#emc | Logs for 2007-12-07

[00:12:48] <BigJohnT> jepler: I'm not getting it to work darn
[00:13:41] <jepler> BigJohnT: to use it that way, you have to set up your .ini to know about .py files, then choose face.py from the axis "open" dialog
[00:15:20] <BigJohnT> jepler: I get near line 2 of /face.py Bad number format. I've made the changes to the ini to point .py to python
[00:15:48] <BigJohnT> And the title of the error was G-Code error in face.py
[00:15:55] <skunkworks> jepler - what is it called when axis is drawing the tool path while running?
[00:18:31] <BigJohnT> jepler: ok I changed the wrong ini file it works now
[00:19:37] <BigJohnT> jepler: SWEEEEET! as they say down in Arkansas...
[00:20:01] <eric_U> jlmjvm: did you find me any 13 tooth timing pulleys for my bridgeport?
[00:20:11] <BigJohnT> jepler: how can we make that standard in AXIS?
[00:23:59] <BigJohnT> jepler: how can we add that to the AXIS menu as well as the bolt circle one?
[00:27:14] <BigJohnT> It took like 10 seconds to generate a facing profile and have EMC doing it WOW!
[00:27:47] <jepler> skunkworks: "backplot"?
[00:28:03] <jepler> BigJohnT: glad you got it to work
[00:28:22] <BigJohnT> jepler: thanks for the help
[00:28:50] <jlmjvm> eric_U:just sent that guy an email,will have it for ya tomorrow
[00:29:25] <BigJohnT> jepler: is there a way to add the g-code generators to AXIS without doing the OPEN routine?
[00:31:25] <jepler> BigJohnT: not at the present time
[00:32:11] <BigJohnT> jepler: ok, I'll document the procedure on EMCWiki so others can do the same
[00:32:17] <jepler> that's great, thanks
[00:32:55] <BigJohnT> your welcome... yummm I smell chicken cooking time to go upstairs
[00:34:21] <jepler> yeah dinner is cooking here too
[00:35:38] <dmess> i have a strawberry frut source bar and a box a beers.... ahhh life is good
[00:36:43] <dmess> lost 24 lbs in 3 weeks...
[00:40:20] <eric_U> jlmjvm: thanks
[00:52:11] <toastydeath> jepler re: earlier
[00:52:15] <toastydeath> yes, i would side with the modchip maker
[02:49:01] <jmkasunich> note to self: do not remove the 90 lb millhead until you have decided where you intend to put it down
[02:49:50] <SWPadnos> I take it you found a spot
[02:50:25] <jmkasunich> no, I
[02:50:33] <jmkasunich> I'm still holding it while I type
[02:50:44] <SWPadnos> heh
[02:51:21] <jmkasunich> moving a machine 2" is almost as much of a pain as moving it 4 feet
[02:51:27] <SWPadnos> yep
[02:51:45] <SWPadnos> the 2" can put it on rollers, so the other 3'8" (plus 2" at the end) is cake
[02:51:48] <jmkasunich> (when I moved it 4 feet last year, I didn't leave quite enough space for the door on my electrical box to open all the way
[02:52:08] <SWPadnos> crowbars to the rescue
[02:52:26] <jmkasunich> yeah
[02:52:37] <jmkasunich> crowbar is in the garage - freezing cold
[02:52:50] <jmkasunich> I guess thats what gloves are for
[02:52:52] <SWPadnos> gloves to the rescue
[02:52:54] <SWPadnos> yep
[02:53:12] <SWPadnos> of course, if the gloves are also in the garage, then it's heater to the rescue
[02:55:01] <jmkasunich> actually, since the dog is doing his "I need to go for a walk NOW" dance, the crowbar will have a chance to warm up...
[02:55:02] <jmkasunich> bbl
[02:56:12] <SWPadnos> see you
[03:04:14] <Guest454> Hey Guys, I have a question regarding a potential hardware issue and the latest live-cd install
[03:05:03] <SWPadnos> ok, shoot
[03:05:10] <SWPadnos> (not that I'll be able to help much)
[03:05:31] <Guest454> The install appears to complete fine, but after running EMC for a few hours, it hangs
[03:05:49] <SWPadnos> does the machine hang if you're not running EMC?
[03:06:00] <Guest454> If I just load Ubuntu, or run completely off of the live CD, the machine does not hang
[03:06:11] <Guest454> the whole machine hangs, caps lock key light not responsive
[03:06:26] <SWPadnos> no, I mean the EMC2 install - once you reboot, does the machine hang if you're not running EMC?
[03:06:33] <Guest454> Yes.
[03:06:46] <SWPadnos> ok, then it's not an EMC problem :)
[03:07:04] <SWPadnos> does a normal Ubuntu 6.06 install also hang?
[03:07:10] <Guest454> Nope
[03:07:18] <Guest454> very strange
[03:07:32] <SWPadnos> well, it could be something in the RT kernel
[03:07:47] <Guest454> I have tried cleaning all of the heatsinks on the board, and all of the fans are running
[03:08:19] <SWPadnos> hmmm. it could be heat related, but may not be. the RT kernel has no power saving options enabled
[03:08:41] <Guest454> I turned off ACPI in my BIOS for both installs
[03:09:01] <SWPadnos> I'm not sure the RT kernel will even idle the CPU
[03:09:10] <SWPadnos> when the system is inactive
[03:09:36] <Guest454> I would expect the Live CD to behave the same as the install, but it only hangs on full install
[03:09:47] <SWPadnos> with updates, or no updates>
[03:09:48] <SWPadnos> ?
[03:10:10] <Guest454> Just the full live image, no updates.
[03:10:20] <SWPadnos> I would also expect them to act the same, though the boot process is a bit different when booting from CD
[03:10:22] <SWPadnos> ok
[03:11:24] <SWPadnos> I don't know if ACPI has to be enabled to allow CPU speed scaling
[03:11:43] <SWPadnos> but again, the CD and HD boots should act more or less the same
[03:12:11] <Guest454> That is the really funny thing. I fully expected the live CD to be hung today when I got home, but it is happy here.
[03:12:27] <Guest454> Usually it hangs after a hour or two, this ran for 12+
[03:12:52] <Guest454> I ran the memtest and all tests passed
[03:13:13] <SWPadnos> for just a couple of hours though, right?
[03:13:38] <Guest454> I ran memtest for about 10 hours
[03:14:07] <Guest454> Also, this machine ran windows NT fine for years.
[03:14:40] <Guest454> AMD 1.1Ghz, 60Gig HD, CDR, worst case latency 12000ns
[03:14:50] <Guest454> vintage 2000 or so
[03:15:01] <SWPadnos> hey - that's about what I'm using right now
[03:15:04] <SWPadnos> for day-to-day work
[03:15:15] <eric_U> I've had a couple machines just like that keel over dead in the last year
[03:15:24] <SWPadnos> please don't say that
[03:15:26] <Guest454> hmm, what dies? HD?
[03:15:31] <SWPadnos> if this one dies, I'm in trouble
[03:15:38] <eric_U> mobo and processor
[03:15:45] <eric_U> possibly due to bad caps?
[03:16:01] <Guest454> I have heard that the electrolytics are usually the first things to go.
[03:16:21] <eric_U> my first suspect in your case would be the power supply
[03:16:48] <eric_U> but it does seem suspicious that the rt kernel is taking it down
[03:17:26] <eric_U> unless someone notices something about this story that contradicts my conclusion
[03:17:27] <SWPadnos> and only when booted from HD, not from the liveCD
[03:17:45] <eric_U> ok, first contradiction, live cd must have rt kernel too
[03:18:41] <Guest454> coincidence?
[03:18:46] <eric_U> Guest454: any motion control hardware attached?
[03:18:53] <Guest454> Not at the moment
[03:19:25] <eric_U> sudo cat /proc/interrupts show any conflicts?
[03:19:36] <SWPadnos> does it matter what time of day you start the machine?
[03:19:37] <eric_U> or rather, shared interrupts?
[03:19:52] <SWPadnos> ie, does it seem to die in the evening, morning, random ... ?
[03:19:55] <Guest454> let me check the interrupts
[03:20:14] <Guest454> I have had it die both in the morning (playing before work) and at night (when I get home.)
[03:20:20] <SWPadnos> ok
[03:20:28] <Guest454> I generally leave all day or all night to time test
[03:20:38] <eric_U> mine died because I made a crummy bracket for my mesa breakout boards
[03:20:44] <Guest454> sudo /proc/interrupts
[03:20:52] <eric_U> sudo cat
[03:21:24] <Guest454> yeah, sorry, checking
[03:22:10] <Guest454> how would a conflict be shown?
