#emc | Logs for 2010-08-04

[00:03:37] <andypugh> davidf: You still there?
[00:04:06] <theorb> theorb is now known as theorbtwo
[00:12:08] <davidf> andypugh, back now.
[00:12:29] <davidf> what's up?
[00:12:41] <andypugh> Another lathe config
[00:12:42] <andypugh> file:///Users/andy2/Desktop/SemiManualLathe/lathe.html
[00:12:49] <andypugh> Not that link.
[00:12:51] <andypugh> Doh!
[00:13:08] <andypugh> http://www.bodgesoc.org/lathe/lathe.html
[00:13:18] <davidf> an oft-repeated expression here...
[00:13:56] <andypugh> That does 95% of my lathe work.
[00:14:33] <davidf> ooooh this is real cool. I'm looking...
[00:14:40] <andypugh> (Though in the actual screenshot it has a different G-code file loaded)
[00:17:09] <andypugh> I am thinking that I might change the way it works, though. I keep hitting "run" with the wrong file loaded (ie not the special one) or boring when I mean to part etc. A better way would be to have G-code routines called by (for example) a "boring" button that reads the values and does the operation regardless of loaded file (this is pretty much how the tool-select buttons work)
[00:20:06] <davidf> I am all jazzed about that... I it is very much like what I was wishing I had.
[00:20:19] <davidf> You wrote all this?
[00:20:27] <andypugh> Yes.
[00:20:33] <davidf> awesome.
[00:20:46] <andypugh> Feel free to re-use any bits that suit your needs
[00:21:09] <davidf> That tach, for instance, does it run in a slow thread, and does it stay fairly stable?
[00:21:19] <davidf> way cool!
[00:21:30] <davidf> thanks!
[00:22:04] <andypugh> It runs in the servo thread, but has a lowpass filter to steady it. As the machine runs CSS in that config it is pretty pointless anyway. I might make it digital to make some more space
[00:23:14] <davidf> can it be used with a stepper config? I have a spindle endoder (quad) and index already working...
[00:23:34] <andypugh> That is a stepper config
[00:23:41] <davidf> neato
[00:23:54] <andypugh> Though it uses a Mesa card
[00:24:12] <skunkworks> andypugh: in comp programming... I see sometimes variables are defined before ;; and sometimes after... does it have to do with what gets kept from thread run to thread run?
[00:24:47] <skunkworks> could it be thought as global and local?>
[00:24:52] <davidf> I can do threading right now (g76) but I have to be careful not to over-rev the spindle to avoid aliasing thenrpm / position. That tach would be helpful
[00:25:00] <andypugh> skunkworks: I think it is to do with variable scope, yes. The ones before the ;; are "storage" variables
[00:27:12] <andypugh> davidf: When you get to the point of running 2 p-port cards the Mesa 7i43 starts to look like $75 well spent. It gives you 48 channels of IO, with hardware stepgens, encoder counters that can count to MHz, PWM to 100kHz etc. It plugs into the parallel port itself.
[00:27:27] <davidf> btw I haven't been using the post-gui custom hal files etc. Can you do most anything you need to add in those files, w/o messing with your mail hal and ini?
[00:27:56] <davidf> dang!
[00:28:09] <skunkworks> so - If the data is stored in a pin - then I could use that data from run to run.
[00:28:25] <davidf> Just spent $35.00 on a 2nd card, and $30 on a second BOB.
[00:28:29] <skunkworks> * skunkworks should just try it ;)
[00:29:33] <andypugh> Yes, that is what they are for, so you can keep the stepconf wizard config and tweak things there, and then keep the custom stuff safe in the custom and postgui. I got bored of all the unlinkp lines neede though, and I just work directly in HAL now, (I don't even pull values from the INI, they are explcit in the HAL)
[00:30:15] <andypugh> skunkworks: Yes, pins count as static storage too. But the pre-;; variables are useful for things you don't want to make a pin for.
[00:31:39] <skunkworks> andypugh: thanks - coming together :)
[00:31:40] <andypugh> OK, 0130 here. Have fun.
[00:36:32] <Valen> so i've been on holidays for a bit over a week what have i missed?
[00:42:25] <davidf> hi valen
[00:42:33] <Valen> mooornin
[00:42:49] <Valen> oh got some posts up about our contact probe
[00:42:50] <Valen> http://www.cnczone.com/forums/showthread.php?t=109146
[00:42:53] <davidf> 7:30 pm yesterday here. :)
[00:43:08] <davidf> How is tomorrow morning looking?
[00:43:16] <Valen> fine and sunny
[00:43:24] <Valen> light clouds
[00:43:29] <Valen> kinda fluffy
[00:43:33] <davidf> 100 F super humid here
[00:43:41] <Valen> you can keep that
[00:44:00] <davidf> just awful, like a sauna been that way for weeks
[00:44:24] <Valen> yeah thats no fun
[00:44:28] <Valen> tis winter here
[00:44:41] <davidf> thank God for AC
[00:45:12] <davidf> Grass in back yard is 18 inches, been too hot and wet to mow it
[00:45:36] <davidf> that's right opposite seasons, forgot that.
[00:45:50] <Valen> havent mowed in months
[00:46:01] <Valen> about an inch and a bit lol
[00:47:05] <Gangsta> hey
[00:48:40] <Gangsta> I got my spindle speed display working :) and had my first involvment with hal!
[00:49:48] <Gangsta> it was a really simple problem too
[00:51:14] <davidf> hi
[00:51:26] <davidf> what prob?
[00:55:28] <skunkworks> Gangsta: hooking the right pin to the right ping? ;)
[00:55:33] <skunkworks> *pin
[00:55:41] <Gangsta> yup
[00:56:29] <skunkworks> davidf: sounds like here. where you in chicago?
[00:56:42] <Gangsta> for some reason stepconf decided that i didnt have a spindle encoder so connected the speed display to commanded rpm instead of actual
[00:56:56] <skunkworks> Gangsta: how is it working?
[00:57:52] <Gangsta> spindle is accurate to within 25 rpm of commanded, verified by my scope
[00:57:54] <davidf> skunkworks, Missouri
[00:57:55] <Gangsta> :)
[00:58:44] <Gangsta> not bad for a pwm -> DAC (0-10v) with only 4 components
[00:58:56] <davidf> was just looking at Valen 's post re that probe. Interesting, looks good
[00:59:12] <Valen> not too bad for the first attempt
[01:00:18] <skunkworks> Gangsta: cool
[01:01:18] <Gangsta> im very new to this, is there something in hal that will smooth my speed display - its a little bit jumpy at the moment, like a filter. im reading the manual but its lots to takee in
[01:02:00] <Valen> I do believe there are filter widgets
[01:02:19] <davidf> yes there is. let me check. it's something like position_interpolated I think
[01:02:22] <skunkworks> what pins did you hook up? the encoder module has a interpolated velocity pin
[01:02:27] <skunkworks> he
[01:02:29] <skunkworks> heh
[01:02:36] <Valen> its a filter module as i recall
[01:02:52] <Gangsta> so is that velocity-interpolated?
[01:03:00] <Valen> I don't think thats what you want
[01:03:01] <skunkworks> probably - es
[01:03:04] <skunkworks> yes
[01:03:04] <Valen> you want a low pass filter
[01:03:13] <Gangsta> will that be easier than a filter?
