#sub:Parameters
[19:14:41] <jepler> but the parameter file is not intended to be edited while emc is running: it may or may not be written or re-read during an emc session
[19:14:47] <jepler> bbl
[19:15:07] <ds2> oh.
[19:22:52] <JymmmmEMC> anonimasu: it's all relative
[19:25:32] <lerneaen_hydra> 'lo all
[19:25:40] <lerneaen_hydra> anyone here?
[19:25:53] <skunkworks> depends on who you want ;)
[19:26:08] <lerneaen_hydra> some hal assistance :p
[19:26:27] <JymmmmEMC> hal? No hal here!
[19:26:33] <skunkworks> figured ;) ask and hopefully someone will know
[19:26:51] <JymmmmEMC> lerneaen_hydra: Try /join #2001
[19:27:47] <lerneaen_hydra> http://pastebin.ca/594121 , emc prints "HAL: ERROR: pin 'halui.feed-override.scale' not found, HAL:110: link failed" when starting
[19:27:51] <lerneaen_hydra> JymmmmEMC, :p
[19:28:22] <lerneaen_hydra> the funny thing is I ripped this file off of a working config
[19:30:12] <JymmmmEMC> lerneaen_hydra: It has LoJack in it, has been deactivated, and the cops are on the way!
[19:30:25] <lerneaen_hydra> O_o
[19:30:40] <JymmmmEMC> lerneaen_hydra: you dont know what lojack is?
[19:30:50] <lerneaen_hydra> fraid not
[19:30:59] <JymmmmEMC> http://www.lojack.com/
[19:31:20] <lerneaen_hydra> could halui be needed perhaps?
[19:31:30] <skunkworks> :) I would think so..
[19:31:47] <JymmmmEMC> hidden vehicle recovery system, satellite tracked and can disbale the car as well as the police cars have radio trackers in them as well.
[19:32:06] <lerneaen_hydra> ah, I see
[19:32:15] <lerneaen_hydra> skunkworks, testing
[19:32:27] <lerneaen_hydra> JymmmmEMC, disable the car?
[19:32:36] <JymmmmEMC> lerneaen_hydra: Yep, from starting
[19:32:52] <lerneaen_hydra> oh, ok
[19:32:57] <lerneaen_hydra> not when theyre running?
[19:33:01] <lerneaen_hydra> that would be too fun for other people
[19:33:25] <JymmmmEMC> lerneaen_hydra: well, I guess you could cut off the fuel, make it look like your out of gas.
[19:33:43] <lerneaen_hydra> heh
[19:33:45] <JymmmmEMC> Mostly installed on high priced/classic cars
[19:36:40] <JymmmmEMC> LOL http://www.lojackforlaptops.com/
[19:43:50] <lerneaen_hydra> well that worked
[19:44:02] <lerneaen_hydra> now however I can't get out of estop mode
[19:44:33] <lerneaen_hydra> oh, wasn't there some software link you had to do if you don't have a hardware estop?
[19:45:02] <alex_joni> lerneaen_hydra: it's in all sample configs
[19:45:22] <lerneaen_hydra> yeah I just thought of that :)
[19:45:43] <skunkworks> net estop-loop iocontrol.0.user-enable-out iocontrol.0.emc-enable-in
[19:46:08] <alex_joni> skunkworks: you're getting better each day
[19:46:18] <skunkworks> I happened to have a hal file open ;)
[19:46:28] <alex_joni> in 1-2 years you'll run the show alone
[19:46:29] <alex_joni> :P
[20:31:15] <lerneaen_hydra> is there a program start hal call?
[20:37:45] <lerneaen_hydra> newsig runsignal bit net runsignal parport.0.pin-12-in-not halui.program.run
[20:37:55] <lerneaen_hydra> doesn't seem to work
[20:38:30] <lerneaen_hydra> I get a "can't do that (EMC_TASK_PLAN_RUN) in manual mode
[20:38:40] <lerneaen_hydra> * lerneaen_hydra pokes alex_joni
[20:39:59] <jepler> first you need to set "auto" mode by asserting halui.mode.auto
[20:40:15] <jepler> just like you have to switch to auto mode in the classic GUIs (tkemc etc)
[20:41:00] <lerneaen_hydra> ah
[20:41:24] <lerneaen_hydra> how do I assert it?
[20:41:34] <jepler> just like you did for program.run
[20:41:47] <JymmmmEMC> lerneaen_hydra: with an assertion tool of course
[20:42:16] <lerneaen_hydra> ah ok
[20:42:23] <JymmmmEMC> lerneaen_hydra: I bet the next thing you're gooing to say is that you don't know what a board stretcher is
[20:42:39] <lerneaen_hydra> JymmmmEMC, ... do I really want to know?
[20:43:02] <JymmmmEMC> lerneaen_hydra: What, you REALLY don't know what a board stretcher is?
[20:43:10] <lerneaen_hydra> jepler, hmm would binding it to the button work? or do I need to bind it to true constantly?
[20:43:22] <lerneaen_hydra> JymmmmEMC, fraid not
[20:43:56] <JymmmmEMC> lerneaen_hydra: Are you being serious? You really don't know what a board stretcher is? You've never used one ever?
[20:44:47] <lerneaen_hydra> as I can't say I connect the word with an the intangible concept of board stretchers I can't say I know either way ;P
[20:45:22] <JymmmmEMC> * JymmmmEMC hides his face from lerneaen_hydra in the shame of it all
[20:47:28] <Sapote_reloaded> Sapote_reloaded is now known as sapote
[20:47:32] <sapote> sapote is now known as Sapote
[20:51:21] <lerneaen_hydra> jepler, works now, thanks muchly :)
[20:52:15] <lerneaen_hydra> 'night all
[21:08:33] <SWPadnos> jepler, ds2: note that if you negate scale (for mirroring), limit switches will be adversely affected
[21:08:51] <SWPadnos> ie, emc will hit the +limit when it thinks it's going -
[21:08:54] <jepler> SWPadnos: oh yeah -- homing too
[21:09:00] <SWPadnos> yep
[21:09:00] <jepler> I hope it didn't sound like I was *endorsing* doing that
[21:09:03] <SWPadnos> heh
[21:09:17] <SWPadnos> well, you mentioned it a couple of times
[21:09:38] <ds2> Hmmmm
[21:09:39] <jepler> ds kept asking me
[21:09:46] <jepler> :-P
[21:10:24] <SWPadnos> indeed. he was pestering almost as badly as cnc_engineer123
[21:10:32] <SWPadnos> well, maybe not that badly
[21:10:53] <ds2> i was just exploring the possibilities....
[22:51:19] <JymmmmEMC> Damn, ATC fuses only rated for 24V =(
[22:51:56] <toastydeath> ONE TRILLION VOLTS
[22:55:26] <anonimasu> eh?