#emc | Logs for 2008-06-25

[00:00:06] <fusion_crank> thanks for your help Big John
[00:00:13] <BigJohnT> ok, maybe we can both learn something
[00:00:18] <BigJohnT> np
[00:00:34] <BigJohnT> time to cook pizza here
[00:00:45] <archivist> time for bed here
[00:12:06] <dmess> http://www.youtube.com/watch?v=Z25PBj3YlCE&feature=related
[00:35:39] <PMinMO> Linux question...I can't get my 8.04 system to power off automatically. I've searched and seen all kinds of answers but none seem to work, any ideas?
[00:36:12] <jepler> PMinMO: yes, that is a known limitation. use the power switch after the shutdown "progress bar" is empty.
[00:36:50] <PMinMO> is it a bug or?
[00:38:02] <jepler> We disable all ACPI in the kernel, because we have found that having it enabled is bad for realtime on some systems. Unfortunately, the linux powerdown method goes through ACPI, so you get no actual powerdown
[00:38:27] <PMinMO> Ok
[01:00:26] <chr0n1c> yo!
[01:02:41] <fusion_crank> SWPadnos are you back?
[01:21:42] <fusion_crank> jepler, are you available to help?
[01:21:51] <dmwaters> {global notice} Good day all, I need to do a bunch of rerouting to prep for an outage tomorrow. If all goes well, this should not take long. Any further messages about this will be given in wallops, '/mode your_nick w' to see them. Thank you for your patience, and thank you for using freenode!
[01:31:58] <fusion_crank> jepler, are you available to help?
[01:33:23] <cradek> fusion_crank: on irc you should always just ask your question. if someone has the answer and the time, they will answer, even if it's much later.
[01:33:49] <cradek> this advice is also known as DATAJA (don't ask to ask, just ask)
[01:34:01] <fusion_crank> thanks
[01:35:13] <fusion_crank> jelper was helping ealier and mentioned "PERMISSIONS AND UDEV" in the hal_input manpage?
[01:35:24] <fusion_crank> I read the man pages
[01:36:02] <fusion_crank> so you create a text file so that I can use the game pad, but I am still not clear
[01:37:02] <chr0n1c> i usually have to read the hal docs about 38 times when i need to do something i haven't done before
[01:37:24] <chr0n1c> and then try and get it wrong a few times...
[01:37:53] <fusion_crank> I thought a USB game pad should have been simple
[01:37:58] <fusion_crank> ha!
[01:38:12] <chr0n1c> this is linux, though...
[01:38:17] <fusion_crank> yeah
[01:38:25] <chr0n1c> ;)
[01:38:50] <fusion_crank> I've tried unsucsessfully for 4 days to get a graphics driver installed, I gave up
[01:42:34] <fusion_crank> so, does anyone know how to use the .rules files to set permissions for the usb game pad?
[01:44:00] <cradek> I see an example in the hal_input manpage. where specifically are you stuck?
[01:44:54] <fusion_crank> I'm trying to get the usb gamepad to work, and it won't
[01:45:37] <cradek> that gives me no information with which to help. what have you tried? then what happened?
[01:46:24] <fusion_crank> I'm looking above to see what SWPadnos did with me
[01:46:34] <fusion_crank> i'll cut and paste
[01:46:42] <cradek> did you read the manpage? did you follow the instruction there?
[01:47:48] <fusion_crank> I read the man page, but I didn't understand the instruction
[01:48:06] <cradek> ok then ask a question to clarify what you didn't understand
[01:48:26] <fusion_crank> from what I understand, my user doen
[01:48:44] <fusion_crank> doesn't have permision to use the usb pad
[01:49:14] <fusion_crank> so I need to write a file in the ruls folder to assign them
[01:49:26] <fusion_crank> what do I call the file?
[01:49:32] <fusion_crank> what goes in the file?
[01:49:43] <fusion_crank> do I write it as root?
[01:50:03] <fusion_crank> its just a text file right?
[01:50:05] <cradek> you can call the file whatever you like; udev reads all of the files in that directory
[01:50:45] <cradek> have a look at the other files to get the format. It looks like the man page's instruction doesn't have exactly the format, which is probably a documentation bug
[01:50:46] <fusion_crank> the xx-desc.rules the xx doesn't matter?
[01:51:00] <cradek> yes it is just a text file that you will have to write as root
[01:51:13] <fusion_crank> how do you do that?
[01:51:30] <fusion_crank> sudo something?
[01:53:07] <fusion_crank> do you have an emc2.rules file?
[01:53:08] <cradek> at the prompt I used this to create the file: EDITOR=gedit sudoedit /etc/udev/rules.d/emc2-input.rules
[01:53:11] <fusion_crank> would the text in the file be
[01:53:28] <cradek> yes
[01:54:03] <fusion_crank> SUBSYSTEM="input", mode="0660", group="plugdev"
[01:54:22] <fusion_crank> sould I have that rules file too? I don't
[01:54:33] <cradek> I see that MODE and GROUP are uppercase in the other files.
[01:55:09] <cradek> you don't have emc2.rules?
[01:55:16] <cradek> that is part of the installed emc2 package
[01:56:20] <fusion_crank> not in the rules.d folder
[01:56:29] <fusion_crank> is that where it should be?
[01:56:32] <cradek> yes
[01:57:20] <fusion_crank> mine all start with numbers too
[01:57:48] <cradek> my system is dapper. what is your install?
[01:59:13] <fusion_crank> the bdi on the emc website
[01:59:34] <cradek> do you mean the ubuntu live cd install?
[01:59:49] <fusion_crank> yes
[02:00:22] <cradek> ok, that is not called bdi; bdi is another distribution that was used mostly with emc1.
[02:01:09] <cradek> I do not have the new 8.04 install here so I don't know if there is a file in udev, but I'm surprised if there isn't. Either way you can do your gamepad thing.
[02:01:09] <fusion_crank> oh, I thought it was, so a brain dead install and a live cd are different?
[02:01:20] <cradek> yes
[02:01:21] <chr0n1c> * chr0n1c looking
[02:01:25] <fusion_crank> ah
[02:01:39] <chr0n1c> ehh where?
[02:01:50] <chr0n1c> where is it supposed to be?
[02:02:25] <cradek> brb
[02:02:33] <chr0n1c> /etc/udev
[02:04:07] <chr0n1c> http://imagebin.org/21528 <- screenshot of my /etc/udev/rules.d
[02:04:37] <chr0n1c> *it's not a a stock install so it may have extras
[02:09:18] <chr0n1c> oh... i'm running 8.04 and emc2 (the script install)
[02:09:32] <chr0n1c> on an ubuntu studio 64 bit base system
[02:11:27] <fusion_crank> I created the rules file
[02:11:34] <fusion_crank> do I have to reboot?
[02:28:18] <mib_zluhzavu> fusion_crank: sometimes you can run: sudo udevtrigger to test changes without rebooting
[02:58:37] <tom1> when you shutdown under hardy 8.04, after the progress bar appears,
[02:58:37] <tom1> you usually have to hold the power btn for several seconmds
[02:58:37] <tom1> this will get a power down immed on pressing the power btn under 8.04
[02:58:37] <tom1> I added
[02:58:37] <tom1> rmmod snd-hda-intel
[02:58:37] <tom1> to end of
[02:58:39] <tom1> /etc/default/halt
[02:58:41] <tom1> it still doesnt shut down by itself, but this is better
[02:59:36] <tom1> err.. immed after the progress bar ends
[03:10:17] <tom1> that is stranger than strange, tom2 should be my system at work and nobodys there. tomp just left & couldnt check his address.
[03:21:41] <LawrenceG> anyone got a copy of bigjohnt's face.py? the one linked to on the wiki is several versions out of date
[03:22:08] <LawrenceG> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Simple_EMC_G-Code_Generators#Facing_Software
[03:37:02] <LawrenceG> hi Ted
[09:56:30] <ALS_> anyone have links to the fest youtube vidios?