[01:03:17] <Valen> to make the display not flicker
[01:03:31] <skunkworks> I would try the interpolated pin first... see how that works. (easy to try)
[01:04:00] <Gangsta> thats true. and ive found a section called low pass too :)
[01:04:14] <skunkworks> you are going to be a hal master soon :)
[01:04:31] <Valen> actually sorry i'm not sure if low pass is what you want
[01:04:40] <Valen> you want a time average
[01:06:47] <Gangsta> its not as bad as it looks tbh
[01:07:23] <Gangsta> i wish i had looked sooner, mabey it wouldnt have taken so long to get going
[04:55:04] <ichudov> I have worked on tuning my PID loop on my Bridgeport Interact. I finally got to the point that with slow or rapid moves, my following error is always udner 0.002", and usually under 0.001". Is that "good enough"?
[05:17:21] <Valen> if your making telescope mirrors then no, if your making cupboards then yes
[05:41:32] <elmo40> +1 for Valen
[05:47:45] <elmo40> I believe this is an Agilent scope. http://qurl.org/Q31 Is it worth the price? I have no idea on the age. 100MHz is a good range, no?
[05:48:36] <Valen> if your measuring AVR serial streams yes, if your debugging USB3 data streams then no
[05:48:38] <Valen> ;-P
[05:49:13] <Valen> that scope seems like overkill in aa number of respects
[05:49:48] <Valen> if you can get a storage scope it'd probably be more usefull
[05:50:28] <elmo40> here is a better pic of the front: http://mark.rehorst.com/AN_USM_338/front.JPG
[05:50:46] <elmo40> Storage scope?
[05:51:10] <Valen> you can have it trigger and it'll keep the trace on the screen
[05:51:19] <elmo40> wow, tiny screen :/
[05:51:29] <Valen> that was my other concern
[05:52:46] <Valen> it really depends on what you want to do with it
[05:53:07] <Valen> if your in the field, then it might be a good one, though the new "pocket" ones might be better
[05:53:08] <elmo40> see the gold circuitry? http://mark.rehorst.com/AN_USM_338/bottom.JPG
[05:53:19] <elmo40> crazy
[05:53:27] <elmo40> anyways... it is from the 70's
[05:53:30] <Valen> its not that expensive to get stuff plated
[05:53:33] <elmo40> don't know how 'burnt' the screen is
[05:53:42] <Valen> scopes don't seem to burn really
[05:54:13] <Valen> I've got valve powered ones that still seem to work ok
[05:54:54] <elmo40> this one is less expensive. http://qurl.org/R31
[05:55:13] <Valen> tek is a good brand
[05:55:14] <elmo40> less features, though.
[05:55:40] <Valen> it really does depend on what you want to do
[05:55:51] <elmo40> I aim to use it for trouble shooting and dissecting cnc components ;)
[05:56:00] <elmo40> and other electronics.
[05:56:51] <elmo40> * elmo40 drools...
[05:56:52] <elmo40> http://qurl.org/S31
[05:58:31] <elmo40> anyone want to lend me $600? :D
[05:58:49] <Valen> that'll do it for ya
[05:59:41] <elmo40> half my paycheque!
[05:59:45] <Valen> I got a 750mhz 4 chan tek
[05:59:59] <Valen> not a storage scope though
[05:59:59] <elmo40> if I made it into a business... then it would be worth it.
[06:00:21] <elmo40> 750MHz? these are only 60MHz
[06:00:37] <Valen> I think its 750
[06:00:43] <Valen> might be 350
[06:03:57] <elmo40> that is awsomely high. you could scope TV transmitters!
[06:04:17] <elmo40> channel 21-36 is bellow 608MHz
[07:48:44] <elmo40> Valen: I see the interest in the portable scopes. they look great. http://toronto.en.craigslist.ca/tor/ele/1818401412.html
[07:53:53] <elmo40> some things just don't add up... http://toronto.en.craigslist.ca/bra/bfs/1865575375.html
[07:54:07] <elmo40> says I need to supply my own computer with 'serial port' ?
[07:54:30] <elmo40> plus, Large and NEMA23 just don't go together ;)
[07:56:33] <celeron55> it uses this http://deskcnc.com/controller.html
[07:57:16] <elmo40> only when I install the software on my computer...
[07:57:34] <elmo40> http://toronto.en.craigslist.ca/yrk/tls/1860565172.html anyone need 25 year old motors? :-/
[08:05:23] <elmo40> depending on the spindle taper and table size, this looks like a deal http://toronto.en.craigslist.ca/yrk/bfs/1833673218.html
[08:29:14] <Fox_M|afk> Fox_M|afk is now known as Fox_Muldr
[08:54:50] <alex_joni> Jymmm: still around?
[09:00:37] <maximilian_h> Will emc2's realtime run for programming purpose run unter vmware or virtualbox ?
[09:01:17] <maximilian_h> Shopfloor is too noisy for me to concentrate :(
[09:01:33] <alex_joni> yup
[09:01:42] <alex_joni> I run emc2 under vmware server all the time
[09:02:07] <alex_joni> but you can also configure emc2 to run in simulation mode (you don't get access to drivers - but using VM you don't anyways)
[09:02:17] <alex_joni> that way you can run on any distro with the stock kernel
[09:03:06] <maximilian_h> no I'd like to be able to run "realtime" even if it's just inside a virtual machine
[09:03:25] <maximilian_h> did you try virtualbox ?
[09:03:34] <maximilian_h> or just vmware server ?
[09:03:48] <alex_joni> never tried virtualbox
[09:04:19] <maximilian_h> to bad, but if it runs under vmware, that I'll give it a try
[09:04:48] <alex_joni> maybe someone else has
[09:05:03] <alex_joni> I vaguely recall jthornton mentioning a couple days ago in #emc-devel
[09:05:06] <alex_joni> check the logs
[09:05:35] <maximilian_h> thanks Alex
[09:05:46] <maximilian_h> how's life in Timisoara going ?
[09:07:09] <alex_joni> 2010-08-02 13:51:29 <JT-Work> my last install on the zotac was stock 10.04 + your script... all working fine and playing nice with everything else including virtualbox ose
[09:07:39] <alex_joni> maximilian_h: just came back to work today, a big pile was waiting for me :D
[09:07:51] <Fox_Muldr> Fox_Muldr is now known as Fox_M|afk
[09:08:03] <maximilian_h> Haha, no rest for the wicked :)
[09:08:23] <maximilian_h> You were on vacations ?
[09:10:15] <alex_joni> took a couple days off
[09:12:54] <maximilian_h> well, enjoy your pile then. I'll probably get back once I have problems with emc in virtualbox :(
[09:13:17] <alex_joni> I played with bubbles :D
[09:20:24] <birdman3131|zzzz> birdman3131|zzzz is now known as Birdman3131
[09:23:22] <alex_joni> Jymmm: http://juve.ro/blog/photography/01280913408
[09:40:12] <Valen> what use is "realtime" if its inside the vm?
[09:41:12] <alex_joni> Valen: it has some use
[09:41:42] <alex_joni> it ensures you run the exact same environment as the real machine
[09:42:23] <Valen> yeah but you can't actually do anything with it though?
