Back
[00:02:12] <PCW> So for example if you machined a convex path the tool center has to move much faster than the cut path to
[00:02:13] <PCW> maintain a constant cutting speed? is that what that form question is about?
[00:02:22] <PCW> forum
[00:04:23] -!-
theorbtwo has quit [Ping timeout: 264 seconds]
[00:04:30] theorb is now known as
theorbtwo
[00:04:47] <andypugh> Yes, I think so.
[00:19:55] -!-
Valen has quit [Quit: Leaving.]
[00:24:09] -!-
bootnecklad has quit [Read error: Connection reset by peer]
[00:31:21] <cradek> andypugh: that question is independent of any cutter comp issue, IMO, and like jepler says, the answer is no
[00:32:21] <cradek> the next answer is that you can use inverse time mode whenever you know something the gcode doesn't about which part of the tool you care about feeding
[00:32:58] <andypugh> I am not sure I understand the meaning of "no" here.
[00:33:00] <cradek> it's true that in comp mode, you may think the control knows something more about which part of the tool is making the important part of the cut
[00:33:28] <andypugh> Surely it does, as that is the programmed arc?
[00:33:34] <cradek> whether or not you're using cutter comp, it feeds the tool center (controlled point) at the specified rate
[00:34:14] <cradek> whenever you feed an arc, you're leaving two arcish cuts that were "fed" at different rates, one on your left, one on your right
[00:34:14] <andypugh> OK, so it is definitely the tool centre velocity that is controlled, not the position on the programmed arc?
[00:34:33] <cradek> one is faster than the requested F, the other is slower
[00:34:46] <cradek> yes definitely the tool center
[00:35:02] <andypugh> That seems a little odd.
[00:35:36] <andypugh> If I program a move of length N, then I would expect it to take N/F seconds.
[00:35:45] <cradek> for arcs very small compared to the tool, you've gotta be careful. I think cam packages do this compensation
[00:35:57] <cradek> well obviously the length changes when you comp the arc
[00:36:20] <andypugh> The length of my G3 doesn't change.
[00:36:42] <cradek> that is true
[00:36:51] <andypugh> The axis movement does.
[00:36:57] <cradek> this is even more swirly when you think of how inverse time mode should work
[00:37:10] <cradek> there you DO specify the time, but the length changes
[00:37:28] <andypugh> I am tempted to take a stopwatch to a sim to check this.
[00:38:13] <cradek> comp also cuts around corners, so you don't even need arcs to get length changes
[00:39:20] <andypugh> Hmm, yes, outside corners are 0 distance of cut, finite too-centre move
[00:39:53] <cradek> when you're thinking about feed adjustment when using comp, consider comp of negative diameters, or diameters that are really actual minus nominal
[00:40:03] <cradek> there's not one right answer
[00:40:30] <andypugh> I feel that, in an ideal world, if I program a 1" arc at 1"/s it should take 1s, regardless of the tool I am using.
[00:40:31] <cradek> that being said, I think if you can make a stronger argument for our G93+comp handling being wrong
[00:41:17] <cradek> if you use a 1" cutter to cut inside it, it should stop in the corner for 1 second?
[00:42:03] <cradek> I don't think keeping that time constant is the important thing at all
[00:42:33] <cradek> er I confused arc length and radius, but you get my drift hopefully
[00:43:13] <andypugh> Well, if you use a 0.999" cutter, yes.
[00:43:32] <andypugh> As I think that equates to constant chip volume?
[00:43:57] <andypugh> 1" cutter in 1" radius is wrong :-)
[00:44:25] <cradek> now I think you're confusing radius and diameter, but I understand what you mean
[00:46:36] <cradek> fwiw see interp_convert.cc:379
[00:46:58] <cradek> this is the place you could make a strong(er) case for the current behavior being wrong
[00:47:17] <cradek> and iirc it's different than historical emc versions
[00:48:27] Computer_ is now known as
Computer
[00:48:33] -!-
Computer has quit [Changing host]
[00:49:18] <andypugh> you can also make an argument that macro-substitutions can make code hard to read.