[03:22:21] <eric_U> two drivers on the same interrupt
[03:22:42] <Guest454> 0 : timer, rtai_jif_chk
[03:22:50] <eric_U> has to be there
[03:22:55] <Guest454> 9 : eth0, VIA8233
[03:23:09] <eric_U> that's the same thing I think
[03:23:37] <Guest454> ok, then I don' think there is a conflict
[03:23:58] <eric_U> longshot anyway
[03:24:01] <Guest454> latency test returns no problems (other than it too freezes after a few hours)
[03:24:17] <Guest454> hmm, power supply potentially?
[03:24:29] <eric_U> that's always my first suspicion
[03:24:57] <eric_U> you might wanna look for domed caps on your mobo
[03:25:01] <Guest454> Ok, I may give that a try. The machine appears to work great other than this freze problem.
[03:25:14] <eric_U> the top of the caps should be pretty flat
[03:25:32] <Guest454> Ok, the bulge probably means short?
[03:25:39] <Guest454> I can do that too.
[03:25:40] <eric_U> bulge is bad
[03:26:07] <SWPadnos> whmmm. just to be very clear, does the machine still freeze if you boot it up and never run EMC?
[03:26:16] <eric_U> good question
[03:26:17] <Guest454> Any ideas where else to get old machines for cheaps? Craig's list is letting me down. I am checking out intechra outlet.
[03:26:21] <SWPadnos> just boot, log in, and let it sit at the desktop
[03:26:40] <SWPadnos> that's what retrobox turned into?
[03:26:52] <Guest454> Yeah, it freezes if I install the EMC2 live-cd and just log in.
[03:26:59] <Guest454> Just Ubuntu 6.06 does not freeze
[03:27:12] <LawrenceG> turn off the screen saver.... there were some 3d screen savers that would crash the system
[03:27:15] <SWPadnos> ok. that's quite odd, since the RT kernel isn't too different from the normal kernel, when there are no RT apps running
[03:27:24] <SWPadnos> yep, that's a good idea as well
[03:27:25] <Guest454> retrobox redirects to intechra outlet
[03:27:27] <SWPadnos> ok
[03:27:31] <eric_U> it has acpi disabled
[03:27:39] <SWPadnos> that's a good thing
[03:27:45] <eric_U> maybe
[03:27:54] <Guest454> I disabled acpi in the bios for both installs
[03:27:55] <SWPadnos> absolutely, for an RT system
[03:28:16] <eric_U> of course for rt
[03:28:31] <Guest454> I have turned screen savers off on the freezing install. The non-emc install runs fine with and without screensavers
[03:29:21] <eric_U> seems like that has to be a kernel issue
[03:30:03] <SWPadnos> ok, install normal Ubuntu 6.06 and make sure it doesn't freeze. then install EMC2 on that install according to the wiki instructions
[03:30:11] <eric_U> too bad looking through dmesg is so tedious
[03:30:14] <SWPadnos> see if it freezes then
[03:30:36] <SWPadnos> if so, then boot from the old kernel (using the grub menu to select), and see if it freezes
[03:30:46] <Guest454> what is dmsg?
[03:30:54] <eric_U> the kernel logs
[03:30:58] <eric_U> boot logs
[03:31:03] <Guest454> how does one access?
[03:31:14] <eric_U> dmesg
[03:31:23] <eric_U> dmesg | more
[03:31:39] <SWPadnos> installing EMC2 onto a normal install will give you a single installation with multiple bootable kernels, which may be more useful for testing
[03:31:40] <Guest454> ok, I will poke through there while waiting for it to freeze.
[03:31:56] <eric_U> probably faster to reinstall ubuntu than read the whole output
[03:32:28] <Guest454> Ok. My EMC box is not currently hooked to the internet. Is there a way to install the EMC portion to a normal 6.06 install from the live-cd?
[03:32:38] <SWPadnos> not really
[03:32:58] <Guest454> ok, step one get a hub.
[03:33:00] <eric_U> cant you do some voodoo where the cd is one of your supositories?
[03:33:01] <SWPadnos> you can download all the debs from linuxcnc.org and stick them on a CD/USB stick
[03:33:08] <SWPadnos> not with the liveCD
[03:33:26] <SWPadnos> that's a live filesystemm that more or less gets copied to the HD when installed
[03:33:35] <SWPadnos> it doesn't install from debs on the CD
[03:33:47] <eric_U> makes sense
[03:34:17] <Guest454> would I have two partitions then?>
[03:34:33] <Guest454> Or would grub just figure out that I may want to boot with and without the latest add-on?
[03:34:38] <SWPadnos> well, sort of
[03:34:46] <eric_U> no, you choose your kernel on boot
[03:34:47] <eric_U> if you want
[03:34:50] <SWPadnos> the default install makes one data partition and one swap partition
[03:35:15] <SWPadnos> you would install Ubuntu to a partition, then install the EMC2 packages onto that system (same partition)
[03:35:41] <Guest454> Ok, and grub would let me boot either the original or original with packages?
[03:35:52] <SWPadnos> EMC2 will happen to install another kernel, but that's a detail (it's more or less the same as when the Ubuntu team puts out an updated kernel package)
[03:35:56] <SWPadnos> yes
[03:36:14] <SWPadnos> you may have to hit ESC or something at boot time
[03:36:27] <Guest454> Ok, I will give that a shot.
[03:36:39] <Guest454> Would you guess power supply first?
[03:36:51] <Guest454> Or guess kernel first?
[03:36:57] <SWPadnos> no, because the system runs fine from CD, which is the same kernel
[03:37:10] <SWPadnos> I don't think it's the kernel, beacuse the installed and CD kernels are the same
[03:37:45] <SWPadnos> it's possible that it's a video driver thing of some sort, but I'm not sure how the X settings get sorted out either for install or at liveCD boot time
[03:38:27] <Guest454> how does one figure out which video card and driver are running on the system?
[03:39:03] <Guest454> I am not seeing a splash screen I recognize and the physical card has no marks I recognize either.
[03:39:20] <eric_U> sudo lspci -v |more tells you what video card is running
[03:39:35] <SWPadnos> I'd look at /etc/X11/xorg.conf (I think)
[03:40:17] <eric_U> xorg.conf tells you what driver you are running in X
[03:40:19] <Guest454> via technologies?
[03:40:26] <eric_U> where did you see that?
[03:40:31] <eric_U> in lspci?
[03:40:36] <Guest454> yeah
[03:40:41] <SWPadnos> is that the line that says "VGA graphics controller" or similar?
[03:40:45] <eric_U> via makes everything
[03:41:16] <Guest454> oh, no. My bad. It is an nVidia GeForce2 Ti.
[03:41:29] <Guest454> I have read that the nVidia cards are evil
[03:41:40] <eric_U> not so bad
[03:41:46] <eric_U> but not good for rt
[03:41:48] <SWPadnos> that old one may not be, and it's the accelerated driver that's generally the problem
[03:41:51] <eric_U> run the vesa driver
[03:42:06] <eric_U> look in /etc/X11/xorg.conf and see what driver
[03:42:18] <eric_U> if it's nv, change it to vesa
[03:42:47] <eric_U> probably wouldn't cause freezeups though
[03:42:53] <eric_U> rt delays, yes
[03:43:05] <Guest454> driver is "nv"
[03:43:18] <Guest454> q
[03:43:41] <eric_U> you could try changing it
[03:44:03] <Guest454> I am working on that now
[03:48:47] <Guest454> ok, thanks for the linux and EMC info. I will tinker with the ideas we have discussed.
[03:49:15] <Guest454> (and silently hope the vesa driver change solves it.)
[03:49:31] <SWPadnos> good luck with it
[03:51:10] <eric_U> it seems like I should be able to buy a decent computer for rt brand new
[03:51:15] <eric_U> not sure what though
[03:51:28] <eric_U> ddr2 memory is nearly free right now
[03:51:41] <SWPadnos> unfortunately, inexpensive PCs are usually not good for RT
[03:51:53] <SWPadnos> integrated video, no parport, no PCI clots, etc
[03:52:00] <SWPadnos> slots
[03:52:06] <eric_U> I'm not talking dell
[03:52:29] <SWPadnos> no, it's a general problem. inexpensive motherboards often don't have parports and/or PCI slots
[03:52:43] <eric_U> expensive mobos are not that much money
[03:52:46] <SWPadnos> not that they can't be found, but they're getting less and less common
[03:52:51] <SWPadnos> true enough ;)
[03:53:19] <SWPadnos> I just bought two of the google OS devkits, but my first tests look like it's not good for EMC/ET
[03:53:21] <SWPadnos> RT
[03:53:34] <SWPadnos> which is too bad, the MB is $60 including a 1.5GHz eden CPU
[03:53:35] <eric_U> what is the board?
[03:53:43] <eric_U> it's probably the eden
[03:54:14] <eric_U> didn't hear about that deal, is it open to all comers?