[09:44:25] <alex_joni> no, you can only run sim configs
[09:44:33] <alex_joni> and test the kernel, etc
[09:44:41] <Valen> but the sim doesn't use realtime anyway?
[09:44:58] <Valen> I can see the kernel testing being handy inside a VM
[09:54:16] <alex_joni> sim uses RT if emc2 isn't compiled without RT
[10:17:22] <maximilian_h> the first line in a tbl file
[10:17:28] <maximilian_h> does it have to be a comment
[10:17:42] <maximilian_h> I am trying the simulation inside virtualbox
[10:17:46] <maximilian_h> and getting this error
[10:18:02] <maximilian_h> Starting EMC2...
[10:18:02] <maximilian_h> Unrecognized line skipped: POC FMS LEN DIAM COMMENT
[10:18:34] <maximilian_h> this looks like the first line in a tbl file
[10:18:53] <maximilian_h> Example:
[10:18:54] <maximilian_h> POC FMS LEN DIAM COMMENT
[10:18:54] <maximilian_h> 1 1 17 5.0 first tool
[10:18:54] <maximilian_h> 2 2 18 2.0
[10:18:54] <maximilian_h> 3 3 13 25.0
[10:18:54] <maximilian_h> 4 4 18 25.0
[10:18:54] <maximilian_h> 550 10.0
[10:18:55] <maximilian_h> 32 32 13 0.0 last tool
[10:19:06] <maximilian_h> ?
[10:23:21] <maximilian_h> And this ?
[10:23:21] <maximilian_h> libnml/buffer/physmem.cc 143: PHYSMEM_HANDLE: Can't write 10748 bytes at offset 60 from buffer of size 10208.
[10:23:21] <maximilian_h> libnml/cms/cms_in.cc 1383: CMS:(emcStatus) Error writing 10748 bytes to global memory at offset 81F8CB0
[10:23:21] <maximilian_h> (See libnml/cms/cms_in.cc line 1386.)
[10:23:30] <maximilian_h> ?
[10:25:17] <alex_joni> maximilian_h: the first issue is a warning, disregard it
[10:25:21] <alex_joni> the second is an error
[10:25:25] <maximilian_h> ok
[10:25:41] <alex_joni> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?UPDATING#emc_nml_changes
[10:25:56] <maximilian_h> any ideas on the error ? I am using axis_mm.ini from the sim examples
[10:26:02] <alex_joni> see ^^
[10:26:14] <maximilian_h> sorry, moment have a read your link
[10:28:25] <maximilian_h> Ah Alex, your a gem!
[10:29:07] <maximilian_h> It starts, complains about a realtime delay, which is ok inside a vm, and then shows axis in all its glory :)
[10:30:26] <maximilian_h> Who do I notify that the axis_mm.ini sim example needs to have this line removed ?
[10:50:12] <alex_joni> it's more likely your axis_mm was from 2.3.x
[10:51:09] <alex_joni> see http://git.linuxcnc.org/gitweb?p=emc2.git;a=blob;f=configs/sim/axis_mm.ini;h=5350cbec4fe114e42e010bef7ba7beb04b7b299d;hb=be6381b74c7b491988f9836baf49f690252a594d
[10:51:45] <maximilian_h> Hey Alex, I updated to v2.4 right after installing
[10:51:53] <maximilian_h> and then started emc2
[10:52:14] <maximilian_h> but the original version of the cd image contained either 2.2 or 2.3
[10:52:16] <alex_joni> dunno, but the emc.nml reference was removed in 2009.05.04 from axis_mm.ini
[10:52:22] <alex_joni> probably 2.3
[10:52:22] <maximilian_h> I honestly don't know
[10:52:29] <alex_joni> ah, I bet I know what happened :)
[10:52:53] <maximilian_h> ? Enlighten me please :)
[10:53:00] <alex_joni> we moved sample config location from /etc/emc2/sample-configs to /usr/share/doc/emc2/sample-configs/
[10:53:18] <alex_joni> but stuff in /etc/emc2 doesn't get removed when updating the package (debian policy)
[10:53:31] <alex_joni> so it's likely that you picked the /etc/emc2/sample-config (which was 2.3)
[10:53:38] <maximilian_h> yep
[10:53:40] <alex_joni> or was it 2.2.x ..
[10:53:42] <maximilian_h> quite probable
[10:53:51] <alex_joni> you can check if the files are there
[10:53:58] <maximilian_h> so I should copy from /usr/share/doc/emc2/sample-configs/
[10:54:36] <alex_joni> the config picker should do that
[10:55:56] <maximilian_h> /etc/emc2/sample-configs does not exist
[10:56:11] <maximilian_h> /usr/share/doc/emc2/sample-configs/ does exist
[10:58:25] <alex_joni> check sim/axis_mm.ini for emc.nml
[10:58:47] <maximilian_h> yep, I already commented it out
[10:59:05] <maximilian_h> runs, now. (Me is happy :) )
[10:59:45] <maximilian_h> How do I get a screenshot (png) into a wiki page, I have to upload it first
[11:30:44] <maximilian_h> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?EMC2RunningInsideVirtualbox
[11:31:10] <alex_joni> maximilian_h: so you say that /usr/share/doc/emc2/sample-configs/sim/axis_mm.ini contains emc.nml ?
[11:34:35] <maximilian_h> no, http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?EMC2RunningInsideVirtualbox is without emc.nml, but ~/emc2/configs/sim/axis_mm.ini did
[11:34:47] <maximilian_h> ups copy paste problems
[11:35:35] <maximilian_h> I meant /usr/share/doc/emc2/examples/sample-configs/sim/axis_mm.ini is without emc.nml, but ~/emc2/configs/sim/axis_mm.ini did contain the line
[11:36:16] <maximilian_h> honestly, it might be that I tried starting it just after installing, but before upgrading to 2.4
[11:57:59] <piasdom> hi
[11:58:03] <Fox_M|afk> Fox_M|afk is now known as Fox_Muldr
[12:00:16] <piasdom> what does this dmesg mean http://paste.org/pastebin/view/21154
[12:00:42] <maximilian_h> @Alex: Is there a way to set the absolute feed rate through the hal layer, not just a factor to it through adaptive_feed. I want to have a fixed feed rate depending on the tool in use
[12:01:10] <maximilian_h> @piasdom: Your realtime setup is somehow not good
[12:01:34] <piasdom> maximilian_h: thanks
[12:01:47] <maximilian_h> I don't know your background: Might me either smi problems, graphics card, bad hw for realtime
[12:02:10] <piasdom> new graphic card
[12:02:29] <maximilian_h> steppers or servo
[12:02:29] <maximilian_h> ?
[12:02:29] <maximilian_h> which driver are you using
[12:02:29] <maximilian_h> ?
[12:02:30] <piasdom> must be that .... can't get both to work at the same time
[12:02:37] <maximilian_h> graphics driver I mean
[12:02:51] <piasdom> fglrx
[12:03:10] <piasdom> can't get vesa to work at all
[12:03:16] <maximilian_h> does your monitor support one of the vesa resolutions ?
[12:03:19] <maximilian_h> ok
[12:03:36] <maximilian_h> have you tried the latency test ?
[12:03:53] <piasdom> but that just be "ME" that is not corect :)
[12:04:01] <piasdom> will try now
[12:04:13] <maximilian_h> have you tried with xforcevesa on the grub loader ?