[00:49:36] <cradek> heh, we could rightfully make many arguments
[00:50:07] <cradek> oops sorry, I meant line 337
[00:50:11] <cradek> I don't know how I got that wrong
[00:51:08] -!-
danielfalck [
[email protected]] has joined #emc-devel
[00:56:20] <andypugh> "The
[00:56:20] <andypugh> feed is still set to the original programmed arc length divided by the F
[00:56:21] <andypugh> number" seems to say the converse of what you have said...
[00:56:39] <cradek> hmmm
[00:56:45] <andypugh> No, I need to learn to read
[00:56:55] -!-
YK has quit [Quit: Page closed]
[00:56:56] <cradek> I wonder: was I smarter when I wrote that comment, or now?
[00:57:06] <cradek> I'm changing my answer to I don't know how it works
[00:57:35] <andypugh> Inverse time mode is not the normal mm/s or in/s mode, is it?
[00:57:38] <cradek> I don't know if your user was talking about G93 at all
[00:57:52] <cradek> do you mean /min? no it is not
[00:58:40] <andypugh> I have never used inverse time, I don't think. /rev and /sec yes.
[00:59:20] <andypugh> This is the (short) forum thread:
http://www.linuxcnc.org/component/option,com_kunena/Itemid,20/func,view/catid,38/id,12707
[01:00:12] <cradek> drilling holes in cutter comp mode is utterly crazypants
[01:01:28] <cradek> and for the regulating "speed" (I think he means feed?) I don't think there's enough detail there for me to understand what he's asking about
[01:02:34] <cradek> I am still pretty sure you don't know enough to get it right in the general case, since cutter comp can be used as full dia, or offset from nominal dia in either direction
[01:14:31] <andypugh> Aye, though folk using it for wear comp are probably a small minority, and a wierd one.
[01:17:16] <danielfalck> Hi Chris, Andy- I'm starting to help a friend with a new gui for a lathe and we are looking into 'load_tool_table' while emc is running a program (before the appropriate tool change). Do you see any show stoppers that would prevent me from hacking this in- we want to keep the machine running while measuring parts (high volume stuff) and edit the tooltable prior to a toolchange and have the new tool diameter value reloade
[01:17:57] <danielfalck> my friend just bought a lathe that is identical to John T's Hardinge
[01:18:43] <andypugh> AFAIK toolchange reloads the tool table, it will probably just work.
[01:18:59] <cradek> I'm afraid the whole architecture is against you being able to "just hack it in"
[01:19:21] <cradek> andypugh: I don't think that's true
[01:19:35] <danielfalck> we did a crude hack using G10 L1 Pxxx :) and subroutines last night
[01:19:37] <cradek> on some machine it *writes over* the tool table
[01:19:39] <andypugh> AFAIK is often not very far
[01:20:05] <cradek> sure you can modify it with G10 during a program run
[01:20:21] <cradek> but I don't understand how that's related to what you are asking about
[01:20:36] <cradek> can you explain more?
[01:20:42] <danielfalck> we created a separate file with named variables that we read with the G10- and were able to change the values in the tool table with that
[01:21:03] <andypugh> A G10 based on a measurement of the previous part, at the start of each part might be interesting.