[03:54:27] <SWPadnos> I think so. I'm looking for the link now
[03:55:30] <SWPadnos> article: http://www.linuxdevices.com/news/NS5305482907.html
[03:55:55] <SWPadnos> purchase: http://www.clubit.com/product_detail.cfm?itemno=A4842001
[03:57:31] <eric_U> what is it with companies that have tiny pictures?
[03:57:37] <eric_U> bits becoming more expensive?
[03:57:42] <SWPadnos> myopic ad people
[04:03:56] <eric_U> newegg had a mobo for $7 after rebate
[04:04:06] <eric_U> don't think I want it
[04:04:29] <SWPadnos> heh
[04:04:52] <SWPadnos> I was just noticing the Phenom 2.3GHz quad-core CPU at $270
[04:05:14] <eric_U> I was moderately tempted, but it's a power hog
[04:05:38] <eric_U> intel outperforms
[04:05:49] <eric_U> at least that's my understanding
[04:05:51] <SWPadnos> 95W TDP
[04:06:13] <SWPadnos> the Intel quads are 65W I think, plus the memory controller
[04:06:52] <eric_U> so it's 33% more heat escaping from that poor little overheated chip
[04:07:06] <SWPadnos> well, yes, but systemwide, it's much closer to parity
[04:07:16] <eric_U> I'm gonna assert that's a bad thing
[04:07:33] <SWPadnos> if you put a fan/heatsink on it, it's just fine
[04:07:39] <SWPadnos> same as the Intels
[04:07:47] <Guest454> Has anyone tried the $200 linux box from walmart for EMC?
[04:08:00] <eric_U> dont' do it
[04:08:00] <SWPadnos> not that I've heard
[04:08:04] <eric_U> but no
[04:08:14] <Guest454> its a trap?
[04:08:22] <eric_U> I'm guessing it has the video on the mobo, that's always bad
[04:08:37] <Guest454> good to know
[04:09:02] <Guest454> Does conventional wisdom reccomend any particular new machine or vendor?
[04:09:19] <eric_U> we were talking about that
[04:09:27] <eric_U> solution is not obvious
[04:09:47] <SWPadnos> oh wait, the Intel quads are also 95W, plus the memory controller
[04:10:08] <eric_U> that's because they compute faster, bit for bit
[04:11:07] <eric_U> dual or quad core is the way to go for rt
[04:11:13] <eric_U> you can dedicate a core to rt
[04:11:39] <eric_U> or at least that's what Paulo said today
[04:11:47] <cradek> if you can get the kernel and rtai built right...
[04:11:55] <SWPadnos> the Intel dual-cores have the strange property of having exceedingly good RT timing when you have a do-nothing task on the other core
[04:13:07] <eric_U> interesting
[04:13:15] <eric_U> I haven't been following the rtai mailing list
[04:13:30] <eric_U> they were building a database of known good systems
[04:13:32] <SWPadnos> that's personal experience, not RT mailing list
[04:14:02] <eric_U> was the rt task small enough to fit in the cache?
[04:14:19] <SWPadnos> possibly
[04:14:37] <SWPadnos> well, probably, I'm sure it wasn't 2M or whatever the L3 cache is
[04:14:43] <eric_U> the amd have been better about cache usage recently, not sure about the current situation
[04:15:34] <SWPadnos> it seems strange that this script on core 1 helps RT on core 2: `while true ; do echo "nothing" > /dev/null ; done`
[04:15:55] <SWPadnos> that is the exact script I use to improve RT latency
[04:16:11] <eric_U> so you built a kernel that dedicated one core to rt?
[04:16:24] <SWPadnos> boot with "isolcpus=1"
[04:16:24] <eric_U> or this was regular smp
[04:16:39] <SWPadnos> it's an RT kernel with the isolcpus boot time option
[04:16:49] <SWPadnos> RT+SMP
[04:16:52] <eric_U> ok
[04:17:04] <SWPadnos> the one chris built, which is in the /experimental dir on linuxcnc.org
[04:17:20] <eric_U> dell had a dual quad core for sale a while back
[04:17:43] <eric_U> if they had a full video slot I would have bought it
[04:18:24] <eric_U> I bought dual quad core for my boss, 8 GB ram
[04:18:40] <SWPadnos> I'll probably get something similar for my mother
[04:18:52] <eric_U> mom should get 16 Gb
[04:19:03] <eric_U> don't hold back
[04:19:18] <SWPadnos> she'd probably prefer that, but sinec I'm buying the computer, she's probably only getting 8GB
[04:19:49] <eric_U> does anything really use that much ram now?
[04:20:10] <SWPadnos> programmers can write programs that do
[04:20:19] <eric_U> yabut they don't
[04:20:23] <SWPadnos> for things like training 100-neuron neural nets
[04:20:26] <SWPadnos> my mother does
[04:20:40] <SWPadnos> or would if she had something better tahn a P-133 with 64M
[04:20:46] <cradek> oh blame it on me why don't you
[04:20:51] <eric_U> that's cool, I have problems that are definitely memory bound
[04:21:27] <eric_U> I would like to have 32GB or so
[04:21:27] <cradek> actually I'm using that one too, but not with isolcpus. I need them both for compiling.
[04:21:33] <cradek> "need"
[04:21:53] <SWPadnos> when you expect to need 100-000 test points, each of which is a 35-element vector for each of 100 neurons, and you need to run the test data through the network several thousand times, you get to need more memory than even a video game
[04:22:13] <SWPadnos> err - 100,000 test points
[04:22:49] <eric_U> I've trained neural nets
[04:22:56] <eric_U> not my favorite technique
[04:23:03] <SWPadnos> (even that only comes out to about 6GB, so 8G should be enough) :)
[04:23:11] <cradek> I really need a quad quad at work, but I can't talk my boss out of the cash
[04:23:22] <SWPadnos> it's only $8000 or so
[04:23:26] <eric_U> bastids
[04:23:43] <eric_U> my boss's computer was 8-10k, forget exactly
[04:23:50] <eric_U> 2 24 inch monitors
[04:23:57] <eric_U> I shoulda got him 2 30"
[04:24:11] <SWPadnos> heh. exactly what my mother wants (the dual 24's)
[04:24:21] <eric_U> it takes a while to get used to it
[04:24:26] <SWPadnos> dual 30's are just too much area to look at
[04:24:43] <eric_U> when I was using it at first, I'd lose things
[04:25:02] <eric_U> the computer I use most has 24" and 42"
[04:25:22] <eric_U> do not surf porn on the 42"!!!!!!
[04:25:37] <SWPadnos> I'm all for lots of pixels, but the physical arrangement of dual 30's just doesn't seem all that good for me
[04:25:47] <eric_U> true
[04:25:50] <SWPadnos> I'd probably get used to it, but I don't like it from what I've seen
[04:25:58] <SWPadnos> (and I've seen a couple of setups like that)
[04:26:00] <eric_U> I like to doc things over there
[04:26:04] <eric_U> dock
[04:26:25] <SWPadnos> sure, if there's a lot of stuff you might look at from time to time, even manuals or the like, then it's useful
[04:26:45] <eric_U> I usually have 20 windows open at least
[04:27:04] <SWPadnos> but if you have 37 windows open which you're actively using (like stupid LabView, or a lot of design files), then it's hard to move your head that far all the time
[04:27:08] <eric_U> I hate having to sort through figures from matlab
[04:27:13] <SWPadnos> heh
[04:27:24] <SWPadnos> try labview some day
[04:27:30] <eric_U> I don't like it
[04:27:39] <SWPadnos> me either, but it does like to open many manhy windows
[04:28:17] <SWPadnos> in fact, twice as many as any sane programming environment, since you have the front panel and diagram for every friggin module
[04:28:27] <eric_U> most arent' useful
[04:28:32] <eric_U> as I recall
[04:28:52] <eric_U> I was trying to upgrade to the most recent labview
[04:29:00] <SWPadnos> I had a stupid app that I was foolishly required to do in LV, which had something like 10 screens the user had to interact with
[04:29:08] <SWPadnos> plus some dialogs
[04:29:17] <eric_U> we can get it for free, but have to download all 35 1/2 disks
[04:29:24] <SWPadnos> each of which is, of course, two "source" windows
[04:29:26] <SWPadnos> heh
[04:29:39] <SWPadnos> that's CDs, not floppies, right?
[04:29:45] <eric_U> can't you compile it to look like a program?
[04:29:49] <eric_U> yeah, that's cds
[04:30:01] <SWPadnos> oh sure, it looks like a program once it's compiled. even has an installer
[04:30:15] <eric_U> ok, I've never gotten that far
[04:30:27] <SWPadnos> but writing it means that each of the user windows has two source modules, plus backend stuff, etc
[04:30:27] <eric_U> some really simple stuff is a real pain
[04:30:31] <SWPadnos> yes
[04:30:41] <eric_U> like sending a packet over tcpip
[04:30:44] <SWPadnos> like most GUI apps, anything they thought of is trivial, anything else is impossible
[04:31:19] <eric_U> the c++ stuff is nice
[04:31:26] <SWPadnos> inLV?