[12:04:29] <piasdom> never heard of it
[12:04:53] <maximilian_h> xforcevesa forces the graphics card to use the vesa driver
[12:05:13] <maximilian_h> have you tried to use the open source radeon driver ?
[12:05:23] <piasdom> when i ran latency before it was about 39000 and i set base 45000. worked but picture was worng
[12:05:47] <piasdom> yes i have(open) no good
[12:06:38] <maximilian_h> please try the vesa graphics driver then.
[12:06:45] <piasdom> i'll look for xforcevesa
[12:06:54] <piasdom> maximilian_h: Thanks
[12:07:09] <maximilian_h> I have to go, the dog is waiting for its midday walk
[12:07:15] <maximilian_h> good luck
[12:07:41] <piasdom> maximilian_h: you too
[12:11:59] <alex_joni> piasdom: what maximilian_h said
[12:12:24] <alex_joni> piasdom: try running the latency test from a terminal
[12:12:34] <piasdom> alex_joni: looking to all xforce
[12:12:40] <piasdom> alex_joni: ok
[12:12:46] <alex_joni> (the RTAI latency test -> found in /usr/realtime-*/testsuite/latency/kern iirc)
[12:13:02] <piasdom> at 31000 now bu thru icon
[12:13:04] <alex_joni> then try loading some opengl apps, and see if you get overruns all the time, or only at startup
[12:13:23] <alex_joni> if you don't get overruns, or only at startup then you *might* be fine
[12:14:24] <piasdom> what's the command to run latency ?
[12:14:43] <piasdom> it is usually at startup only
[12:23:15] <piasdom> i tried this ....
[12:23:17] <piasdom> itt@CNCmill:/usr/realtime-2.6.24-16-rtai/testsuite/user/latency$ ./latency
[12:23:17] <piasdom> Segmentation fault
[12:23:38] <piasdom> and got seg
[12:25:55] <piasdom> and now this when i ./run latency ..... http://paste.org/pastebin/view/21154
[12:38:41] <piasdom> finally ..... http://paste.org/pastebin/view/21154
[12:39:49] <alex_joni> sudo ./run
[12:40:05] <alex_joni> cd to the folder, then sudo ./run
[12:41:20] <piasdom> got it running with this cd /usr/realtime*/testsuite/kern/latency; ./run
[12:41:58] <piasdom> last paste.org is the results
[12:42:19] <alex_joni> I doubt that
[12:42:23] <piasdom> oops
[12:43:11] <piasdom> http://paste.org/pastebin/view/21157
[12:43:31] <alex_joni> piasdom: looks good so far
[12:43:37] <alex_joni> let it run a couple minutes/hours
[12:43:46] <alex_joni> and really stress the system as much as you can
[12:43:58] <alex_joni> if you see that the last column is 0, then you're golden ;)
[12:45:27] <piasdom> alex_joni: Thanks ok
[12:55:03] <Valen> alex_joni: it'd be handy to see latency errors in EMC more than just at startup
[12:57:24] <alex_joni> Valen: yeah, but also tricky
[12:57:37] <alex_joni> if your system has overflows, then you get about 1000 messages/second
[13:09:41] <alex_joni> Valen: so.. how often would you like the message?
[13:09:42] <Fox_Muldr> Fox_Muldr is now known as Fox_M|afk
[13:09:58] <alex_joni> every time? 1000/sec, once / second? once every 10 seconds?
[13:10:09] <alex_joni> I'm sure it gets annoying really fast either way
[13:10:21] <alex_joni> a reliable sytem shouldn't have _any_ of these errors
[13:10:50] <cradek> I think there is a count of them, and you can view it with halmeter/halscope
[13:10:50] <Valen> yes, but finding out if something is wrong with your system after you start adding loads of stuff to hal etc
[13:11:04] <cradek> alex_joni is right that you don't really want to have a bunch of popups.
[13:11:22] <alex_joni> one popup means you have something wrong with your system
[13:11:22] <cradek> not sure I understand. you do get the error message the first time, don't you?
[13:11:23] <Valen> normal log stuff should be sufficent, popup once every N minutes if there is a problem
[13:11:33] <Valen> alex_joni: I get that message every time I start axis
[13:11:42] <alex_joni> then your setup is broken ;)
[13:11:43] <cradek> then you've got a problem
[13:11:48] <Valen> starting openGL windows causes a latency spike
[13:11:53] <cradek> then you've got a problem
[13:12:06] <Valen> no, I just don't start opengl windows while i'm milling
[13:12:24] <cradek> I'm not sure your assumption that it's OK when the windows are already up is correct
[13:12:43] <Valen> well with latency test running for 48 hours the max latency is ~4500 or so
[13:12:54] <Valen> with glxgears running
[13:13:01] <cradek> ok, interesting
[13:13:07] <Valen> exiting and restarting glxgears causes a 30K spike
[13:13:13] <cradek> I haven't seen that particular behavior before
[13:13:46] <cradek> maybe you could fix it by switching to the vesa video driver
[13:13:59] <Valen> I'd rather keep the GL acceleration
[13:14:22] <alex_joni> chose your poison
[13:15:21] <Valen> there is no poision as far as I can tell
[13:15:33] <Valen> everything works fine, there is a spike when you start openGL
[13:15:41] <Valen> this is with SMP kernel and intel drivers
[13:15:52] <Valen> I had the same behaviour with NVIDIA binary and SMP
[13:16:50] <elmo40> does 10.04 use Grub2?
[13:17:05] <elmo40> and is EMC with 10.04 still going to use Gnome?
[13:17:57] <Valen> I havent seen a count of latency errors in halscope anywhere
[13:18:12] <Valen> would be handy
[13:38:25] <piasdom> is there a way to see "what" is Broken on my sys
[13:39:24] <JT-Work> synaptic package manager
[13:40:10] <piasdom> broken packages ?
[13:40:17] <piasdom> thanks
[13:40:30] <piasdom> thought ya'll meant hardware
[13:41:02] <JT-Work> I just got here...
[13:41:45] <piasdom> hi
[13:43:17] <JT-Work> hi
[13:43:30] <mozmck> elmo40: yes, yes.
[13:43:45] <piasdom> trying to fix realtime error
[13:44:13] <piasdom> latency test looks ok for now
[13:44:27] <piasdom> just don't like that message :)
[13:48:34] <piasdom> in the latency test, do i have to look at "all" the overruns or just the last results ?
[13:51:40] <SWPadnos> when you're machining, do you want "all" the g-code to execute properly, or just the last few segments?
[14:00:41] <elmo40> mozmck using gnome only because ubuntu comes with it?
[14:00:42] <piasdom> hahahahhahah
[14:00:57] <piasdom> ok i get the point
[14:01:01] <elmo40> many people have asked why not a lightweight manager such as *box
[14:01:28] <cradek> elmo40: yet none of them have made a cd
[14:02:11] <elmo40> true.
[14:02:31] <elmo40> you know people... they want the simple way out ;)
[14:02:31] <piasdom> how do i check the buffer to see if one went by ?
[14:04:04] <cradek> piasdom: motion.servo.last-period, motion.servo.overruns
[14:04:55] <piasdom> cradek: what do i do with that ?