[10:44:42] <anonimasu> h
[10:44:43] <anonimasu> hmm
[10:44:47] <anonimasu> what dist should I choose
[10:44:52] <anonimasu> dapper drake or hardy heron?
[10:49:05] <anonimasu> cool my can adapter runs on linux.
[11:29:31] <micge1> hello
[12:54:44] <skunkworks_> Good morning
[12:55:10] <archivist> tis afternoon
[13:13:00] <skunkworks_> Heh
[14:34:48] <alSMT> where are the fest videos?
[14:39:17] <jepler> alSMT: the only two videos I know of are http://www.youtube.com/watch?v=1EYaM4FkASA http://youtube.com/watch?v=_tA-W7nTNSc
[14:41:11] <alSMT> thanks
[15:19:03] <lewin1> lewin1 is now known as lewing
[15:22:49] <Hugomatic> I'm having trouble with my A axis: when I try to combine a linear (XYZ) move with a rotation in a G1 command, my axis rotates too fast. Should I change something in my ini file?
[15:27:02] <Hugomatic> Or can I specify a different feed for the A axis in my G1 code? I think the problem is that my rotary axis has a velicity of 900 while my X axis only goes 20 inches per min
[15:37:31] <cradek> Hugomatic: 900 degrees/second does seem pretty darn fast to me
[15:39:29] <jepler> Hugomatic: you're saying that in a G1 X- A- move, the cutter is moving over the material faster than you want?
[15:39:53] <Hugomatic> cradek: that's 900 deg/min. But I think my problem is that I am asking for too much feed
[15:40:06] <cradek> oh I see what you're asking, I misunderstood
[15:40:10] <jepler> Hugomatic: emc is probably obeying the rules set out in the documentation regarding feeds for combined rotary and linear moves, but that's not what you desired.
[15:40:24] <cradek> it's possible you should use inverse time mode when doing this.
[15:40:41] <x3rox> What is the actual version of EMC2, please? I didn't watch it long time because I had no time, but I hope that I can continue with my mill in next time...
[15:40:52] <Hugomatic> I did not find that section... I'll have a read
[15:41:03] <jepler> x3rox: the latest release is EMC 2.2.5
[15:41:17] <Hugomatic> I have 2.2.5
[15:41:22] <x3rox> Many changes to 2.2.2?
[15:41:55] <Hugomatic> Its stock.
[15:42:07] <jepler> x3rox: on wiki.linuxcnc.org front page, click the link "view a list of bugs fixed in each release" to get this page: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Released
[15:42:31] <jepler> it looks like there are several dozen items listed in 2.2.3, 2.2.4, and 2.2.5.
[15:42:53] <x3rox> So this is a bugfixed version but not a feature-added one?
[15:43:09] <jepler> That's right.
[15:44:03] <x3rox> And still based on ubuntu (as distro)?
[15:44:13] <jepler> x3rox: here's our version number scheme, in short: from 2.2.x to 2.2.(x+1) there are bugfixes and occasionally new hardware drivers, very rarely any new features. From 2.x.y to 2.(x+1).0 there are new features.
[15:44:36] <x3rox> Great, thanks.
[15:44:38] <jepler> x3rox: we compile emc 2.2.x for Ubuntu 6.06 and 8.04
[15:44:56] <jepler> we have installable/live CD images for 6.06 and 8.04
[15:45:08] <jepler> in principle, emc2 is portable to many other linux distributions
[15:45:32] <x3rox> Very good. :D
[15:45:37] <jepler> Hugomatic: http://linuxcnc.org/docs/html/gcode_main.html#sub:Feed-Rate http://linuxcnc.org/docs/html/gcode_main.html#sub:G93,-G94:-Set
[15:46:24] <Hugomatic> thanks jepler
[15:47:18] <jepler> Hugomatic: in "inverse time feed" mode (G93) you will figure out how long a move should last to have the material removal rate you want
[15:49:03] <Hugomatic> jepler: this "inverse time" thing is nice... I didn't know it existed
[15:49:18] <x3rox> Last time I worked on my mill, was ~ 1 year ago. I am still not finished, because I had no time. At that time I saw the primary problem in how to easily get G codes for 2.5D milling from a CAD system, i.e. AutoCAD in an EASY way.
[15:49:25] <x3rox> Something new here?
[15:49:56] <BigJohnT> EASY cost $$$$
[15:50:06] <jepler> x3rox: emc is fundamentally a CNC system, not a CAD or CAM system. However, sheetcam TNG (commercial software) now has a linux version, and gcam (free software) seems to be progressing every day.
[15:50:54] <x3rox> Some conversion program from plotter or Postscript -> gcode does not exist?
[15:51:26] <x3rox> (Where every pen/color has a preset Z-position)?
[15:54:38] <jepler> the internet finds such programs, among them http://www.chiark.greenend.org.uk/~alanb/ps2gc.html -- but I don't have specific experience with that program
[15:54:45] <jepler> (I notice it has a stupid license statement)
[15:55:09] <cradek> if you have actual AutoCAD, REALIZE does what you are asking for
[15:56:16] <cradek> (and REALIZE is real free software)
[15:58:16] <Hugomatic> jepler: I can't get G93 to work in the MDI. I just typed in G93 and a few g1 commands, but slowing the F parameter slows the tool
[15:58:57] <cradek> are you saying a lower F number gives a slower move?
[15:59:53] <Hugomatic> cradek: yes... I see no change
[16:00:30] <cradek> a lower F number gives a slower move in inverse time [g93] mode
[16:00:43] <cradek> http://www.linuxcnc.org/docview/html//gcode_main.html#sub:G93,-G94:-Set
[16:01:22] <Hugomatic> thanks cradek
[16:01:54] <jepler> if you're at X0 in G93 mode and command X1 F1 the feed rate will be 1 inch/minute so that the move takes 1 minute. If you command X2 F1 the feed rate will be 2 inch/minute so that the move takes 1 minute. If you command X2 F3 the feed rate will be 6 so that 3 moves of that length would complete in 2 minutes. (2*3=6)
[16:02:18] <jepler> er, "so that 3 moves of that length would complete in one minute"
[16:02:35] <jepler> cradek: is that accurate?
[16:03:02] <cradek> accurate but not very clear, I think
[16:03:08] <jepler> hah!
[16:05:16] <Hugomatic> You are probably right, but I would expect to hear a different noise when I'm in G93. Is it an error to give "g93 g1 x1 f20" and "g94 g1 x0 f20" in the MDI window?
[16:06:44] <cradek> those moves will have the same feed because they are one inch long and each take 1/20 of a minute
[16:06:57] <cradek> http://www.linuxcnc.org/docview/html//gcode_main.html#sub:G93,-G94:-Set
[16:07:25] <cradek> read this a few more times and you will see it :-)
[16:07:30] <Hugomatic> ok.. I'll get another coffee, read things slowly, and come back when I get everything. thanks
[16:14:23] <kirk_wallace> Hello. Someone on DIY-CNC may be having trouble with Vista. I was going to respond by suggesting running the EMC LiveCD, but I recall realtime features don't work unless you install to a hard drive. Does anyone have a way to run real-time without affecting the hard drive?
[16:15:40] <cradek> realtime runs from the cd just fine.
[16:16:30] <kirk_wallace> I wonder were I got that notion?
[16:16:41] <jepler> kirk_wallace: because I sit on IRC all day and tell lies
[16:16:53] <jepler> ooh here's one now: actually, the emc live cd *is* vista
[16:17:04] <jepler> so of course real time doesn't work
[16:18:13] <kirk_wallace> Okay, so I am safe in suggesting the LiveCD?
[16:18:42] <cradek> depends what you mean by safe
[16:20:30] <fenn> i'm so gonna sue you for suggesting that
[16:21:30] <fenn> btw what is "DIY-CNC"?
[16:21:43] <kirk_wallace> Safe by running the LiveCD to run steppers and find out how wonderful EMC2 is, but still be able to boot without the CD and be back to the normal MS wasteland.
[16:21:54] <archivist> DIY-CNC we does it :))
[16:22:52] <archivist> kirk_wallace, I put hard disks in caddies to swad windaz to real
[16:22:58] <archivist> swap
[16:23:06] <kirk_wallace> DIY-CNC is a forum, mail list and desert topping.