[04:31:32] <eric_U> measurement studio
[04:31:43] <eric_U> uses LV somehow
[04:31:48] <SWPadnos> ok, that stuff may be OK. I've never had the "pleaseure" of using it
[04:32:01] <eric_U> that's the nice thing about educational licenses
[04:32:05] <SWPadnos> I didn't think measurement studio used C++
[04:32:16] <eric_U> I'm pretty sure what they call it
[04:32:25] <SWPadnos> I thought it was a "simple" way to log data from various instruments
[04:32:44] <SWPadnos> there's another app called TestStand, which LV may actually be suited for
[04:33:00] <SWPadnos> measurement studio is also in its realm of usability
[04:33:22] <eric_U> I'm pretty sure it's a takeoff on visual studio
[04:33:31] <SWPadnos> nope
[04:33:36] <SWPadnos> LV came way before VS
[04:33:54] <eric_U> don't see how that is important
[04:34:11] <SWPadnos> well, if one was written before the other, then it's not a takeoff of the other, right?
[04:34:22] <eric_U> talkin' about the name
[04:34:28] <eric_U> measurement studio
[04:34:35] <SWPadnos> oh, hmmm. I don't know - could be
[04:35:09] <eric_U> National Instruments Measurement Studio, winner of E&E's 2006 Best Product Award for Design Tools and Software, is an integrated suite of classes and controls for test, measurement, and automation applications in Microsoft Visual Studio 2005,
[04:35:22] <SWPadnos> actually, it looks like measurement studio is meant to plug into visual studio for measurement apps
[04:35:26] <SWPadnos> right :)
[04:35:45] <eric_U> yes, and labview connection may mostly be in my head
[04:35:52] <SWPadnos> well, it is by NI
[04:36:11] <eric_U> but if you read their documentation for their c++ stuff, it's the same as the labview docs
[04:42:08] <SWPadnos> man. you know the world is crazy when a mouse costs more than 2GB or memory
[04:42:14] <SWPadnos> of
[04:42:37] <eric_U> pretty good mouse for $10
[04:43:17] <SWPadnos> Logitech G5 laser mouse, $52.99. Corsair XMS2 2x1GB DDR2 DIMMs, $47 after rebate
[04:43:43] <eric_U> takes more sand and plastic to build the mouse
[04:44:06] <SWPadnos> more plastic for sure, but it still seems wrong
[04:46:44] <SWPadnos> ok, bedtime for me. good night
[05:20:19] <christel> [Global Notice] Hi all! One of our sponsors experienced a rather large ddos attack just now, which in turn caused disturbance on the network. As a result that leaf server is not accepting client connections atm. Apologies for the inconvenience and have a good day.
[05:24:07] <renesis> omg
[05:24:13] <renesis> axis only displays g54?
[05:25:13] <renesis> nm
[05:25:51] <renesis> changing zero coordinate system doesnt change the machine g5x mode
[05:26:17] <renesis> oh and theres should totally be a confirmation on the home axis button
[05:26:29] <renesis> accidental click = many sadfaces
[05:38:46] <renesis> see there i just did it
[05:38:54] <renesis> i hit home instead of zero
[05:38:58] <renesis> fuck
[05:39:19] <renesis> like, zero actually has some sort of settings/confirm dialog
[05:39:42] <renesis> home is silently like 'haha youre fucked now!'
[05:39:54] <renesis> and why doesnt that overide limits thing work?
[05:39:57] <JymmmEMC> ESC for the win!
[05:40:07] <JymmmEMC> hint BIG FAT RED BUTTON
[05:40:13] <JymmmEMC> push it!!!!!!!!!!!!!!!!!
[05:40:14] <JymmmEMC> lol
[05:40:50] <renesis> yeah and then?
[05:41:10] <renesis> machine estop doesnt do anything
[05:41:41] <renesis> it rehomes my machine axis and doesnt let me travel past the soft limit so i can reset it
[05:42:23] <renesis> it doesnt have home/limit switches, so home just sets machine zero
[05:42:58] <renesis> which resets all your coordinates
[05:43:05] <JymmmEMC> yeah that sounds right, dont recall how to override those though.
[05:43:08] <renesis> so basically all my setup is trash
[05:43:17] <renesis> i have to restart emc
[05:43:31] <JymmmEMC> in the config, those are to be the HARD limits iirc.
[05:43:40] <renesis> also im not sure changing workset coordinate system in the axis menu actually does anything
[05:44:13] <renesis> well they are hard, but i accidentally hit the home button so now hard limit is like 8" short of actual
[05:44:16] <JymmmEMC> you're talking to the wrong peron on those... try cradek jepler or jmkasunich
[05:44:22] <renesis> and home doesnt have any kind of confirm dialog
[05:44:47] <JymmmEMC> there's a way to override, I just dont what it is.
[05:45:04] <renesis> theres a checkbox, which unchecks itself, basically does nothing
[05:45:09] <renesis> ill google, ty
[05:47:49] <renesis> should i just not use soft limits?
[05:48:23] <renesis> cuz seriously, having an overide limits button that doesnt do anything is really, really frustrating
[05:49:23] <renesis> this is the most reccurring annoying thing about emc since ive started using it =(
[05:49:50] <JymmmEMC> just set your har limits outside what they really are for now.
[05:49:56] <JymmmEMC> hard
[05:54:17] <renesis> meh, ill just do it over
[05:55:48] <renesis> or how do you unhome an axis
[05:57:17] <renesis> oh look i am first hit on google fof unhome complaining about fucking up like this before
[06:11:59] <fenn> override limit only lets you jog in one direction so you can get off the limit switch
[06:12:30] <renesis> okay
[06:13:15] <renesis> like, axis coordinate system somehow got more confusing instead of less
[06:13:31] <fenn> i think that's just what coordinate system is displayed on the DRO
[06:14:48] <renesis> yeah but it doesnt match the zero button
[06:14:54] <renesis> or at least im not sure
[06:15:07] <renesis> really i thought i figured it out, but i ended up feeling stupid
[06:15:24] <renesis> im just gonna mdi it, heh
[06:16:17] <fenn> what's the 'zero button'?
[06:17:13] <renesis> 'touch off' button
[06:17:25] <renesis> g92 button
[06:45:36] <renesis> okay now something weird just happened
[06:46:00] <renesis> i did g55, then g92 x.1, then did the same thing for g54 and g65
[06:46:12] <renesis> and now my g55 x0 is in a totally diff place
[06:47:38] <renesis> and the DRO is set by the current coordinate system
[06:47:55] <renesis> so i guess the menu sets the touch off system?
[06:48:03] <renesis> but thats also set by a drop down in that dialog
[07:18:52] <fenn> am i wrong in thinking there were emc2 simulator packages?
[07:20:11] <fenn> aha! http://linuxcnc.org/emc2/dists/dapper/emc2.2-sim/binary-i386/emc2-sim_2.2.2_i386.deb
[07:20:38] <fenn> now why doesnt apt see that
[07:33:34] <alex_joni> fenn: do you have emc2.2-sim in /etc/apt/sources.list ?
[07:38:00] <fenn> no
[07:38:13] <fenn> are there instructions somewhere that say to do that?
[07:39:28] <fenn> for some reason it doesnt like the new libpth20 2.0.7-8 (debian lenny) so i downgraded to 2.0.7-6 (etch)
[07:47:50] <fenn> seems to be some 9-axis stuff going wrong, maybe it's my fault
[07:48:35] <fenn> ah maybe its junk left in /usr/local/lib/
[07:48:50] <fenn> is there a "make uninstall"? :)
[07:50:23] <alex_joni> fenn: not unless you submit a patch
[08:45:06] <fenn> aha - /usr/lib/python2.4/site-packages/ has old .py files (and new .pyc files from the package)
[09:12:11] <JymmmEMC> alex_joni: My new phone http://www.phonescoop.com/phones/phone.php?p=1008
[09:19:57] <renesis> http://www.darkertechnologies.com/image/vfdclk_psu_top_milled2.jpg
[09:20:07] <renesis> vfdclk_psu_top_milled1.jpg
[09:20:16] <renesis> vfdclk_psu_board_top.png
[09:20:23] <renesis> cnc machines are so sweet
[10:16:51] <fenn> alex_joni: i dont know how to use GNU make very well, so this is what i came up with: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Installing_EMC2#Uninstalling_EMC2_from_source
[10:32:40] <fenn> why is emc2.2-sim a separate repository?