[14:05:15] <cradek> brb
[14:46:46] <Jymmm> alex_joni: very cool
[14:55:00] <Fox_M|afk> Fox_M|afk is now known as Fox_Muldr
[14:56:11] <elmo40> maybe I will make a lightweight version of the next EMC. Are there 64-bit RTAI kernels?
[15:05:44] <JT-Work> how much does it weigh now?
[15:09:26] <skunkworks> depends on if it's on my keychain drive or a cd.
[15:09:37] <skunkworks> I don't know which one is lighter...
[15:09:57] <skunkworks> when it is installed on the hd - that weighs quite a bit more.
[15:11:16] <JT-Work> LOL
[15:11:30] <JT-Work> mine all seem to be quite heavy
[15:11:47] <JT-Work> one weighs 6k lbs
[15:14:36] <skunkworks> heh - I think ours is 15ton
[15:16:15] <JT-Work> that's heavy for sure
[15:23:02] <wobblybootie> ...
[15:31:28] <ichudov_> Been working on my Bridgeport iNteract last night to improve servo tuning. Went from 0.004-.005" following error to usually under 0.001", almost always under 0.002 even during rapids. Is there some kind of a guideline on how much following error is acceptable.
[15:36:12] <JT-Work> I've heard 10-20 counts on your encoder
[15:37:40] <ichudov_> JT-Work: works out to about 0.0005". So I am close. I would leave it as is right now
[15:38:12] <ichudov_> Do you guys use HeeksCNC? What CAM solutions do you use? I am getting a 4th axis (fixing up an old CNC Troyke rotary table) and would like to get some 4D CAM software
[15:39:07] <JT-Work> I use onecnc for my mill when I don't program by hand
[15:39:40] <ichudov_> JT-Work: this one? http://www.qarm.com.au/
[15:40:20] <ichudov_> does that thing run under Win, just curious. I am not totally hung up on Using Linux, only almost totally
[15:40:26] <ichudov_> under Wine
[15:45:27] <JT-Work> yes, it is windows but atm I don't recommend it...
[15:46:23] <Fox_Muldr> Fox_Muldr is now known as Fox_M|afk
[15:47:02] <JT-Work> it's too constrictive and difficult to change things and has some annoying things
[15:47:12] <ichudov_> got it
[15:47:19] <mshaver> ichudov_: You'll also want to go hang out on #cam
[15:47:40] <ichudov_> I will /join
[15:47:50] <JT-Work> yea those guys over on #cam use heekscad
[15:48:05] <ichudov_> I will talk to them
[15:53:37] <piasdom> what do i do with this --> motion.servo.last-period, motion.servo.overruns
[15:53:52] <piasdom> can't find anything helpful online
[15:59:31] <awallin> I guess you scream in horror if it ever shows a very big value
[16:00:25] <piasdom> awallin: hhahahha thanks
[16:01:44] <piasdom> i'll try to scream quietly
[16:12:01] <skunkworks> killing me softly with overruns - killing me softly... with overruns..
[16:13:22] <piasdom> i want to stay away from that command
[16:15:28] <maximilian_h> JT-Work: Hello, Alex pasted me a line from the emc-devel irc that you use virtualbox
[16:24:31] <piasdom> is ovl max 71782 good,bad,doesn't matter for emc? ovl max just repeats that number
[16:30:46] <piasdom> well, thanks anyway. i do appreciate all ya'll help. <off to search>
[16:51:32] <JT-Work> maximilian_h: yes I use virtualbox ose
[16:54:30] <IchGuckLive> foxtrot_lab: ?
[16:54:44] <IchGuckLive> good evening from germany
[17:36:46] <foxtrot_lab> IchGuckLive, still around?
[17:37:09] <foxtrot_lab> i think ive got my current settings alright now
[17:37:44] <foxtrot_lab> but my problem seems to be now just shoddy couplings and just a crappy design in general
[17:39:05] <foxtrot_lab> http://epicstuff.50webs.com/cnc/cnc3.JPG
[17:39:50] <foxtrot_lab> im going to try extending the X and Y plates to have more travel and put a support bracket on the end of the threaded rod so it wont wobble so much
[17:40:17] <alex_joni> and remember t
[17:40:19] <elmo40> dude, stand still :P
[17:40:22] <alex_joni> and remember a better camera
[17:40:38] <alex_joni> elmo40: earthquake area
[17:42:39] <foxtrot_lab> yeah
[17:42:43] <foxtrot_lab> maybe http://epicstuff.50webs.com/cnc/cnc5.JPG is clearer
[17:45:37] <foxtrot_lab> any ideas for how to improve it
[17:45:54] <L84Supper> it's surprising how much pipe tends to wobble
[17:47:56] <foxtrot_lab> maybe i'll put spacers between all the U channel aluminum rails so that I can slide past the motors instead of them being the limits
[17:58:39] <IchGuckLive> foxtrot_lab: yes i lookt at the frechchees mill you made
[18:00:00] <IchGuckLive> foxtrot_lab: did the tource increase
[18:06:06] <alex_joni> foxtrot_lab: a tripod would work
[18:11:36] <foxtrot_lab> tripod?
[18:11:40] <foxtrot_lab> yes the torque did increase
[18:11:50] <foxtrot_lab> im not losing steps anymore i dont think
[18:11:55] <alex_joni> foxtrot_lab: to fix the camera
[18:11:58] <foxtrot_lab> just lots of problems with friction/rubbing/slidding
[18:12:07] <alex_joni> http://www.gearfuse.com/wp-content/uploads/2009/04/lightweight_tripod.jpg
[18:14:09] <IchGuckLive> foxtrot_lab: what about the head at 24V
[18:14:21] <foxtrot_lab> heat?
[18:14:24] <foxtrot_lab> seems okay
[18:14:31] <foxtrot_lab> the drivers have auto thermal shutdown
[18:14:53] <foxtrot_lab> and i ran it on jog +/- on each driver for 20 minutes each
[18:15:05] <IchGuckLive> B)
[18:15:05] <foxtrot_lab> only things that happened was a wobbly threaded rod fell out
[18:15:41] <foxtrot_lab> currently the threaded rods are only attached by a rubber grommet on the motorshaft
[18:15:47] <IchGuckLive> nice now on with EMC mill the main file into wood
[18:16:01] <foxtrot_lab> ok
[18:16:15] <foxtrot_lab> problem is my travel is currently 8"x3.8"
[18:16:31] <IchGuckLive> you can scale the main file
[18:17:17] <IchGuckLive> axis.ngc
[18:17:28] <IchGuckLive> in the example folder
[18:17:35] <IchGuckLive> it is protected
[18:17:54] <IchGuckLive> copy it to NC-code
[18:18:10] <IchGuckLive> open with the editor and scale it down
[18:18:28] <IchGuckLive> to fit into the trevel limits
[18:18:48] <IchGuckLive> sorry for my bad english im german
[18:19:43] <foxtrot_lab> na your english is good
[18:19:49] <foxtrot_lab> ok i will try
[18:19:58] <foxtrot_lab> what RPM speed do you think ishould try on my dremel
[18:20:01] <foxtrot_lab> for plywood
[18:20:05] <foxtrot_lab> ive only ever milled foam
[18:20:24] <IchGuckLive> less then 4inch
[18:20:56] <IchGuckLive> the numpad will increase it to 120%
[18:21:02] <foxtrot_lab> im going to go to the hardware store to get some supplies, what wood sized/thicknesses would you recommend i get for this
[18:21:07] <IchGuckLive> if you see it working good
[18:21:20] <IchGuckLive> stop the file and double the feedrate
[18:21:54] <IchGuckLive> take the chepest
[18:22:31] <IchGuckLive> if you got a Modell maker intown ask for Silikone stuff like UREOL
[18:22:47] <IchGuckLive> Lecura600 etc.