[16:23:16] <kirk_wallace> dessert?
[16:23:34] <archivist> hmm Lemon Merangue Pie
[16:23:44] <jymm> dessert = double s' as in you want more.
[16:23:44] <kirk_wallace> I suppose both could use it.
[16:23:54] <fenn> Sweet Stuff
[16:24:13] <kirk_wallace> Clever.
[16:27:58] <kirk_wallace> http://groups.yahoo.com/group/DIY-CNC/?v=1&t=search&ch=web&pub=groups&sec=group&slk=1
[16:33:23] <jepler> hi ray, ray
[16:33:59] <jepler> hi ray
[16:34:04] <jepler> er, bye ray
[16:34:11] <jepler> I'm getting dizzy
[16:34:47] <lewing> lewing is now known as lewin1
[16:34:48] <rayh> Yea me too
[16:35:32] <lewin1> lewin1 is now known as lewing
[16:41:22] <dmwaters> {global notice} Good day all, In a few minutes I'm going to be rebooting one of our main us hubs. This will mean some splitting, but things should come back together rather quickly. Thank you for your patience, and thank you for using freenode!
[16:48:40] <jymm> * jymm hands jepler the dramamin
[16:49:24] <chr0n1c> e*
[16:50:37] <chr0n1c> i was like what's dramamin, then i waw like oh.. add an e and you know what that is
[16:52:47] <jymm> Aoccdrnig to rscheearch at an Elingsh uinervtisy, it deosn't mttaer in waht oredr the ltteers in a wrod are, olny taht the frist and lsat ltteres are at the rghit pcleas. The rset can be a toatl mses and you can sitll raed it wouthit a porbelm. Tihs is bcuseae we do not raed ervey lteter by ilstef, but the wrod as a wlohe.
[16:56:31] <rayh> dexlsiya rnus rapmnat
[17:04:54] <chr0n1c> i have a mild case of "only when i type in the dark" dyslexia...
[17:21:35] <kirk_wallace> And if my e-mail is an indicator, Lessdiksia too.
[17:58:26] <chr0n1c> they trying to seel you mor v1@gr@?
[17:58:30] <chr0n1c> sell*
[18:08:35] <jymm> chr0n1c: he might have run out and need more
[18:17:00] <LawrenceG> BigJohnT, hi John, I notice on the wiki, that the face.py version seems to be out of date compared to the text description... ver 1.3.1 vs 1.3.3
[18:20:35] <kirk_wallace> When I run "sudo comp --document turret", I get a "Permission denied: 'turret.9'". I thought sudo would allow turret.9 to be written. "sudo comp --install turret.comp" seems to work fine. Any hints?
[18:22:56] <kirk_wallace> Correction "sudo comp --document turret.comp"
[18:37:16] <jepler> "comp --document turret.comp" should just write the manpage source ("turret.9") to the current directory. "[sudo] comp --install-doc turret.comp" should install it to the same place as the documentation for other components, with sudo required if you use a packaged/run-installed version and not required if you use a self-built, --enable-run-in-place version.
[18:37:21] <jepler> I'm surprised at the behavior you report
[18:56:16] <kirk_wallace> jepler: Thank you, "sudo comp --install-doc turret.comp" gives me what I wanted. I can type "man turret" and get the man page. I guess this shows that sudo is not exactly like root?
[18:56:41] <jepler> kirk_wallace: I'm not sure what it shows.
[19:00:44] <chr0n1c> why not do a "sudo passwd" then set a passwd and then just log in as root with su from the terminal?
[19:01:14] <cradek> sudo -s gives you a root shell
[19:01:24] <chr0n1c> ohh.. didn't know that
[19:01:41] <chr0n1c> i been doing it the su way for things i need to be root for...
[19:04:35] <gefink> or sudo bash
[19:18:03] <BigJohnT> hi LawrenceG
[19:19:53] <kirk_wallace> I think using using "su -" is how I got the idea that my original command should work. Using Ubuntu seems to be getting me away from doing things as root, which might be a good thing.
[19:20:44] <BigJohnT> LawrenceG: I'll take care of that when I get home, thanks for catching that
[19:22:02] <gefink> its easy to use sudo. The Userpasswrd is only onetime necessary
[19:25:56] <anonimasu_> hmm..
[19:25:59] <anonimasu_> sudi su works too..
[19:26:09] <anonimasu_> sudo su - :p
[19:32:38] <LawrenceG> BigJohnT, no problem.... was admiring your work on the python scripts... I would like to port some of the cp1 objects I wrote over to python
[19:32:55] <anonimasu_> hm..
[19:32:56] <BigJohnT> LawrenceG: cool
[19:33:04] <anonimasu_> I wonder what the hell is up with the rutex guys...
[19:33:09] <BigJohnT> what is a cp1 object
[19:33:42] <LawrenceG> the cp1 link you have as other gcode generators
[19:34:35] <LawrenceG> I wrote about 1/2 the objects and have another that never got into the distribution version
[19:34:40] <BigJohnT> Sorry, I've slept since looking at that
[19:35:28] <LawrenceG> I like the python method..
[19:35:47] <BigJohnT> I have the Anliam conversational manual if you would like to see how they did it...
[19:35:51] <BigJohnT> I like python too
[19:36:14] <BigJohnT> Anilam is very easy to use
[19:36:17] <LawrenceG> tcl is very uncomfortable to me... right up there with perl!
[19:36:42] <BigJohnT> I tried tcl too and thumbs down from me
[19:37:12] <LawrenceG> what sort of objects does anilam have? any lathe stuff?
[19:37:29] <BigJohnT> yep they have both lathe and mill.
[19:37:37] <BigJohnT> I'm not sure I have the lathe manual
[19:38:09] <BigJohnT> for simple things it is very fast to program
[19:38:29] <anonimasu_> * anonimasu_ loves the heidenhain for that
[19:38:47] <LawrenceG> I wrote an oword program for turning bearing mounts on the end of ballscrew shafts, but a picture would be nice with real data entry fields
[19:39:57] <BigJohnT> I agree, that was my goal with the g code generators but I have not got there yet
[19:40:49] <LawrenceG> your arc buddy application is close to what I had in mind
[19:41:11] <kirk_wallace> anonimasu_: You have Rutex drives?
[19:41:16] <BigJohnT> that was my latest and best IMHO
[19:41:17] <anonimasu_> no
[19:41:21] <anonimasu_> I tried to buy a thc off them
[19:41:36] <anonimasu_> for the oxyfuel table
[19:43:34] <kirk_wallace> I have drives and motors I haven't used yet, I missed the finer points of using them with EMC when I bought them. Oh well.
[19:44:00] <anonimasu_> but it seems like I have to cook something up
[19:44:01] <anonimasu_> :)
[19:44:03] <BigJohnT> LawrenceG: I can send you a copy of the Anilam manual if like
[19:44:54] <LawrenceG> sure.... email ve7it at shaw dot ca
[19:46:12] <kirk_wallace> anonamousu_: I wish I had more time. EMC seems ripe for a simple low cost THC.
[19:46:35] <chr0n1c> what about a THC with a PIC chip?
[19:46:45] <anonimasu_> wtf..
[19:46:52] <anonimasu_> the problem is that I need a capacitive sensor..
[19:46:54] <chr0n1c> i got 3 samples of 3 different pics and a programmer coming...
[19:46:58] <anonimasu_> with a ring pickup..
[19:47:03] <BigJohnT> LawrenceG: it's on the way
[19:47:15] <BigJohnT> most of the screens are in chapter 4
[19:47:29] <gefink> THC ?
[19:47:34] <anonimasu_> torch height control
[19:47:45] <gefink> thanks
[19:47:52] <anonimasu_> im thinking of using a ultrasonic sensor.. as a makeshift solution
[19:48:03] <anonimasu_> and sense a bit away from the torch
[19:48:13] <BigJohnT> anonimasu_ how flat is your sheets?