[01:21:15] <danielfalck> so, you edit the external file as you are running, place that G10 call just before a toolchange and it changes the tooltable
[01:22:10] <danielfalck> that's exactly what we want- my friend and I work at a shop that is full of Fanuc controlled lathes and we deal with multi multi thousand piece runs
[01:22:30] <danielfalck> so, the operators have to measure parts as parts are running and they chase around the offsets
[01:22:37] <andypugh> I just noticed the time (0221) and remembered I have a driving course tomorrow (skid-pan, that sort of thing). I have to go,
[01:23:03] <cradek> many people have asked for reading the tool automatically and more often
[01:23:11] -!-
andypugh has quit [Quit: andypugh]
[01:23:29] <danielfalck> my friend's initial question on the forum was here , for reference :
[01:23:30] <danielfalck> http://www.linuxcnc.org/component/option,com_kunena/Itemid,20/func,view/catid,38/id,12518/lang,english/
[01:24:32] <cradek> that's two issues (adding wear offsets, and reading them frequently)
[01:25:04] <danielfalck> we thought it would be good to approach things one at a time
[01:25:27] <danielfalck> the wear offsets could be dealt with later
[01:26:10] <danielfalck> I've been grepping around in the source looking for references to reloading tool table
[01:26:24] <cradek> well the trick is it's done in manual mode
[01:26:49] <danielfalck> 'mini' didn't complain, when I reloaded an externally edited tool table while running in auto mode
[01:27:01] <danielfalck> all the other guis did though
[01:27:18] <cradek> ok but did it work?
[01:28:02] <danielfalck> I didn't actually create a toolpath with cutter comp, when I tested it. Since that test, I can't seem to bring up mini or tkemc, for some reason
[01:28:17] <danielfalck> that's another story, I guess
[01:28:20] <cradek> what does cutter comp have to do with it?
[01:28:35] <cradek> thought you were talking about tool offsets (g43)
[01:28:50] <danielfalck> reloading a tool table with a cutter that has been changed from .75 to .001 will show you very quickly if you made a real change
[01:28:52] <cradek> or do you mean diameter too?
[01:29:19] <danielfalck> ok, let's back up a bit- let's start with diameter first
[01:30:29] <danielfalck> I run mills in my home shop and don't have a cnc lathe, so I will be concerned about diameter first and foremost. So, I'm a little biased on getting that to reload first
[01:30:53] <cradek> rereading diameter from tool table at ever g41/g42 is a semi-common request
[01:31:05] <cradek> at every
[01:31:22] <danielfalck> in the production shop, it's a must have feature
[01:31:50] <danielfalck> I'm surprised that Stuart didn't request it already
[01:32:07] <danielfalck> but, his parts are huge and take a while to load onto the table :)
[01:32:59] <danielfalck> my friend just picked up the Hardinge for a song and is in the process of ripping the guts out of the GE Fanuc control. He just got some Mesa cards and is going the emc route
[01:33:32] <danielfalck> he was an operator from our shop floor, so he's used to the Fanuc way of doing things- so we are at a 'how can we do this' point
[01:33:51] <danielfalck> he's just getting into Touchy and likes it
[01:34:24] <cradek> it always seems tough on people when they expect a new control to work like an old control
[01:34:51] <cradek> I hope you can get it going how you want, but I'm afraid what you ask for isn't just a simple hack
[01:35:04] <danielfalck> we're more than willing to spend time on working on it though.
[01:35:10] <cradek> excellent
[01:35:33] <danielfalck> Sorry about the hack reference. I've been working on Heekscad /cnc for the last couple years- I know it can be a can of worms
[01:36:15] <cradek> nah, no offense taken or anything. just that I can see how it should seem simple, but I also happen to know it's a rabbit hole and I'm trying to warn, not discourage, you
[01:36:38] <danielfalck> Justin is getting into Glade too- he wants to work a bit on the gui himself - I told him I could help hook it up to some of the underlying logic
[01:37:26] <cradek> are either of you fluent in git yet?
[01:37:43] <danielfalck> I'm not fluent, but we are using it on the HeeksCAD project now
[01:37:50] <danielfalck> mhabler is helping us
[01:37:56] <cradek> that's great
[01:38:07] <danielfalck> it's been very good for us
[01:38:37] <danielfalck> I love being able to switch branches and do little things
[01:38:50] <cradek> yeah, our lives would all be harder if we didn't have it.