[18:23:04] <IchGuckLive> then you can mill with 1200mm/min
[18:23:25] <IchGuckLive> 50"/min
[18:23:52] <IchGuckLive> at 3mm with and 5mm depth with your chees mill
[18:24:55] <IchGuckLive> http://www.youtube.com/watch?v=lBMkT39LtEU
[18:25:52] <IchGuckLive> so its late here got to get up early By
[18:26:04] <foxtrot_lab> ok
[19:11:25] <Guest187> Hello, is there anyone home?
[19:11:54] <foxtrot_lab> hi
[19:12:59] <foxtrot_lab> sigh, now my spindle holder isnt holding up
[19:13:07] <foxtrot_lab> dremels going all sideways and shit in wood
[19:15:48] <foxtrot_lab> of course im also using a tiny needletipped dremel attachment
[19:15:55] <foxtrot_lab> not a router bit or anything
[19:16:24] <JT-Work> no I'm at work
[19:18:07] <archivist> "wherever I lay my lathe is my home" sung to wherever I lay my hat :)
[19:20:44] <Guest187> Well, Its been a long time since my last visite to a "chat" room...
[19:21:48] <Guest187> I have built a 4 axis hot wire foam cutter, and waiting for the control electronics from ebay
[19:21:53] <cradek> this is an "irc channel"
[19:22:06] <cradek> cool
[19:22:39] <archivist> me points to the C in irc :)
[19:23:07] <Guest187> Im looking at EMC as the control software... (what does IRC stand for)
[19:23:17] <ichudov_> internet relay chat
[19:23:21] <archivist> internet relay chat
[19:24:15] <ichudov_> EMC s great and very versatile
[19:25:59] <Guest187> By the way, Name here is Matthew. :) can I ask what the basic steps are
[19:26:16] <Guest187> to gettinf from a cad drawing to EMC
[19:27:03] <OoBIGeye> Guest187: well, there are 2 ways. 1 program the g-code by hand
[19:27:37] <OoBIGeye> or 2. get a program to convert say a .dxf file to g-code
[19:28:13] <Guest187> Im aware of the need for G-code, but am doing aerofoil tapered wings, hence probably rather harder to do straight coding
[19:28:50] <OoBIGeye> yea, thats just silly to do by hand...
[19:29:23] <OoBIGeye> i am searching for a free program to convert drawing to g-code myself, so i can help you there...
[19:29:31] <Guest187> Yup, I've found a few links for DXF-G-Code software in the knowledge base
[19:29:59] <OoBIGeye> *can't help you there...
[19:32:29] <Guest187> when doing the cad drawings, might I need to extrapolate the cutting line (hot wire) out from the foam to the X-Y axis, how might this be done
[19:34:20] <cradek> the basic problem is that a cad drawing is flat, so there's not enough information in it to actually make a part
[19:34:34] <cradek> well sometimes there is - for instance a waterjet cutting stuff out of a sheet
[19:35:06] <cradek> some programs START with a cad drawing and then let you add the rest of the information some other way
[19:35:11] <cradek> I don't know what does this well for wire cutting
[19:35:38] <cradek> depending what parts I need to make, I think writing gcode by hand is sometimes the easiest way
[19:36:00] <cradek> http://timeguy.com/cradek/01280719622
[19:36:14] <cradek> for this project I wrote a simple python program that generated the gcode
[19:37:32] <ichudov_> cradek, awesome set, what are the domensions of raw material
[19:37:44] <ichudov_> I write perl scripts that generate g codes
[19:38:04] <cradek> ichudov_: the dominoes are .625 x 1.25
[19:38:17] <ichudov_> what about the raw material pece
[19:39:02] <cradek> it was ~ 2.5" wide piece of scrap .25 plate
[19:40:05] <ichudov_> I would like to make the same thing, that's why I am asking
[19:40:07] <ichudov_> I love it
[19:40:26] <cradek> thanks
[19:40:29] <cradek> they are fun to play
[19:40:41] <cradek> they feel nice
[19:40:45] <Guest187> I like the dominoes... ;) I can get aerofoil plots (X-Y co-ordinates) online - so it might be an idea to look into generating the g-code
[19:40:48] <ichudov_> So I would like to know the raw material and the python if it is possible
[19:41:59] <cradek> yes if you have coordinates, maybe you can generate the gcode yourself.
[19:42:11] <ichudov_> I was just lazy
[19:42:23] <ichudov_> Anyway, nice job
[19:42:46] <cradek> ichudov_: I meant that to Guest
[19:42:51] <cradek> brb
[19:45:15] <Guest187> for a tapered cone as an example with the pointy end cut off, i.e. where the foam ends, I would have the imaginary "point" fixed at the center point of the foam X-Y on say the right hand side
[19:46:10] <Guest187> and then simply draw a circle in the air big enough to create the mouth of the cone on the left
[19:46:51] <SWPadnos> that requires knowledge of the material thickness, among other things
[19:46:59] <andypugh> Ever feel you have entered a conversation at a confusing point?
[19:47:22] <SWPadnos> XY/UV hot wire cutting is the subject, along with CAM for such a machine
[19:48:45] <andypugh> Ah, yes.
[19:49:24] <andypugh> I like the german one made from string, the creator of which I think we see on here occasionally?
[19:49:29] <Guest187> say the foam block is 1/3 the length of the cutting table, I would then need to make the LHS Circle say twice as big to allow for the cone mouth on the foam to be the correct size....
[19:50:35] <andypugh> It depends where the faces of the block are relative to the XY and UV planes
[19:51:35] <Guest187> Ive seen the german one, no. I finished building mine last week - minus electronics (ebay) its two independant x-y axis built from 3mm steel sq box
[19:51:42] <andypugh> However, I can conceive of a kinematics module that, given the distance from the faces to the axes, and the distance between the axes, could virtually move the XY and UV planes to the surface of the block.
[19:52:23] <SWPadnos> that may be more difficult than you think
[19:52:27] <SWPadnos> or possibly easier than I think
[19:52:33] <SWPadnos> or both
[19:53:02] <andypugh> It ought to just be trigonometry, and stuff that has already been solved by raytracing types.
[19:53:44] <andypugh> The wire is at least straight.
[19:53:47] <SWPadnos> that may be true in theory, though I'm not sure it is
[19:53:47] <Guest187> Is Kinematics a software package?
[19:54:08] <SWPadnos> kinematics is how you translate between the machine's actuators and the cartesian coordinate system that G-code uses
[19:54:29] <andypugh> No, it is the part of EMC2 that converts coordinates to actuator positions.
[19:54:55] <andypugh> For many machines it is a 1:1 mapping.
[19:55:32] <andypugh> That string-based hotwire cutter uses non-trivial kinematics to convert XY UV to lengths of string.
[19:56:24] <Guest187> I suppose 1:1 mapping relates to the exact position of the drill / router bit tip
[19:56:42] <SWPadnos> no, to the way that the joints/actuators of the machine move
[19:56:51] <andypugh> The problem is that non-standard kinematics involves writing C code.