[19:48:23] <anonimasu_> pretty flat
[19:48:27] <anonimasu_> 2x1m sheets
[19:48:42] <anonimasu_> 20mm thick
[19:48:55] <BigJohnT> why not use a floating head to find the material top like I do on my plasma
[19:49:04] <BigJohnT> have you seen it?
[19:49:18] <LawrenceG> BigJohnT, Thanks
[19:49:19] <anonimasu_> not yours, but I've seen the arrangement
[19:49:35] <BigJohnT> one sec
[19:49:53] <kirk_wallace> In my oppinion, PIC's only add remote intelligence, why not implement it in EMC?
[19:50:01] <BigJohnT> http://i47.photobucket.com/albums/f163/johnplctech/Plasma%20Cutter/Plasma012.jpg
[19:50:12] <anonimasu_> I have plc's to spare..
[19:50:20] <anonimasu_> why implement it in something stupid like ladder..
[19:50:39] <anonimasu_> :)
[19:50:40] <BigJohnT> anonimasu_ I use a probe move to locate the top of the material using the switch on the lower left
[19:50:47] <anonimasu_> ok
[19:50:50] <gefink> Pics or AVR are much faster in "turnaround" then EMC
[19:50:58] <anonimasu_> turnaround?
[19:51:01] <chr0n1c> wow, the machines you guys build amaze me ;) i'm just here with my mini mill in awe
[19:51:26] <BigJohnT> then I can go to whatever pierce height and then down to cut height
[19:51:40] <anonimasu_> ok
[19:51:49] <anonimasu_> :)
[19:51:52] <gefink> with a AVR you can sample an encoder easey 100khz, emc only 40
[19:52:08] <BigJohnT> real simple to build
[19:52:19] <chr0n1c> i have two old old AB plcs in my garage they weigh about 40#'s each
[19:52:27] <gefink> you cyn use PWM with a us resolution ...
[19:52:38] <anonimasu_> I think you missed something
[19:52:41] <BigJohnT> slc500's or 5's
[19:52:53] <anonimasu_> gefink: that's because encoders are usually interfaced to a card with encoder inputs
[19:53:11] <chr0n1c> i'd have to look, i got them from my electronics buddy when he couldn't store them anymore
[19:53:16] <anonimasu_> * anonimasu_ nods
[19:53:31] <chr0n1c> i never have programmed PLCs but i have wired plenty of panels for them
[19:53:33] <anonimasu_> * anonimasu_ is a big fan of plcopen(iec-61131-3)
[19:53:37] <gefink> anonimasu the Avr can sometime replace such hardware
[19:53:40] <BigJohnT> the size will tell all chr0n1c
[19:53:52] <chr0n1c> my first job ever was wiring panels for PLC/automation cells
[19:54:31] <BigJohnT> about 6x6xhowlong is a 500, about 12x9xhowlong is a 5
[19:55:28] <anonimasu_> :)
[19:56:05] <chr0n1c> ok, i just walked out to check them out.... 1771 i.o chassis
[19:56:13] <chr0n1c> i guess there is no actual polc in them...
[19:56:27] <chr0n1c> plc*
[19:56:31] <BigJohnT> just a power supply and a backplane?
[19:56:34] <chr0n1c> yeah...
[19:56:42] <chr0n1c> never even looked at them before now really
[19:56:51] <kirk_wallace> BigJohnT: How much splatter and how hot comes off the tourch?
[19:57:19] <gefink> anonimasu: you can sure solve such things with EMC PID
[19:57:31] <BigJohnT> kirk_wallace: with a water bath it is minimal and I assume you mean my plasma
[19:57:58] <BigJohnT> dry splatters crap everywhere and a fine dust covers everything
[19:58:49] <anonimasu_> gefink: what things?
[19:58:59] <gefink> THC
[19:59:22] <kirk_wallace> I suppose a big mess isn't a necessay evil.
[19:59:51] <anonimasu_> gefink: I have a plc that's acting as a operator interface for emc..
[19:59:56] <BigJohnT> If you use a water bath it is not bad
[20:00:04] <anonimasu_> gefink: it does it pretty well
[20:00:13] <gefink> ah
[20:00:45] <anonimasu_> well it will do it..
[20:01:03] <anonimasu_> :)
[20:02:34] <kirk_wallace> Darn, I just re-read my question, and I guess it _could_ be taken badly, oops.
[20:06:08] <BigJohnT> LawrenceG: did you get the info?
[20:12:56] <BigJohnT> * BigJohnT is headed out to do some mindless activities so my synapse can recharge
[20:30:43] <cradek> is there a good way to test a square without a reference square?
[20:32:11] <cradek> without the step "first make a cylindrical reference square"
[20:33:09] <archivist> methinks there is an optical way
[20:34:19] <cradek> I have a surface plate, gage blocks, etc, but no reference square
[20:34:59] <OoBIGeye> are you trying to check if the machine is square?
[20:35:18] <archivist> I think a prism is involved
[20:35:19] <cradek> no, say I have a square in my hand. I want to know if it's square.
[20:36:01] <cradek> somehow I bet "now turn it around and put it against the thing that previously sat against it" will be part of the procedure
[20:36:15] <chr0n1c> cradek, there is a formula that the egyptians used to build the pyramids
[20:36:36] <chr0n1c> oops, soory boutthe caps
[20:36:56] <cradek> yeah that would depend on my being able to measure several lengths accurately. I hope there's something simpler.
[20:37:01] <OoBIGeye> cradek do you have a mill at your disposal, that you know is square?
[20:37:34] <cradek> OoBIGeye: no, I'm looking for a method that doesn't require having any other square-like thing
[20:37:39] <chr0n1c> if a line from 2 corners is the same length of the other two corner's line it's square
[20:38:04] <OoBIGeye> chr0n1c: may be on to something...
[20:38:27] <cradek> I need to be more precise - I have a thing with a ~ 90 degree angle and I want to know how close it is to actually being 90
[20:38:33] <archivist> 3 4 5 triangle has a square
[20:38:53] <cradek> it's not a square [four sided shape], it's a square [two bars at right angles used for checking whether corners are 90 degrees]
[20:39:20] <chr0n1c> cradek, do you have a surface plate and a height gauge w/an indicator?
[20:39:23] <OoBIGeye> are you guys looking for the a2=sqrt(B2 +C2)
[20:39:44] <ds2> make or get a cylindrical square
[20:39:46] <cradek> chr0n1c: yes, but I don't want to trust the height gauge to be square to the plate
[20:39:51] <chr0n1c> an angle to clamp the square too would be handy as well
[20:40:07] <ds2> cradek: do you have 2 of these?
[20:40:12] <archivist> time to make a cylindrical
[20:40:14] <cradek> ds2: nope
[20:41:14] <OoBIGeye> cradek: if you have calippers you can get within 0.02mm of sqare
[20:41:31] <ds2> start by making a pair of angle blocks
[20:41:45] <ds2> then compare it against that
[20:43:26] <chr0n1c> if you have a surface grinder you could always just touch it up and know it's square
[20:43:35] <archivist> how about a good level and a rectangular block, test parallelism with the level turn 90 deg and test 90 any offset from square will show on the level, if good use as reference
[20:43:45] <chr0n1c> *if you are good with the surface grinder that is
[20:46:16] <lerman> Do you have a flat (that you trust)? And a parallel?
[20:46:31] <cradek> flat yes
[20:46:39] <cradek> parallel ... I can check with a micrometer.
[20:46:43] <cradek> so yes
[20:47:12] <archivist> a matched pair of 3,4,5 blocks?
[20:47:38] <lerman> Take the parallel and clamp it on a block so it is standing upright on the flat. Adjust it using the square so that one side is square to the flat.
[20:48:30] <lerman> Then check the other side of the parallel with the square. Any space near the top or bottom shows how far out of square the square is. (Or how unparallel the parallel is).
[20:49:18] <lerman> That procedure uses the square to create a second square using a parallel, a flat, a block, and a clamp.