[12:14:36] <alex_joni> fenn: I would do this:
[12:14:47] <alex_joni> export DESTDIR=/tmp/foo
[12:14:50] <alex_joni> make install
[12:15:05] <alex_joni> then you'll get them all installed as if /tmp/foo would be /
[12:32:15] <fenn> d'oh shoulda used checkinstall
[12:33:24] <fenn> um, it didnt work
[12:33:42] <fenn> it just installed everything in the usual places
[12:34:51] <alex_joni> maybe export DESTDIR=/... & make install ?
[12:35:06] <alex_joni> it's been a while since I played with this stuff
[12:36:43] <fenn> it makes the directories usr/local/bin usr/local/etc usr/realtime..magma in DESTDIR
[13:13:48] <fenn> possible bug if anyone's interested: run the scara config, switch to teleop mode, set angular jog speed to some large value
[13:14:51] <fenn> everything's ok until you set linear jog speed to a large value, and then it starts acting strange.. positive jogs in C dont act the same as negative jogs
[13:16:38] <fenn> its like positive jogs start with low acceleration and stop with infinite accel, and negative jogs start with infinite accel and stop with low accel
[13:18:44] <fenn> just get a following error in joint mode
[13:26:48] <alex_joni> fenn: sorry.. you need to make DESTDIR=/path/foo/ (with a trailing /)
[13:31:29] <fenn> export DESTDIR=~/sandbox/emc-install/ && make install
[13:31:32] <fenn> is what i did
[13:32:11] <fenn> oh, and then it whined about permissions, so like a fool i did && sudo make install
[13:32:26] <cradek> fenn: teleop is all screwy and I don't think anyone has looked into it yet
[13:33:14] <fenn> i'll add it to the pile then :P
[13:33:44] <cradek> also there's no support for uvw
[13:34:08] <cradek> jmk/alex/I talked briefly about using the TP instead of all that code for teleop
[13:34:35] <cradek> a real limitation of the current teleop IMO is no incremental moves
[13:35:19] <fenn> well, it was for moving cranes i guess
[13:35:35] <cradek> assuming all that is 'easy' that still leaves jogwheel as the hard problem
[13:36:14] <fenn> i think you'd want to be able to feed it a stream of 6d vectors from a joystick or something
[13:36:41] <fenn> jogwheel is useful too
[13:38:01] <cradek> honestly I'm less concerned about teleoping in any direction. I'd trade it for something just like trivkins jogging (along axes, continuous, incremental, wheel)
[13:39:15] <cradek> but trivkins machines can jog along more than one axis simultaneously which leaves us with pretty much the same thing (like with a joystick)
[13:39:26] <cradek> so I guess I don't know what I want or how to do it.
[13:39:54] <cradek> but I do need to go to work. bbl.
[13:57:14] <jepler> fenn: the main reason that emc2.2 and emc2.2-sim are in separate repositories is that the -dev package has the same name for both. maybe this was what they call a "bad move".
[13:59:29] <alex_joni> maybe it should be emc2.2-sim-dev
[14:01:53] <fenn> in any case, there should probably be a script to install the sim packages
[15:01:23] <Guest916> Guest916 is now known as skunkworks__
[15:15:22] <seb__> seb__ is now known as sebjames
[16:10:45] <steve_stallings> steve_stallings is now known as steves_logging
[16:57:10] <hmtr> Before I get too far into this project (my first) and find out I'm in over my head, is this configuration possible? 5 analog axis + analog spindle, MPG and lots of I/O?
[16:57:28] <cradek> of course
[16:57:32] <alex_joni> hmtr: how much are you willing to spend?
[16:57:42] <cradek> you can add interfaces until you run out of slots
[16:57:42] <alex_joni> because virtually anything is possible.. :)
[16:57:46] <hmtr> money is really not problem
[16:57:57] <alex_joni> cradek: even then.. bigger mobo's, other links, etc
[16:57:59] <hmtr> It's either EMC or a Siemens 840d
[16:57:59] <cradek> mesa cards have lots and lots of IO. you could use several of them.
[16:58:07] <alex_joni> I think 4 is the max
[16:58:24] <cradek> oh ... still that's quite a few.
[16:58:49] <hmtr> SIemens 840di is going to be around $20k
[16:59:03] <hmtr> so is mesa the way to go?
[17:00:35] <cradek> don't think we have enough information to say. mesa is very popular and well supported.
[17:00:59] <alex_joni> the mesa is digital though, and the outputs are converted to analog
[17:01:05] <cradek> right, and mesa does not have ADCs
[17:02:25] <alex_joni> there used to be another board called Motenc-100
[17:02:32] <alex_joni> http://www.vitalsystem.com/web/motion/motion100.php
[17:02:39] <hmtr> So using MESA doesn't allow for analog feedrate pots?
[17:02:56] <alex_joni> they say it's "only available for high volume oem orders"
[17:03:01] <cradek> with emc you would use an encoder/mpg, not pot, for feed rate override
[17:03:06] <alex_joni> maybe 2 x Motenc Lite would be ok
[17:03:25] <cradek> that's because you can also change feed override in the gui, and it allows them not to fight
[17:04:13] <alex_joni> http://www.vitalsystem.com/web/motion/motionLite.php
[17:04:23] <alex_joni> do you need an encoder on the spindle?
[17:04:31] <hmtr> no encoder on the spindle
[17:04:44] <cradek> no need for rigid tapping?
[17:04:46] <alex_joni> so no rigid tapping?
[17:04:49] <cradek> or spindle orient?
[17:04:57] <alex_joni> then maybe you can get away with only one Motenc Lite
[17:05:06] <alex_joni> 4 encoder counters, 8 DACs, 8 ADCs
[17:05:16] <alex_joni> 48 IOs
[17:05:16] <jepler> I thought the machine was to be 5 axis
[17:05:34] <alex_joni> right.. darn :)
[17:05:38] <hmtr> 5 programmable axes + 1 open loop spindle
[17:05:48] <hmtr> 20v 5 axis Cincinnati to be axact
[17:06:08] <alex_joni> what kind of analog input do you need?
[17:06:13] <cradek> that's very ambitious for a first emc project
[17:06:21] <hmtr> I thought about trying the EMC first, and if crash and burn, i'll hang a siemens on it :)
[17:07:01] <hmtr> I thought using analog feedrate pots would be less hassle than BCD switches
[17:07:16] <cradek> not bcd. quadrature.
[17:07:50] <hmtr> using encoders for feedrate control is something I've not seen before
[17:08:07] <hmtr> gives me something else to research
[17:08:22] <cradek> it allows you to use the gui and also have a feedrate control (or more than one)
[17:08:34] <cradek> imagine one knob on the panel and another on a pendant. if they are pots this is impossible.
[17:09:22] <hmtr> so if you use an encoder, does it feed into an actual encoder channel or use it as I/O for the feedrate override?
[17:10:08] <cradek> you can "count" encoder signals in several ways, with dedicated hardware like a mesa/motenc encoder channel, or in software using two general purpose IO lines
[17:10:33] <cradek> if it is slow (like a knob you turn) IO lines are generally OK. of course you need dedicated hardware to count encoder pulses on a fast machine axis.
[17:10:38] <alex_joni> for manual turned feed-override thingies, it's enough to use software counting
[17:11:47] <jepler> "slow" being a few hundred pulses per second or less
[17:12:11] <alex_joni> jepler: I think you can read mesa IO faster than that
[17:12:14] <jepler> oh can you?
[17:12:21] <cradek> depends how fast you read your IO. on many cards like mesa/motenc it is read only at the servo cycle
[17:12:47] <cradek> I bet you could have one fast mesa card (base period) and one slow (servo cycle)
[17:12:48] <alex_joni> hmm.. I would have thought it's faster than a parport
[17:12:58] <jepler> yes I bet you can too
[17:13:30] <jepler> you could probably even modify the driver to read/write some IO at the fast rate and the rest at the slow rate
[17:13:31] <hmtr> ack! gotta take a customer to lunch! Thanks for your tiime!!!! I'll ask some more questions a little later !!!
[17:13:40] <alex_joni> see you
[17:13:51] <jepler> come back anytime
[17:14:38] <jlmjvm> what would be required to male emc2 work with a galil
[17:14:46] <skunkworks__> heh
[17:15:04] <skunkworks__> galil had ist own motion controller internally - correct?
[17:15:06] <alex_joni> jlmjvm: talking people or money?
[17:15:30] <jlmjvm> both possibly
[17:16:01] <cradek> sounds like a bad match for emc then
[17:16:22] <jlmjvm> how so?