[19:57:20] <SWPadnos> a knee mill like a Bridgeport has a 1:1 mapping - it has an X, Y, and Z actuator (the table, saddle, and quill respectively)
[19:57:37] <Guest187> Ah, ive done that before for robots at uni, pain in the rear
[19:57:50] <SWPadnos> a SCARA robot, which is an arm on a pivoting base, has to be controlled in polar coordinates, but the input G-code is still cartesian
[19:58:18] <andypugh> There are a couple of pre-built kinematics modules for robots, so you just need to put in the lengths and offsets.
[19:59:40] <andypugh> But for something that automatically projects the cartesian coordinates off of the faces of the foam onto the physical axes (which I think is the conceptually easier way to look at it) you would need to write a custom module.
[20:01:06] <Guest187> What I am beginng to see is, I need to do some more reading to get up to speed on G-code etc. so for the mean time Ill get bact to the EMC knowledgebase, and check back here later in the week
[20:02:04] <Guest187> Yes, "automatically projects " thats what Im thinkig off .....
[20:04:11] <andypugh> It's do-able, I think. And easier than the last time I persuaded somebody that kinematics was the answer.
[20:06:44] <andypugh> One problem, though. I am not entirely sure that G-code can handle simultaneous arcs in XY and UV
[20:08:06] <skunkworks> arcs can only be created in xyz planes. I would think xy and uv moves would be made up of short line segments.
[20:10:19] <andypugh> Adding L, M, N words to the G2 and G3 words would have an obvious difficulty.
[20:14:11] <skunkworks> * xy xz and yz planes
[20:16:20] <andypugh> Wierd.. http://cgi.ebay.co.uk/KURT-ANGLOCK-VISE-D688-30075-/150473149227
[20:16:48] <cradek> well it's blue
[20:16:51] <skunkworks> that doesn't seem right...
[20:22:12] <Fox_M|afk> Fox_M|afk is now known as Fox_Muldr
[20:29:51] <alex_joni> andypugh: obviously it's a Baldor Vise
[20:30:22] <andypugh> I am trying to work out if it is worth more or less than what it is masquerading as.
[21:00:53] <DaViruz> i kind of bought a huge vertical cnc mill
[21:01:30] <DaViruz> sinumerik 810 control, i tried to get the thing running on the site to test servos etc but i couldn't figure it out
[21:02:40] <Dave911> kind of bought?? .... ;-)
[21:04:33] <DaViruz> moving a 4000kg machine turns out to be a little harder than i initially thought
[21:04:47] <DaViruz> well, moving it is simple, geting it up on the vehicle on the other hand
[21:05:40] <Dave911> That is a good size mill.. Big forklift time..
[21:06:24] <DaViruz> with the right wallet i suppose it's simple enough
[21:06:48] <Dave911> Yes .. money does help.. What do you to move it with?
[21:07:03] <Dave911> do you have ...
[21:09:03] <DaViruz> i can borrow a tractor with suitable trailer, the site i intend to place it on has an overhead crane so getting it off is simple
[21:09:11] <Dave911> Some people have used rollback wreckers to load mills, but that might be beyond their limits. If there is a rigger nearby you might be able to get them to load it for not much $.
[21:09:23] <DaViruz> but i suppose i'll just have to pay someone with a truck with a crane..
[21:10:15] <DaViruz> i have a feeling that moving the machine will end up costing more than the machine itself
[21:10:29] <Dave911> I hired a rigger a while back for $150/hr which included two guys and a 25,000 lbs lift capacity forklift.. it was well worth the cost..
[21:11:19] <DaViruz> i only paid $1000, i suppose that leaves more money for moving it
[21:11:49] <Dave911> moving will cost more than the machine ... yep. Not uncommon..
[21:12:21] <DaViruz> amazingly cheap for such a machine, 1997 model
[21:12:27] <Dave911> Your cost per pound of machine is pretty cheap....
[21:13:52] <Dave911> Once you get much above a bridgeport size the cost of machines often falls on broken ones quite rapidly as the cost of movement of them is expensive. Even scrapping them is expensive.
[21:14:00] <DaViruz> hmm, it's a key way milling machine, that explain the 1300mm X travel and 150mm Y travel..
[21:15:37] <Dave911> Good riggers usually earn their money ... Moving machines is tough, risky work.
[21:16:00] <DaViruz> yeah
[21:19:25] <ds2> or look for old riggers who don't look thinner then a pancake? ;)
[21:25:51] <ds2> 8
[21:27:27] <andypugh> We hired a man and a big HIAB for £300 I think.
[21:30:38] <andypugh> They had a 72 tonne-meter HIAB, so if they can get the truck to within 15m of your machine, it is liftable...
[21:31:48] <andypugh> Even a little 20 tonne-meter HIAB could lift your mill from 5m away.
[21:39:47] <DaViruz> moving stuff around locally is no problem, i have plenty of connections for that
[21:59:48] <DaViruz> anyone has any sinumerik 810 experience? i couldn't get the controller started up, i think it was because the plc reported a low battery level..
[22:04:28] <JT-Hardinge> like on a Hardinge?
[22:04:57] <DaViruz> like on a busch NCNF1 keyway vertical milling machin
[22:08:00] <JT-Hardinge> look anything like this one http://i47.photobucket.com/albums/f163/johnplctech/Hardinge/P1010023.jpg
[22:20:48] <DaViruz> yeah, pretty much exactly
[22:21:17] <DaViruz> the red light with a question mark is lit, the other dark
[22:21:25] <DaViruz> when i power it up i end up in some kind of plc setup menu
[22:21:47] <DaViruz> ican do things like "erase plc flags" "erase plc memory"
[22:22:00] <DaViruz> and i don't know how to get out of that menu..
[22:22:32] <JT-Hardinge> I had a short on my 24v buss that drew mine down to 4-6 volts and that prevented me from booting up the plc
[22:22:50] <JT-Hardinge> I also had a dead battery and it was on the back side of the crt box
[22:23:21] <DaViruz> i get a error message on the battery, but i don't know if that is hindering startup
[22:23:26] <DaViruz> that is the only error message
[22:24:05] <JT-Hardinge> I think if you loose the battery (shut down for a long time) you have load the plc back into the control
[22:24:19] <JT-Hardinge> I have the instructions for that if you don't
[22:25:40] <DaViruz> oh, i'd like that
[22:25:50] <DaViruz> there was a ton of documentation with the machine, it's probably in there somewhere
[22:25:56] <JT-Hardinge> but it might not be the same for a vertical milling machine
[22:26:10] <DaViruz> where do i find the program to load?
[22:26:39] <DaViruz> or is it still in there?
[22:27:04] <JT-Hardinge> I'm not sure what you loose when the battery goes dead
[22:27:11] <DaViruz> http://cncforums.machinetoolhelp.com/machine-repair-troubleshooting/1221-siemens-sinumerik-810c-memory-erased.html
[22:27:14] <DaViruz> seems related
[22:27:17] <JT-Hardinge> robh__: is the expert on them lol
[22:27:50] <DaViruz> someone need to slap siemens on the nose for using volatile memory to store programming..