[20:49:20] <cradek> ahhhh I think you have it
[20:49:31] <cradek> use a parallel, since it can be verified
[20:50:55] <archivist> level can also verify the table and parallel
[20:52:53] <lerman> Oh Python gurus: How do I invoke an external script, program, etc from python, passing arguments and getting return values, stdout...
[20:53:38] <lerman> At Fest, Dale convinced me that GWiz should permit external programs written in any language to be wizards.
[20:55:01] <cradek> yay
[20:55:03] <lerman> In "C", I could use popen.
[20:55:36] <seb_kuzminsky> lerman: i've used twisted python for that, though it may be too heavy for you
[20:55:49] <seb_kuzminsky> heavy-weight i mean, it does a lot you dont care about
[20:55:54] <cradek> lerman: I'm sure there's an os.popen
[20:56:13] <lerman> The present plan is python and wxPython. I already have to use more than I want to.
[20:56:26] <cradek> I bet it will suck in all the ways that popen sucks though
[20:57:00] <cradek> like in C, you ought to use execv* instead of popen
[20:58:06] <jepler> there is a module called subprocess
[20:58:24] <jepler> it lets you choose how each of the standard file descriptors are redirected in the child, and it lets you pass an argument list (not string)
[20:58:30] <lerman> That would work if there is a python interface that let's me connect the pipes. I want to be able to pass arguments and get error messages in case of failure.
[20:59:00] <lerman> Thanks, jepler. That should do the trick.
[20:59:16] <jepler> you can get stdout and stderr to separate Python file objects
[20:59:55] <seb_kuzminsky> jepler: do you have a minute to talk about outside dependencies on #emc-devel?
[21:00:22] <jepler> here's a snippet of related python: http://rafb.net/p/38vYaj24.html
[21:07:13] <lerman> Thanks, jepler -- a sample is always welcome.
[21:08:51] <lerman> As you can probably tell from the previous,I'm now back to python for GWiz. The reason I had given up on it is that I was getting seg faults on some of my code. I didn't think that interpreted languages were supposed to do that.
[21:09:16] <lerman> Anyway, I found the problem in my use of wxPython and I'm back to work on that version.
[21:09:18] <jepler> nope
[21:10:22] <lerman> The seg fault came from within the wxWidgets code which is wrapped in some C++ stuff that I don't need to understand (yet).
[21:10:46] <jepler> it's very easy for C++ code wrapped into Python to segfault -- no harder than it is for C++ code on its own
[21:11:31] <jepler> but lay the blame on wxpython, not python
[21:11:34] <lerman> So I've seen. And it's probably harder to find the bug.
[21:11:54] <lerman> Agreed. Not a python problem.
[21:14:57] <lerman> My test program for this facility might be a generic flat bottom (parallel to XY plane) pocketing program. It should support N sides, radiused corners, islands, etc.
[21:15:30] <jepler> you realize you're biting off quite a bit there, right?
[21:16:02] <anonimasu_> islands are not easy.
[21:16:10] <anonimasu_> -_-
[21:16:11] <jepler> nonconvex shapes aren't easy
[21:16:17] <jepler> islands are beyond this mortal brain
[21:16:19] <lerman> I've done most of it before, as part of a graphics area fill program.
[21:17:14] <lerman> I would start by tracing the inside boundary. Then convert the inside to a raster image. Finally, run around the inside erasing bits.
[21:17:28] <lerman> bits == pixels
[21:18:09] <jepler> using rasters == cheating
[21:18:14] <ds2> cradek: why about using your mill to verify it after tramming it first?
[21:18:29] <anonimasu_> hm
[21:18:40] <anonimasu_> this is where voronoi diagrams come into play
[21:18:43] <lerman> Cheating == good.
[21:19:00] <ds2> indicator on tool holder and just jog it up and down on the peice while sitting on the table
[21:19:21] <lerman> That's why it needs to be an external program. Writing raster scan algorithms in gcode would be slow and difficult.
[21:19:31] <archivist> ds2, why should a machine be any better
[21:19:41] <ds2> archivist: not better, just more available
[21:19:51] <ds2> archivist: when one lacks tools...
[21:20:16] <archivist> ds2, he was looking for first principles
[21:20:26] <ds2> ah I miss understood
[21:20:41] <ds2> nevermind... back into the tunnels for me ;)
[21:21:06] <lerman> Tramming it assures that the spindle is square to the table at a single height. It doesn't assure that the quill motion is perpendicular to the table.
[21:21:23] <lerman> To check that, you need a ... square.
[21:21:45] <ds2> you are thinking of bridge port
[21:22:09] <ds2> I am thinking of something like the haas minimills or the sherline/taig stuff where the head moves and there is no quill
[21:22:22] <ds2> so the only assumption I am really making there is the Z ways are straight
[21:22:41] <lerman> You have the same problem. The spindle might not be parallel to the Z ways.
[21:23:02] <archivist> and my mill wasnt
[21:23:33] <lerman> Of course, it doesn't have to be. The kinematics could correct for that error.
[21:23:33] <ds2> Hmmmm that makes it nasty problem to solve
[21:23:54] <ds2> it would mean fly cutting any height other then the trammed height could be problematic
[21:24:32] <archivist> I got a saw tooth form, got rid of the mill as soon as I could
[21:24:34] <lerman> That's why you need a square. With a square you can check that the ways are parallel to the spindle.
[21:24:50] <ds2> *nod*
[21:25:10] <anonimasu_> why not just get a set of paralells?
[21:25:15] <anonimasu_> or one..
[21:25:54] <fenn> you can measure the squareness of a mill spindle to the table by sweeping a test indicator in a circle
[21:26:06] <lerman> Sure. You can check by tramming at a variety of heights.
[21:26:42] <anonimasu_> that still dosent help you determinate that the column is straight
[21:27:15] <anonimasu_> unless you do it at several places
[21:27:48] <archivist> I have a nice book at home on machine testing
[21:28:03] <lerman> Three places should be enough to determine the radius of curvature if it is not straight.
[21:28:12] <archivist> levels are very useful
[21:29:45] <lerman> I have a precision level and a precision cylindrical square. I have to admit that I've used the level more than I've used the square.
[21:30:55] <archivist> testing dividing heads is entertaining
[21:31:12] <lerman> Have you ever made a hirth coupling?
[21:31:28] <archivist> no
[21:31:48] <ds2> what do you use the level for?
[21:31:52] <lerman> So, how do you test a dividing head?
[21:32:39] <lerman> Use the level to level your machine. If the table is level then the quill should be upright.
[21:32:55] <archivist> polygon and colimator
[21:33:18] <anonimasu_> lerman: you didnt tram a chinese mill before did you?
[21:33:22] <lerman> I suppose you could test the head by stepping around, drilling holes and measuring the space between the holes.
[21:33:43] <lerman> Nope never trammed one (whoops, I lie, forgot about my smithy).
[21:33:59] <archivist> lerman, basicly how I found the first errors here yes
[21:34:04] <lerman> Sure, everyone has a precision polygon and colimator.
[21:34:40] <lerman> That's on my list of tools to get after the laser interferometer is working.
[21:34:46] <archivist> vernier over a number of gear teeth
[21:35:21] <archivist> I want to play with lasers one day
[21:35:40] <lerman> After seeing cradek's 5 axis mill, I decided I want to build a lathe with live tooling and servo C axis.
[21:36:27] <lerman> I would use the same servo for spindle and C axis with gearing (actually belts) to change speeds. For a clutch I was thinking of using a hirth coupling.
[21:36:52] <anonimasu_> Hmm..
[21:36:53] <archivist> after seeing the Citizen sliding heads I want to mod a Bechler we have here
[21:37:08] <anonimasu_> hm.. hirth couplings arent involute right?
[21:37:11] <lerman> I could cut the first ones on my mill. Then cut the next generation on the lathe.
[21:37:47] <lerman> Nope. Hifth couplings are Vee's radially from the center to the perimeter.
[21:38:54] <lerman> I was thinking I could use the mill or the lathe as a shaper with a 60 degree carbide insert as a tool.