[17:16:33] <cradek> 'has its own motion controller'
[17:16:40] <jlmjvm> isnt it basically like a mesa
[17:16:41] <alex_joni> jlmjvm: you don't have much left then
[17:17:03] <alex_joni> jlmjvm: it's basicly: mesa+driver+motion controller+sequencing logic
[17:17:11] <alex_joni> you only keep the GUI & interpreter
[17:17:20] <alex_joni> and a whole bunch of problems sorting the mess up
[17:17:59] <jlmjvm> i better just stay with parport stuff for now
[17:18:06] <anonimasu> why not a mesa?
[17:19:54] <jlmjvm> may look into that,but it would probably be a bigger job than i can do
[17:20:29] <anonimasu> eh?
[17:20:30] <anonimasu> why
[17:20:34] <anonimasu> or a motenc..
[17:20:45] <anonimasu> i'd bet you it's less work then with the galil stuff
[17:21:03] <jlmjvm> it has a tool changer,never done one before
[17:21:16] <anonimasu> well, it's not like you can get away from that
[17:21:31] <jepler> starting by buying hardware that isn't now supported by emc won't make it easier to make emc work with your toolchanger
[17:21:50] <anonimasu> jlmjvm: you are making up stupid excuses.
[17:21:59] <anonimasu> jlmjvm: lots of people run emc with motenc's..
[17:22:01] <anonimasu> and it works.
[17:22:09] <jlmjvm> it already has the galil in the machine
[17:22:41] <anonimasu> does it perform?
[17:22:51] <jlmjvm> thats why i asked about that
[17:23:00] <anonimasu> well, does it?
[17:23:18] <jlmjvm> yes it will actually run,but its mitsubishi conversational
[17:23:26] <jepler> galil is unsupported by emc, and probably the galil control model is not terribly well-suited to interfacing with emc.
[17:23:42] <anonimasu> well, if it makes parts and does so well, keep it :)
[17:23:56] <jepler> you'll spend less money and time throwing out the galil part (or selling it if they sell on ebay) and buying a motion control card that already works with emc
[17:24:10] <anonimasu> if you dont feel like replacing it.
[17:24:20] <jlmjvm> nobody knows how to program it,they want a machine that will run a g code program
[17:24:21] <anonimasu> you wont be making any parts with the galil without considerable effort
[17:25:21] <anonimasu> hm, maybe you should sell it on ebay if money is a issue
[17:25:42] <anonimasu> the toolchanging is something you need to address if you plan on switching to emc..
[17:25:50] <anonimasu> either way.. with or without galil
[18:26:02] <skunkworks__> but wait - galil works with mach... ;) (so I am told on many occations)
[18:26:45] <jlmjvm> they dont have the bugs worked out yet
[18:28:05] <jlmjvm> camsoft claims they have a program that will run it and the tool changer also,but i find that hard to believe
[18:28:14] <skunkworks__> heh
[18:29:57] <skunkworks__> mach has a few external controllers that do exactly this.. (pretty much send the program to the external device and it does the motion). So it should be a better fit. I myself Like the idea of the cheap computer being the motion controller
[18:32:41] <SWPadnos> hmtr, I designed a board that has 8 analog outputs and 6 analog inputs
[18:32:47] <SWPadnos> which connects to the Mesa card
[18:33:02] <SWPadnos> the analog I/Os are 16-bit resolution
[18:33:23] <SWPadnos> and there's a HAL driver (though I'd need to fiddle with it a bit before general release)
[18:44:54] <skunkworks__> SWPadnos: are you going to start making runs?
[18:45:35] <SWPadnos> I hope not. a cold is bad enough
[18:45:57] <SWPadnos> err - I have a few in stock, and I can make more in a reasonable hurry
[19:27:02] <jlmjvm> http://imagebin.org/12256
[19:27:40] <jlmjvm> http://imagebin.org/12257
[19:28:35] <jlmjvm> skunkworks_:heres a couple of pics of the mill i was talking about
[19:29:52] <anonimasu> jlmjvm: I'd liberate it anytime
[19:29:53] <anonimasu> .
[19:29:54] <anonimasu> ^_^
[19:30:31] <cradek> nice but what are those strange things sitting on the table?
[19:30:48] <anonimasu> arent they covers?
[19:30:58] <anonimasu> for the ways..
[19:30:59] <cradek> (I meant the manuals)
[19:31:06] <jlmjvm> both
[19:31:07] <anonimasu> OH
[19:31:07] <anonimasu> :P
[19:31:17] <anonimasu> thoose are defenetively covers.
[19:31:39] <cradek> 12x24 travel or so? looks pretty small for such a machine
[19:31:55] <cradek> a nice size for the garage
[19:32:01] <anonimasu> yep
[19:32:24] <anonimasu> I cant stop watching theese fraisa movies
[19:32:29] <cradek> does it have a quill or does just the head move?
[19:32:31] <jlmjvm> its got brushless ac servos
[19:32:33] <jlmjvm> head
[19:32:41] <cradek> that's great to have so much Z travel available
[19:32:56] <anonimasu> yep
[19:33:11] <jlmjvm> pics dont do it justice,its a nice mill
[19:33:12] <cradek> although there's only 11? inches under the control
[19:33:14] <archivist> * archivist dribbles
[19:33:20] <cradek> yeah looks nice
[19:33:20] <anonimasu> :/
[19:33:26] <anonimasu> I'd like more Z travel then that
[19:33:37] <anonimasu> http://www.youtube.com/watch?v=uAcBU2iqkKE
[19:34:52] <toastydeath> that's actually pretty common on good hardware
[19:34:59] <toastydeath> looks like it should have more travel than it does
[19:36:53] <jlmjvm> did ya notice the desktop computer with no cover and the crt moitor on the shelf,factory setup
[19:37:19] <toastydeath> i'm just talking about the machine, not the control
[19:37:23] <toastydeath> but yes i noticed
[19:37:57] <dmess> hi all
[19:38:16] <toastydeath> hai
[19:38:31] <anonimasu> can any of yoy-_-
[19:38:33] <anonimasu> err
[19:39:04] <toastydeath> lol epic
[19:39:24] <anonimasu> ?
[19:39:31] <anonimasu> my biggest typo today
[19:39:31] <anonimasu> :)
[19:39:34] <toastydeath> yeah
[19:40:04] <dmess> braincramp??
[19:40:08] <anonimasu> http://www.youtube.com/watch?v=ZPqJTPvvFMI&feature=related
[19:40:09] <anonimasu> yeah
[19:40:53] <jlmjvm> cradek:do you think that mill would be a good mesa candidate
[19:42:24] <anonimasu> nice mill
[19:43:39] <jlmjvm> wish it was in my garage
[19:44:06] <dmess> he's doing well with it..
[19:44:13] <anonimasu> dmess: the one on the video
[19:44:13] <anonimasu> yeah
[19:44:21] <dmess> yes..
[19:44:30] <anonimasu> im seriously doubting how much spindle power you need
[19:44:31] <dmess> nice chip coming off
[19:44:45] <anonimasu> 1.5kw is sufficient for most stuff..
[19:45:17] <anonimasu> and a healthy 8krpm..
[19:45:28] <dmess> for aluminum its minimal if you want to tickle at warp stupid... STEEL needs a little more but not really that much
[19:45:50] <anonimasu> the mill at work has lots of spindle power, thouh the cutters cant handle it..
[19:45:53] <anonimasu> not smaller ones
[19:46:12] <toastydeath> facemilling steel is the big consumer of hp
[19:46:15] <anonimasu> 1.2-1,5kw seems like the norml..
[19:46:15] <anonimasu> yeah
[19:46:21] <anonimasu> and im a wuzz..
[19:46:22] <dmess> i ran a 50mm face mill with 4 inserts at 4mm feed per tooth and 500 rpm
[19:46:26] <anonimasu> I take too small cuts..
[19:46:35] <anonimasu> dmess: wtf!"#
[19:46:40] <anonimasu> dmess: that scares me
[19:46:49] <dmess> at a trade show no less...
[19:46:52] <toastydeath> lol
[19:46:53] <anonimasu> ah.
[19:46:59] <anonimasu> what do you run for production?
[19:47:00] <archivist> bluse chips
[19:47:03] <archivist> blue
[19:47:10] <anonimasu> and machine life
[19:47:20] <toastydeath> the machines can handle that
[19:47:25] <toastydeath> it's not a problem for them
[19:47:33] <dmess> the mazak beside me blew up trying to keep up in chip production ge had on an 8 : f/c
[19:47:43] <anonimasu> f/c?
[19:48:07] <SWPadnos> flycutter?
[19:48:10] <dmess> pretty purrple and blueish pink 6's and 9's
[19:48:11] <anonimasu> ah
[19:48:13] <dmess> yes
[19:48:24] <anonimasu> :)
[19:48:43] <dmess> they stalled it and SMOKED the place out when the blew the belt drive... LOL
[19:49:30] <dmess> so obviously they didnt have the HP
[19:49:45] <anonimasu> hehe
[19:49:51] <dmess> or it coupled properly at least
[19:50:00] <anonimasu> I hope I ever can understand feedrates.