[22:28:32] <JT-Hardinge> if you have a couple of sheet that have numbered steps with things like "27. PRESS THE SERVICE/DIAGNOSIS KEY ,THEN PRESS THE INITIAL CLEAR SOFT KEY"
[22:28:42] <JT-Hardinge> that is the start up instructions
[22:29:50] <JT-Hardinge> if it is like mine was (before I converted it) if you didn't have 24v buss, air, etc it would not boot up
[22:30:15] <robh__> DaViruz, siemns manuals are free online for there controls, it does have all steps listed in there to reload, do you have the files with paramiters etc
[22:30:39] <DaViruz> i'm not really sure what i have
[22:30:45] <JT-Hardinge> hey there is Rob! just when you need him
[22:31:25] <DaViruz> the guy had half a dozen cnc machines and the documentation was kind of mixed up
[22:31:30] <robh__> should have the PLC program also
[22:31:39] <DaViruz> robh__: are those files supposed to come with the machine in some form?
[22:31:44] <DaViruz> (floppy, cd..)
[22:32:18] <robh__> was it a machine that had been left for some time, as if the battery goes flat your right it does loose everything most time PLC stays, its just paramiters that go missing, ie what the machine has installed
[22:32:47] <robh__> no you dont always get them, but who ever made the machine should be able to give them to you.
[22:33:05] <DaViruz> yeah it's probably been without power for a month or two
[22:33:24] <DaViruz> the incoming power line had been hacksawed off..
[22:33:25] <robh__> only 810 i have is a turning control on hardinge no mill but they are the same inside
[22:33:53] <robh__> o the normal way for people to disconnect there machine, chop any and all cables
[22:34:01] <DaViruz> i guess the price tag if i want any help from siemens is huge?
[22:34:50] <robh__> they are good on support most time i find just need to find someone who knows there stuff.
[22:34:53] <robh__> who made the machine
[22:34:59] <DaViruz> busch
[22:35:13] <DaViruz> (never heard of them before today)
[22:35:57] <JT-Hardinge> DaViruz: is this your machine?
[22:36:25] <DaViruz> yes, it still resides in the previous owners shop though
[22:36:53] <robh__> are they still going? siemns will send you any files to reload to there default state, you will realy need the one the machine builder used, when they wrote the PLC etc if its fully dead
[22:37:09] <robh__> does it turn on? and bring up the operator screen, or does it come up with some error
[22:37:25] <DaViruz> i get some kind of plc setup screen
[22:37:34] <DaViruz> there are options like "erase plc memory" "erase plc flags"
[22:38:01] <DaViruz> the guy i bought it from had bought it as part of a package from a bankrupt business
[22:38:04] <Fox_Muldr> Fox_Muldr is now known as Fox_M|afk
[22:38:07] <DaViruz> h had never used it himself
[22:38:46] <robh__> id try siemens if you cont find busch (google didt find them quickly) see if they can help
[22:39:02] <robh__> like i said if they can send you the PLC and stuff its not too long job to reload it in
[22:39:28] <DaViruz> while messing around with the control someone decided it was a good idea to hit "erase plc flags"
[22:39:40] <DaViruz> so i guess even if memory was intact before it isn't now..
[22:39:41] <robh__> johns had bad plc memory we think as we tryed to reload it, now its on EMC and probly runs alot quicker too
[22:40:12] <JT-Hardinge> a lot better
[22:40:14] <robh__> depends if they went on to erase all plc memory then yes it will be just a paper weight now
[22:40:15] <DaViruz> i thought the control looked pretty nice, i had no intention of converting it
[22:40:41] <DaViruz> seems a bit odd though that you could do that from the operator console?
[22:40:46] <DaViruz> with all keys out
[22:40:50] <JT-Hardinge> DaViruz: if you determine you need parts for the control I still have my old one
[22:40:56] <DaViruz> oh
[22:41:08] <robh__> we are used to fanuc here so programing siemns is a task on its own, untill you get back into it as it has its own cycles etc
[22:42:45] <JT-Hardinge> and for the price of shipping it to you you can have it
[22:43:06] <robh__> DaViruz, if you get some files from siemns and your stuck finidng the manual ill gladly help you out where i can, they take abit of finding online unless you have them already in paper form
[22:43:25] <DaViruz> i believe they exist in paper form
[22:44:30] <DaViruz> is it possible to "customize" it myself to suit the machine if i'm only able to get the files from siemens and not from busch?
[22:45:25] <robh__> PLC is all compiled into its own, so with out the source and program to make it you cant alter it
[22:45:35] <robh__> you cant even alter it on the machine
[22:47:53] <robh__> siemens will probly talk you through on the phone also what to do
[22:48:11] <DaViruz> oh.
[22:48:31] <DaViruz> will he base files offer any functionality at all?
[22:48:38] <DaViruz> servos and spindle would be a good start
[22:49:53] <robh__> once you put the PLC back in, the main nc should load up then
[22:50:15] <robh__> then you can load in any machine paramiters (how many spindles, axis etc) then you are pritty good togo
[22:50:58] <DaViruz> how do i load those parameters?
[22:51:03] <DaViruz> by hand via the operator console?
[22:52:38] <DaViruz> or do i need to get some file from busch for that?
[22:53:20] <robh__> u can do by hand
[22:53:26] <robh__> PLC you will need RS232
[22:53:48] <robh__> parameters you can do on RS232 also, there are some good free DNC programs about linux and windows
[22:54:50] <DaViruz> oh
[22:56:05] <robh__> any PC with a comport will do
[22:57:12] <JT-Hardinge> it's actually easier to convert sometimes than to fix http://i47.photobucket.com/albums/f163/johnplctech/Hardinge/computer04.jpg
[22:57:18] <DaViruz> http://daviruz.meeep.net/sinumerik/DSCF0351.jpg
[22:57:24] <DaViruz> my control looks exactly like that
[22:58:23] <DaViruz> i'd be happy to convert if there was some real damage to the default control, or if it was hopelessly outdated
[22:59:12] <ichudov> I junked my old broken Heidenhain control, sold it on ebay (with disclosure), retrofitted to EMC and I COULD NOT BE HAPPIER
[22:59:30] <ichudov> that stuff looks like a VCR
[22:59:56] <JT-Hardinge> I've never regretted converting the CHNC to EMC
[23:00:43] <DaViruz> http://daviruz.meeep.net/sinumerik/diagnostic-mode.jpg
[23:00:49] <DaViruz> that's exactly the screen i get
[23:00:55] <DaViruz> except mine is in swedish instead of english..
[23:01:00] <DaViruz> err, instead of italian
[23:01:24] <ichudov> I recognoze "macchina" and "password"
[23:01:43] <DaViruz> i recognize plc also
[23:02:04] <ichudov> so it is not bad
[23:02:13] <robh__> DaViruz, in that case you just need its TEA1 settings, TEA2 settings
[23:02:22] <robh__> and SEA and it shiould come alive
[23:02:26] <robh__> should
[23:02:57] <robh__> ours comes up with that screen if we leave it afew months being off and have to reload it
[23:03:51] <DaViruz> also, there is a error message at the top saying the battery level is low or words to that effect
[23:04:58] <robh__> they are pritty pricy it will charge up id say if left on all day
[23:05:56] <DaViruz> what kind of battery is it?
[23:08:10] <robh__> some siemens thing plugs into the back of the control