[21:39:02] <archivist> anonimasu, http://www.amcprecision.com/v-tooth.htm
[21:39:24] <anonimasu_> I see
[21:39:39] <anonimasu_> so the only way to make them is with a v bit..
[21:39:53] <anonimasu_> easily..
[21:39:59] <anonimasu_> :)
[21:40:53] <lerman> They could be milled with a rotating V, but I think shaping would give a better finish.
[21:41:57] <lerman> It would help to have a motorized indexer. That way, a cut could be made on each tooth going around the circle. Tool wear would then still leave all the teeth pretty much the same.
[21:44:38] <archivist> centering to shaft axis has got to be good
[21:45:14] <archivist> which is my current problem
[21:46:07] <lerman> Why is that difficult?
[21:46:55] <archivist> not knowing the center line of cutter
[21:48:50] <lerman> What are you making? What type of machine? What type of tool?
[21:48:53] <archivist> gear cutters have some runout, so does the arbor, the arbor does not have a register (collet mounted)
[21:49:04] <archivist> gear milling
[21:50:03] <lerman> Cut a groove with the mill. The center of the cutter is aligned with the center of the cutter. :-)
[21:50:17] <lerman> S/B center of the groove.
[21:50:26] <archivist> Im still using the 1956 Safag swiss machine for 3mm pinions as its easier to stuff a microscope on
[21:51:16] <lerman> How may teeth on a 3mm pinion?
[21:51:29] <lerman> What diameter cutter?
[21:51:36] <archivist> 6,7,8,9 ish
[21:52:01] <archivist> little over a 1/4"
[21:52:49] <archivist> I need to get within a few tenths (trial and error at the moment)
[21:53:36] <lerman> Are those cycloidal teeth on the small pinion?
[21:53:44] <archivist> yes
[21:54:33] <lerman> Where are you located?
[21:54:38] <archivist> UK
[21:55:14] <lerman> Too far for a day trip to your shop to see this tiny stuff.
[21:55:46] <archivist> like the fest was too far for me
[21:56:28] <lerman> It may be a small world, but not quite small enough sometimes.
[21:56:34] <archivist> get cradek to show you a watch pinion
[21:56:51] <lerman> Does he do that stuff?
[21:57:15] <lerman> I've seen the inside of watches. Hasn't every kid taken one apart?
[21:57:27] <lerman> I've just never seen one made.
[21:57:46] <SWPadnos> it's putting them back thogether that's the hard part
[21:57:53] <SWPadnos> (and having them still work)
[21:57:58] <archivist> we make one of replacement stuff mainly
[21:58:23] <lerman> SWPadnos: that's what my twin brother says. He's a veterinarian.
[21:58:47] <SWPadnos> heh
[21:59:05] <lerman> Carving a turkey in front of him usually gets the comment: now let's see you put it back together.
[21:59:49] <lerman> SWPadnos: Did the bracket get made? How did it work out?
[22:00:42] <SWPadnos> yep. we (meaning cradek and jmkasunich) made a few, a couple were given away aas raffle prizes, and Kim stood there and made a bunch while we weren't looking :)
[22:01:00] <SWPadnos> I think there was a stack of 10 or so complete ones when we left
[22:01:05] <anonimasu_> watches?
[22:01:18] <lerman> So I guess this CNC stuff will actually catch on.
[22:01:26] <SWPadnos> some people seem to think so
[22:02:07] <chr0n1c> * chr0n1c goes back to a manual mill and a trig calculator
[22:02:38] <lerman> You've got a calculator? We have to use a pencil and a set of trig tables.
[22:02:48] <chr0n1c> oh my!
[22:03:01] <chr0n1c> hey i finally found a spellchecker for xchat ;)
[22:03:23] <archivist> heh it would segfault here
[22:03:36] <lerman> And it keeps telling you that chr0n1c is spelled wrong?
[22:03:54] <archivist> mazter of the tyop
[22:03:58] <chr0n1c> ehh... i hit the ignore button on chr0n1c
[22:04:14] <chr0n1c> it should go into the next dictionary revision
[22:07:04] <ds2> Mmmmm trig tables
[22:07:40] <ds2> trig tables are nicer in some ways
[22:07:59] <archivist> they have the random errors
[22:08:24] <ds2> for real accuracy, use a calc but quick estimates, trig tables work better
[22:08:36] <ds2> and none of this radian crap either ;)
[22:08:54] <anonimasu_> hehe
[22:09:28] <archivist> some of the early calculators had nicely broken trig
[22:10:51] <chr0n1c> there is a plain calculator at work they bought for $0.50 at a garage sale that only gives you the right answer sometimes.. and the guy still uses it
[22:11:12] <chr0n1c> amazingly everything he builds comes out awesome
[22:11:16] <anonimasu_> lol
[22:11:25] <anonimasu_> he's lucky
[22:11:39] <anonimasu_> doing trig with a broken calc -_-
[22:11:43] <anonimasu_> makes for some scary errors
[22:19:02] <ds2> before calcs, they teach you little short cuts and tricks to avoid having to deal with sin/cos/tan tables
[22:22:50] <Larks> hello
[22:23:23] <Larks> I am interested in obtaining a small scale cnc mill for hobbying in my garage. Where do I start?
[22:23:48] <ds2> define small
[22:24:09] <Larks> convenient =). I was trying to decide between x2 and x3 size.
[22:24:27] <Larks> something benchtop preferably.
[22:24:30] <ds2> small enough for a Chevy V8? small enough for a lawn mower? small enough for a 0.050cc ICE, or small enough to require microscope for work ;)
[22:24:31] <anonimasu_> bigger is always better ^^
[22:24:48] <ds2> IMO, I'd get a ready to go taig
[22:25:02] <Larks> yes? and why?
[22:25:14] <jmkasunich> bigger is always better, until you can't find space for it, or move it into the space
[22:25:16] <ds2> less things to irritate and tick you off when using it
[22:25:30] <anonimasu_> a more stable machine makes better parts
[22:25:32] <ds2> problem with the X2 is, there is an endless array of problems to deal with
[22:25:36] <jmkasunich> note that if you are in a northern climate, you might prefer smaller if you can get it into a heated space
[22:25:44] <anonimasu_> and you wont be fighting with setups
[22:25:47] <anonimasu_> as much
[22:26:06] <Larks> I am very novice (even amateur)
[22:26:30] <ds2> i've used the 7x12 lathe and a taig mill... the 7x12 always has odd ball gotchas.. the taig mostly just work (not as well as a nice bridgeport or haas)
[22:26:46] <Larks> My only experience is on an old (sharp I think) bridgport type manual built in probably the 40's or 50's
[22:26:47] <ds2> so no hogging out 1/4" steel on the taig ;)
[22:27:35] <Larks> probably mostly cutting softwer materiels such as aluminum, plastic, and even wood.
[22:28:22] <Larks> I would love to have a large one (mill that is =) ), but at this point in my life I am looking for something to compliment my hobbies and not get in the wife's way
[22:28:55] <Larks> price is also an important consideration.
[22:29:03] <Larks> how much are Taig setups running?
[22:29:09] <ds2> what's the target price?
[22:29:14] <Larks> cheap.
[22:29:15] <Larks> lol
[22:29:27] <Larks> acap = as cheap as possible
[22:29:36] <Larks> like I said, its to hobby, not make things and sell them
[22:29:58] <Larks> r/c parts, guitar inlays, odd creations I come up with.
[22:30:09] <anonimasu_> that also affects precision and stuff
[22:30:10] <ds2> yes but it is also great for making odd ball parts for repairs
[22:30:18] <anonimasu_> $$$ is a big factor
[22:30:24] <Larks> yes it is.
[22:30:40] <Larks> so lets say around the $2000 mark.
[22:30:45] <ds2> way too much potmetal is used for things... some times it is better to machine a replacement part out of aluminum then to buy yet another potmetal part
[22:31:18] <ds2> taigs go for around $1200-$1500 ready to run, the last time I check... or you can get a manual one and add your own controller for less
[22:31:41] <Larks> hmm that sounds pretty good price range.
[22:31:50] <Larks> what kind of travel do they have?