[19:50:10] <anonimasu> I feel at the mercy of tool manufacturers.
[19:50:13] <toastydeath> ?
[19:50:16] <anonimasu> when I pick feed/speed off their charts..
[19:50:18] <dmess> i was running on a 12 inch extension to btw
[19:50:38] <dmess> why so??/
[19:50:39] <toastydeath> what problem are you having picking your own feeds/speeds
[19:50:53] <anonimasu> I lack experience.
[19:50:58] <toastydeath> oh
[19:51:07] <dmess> its all in the material... let it talk to YOU
[19:51:22] <anonimasu> seems metal cuts worse at lower feedrates..
[19:51:25] <archivist> and the stiffness of the machine
[19:51:37] <toastydeath> anonimasu: metal cuts worse when you aren't cutting, rather shaving
[19:51:52] <anonimasu> toastydeath: the recomendations are 10 times what I feel like cutting
[19:51:55] <anonimasu> by head..
[19:51:55] <archivist> it work hardens
[19:51:55] <toastydeath> a lot of people use really high spindle speeds
[19:52:01] <dmess> yes.. to slow is always more detrimental than a little too aggressive
[19:52:03] <toastydeath> but use low table feeds
[19:52:07] <anonimasu> it sounds a bit _tougher_ on the machine
[19:52:10] <anonimasu> but it cuts better.
[19:52:14] <anonimasu> and leaves better finish
[19:52:21] <toastydeath> that tough sound is what good cutting sounds like
[19:52:27] <anonimasu> hm really
[19:52:30] <anonimasu> it sounds frrrrrrrrp
[19:52:33] <toastydeath> yes.
[19:52:38] <toastydeath> like a triangle wave kind of sound.
[19:52:41] <archivist> luvly
[19:52:44] <dmess> but ruins YOUR ears... you only have 2 keep em working as long as possible
[19:52:56] <anonimasu> :)
[19:53:03] <anonimasu> also depth of cuts.
[19:53:05] <cradek> squeaking is bad
[19:53:05] <anonimasu> they seem insane
[19:53:07] <toastydeath> people don't understand that top spindle speed is actually dictated, in many cases, by maximum table feed rate
[19:53:14] <dmess> whatch the chips...
[19:53:19] <toastydeath> especially in aluminum
[19:53:21] <anonimasu> toastydeath: I understand that
[19:53:27] <archivist> lean to one side
[19:53:28] <toastydeath> excellent!
[19:53:33] <anonimasu> that much atleast..
[19:53:34] <anonimasu> :p
[19:53:39] <toastydeath> lean to one side?
[19:53:46] <anonimasu> aluminium :p
[19:53:48] <archivist> chips flying
[19:54:00] <anonimasu> also, depth of cuts in aluminium is a issue I have..
[19:54:00] <dmess> they stick when HOT
[19:54:00] <anonimasu> :p
[19:54:09] <toastydeath> what depths do you use
[19:54:13] <dmess> max for your HP
[19:54:21] <anonimasu> eh.. that's brutal.
[19:54:30] <anonimasu> I ahve 7.5kw..
[19:54:39] <anonimasu> if I remember it right..
[19:54:40] <cradek> if you use a roughing mill and coolant, you can cut with pretty much the whole thing...
[19:54:52] <anonimasu> yep
[19:55:03] <cradek> 10hp? that's a lot of hp
[19:55:15] <toastydeath> depends if it's a geared head or not
[19:55:24] <anonimasu> it's 7.5kw..
[19:55:26] <anonimasu> and then geared
[19:55:29] <anonimasu> err no
[19:55:30] <toastydeath> cool
[19:55:31] <anonimasu> 5kw..
[19:55:40] <anonimasu> 6.8hp..
[19:55:42] <anonimasu> sorry
[19:55:43] <anonimasu> :)
[19:55:51] <anonimasu> that's the motor rating..
[19:55:56] <anonimasu> then it's geared..
[19:56:10] <dmess> geared is nice for power... but you usually sacrifice speed
[19:56:24] <anonimasu> 2krpm max..
[19:56:32] <dmess> see
[19:56:38] <anonimasu> yep
[19:56:41] <anonimasu> I wouldnt mind 22krpm
[19:56:54] <anonimasu> it's a pretty slow machine
[19:57:00] <anonimasu> 3m/min max
[19:57:07] <cradek> has everyone seen my graphs? I think they're very useful, I have them on the wall by the mill: http://timeguy.com
[19:57:15] <dmess> even though... look up the Hitatchi high-feed mill sometime
[19:57:51] <anonimasu> cradek: would it be awesomely hard for you to re-make them to mm`?
[19:57:54] <toastydeath> anonimasu: 3m/min is plenty for a 2000 rpm machine
[19:58:00] <anonimasu> yeah
[19:58:11] <dmess> we are bringing in a new Fidia today... 30 M/min max
[19:58:14] <cradek> anonimasu: the source is right there - it would be easy for someone to do
[19:58:26] <anonimasu> dmess: I hope to afford one someday
[19:58:27] <anonimasu> :p
[19:58:36] <anonimasu> rather find/invent something that needs me to have one
[19:58:36] <dmess> you and me both
[19:58:49] <dmess> correct
[19:59:32] <anonimasu> toastydeath: I thought cutting should be fairly silent
[19:59:49] <dmess> 2m x 6m x 1.5m travel 5 + 1 axes
[19:59:54] <toastydeath> anonimasu: cutting is loud
[20:00:08] <toastydeath> not squealing/chattering
[20:00:13] <toastydeath> but it sounds like a loud triangle wave
[20:00:29] <anonimasu> frp frp frp
[20:00:30] <anonimasu> like that?
[20:00:32] <dmess> if its done right its a sweet sorta sound like nickels falling on the floor
[20:00:41] <anonimasu> hehe
[20:00:48] <toastydeath> vvmmmmmmm
[20:00:49] <archivist> or frying
[20:00:53] <toastydeath> it should be a constant hum
[20:01:03] <toastydeath> regardless of what the chips hitting the enclosure sound like
[20:01:03] <dmess> yes one tone...
[20:01:07] <anonimasu> I know that sound..
[20:01:20] <dmess> not yeeeeee
[20:01:23] <anonimasu> it's the one I get from the dormer recomended freedrates..
[20:01:47] <toastydeath> http://www.youtube.com/watch?v=E7yEQGDl1ew
[20:01:57] <toastydeath> the triangle wave is cutting with no chatter
[20:02:16] <anonimasu> I should hook up my scope.
[20:02:24] <anonimasu> *grins*
[20:02:25] <anonimasu> just kidding
[20:02:28] <toastydeath> obviously at different, lower frequencies
[20:02:30] <toastydeath> lol
[20:02:56] <anonimasu> chatter's something I know
[20:02:56] <anonimasu> :p
[20:03:06] <anonimasu> as im a wuzz.
[20:03:35] <archivist> chatter, work resonance, machine resonance, and all of the above
[20:03:41] <dmess> ive had scopes on high speed spindles in the past... magnetic pickup on the head.. and lets see what its doing
[20:03:43] <anonimasu> 900mm/min and 2krpm into a 8mm cut just seems scary.
[20:03:55] <anonimasu> or something along thoose numbers
[20:04:01] <toastydeath> i have to convert those to inches
[20:04:02] <toastydeath> hold pls
[20:04:23] <cradek> yikes. here is what it should NOT sound like: http://youtube.com/watch?v=gwixalHuZPc
[20:04:30] <anonimasu> haha
[20:04:37] <dmess> 45 degree lead on the cutter... go for broke
[20:04:38] <anonimasu> is that the video I think it is?
[20:04:48] <cradek> I've never seen it before
[20:04:52] <anonimasu> ah ok
[20:05:00] <toastydeath> anonimasu: with a 2 flute cutter?
[20:05:02] <anonimasu> no
[20:05:04] <anonimasu> 4
[20:05:06] <archivist> bah all these vids /me fires up windaz irc
[20:05:15] <toastydeath> that's actually a mild cut
[20:05:31] <anonimasu> he needs more feed :p
[20:05:35] <toastydeath> ouch that one video is annoying
[20:05:41] <anonimasu> yeah
[20:05:43] <anonimasu> squeee
[20:05:52] <anonimasu> I usually dont take that deep cuts..
[20:05:55] <anonimasu> but I dont know why.
[20:06:03] <anonimasu> <-
[20:06:30] <toastydeath> lol
[20:06:40] <anonimasu> I think that machine's too flimsy for that cut
[20:06:42] <toastydeath> depends on the length/diam of the spindle!