[01:39:03] <danielfalck> oh, by the way - waterline machining works well in heekscnc now
[01:39:09] <cradek> nifty
[01:39:22] <danielfalck> http://www.youtube.com/watch?v=IOWVesqJhkA
[01:39:34] <danielfalck> a crude video I did of a part I was machining Sunday
[01:39:35] <cradek> after my few fixes (and one project I cut) I completely lost steam working on/with heeks
[01:39:57] <cradek> does the V tool engraving stuff work right yet?
[01:40:09] <danielfalck> we haven't worked on that one
[01:40:14] <danielfalck> I wish it did
[01:40:18] <cradek> me too
[01:40:34] <danielfalck> I've been using it for 2.5 D stuff mostly and can trust it now
[01:40:49] <cradek> I recall I changed it from not really working to sort of working
[01:41:16] <danielfalck> that part in the video was done with it- I had a bunch of drills and end mills and a reamer for the part - 14 tools
[01:41:21] -!-
theorbtwo has quit [Ping timeout: 258 seconds]
[01:41:31] theorb is now known as
theorbtwo
[01:41:37] <danielfalck> I'm glad it has python scripting- I would go insane without it
[01:41:42] <cradek> nice. I would mostly hand write something like that.
[01:42:34] <danielfalck> I've had 3 revisions to that part (I'm doing it for a customer) and I have been able to reuse the underlying machining logic and just change the geometry
[01:44:53] <danielfalck> back to the Hardinge and tool table stuff- if I can pull it off, I will let you know (and hopefully not pester you too much )
[01:45:08] <cradek> I'd love to see you do it
[01:45:11] <danielfalck> I am using git with it too
[01:45:31] <cradek> anymore, I bet mhaberler is going to know more about tool table than me
[01:45:37] <cradek> if he's working with you already, bonus
[01:46:19] <danielfalck> that's funny- I've been working with him for a while on heekscad and didn't even think to ask about emc :)
[01:46:51] <danielfalck> he told me he was busy with emc
[01:47:12] <danielfalck> how is the Mori doing for you?
[01:48:05] <cradek> it's a really good machine
[01:48:27] <danielfalck> we have some big Mori cnc lathes- I like them a lot
[01:51:26] <cradek> bbl
[01:51:31] <danielfalck> ok
[01:54:25] <danielfalck> bookmark
[02:08:27] -!-
ve7it has quit [Remote host closed the connection]
[02:19:34] <PCW> So I dont think I've broken other cards with the new card patch (5I20 works, trying more now) 5I25 works and has correct pin#s
[02:34:44] -!-
ries has quit [Quit: ries]
[02:40:17] -!-
i_tarzan has quit [Ping timeout: 252 seconds]
[03:44:04] -!-
mikegg has quit [Quit: Leaving]
[03:50:41] -!-
sumpfralle has quit [Quit: Leaving.]