[22:31:56] <jmkasunich> small
[22:32:11] <Larks> comparable to x2?
[22:32:17] <Larks> or smaller? or more like x3?
[22:32:17] <ds2> yes, comparable to X2
[22:32:28] <chr0n1c> mini drill press frame stripped out, 3 precision linear slides, 3 small steppers, a $100 control card and EMC2 = perfect hobby machine
[22:32:32] <jmkasunich> http://www.taigtools.com/mmill.html
[22:32:40] <ds2> both taig, X2, X3 have a very tight Z clearance compared to a knee mill
[22:32:52] <jmkasunich> http://www.mini-lathe.com/Mini_mill/Versions/versions.htm
[22:32:55] <jmkasunich> google is your friend
[22:33:07] <Larks> i have been googling for 4 hours lol
[22:33:14] <Larks> and read lots.
[22:33:22] <Larks> I want some real opinions
[22:33:56] <ds2> OTH, if you want a frame work to build the machine.. the X2 is good for that
[22:33:57] <jmkasunich> well, you asked about travels - the URLs I just pasted have the travel numbers
[22:34:08] <ds2> I think the taigs don't have enough room for a ball screen retrofit
[22:34:09] <Larks> yes thanks.
[22:34:22] <Larks> problem is, you guys talk about alot of stuff I know nothing about.
[22:34:35] <jmkasunich> I really do think you should get as large as you have space and money for
[22:34:38] <Larks> I am a smart guy, but not a machinist.
[22:34:56] <ds2> Larks: a class at the local JC is a good idea
[22:35:00] <Larks> I don't mind spending coin, but I hope to know why I am.
[22:35:06] <jmkasunich> but a major item in the "have money for" equation is: do you want to get a manual mill and convert it, or do you want a CNC ready mill?
[22:35:15] <ds2> know and understand how things work first then worry about equipment
[22:36:07] <Larks> I could go either/ read or convert. I hear converting sometimes (though rewarding) takes lots of time.
[22:36:18] <jmkasunich> yeah, it can
[22:36:26] <Larks> I do like using things manually however.
[22:36:35] <Larks> so preferable it could do both
[22:36:55] <jmkasunich> the basic (manual) X2 mills are in the $400-500 range
[22:37:15] <jmkasunich> a taig is $750 for manual, $1000 for "cnc ready" which does not include motors or drives
[22:37:30] <jmkasunich> probably $1500 if you bought it with motors and drives, but I didn't see that on the site
[22:38:15] <Larks> does taig make something a little larger, but not real large.Like the x3 for instance.
[22:38:17] <jmkasunich> did you see this page: http://www.littlemachineshop.com/Info/minimill_compare.php
[22:38:24] <jmkasunich> I don't think so
[22:39:00] <Larks> ds2: I have machined several things in college. I built custom xray tube holders.
[22:39:01] <chr0n1c> harbor freight has a medium sized mill, i can't vouch for it's worthiness though
[22:39:56] <ds2> whoa, full taig systems prices have gone up
[22:39:57] <Larks> The other problem I have with converting: most people save money by making some of their parts. I have nothing to do that with now.
[22:40:09] <ds2> http://www.cartertools.com/price.html
[22:40:15] <Larks> yah, I thought those prices sounded low.
[22:40:17] <jmkasunich> if you are new to machining, and even remotely interested in doing manual work, I'd buy a manual mill of a decent size and learn to use it as a manual
[22:40:19] <ds2> those are list prices, subtrace 10% if you order from the guy
[22:40:19] <jmkasunich> then convert it
[22:40:26] <chr0n1c> Larks, if you get the mill to convert you'll have something to use to make parts
[22:40:39] <Larks> good point.
[22:40:43] <ds2> ah I see... the new prices include a Mach license and the ER spindle
[22:41:51] <Larks> Do you guys know of a good walkthrough for doing what I am looking to do and using open source software like EMC?
[22:42:04] <jmkasunich> heh
[22:42:05] <jmkasunich> no
[22:42:11] <Larks> ;)(
[22:42:14] <jmkasunich> the software is the simplest part
[22:42:20] <chr0n1c> when i built my machine i used bits from thousands of websites
[22:42:29] <jmkasunich> learning to be a machinist isn't hard, but its not a cookbook thing either
[22:42:32] <Larks> someone needs to combine them :)
[22:42:32] <ds2> the other consideration is - do you want to do machining or do you want to build a machine
[22:42:34] <chr0n1c> ok maybe only about 20 different websites
[22:42:54] <Larks> I want to do machining.
[22:43:03] <Larks> but I am not opposed to learning either
[22:43:06] <jmkasunich> btw, I would _not_ buy the complete CNC from taig unless you _really_ want a turnkey system
[22:43:13] <fusion_crank> Hello all
[22:43:18] <jmkasunich> they seem to charge about $1000 for motors/drives/computer/sw
[22:43:24] <jmkasunich> and I know you can do loads better than that
[22:43:37] <ds2> jmkasunich: you just think it costs too much?
[22:43:47] <jmkasunich> emc is free, and it can run on a old (free or nearly free) PC
[22:44:06] <ds2> It used to be about $1500 before they went to ER spindle and Mach
[22:44:09] <jmkasunich> motors are $75 each max, drives maybe $50 per axis
[22:44:17] <Larks> ok guys, if I go the route of converting a system. what platform do I start with (maybe x3'ish size)?
[22:44:50] <jmkasunich> Larks: you are asking for a simple answer to a complex question
[22:44:59] <ds2> at the old price, getting it at once place saved a lot of running around
[22:45:02] <jmkasunich> your individual needs, wants, and skills are a big part of the answer
[22:45:24] <chr0n1c> Larks, i have a buncha links i collected when i was reearching my machine here towards the bottom of the pagehttp://ohiopctech.com/dp/?q=node/41
[22:45:28] <chr0n1c> http://ohiopctech.com/dp/?q=node/41
[22:45:38] <Larks> thanks, I'll check them out
[22:46:01] <jmkasunich> the one thing I can tell you - a machine can never be too big or too heavy, except when you need to move it
[22:46:03] <chr0n1c> the links start under the line -> CNC!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
[22:46:18] <jmkasunich> so consider where you are going to put it and how often you might need to move it
[22:46:28] <jmkasunich> then get the biggest thing that you can deal with
[22:46:40] <Larks> thats x3'ish size
[22:46:43] <Larks> for now
[22:46:52] <fusion_crank> Jelper, can you help again with my usb pendant?
[22:46:54] <Larks> later (several years) I might build a shop and go bigger
[22:46:57] <jmkasunich> the machines that you can pick up with one hand are nearly toys - the ones you can pick up with two hands are a little better
[22:47:14] <Larks> I don't mind if I need to call the neighbors to pick it up
[22:47:19] <jmkasunich> get up to a hundred or 200 lbs and you can actually cut some metal without taking all day
[22:47:21] <Larks> my biggest problem is footprint
[22:47:44] <Larks> I need someone to say: "If I had it all to do over..."
[22:47:49] <jmkasunich> heh
[22:47:53] <Larks> starting out that is.
[22:48:10] <jmkasunich> I have a 700 lb 3-in-1 (lathe/mill/drill) in the basement, and an 1800 lb knee mill in the garage
[22:48:24] <jmkasunich> the garage mill is a constant fight against rust, but it can CUT
[22:48:29] <fenn> wouldnt it be great if we had real specs for machine tools regarding rigidity, damping factor, material removal rate
[22:48:44] <jmkasunich> the 3-in-1 is an OK lathe, mediocre mill, and crappy drill press
[22:48:51] <jmkasunich> but I have a real drill press
[22:48:57] <chr0n1c> if i had to do it all over... i would just buy a bridgeport and convert it instead of F*%$ing around with this mini cnc stuff
[22:49:20] <Larks> most of the annoyances I can deal with. Typically when working with the mills I have worked with, it takes about 7x as long to setup as it does to cut.
[22:49:43] <Larks> chronic:mini that bad?
[22:49:59] <Larks> even with small parts?