[04:27:49] -!-
danielfalck [
[email protected]] has parted #emc-devel
[04:47:03] -!-
mhaberler [
[email protected]] has joined #emc-devel
[05:09:59] -!-
WalterN has quit [Read error: Connection reset by peer]
[05:19:01] -!-
Eartaker has quit [Quit: Leaving]
[05:33:30] -!-
psha[work] [psha[work]
[email protected]] has joined #emc-devel
[06:34:22] -!-
vladimirek [
[email protected]] has joined #emc-devel
[06:44:53] -!-
e-ndy [e-ndy!~jkastner@nat/redhat/x-ccqfkqdvhhqnbytk] has joined #emc-devel
[06:50:38] -!-
sumpfralle has quit [Client Quit]
[07:02:35] -!-
Tom_itx has quit [Ping timeout: 264 seconds]
[07:14:45] -!-
Tom_itx has quit [Ping timeout: 246 seconds]
[07:20:15] -!-
psha[work] has quit [Quit: Lost terminal]
[07:40:39] -!-
awallin_ [awallin_!~quassel@2001:708:110:1020:224:7eff:feda:7c7d] has joined #emc-devel
[07:50:15] -!-
robh__ [
[email protected]] has joined #emc-devel
[08:00:40] -!-
mhaberler has quit [Quit: mhaberler]
[08:04:50] -!-
mhaberler [
[email protected]] has joined #emc-devel
[08:08:29] -!-
nicko has quit [Quit: Visitor from www.linuxcnc.org]
[08:10:55] -!-
Danimal_garage has quit [Read error: Connection reset by peer]
[08:14:47] -!-
cevad has quit [Quit: Leaving]
[08:21:02] -!-
bootnecklad has quit [Read error: Connection reset by peer]
[08:26:31] -!-
odiug has quit [Ping timeout: 246 seconds]
[08:49:04] -!-
El_Matarife has quit [Quit: Nettalk6 - www.ntalk.de]
[10:11:52] -!-
mhaberler has quit [Ping timeout: 240 seconds]
[10:40:51] -!-
Poincare has quit [Quit: changing servers]
[10:47:30] -!-
mhaberler [
[email protected]] has joined #emc-devel
[11:00:18] -!-
automata [
[email protected]] has joined #emc-devel
[11:00:51] -!-
psha[work] [psha[work]
[email protected]] has joined #emc-devel
[11:02:47] -!-
mhaberler has quit [Quit: mhaberler]
[11:13:53] -!-
mhaberler [
[email protected]] has joined #emc-devel
[11:43:14] -!-
awallin_ has quit [Remote host closed the connection]
[11:54:06] -!-
awallin_ [awallin_!~quassel@2001:708:110:1020:224:7eff:feda:7c7d] has joined #emc-devel
[12:03:17] -!-
mhaberler has quit [Quit: mhaberler]
[12:09:16] -!-
rooks has quit [Quit: So long, and thanks for all the fish.]
[12:17:38] -!-
awallin_ has quit [Remote host closed the connection]
[12:19:10] -!-
awallin_ [awallin_!~quassel@2001:708:110:1020:224:7eff:feda:7c7d] has joined #emc-devel
[13:20:28] -!-
theos has quit [Disconnected by services]
[13:27:05] -!-
automata has quit [Remote host closed the connection]
[13:59:21] -!-
Loetmichel has quit [Ping timeout: 252 seconds]
[14:04:53] -!-
tom3p [
[email protected]] has joined #emc-devel
[14:30:11] -!-
funkster has quit [Ping timeout: 264 seconds]
[14:37:42] -!-
FinboySlick has quit [Ping timeout: 240 seconds]
[14:55:19] -!-
Valen has quit [Quit: Leaving.]
[15:01:08] -!-
e-ndy has quit [Quit: Ex-Chat]
[15:14:00] -!-
adb [
[email protected]] has joined #emc-devel
[15:34:56] -!-
psha[work] has quit [Quit: Lost terminal]
[16:14:22] -!-
awallin_ has quit [Remote host closed the connection]
[16:20:26] -!-
ve7it [
[email protected]] has joined #emc-devel
[16:20:53] -!-
odiug has quit [Ping timeout: 252 seconds]
[16:22:30] -!-
tom3p [
[email protected]] has parted #emc-devel
[16:27:09] -!-
automata [
[email protected]] has joined #emc-devel
[16:37:24] -!-
e-ndy [
[email protected]] has joined #emc-devel
[16:45:36] -!-
automata has quit [Ping timeout: 260 seconds]
[16:45:44] -!-
automata [
[email protected]] has joined #emc-devel
[16:50:52] -!-
mhaberler [
[email protected]] has joined #emc-devel
[16:52:01] -!-
psha [
[email protected]] has joined #emc-devel
[16:56:44] <psha> mhaberler: should i rebase ontop of v2.5_branch_gladevcp_fixes
[16:56:44] <psha> ?
[16:57:34] <mhaberler> hm, your stuff is in ? what would you want to rebase? try configs/gladevcp emc gaxis.ini - all there
[16:58:01] <mhaberler> I just want the jog buttons and the preset ngc filename
[17:00:14] <psha> i guess there is no gaxis.ini in that branch
[17:00:21] <psha> i'm referring to one you mentioned in your mail
[17:00:47] <psha> ah
[17:00:48] <psha> sorry
[17:00:49] <psha> my failt
[17:15:34] -!-
automata has quit [Ping timeout: 245 seconds]
[17:19:39] -!-
automata [
[email protected]] has joined #emc-devel
[17:27:37] <CIA-11> EMC: 03seb 07v2.5_branch * rebf80e37fdaf 10/src/ (10 files in 10 dirs): build system: honor BUILD_VERBOSE better
[17:27:40] <CIA-11> EMC: 03seb 07v2.5_branch * r08d26c1c4c4a 10/ (debian/control.in src/configure.in): packaging: remove build-dependency on libtk-img
[17:50:13] -!-
psha has quit [Quit: Lost terminal]
[18:03:06] -!-
psha [
[email protected]] has joined #emc-devel
[18:04:12] -!-
sumpfralle has quit [Ping timeout: 240 seconds]
[18:28:12] -!-
kljsdfhklj has quit [Quit: kljsdfhklj]
[18:36:29] -!-
Danimal_garage has quit [Read error: Connection reset by peer]
[18:41:58] -!-
mhaberler has quit [Quit: mhaberler]
[18:54:59] -!-
sumpfralle1 has quit [Ping timeout: 258 seconds]
[19:04:13] -!-
isssy has quit [Quit: Visitor from www.linuxcnc.org]
[19:17:31] -!-
sumpfralle has quit [Ping timeout: 250 seconds]
[19:25:45] -!-
automata has quit [Ping timeout: 250 seconds]
[19:28:22] -!-
andypugh [andypugh!~andy2@cpc2-basl1-0-0-cust1037.basl.cable.virginmedia.com] has joined #emc-devel
[19:35:52] -!-
psha has quit [Quit: leaving]
[19:39:00] -!-
ssloy has quit [Quit: leaving]
[19:48:06] -!-
Danimal_garage has quit [Read error: Connection reset by peer]
[20:06:42] -!-
Danimal_garage has quit [Read error: Connection reset by peer]
[20:37:44] -!-
isssy has quit [Quit: Visitor from www.linuxcnc.org]
[20:38:11] -!-
e-ndy has quit [Quit: Ex-Chat]
[20:40:08] -!-
Mjolinor has quit [Quit: Leaving]
[20:43:47] -!-
syyl has quit [Quit: Leaving]
[20:54:54] -!-
FinboySlick has quit [Quit: Leaving.]
[20:56:39] -!-
PCW has quit [Remote host closed the connection]
[21:03:19] -!-
isssy has quit [Quit: Visitor from www.linuxcnc.org]
[21:16:53] -!-
odiug has quit [Quit: Leaving.]
[21:21:08] -!-
vladimirek has quit [Remote host closed the connection]
[21:55:31] -!-
sumpfralle1 has quit [Ping timeout: 258 seconds]
[22:23:04] -!-
n2diy has quit [Quit: Ex-Chat]
[22:32:59] -!-
andypugh has quit [Quit: andypugh]
[22:34:41] -!-
bootnecklad has quit [Ping timeout: 240 seconds]
[22:51:02] -!-
mhaberler [
[email protected]] has joined #emc-devel
[23:10:21] -!-
mhaberler has quit [Quit: mhaberler]
[23:10:26] -!-
kljsdfhklj has quit [Quit: kljsdfhklj]
[23:27:11] -!-
robh__ has quit [Ping timeout: 240 seconds]
[23:37:23] -!-
n2diy has quit [Quit: Ex-Chat]
[23:43:57] -!-
El_Matarife has quit [Quit: Nettalk6 - www.ntalk.de]
[23:50:14] bootnecklad` is now known as
bootnecklad