[22:50:10] <chr0n1c> Larks, it's very limited in it's potential other than a learning tool or engraving stuff
[22:50:22] <Larks> oh I see.
[22:50:22] <fusion_crank> I started with a Sherline, sold it, then a mill drill, sold it, and now have a 6x26 knee mill-drill, I love it
[22:50:25] <Larks> I didn't realize that
[22:50:29] <BigJohnT> LawrenceG: I've uploaded the most recent face.py file
[22:50:43] <Larks> Ok guys.
[22:50:46] <BigJohnT> fusion_crank: you have any luck with your game pad
[22:50:47] <Larks> thanks for your help
[22:50:55] <Larks> I might just have to build a shop first
[22:50:59] <fusion_crank> Sort of
[22:51:07] <BigJohnT> hows that?
[22:51:12] <Larks> I will stop back by and hound you with more questions soon enough
[22:51:14] <fusion_crank> I added my user to the root group
[22:51:28] <jmkasunich> Larks: one option would be to get a $300 mini mill, use it to learn on
[22:51:33] <BigJohnT> are you running 6.06?
[22:51:40] <jmkasunich> don't put a pile of money into CNCing it
[22:51:53] <jmkasunich> just think of the $300 as machining tuition
[22:52:02] <fusion_crank> and now I have an input listing under the machine show hal configuration
[22:52:14] <jmkasunich> gotta go, back in a few hours
[22:52:20] <Larks> thanks for your help
[22:52:25] <fusion_crank> no its 8.04
[22:52:40] <fusion_crank> I think is has something to do with permissions
[22:52:49] <BigJohnT> Ok I think your right
[22:53:02] <BigJohnT> I don't have 8.04 on any of my machines
[22:53:08] <fusion_crank> when I do that test you mentioned from manfredi? I can see the numbers change
[22:53:23] <LawrenceG> BigJohnT, Thanks... I got the Anilam docs
[22:53:33] <BigJohnT> cool LawrenceG
[22:54:11] <BigJohnT> fusion_crank: if you open up the input pins and put each one in the watch window you can see which button is which
[22:54:32] <fusion_crank> I know I have 8.04 in the machine i'm on now and whatever was on the live CD right before 8.04 is installed in the computer that is contolling my mill
[22:54:41] <fusion_crank> they seem to have the same problems
[22:55:28] <LawrenceG> BigJohnT, any idea if the canvas widget can render a jpg file?... I was thinking of doing a dwg and capturing it as a graphic to show params
[22:55:29] <fusion_crank> I commented out all the lines but the firt one in the postgui.hal file and it still won't work
[22:56:23] <BigJohnT> does input show up now?
[22:56:30] <fusion_crank> yes
[22:56:56] <BigJohnT> check the names of the pins for your device to see if they match the ones on my device
[22:57:12] <fusion_crank> antoher person helping me yesterday with .rules files said he had a rules file for emc
[22:57:17] <fusion_crank> I don't have one
[22:57:20] <fusion_crank> do you?
[22:57:28] <BigJohnT> not that I know of
[22:58:07] <fusion_crank> hmm
[22:58:30] <BigJohnT> It looks like you are at the point of verifying the pin names
[22:59:23] <fusion_crank> how do you do that?
[22:59:35] <BigJohnT> LawrenceG: let me check
[23:00:16] <BigJohnT> fusion_crank: open the hal conf window
[23:00:36] <fusion_crank> I'm there
[23:00:50] <fusion_crank> what do I compare?
[23:00:58] <BigJohnT> click on the + sign for Pins
[23:01:05] <BigJohnT> click on the + sign for input
[23:01:22] <chr0n1c> * chr0n1c wonders if the xbox 360 wired controller would work... (would save me from stretching across the room to jog)
[23:01:46] <fusion_crank> I see them
[23:01:56] <fusion_crank> do I just make sure they are spelled the same?
[23:02:04] <BigJohnT> yes
[23:02:07] <BigJohnT> to start
[23:02:12] <fusion_crank> sounds simple
[23:04:28] <BigJohnT> LawrenceG: yes canvas has an image object
[23:05:13] <BigJohnT> you have to include PIL to use graphic images in a Tkinter app
[23:05:55] <BigJohnT> fusion_crank: if you click on the watch tab then click on each name it will add them to the watch window and as you press each button you can see which one changes
[23:06:52] <fusion_crank> ok
[23:12:59] <fusion_crank> john, I can watch abs x, y and z change when I move the button
[23:13:17] <BigJohnT> cool
[23:13:41] <fusion_crank> but I have the entire halui file comented out
[23:13:48] <fusion_crank> or emc wont' run
[23:14:03] <BigJohnT> do the names match exactly
[23:14:25] <fusion_crank> the names on the input do
[23:14:45] <fusion_crank> I have no pins that start with joy
[23:14:56] <fusion_crank> should I have them?
[23:15:44] <BigJohnT> if you have the lines commented out no
[23:15:58] <fusion_crank> thats what I thought
[23:16:21] <BigJohnT> do the simple one uncomment setp halui.jog.speed 10
[23:16:46] <fusion_crank> I do that, and it crashes emc
[23:17:10] <BigJohnT> go to halui and see if you have a pin jog speed
[23:17:31] <BigJohnT> you have tried to uncomment that one line?
[23:18:46] <fusion_crank> where is halui?
[23:19:02] <BigJohnT> should be under pins
[23:20:30] <fusion_crank> nope, not there
[23:20:44] <fusion_crank> I have axisui
[23:21:12] <BigJohnT> halui is not under hal_manualtoolchange?
[23:22:17] <fusion_crank> nope, input is
[23:26:27] <BigJohnT> * BigJohnT wonders why you don't have halui with 8.04 and I have it with 6.06
[23:26:43] <fusion_crank> I do too
[23:27:08] <BigJohnT> let me look some
[23:27:44] <fusion_crank> is there a repository that would have the 6.06 version on a live cd?
[23:28:22] <BigJohnT> fusion_crank: in your ini file do you have a line HALUI = halui in the [HAL] section?
[23:29:12] <fusion_crank> is that the one that stepconfig makes?
[23:29:28] <fusion_crank> mymill.ini?
[23:29:32] <BigJohnT> yes
[23:30:02] <fusion_crank> no
[23:30:09] <BigJohnT> add that line
[23:30:26] <BigJohnT> I'm sorry I totally missed that
[23:32:28] <fusion_crank> ok! got further
[23:32:36] <BigJohnT> Sweet!
[23:32:52] <fusion_crank> sum2.0 not found
[23:32:57] <fusion_crank> whats that?
[23:33:27] <BigJohnT> that is used if you have to reverse an axis
[23:33:49] <BigJohnT> one moment
[23:35:14] <fusion_crank> its puking on mux4.0
[23:35:17] <fusion_crank> too
[23:38:09] <BigJohnT> comment out everything after sep halui.jog-speed 10 up to net joy-x...
[23:38:40] <BigJohnT> s/sep/setp
[23:39:40] <BigJohnT> the reason it failed on mux and sum is you skipped that part of the instructions on loading them
[23:40:32] <fusion_crank> it works!
[23:40:36] <fusion_crank> woohoo!
[23:40:46] <fusion_crank> do I need the parts commented out?
[23:41:18] <BigJohnT> ok if you want to use the buttons on the joy pad to control the jog speed yes
[23:41:31] <fusion_crank> that would be nice
[23:41:42] <fusion_crank> but the wife is giving me the sting eye
[23:41:50] <fusion_crank> I have to go get the kids
[23:41:55] <fusion_crank> stink eye
[23:42:07] <fusion_crank> thanks so much for your help john!
[23:42:10] <BigJohnT> ok, go back and read the wiki page again later
[23:42:18] <fusion_crank> will do!
[23:42:18] <BigJohnT> Thanks for helping me clear it up
[23:42:26] <fusion_crank> I'll find you later
[23:42:28] <fusion_crank> bye
[23:42:38] <BigJohnT> ok have a cold one on me
[23:42:43] <fusion_crank> will do!
[23:43:05] <BigJohnT> * BigJohnT needs to veg now on the couch