[20:06:44] <toastydeath> er
[20:06:45] <toastydeath> endmill
[20:06:51] <toastydeath> plus if you have a damped spindle or toolholder
[20:06:53] <toastydeath> which probably you do not
[20:06:58] <anonimasu> no
[20:07:00] <anonimasu> I do not
[20:07:06] <anonimasu> arent they very expensive'?
[20:07:13] <cradek> he's cutting at 50sfm. it should be more like 200-300
[20:07:19] <cradek> that's a terrible demo
[20:07:23] <toastydeath> for a hobbyist, yes, they are very expensive
[20:07:31] <anonimasu> hm
[20:07:41] <cradek> err sorry, it's 100
[20:07:47] <toastydeath> they're usually an element you can add to a toolholding system
[20:07:50] <toastydeath> like sandvik's capto
[20:07:55] <anonimasu> * anonimasu nods
[20:07:59] <toastydeath> the base holders for that are like 400 bucks each
[20:08:00] <dmess> ive ran 2" dia. em and cut 10" long .25" doc radially in Ti
[20:08:02] <toastydeath> and that doesn't hold any tools
[20:08:40] <anonimasu> http://youtube.com/watch?v=1AMZlDHwebw&feature=related
[20:09:01] <toastydeath> that's the sound you're looking for!
[20:09:10] <archivist_win> tool resonance
[20:09:13] <anonimasu> I always thought it was scary
[20:09:15] <anonimasu> :p
[20:09:18] <toastydeath> nah
[20:09:27] <toastydeath> that's the mill telling you everything is right with the world
[20:09:45] <anonimasu> hehe
[20:09:48] <anonimasu> I need to mill more.
[20:09:55] <dmess> and she'd be happy to make you a few $$$
[20:10:08] <cradek> yeah that one sounds great
[20:10:23] <anonimasu> oh that's what my milling iron
[20:10:25] <cradek> wow is that dry??
[20:10:27] <anonimasu> with carbide tools sounds like..
[20:10:28] <anonimasu> :p
[20:10:28] <anonimasu> almost
[20:10:37] <anonimasu> (loves carbide)
[20:10:41] <toastydeath> you can't mill that hot and use coolant
[20:10:45] <toastydeath> it shatters the tool
[20:10:49] <anonimasu> yeah
[20:11:02] <anonimasu> same applies to facemills right?
[20:11:07] <toastydeath> yeah
[20:11:16] <anonimasu> seems like when I facemill real heavy cuts they dull/break with coolant
[20:11:19] <toastydeath> but you have to have tooling that can handle the heat
[20:11:24] <toastydeath> in the first place
[20:11:27] <anonimasu> yeah
[20:11:30] <anonimasu> carbide inserts..
[20:11:41] <archivist_win> I love the sparks on that one
[20:11:42] <toastydeath> the right grade of carbide or whisker ceramics
[20:11:51] <anonimasu> * anonimasu nods
[20:11:53] <toastydeath> there are plenty of carbide grades that will not handle dry machining
[20:11:56] <dmess> i had 1 friend who was on a big b/mill where i was programming... he'd look at my programs and tell the guys on adjascent machines to step back while we face this plate... pls.. this will get ugly and be over very fast.. its a denis program...
[20:11:57] <anonimasu> I'd shit myself.
[20:12:09] <anonimasu> LOL
[20:12:16] <toastydeath> lol
[20:12:21] <toastydeath> it's money, man
[20:12:34] <toastydeath> the more volume the machine eats per minute, the more money you make
[20:12:44] <dmess> we machine 53-55 Rc EVERY DAY
[20:12:50] <anonimasu> yeah, but if it involves killing the machine it's a bad deal..
[20:12:56] <toastydeath> the machine will handle it
[20:12:58] <toastydeath> they were designed to
[20:13:04] <toastydeath> and they'll give you years of good service
[20:13:06] <dmess> and have to get thru NDT inspection after machining
[20:13:11] <anonimasu> ndt?
[20:13:15] <anonimasu> ah,
[20:13:25] <anonimasu> (has a old mill)
[20:13:38] <anonimasu> with a quite new control..
[20:13:49] <dmess> non destructive testing... etch magnetic partikle inspection liquid pen
[20:13:54] <anonimasu> ok
[20:14:07] <anonimasu> it's a heidenhain tnc310
[20:14:12] <anonimasu> control :)
[20:14:34] <toastydeath> lol
[20:14:35] <dmess> in case you softened or rehardened the steel
[20:14:48] <cradek> http://www.youtube.com/watch?v=ttZJzauH6CU&NR=1
[20:14:51] <cradek> nice face mill.
[20:14:54] <dmess> its a retro control..
[20:15:11] <anonimasu> the one I have?
[20:15:12] <anonimasu> yes
[20:16:07] <archivist_win> ooo eat that
[20:16:29] <anonimasu> dmess: or what did you talk about?
[20:16:38] <archivist_win> you hear the coils of swarf
[20:16:57] <dmess> i was once asked to design and build a 36" f/c for the face plate of a boring mill.....
[20:17:00] <toastydeath> ahhaha that's the one where the dude
[20:17:01] <toastydeath> near the end
[20:17:05] <toastydeath> touches the chips with a broom
[20:17:07] <toastydeath> and it catches fire
[20:17:23] <toastydeath> or i think it is
[20:17:41] <cradek> yes I see that now
[20:17:48] <anonimasu> yeah
[20:17:48] <anonimasu> wow
[20:18:17] <archivist> wot metal, i didnt feel a thing! says the tool
[20:18:34] <toastydeath> "lol que?"
[20:18:48] <anonimasu> *sighs* im jealous of you guys :P
[20:18:56] <anonimasu> I'd like as much clue :p
[20:19:17] <cradek> anonimasu: haha don't look at me
[20:20:06] <dmess> of who
[20:20:06] <anonimasu> I love machining though :)
[20:20:07] <cradek> I think it takes a lot of experimentation to figure out how to run a certain machine
[20:20:26] <anonimasu> it's so awesome being able to pull custom parts out of it
[20:20:36] <cradek> yeah
[20:21:42] <toastydeath> i like that part too
[20:21:48] <toastydeath> i do not like having to design parts though.
[20:21:49] <dmess> asa much as they are all different guys they are all pretty much the same.... each and every CNC machine/installation has its quirks and habits... you have to get her to tell you her story as soon as possible.. then start lovin'her..
[20:21:52] <anonimasu> machinetime ishm
[20:22:20] <anonimasu> -_-
[20:22:25] <toastydeath> some machine controls are ridiculous and take forver to learn
[20:22:30] <toastydeath> but for the most part i concur with dmess
[20:22:43] <toastydeath> once you figure out where the "feed hold" button is and how to get tool length measurement to work
[20:22:49] <anonimasu> I like the heidenhain
[20:22:55] <dmess> some controls shouild have never left the tes lab
[20:23:08] <anonimasu> the programming is nice
[20:23:15] <anonimasu> they've given thought to loops and stuff..
[20:23:16] <toastydeath> what endlessly bothers me is when the feed control knob doesn't halt the rapid move
[20:23:25] <anonimasu> ouch
[20:23:29] <toastydeath> i love it when that one knob can halt the entire machine
[20:23:32] <toastydeath> makes proving dreamy
[20:23:41] <anonimasu> it does on the heidenhains
[20:23:42] <anonimasu> :)
[20:23:43] <toastydeath> nice
[20:23:45] <dmess> yes us too... we add a circuit to our NEW okk's
[20:23:46] <anonimasu> feed override > all
[20:24:00] <anonimasu> you can feed override anything down to zero
[20:24:09] <toastydeath> i always wondered if it was just a parameter
[20:24:12] <dmess> heidy has NO rapid mode just a feed MAX
[20:24:19] <anonimasu> yeah..
[20:24:22] <anonimasu> but that's rapids..
[20:24:31] <dmess> on some fanucks it is a param
[20:24:34] <toastydeath> i'm tempted to call fanuc and ask them if there's a parameter to change the behavior
[20:24:36] <anonimasu> you usually go over the kink..
[20:24:42] <anonimasu> when doing that kind of moves..
[20:24:58] <dmess> but heidy dont see a RAPID....
[20:25:09] <anonimasu> yeah.. they see FMAX..
[20:25:12] <anonimasu> whatever that ends up as
[20:25:13] <dmess> or change of mode we'll say
[20:25:26] <anonimasu> yep
[20:25:36] <anonimasu> I wouldnt mind a new heidenhain on a fast machine
[20:25:42] <dmess> stil in g1 just af f300
[20:26:00] <anonimasu> L X100 FMAX
[20:26:01] <anonimasu> :)
[20:26:11] <toastydeath> but that means you could get more rapid out of it
[20:26:17] <dmess> they are sweet the 530 on my buddys mecof is awesome
[20:26:27] <anonimasu> toastydeath: me?
[20:26:34] <anonimasu> toastydeath: I think the table/machine is a bit heavy for that
[20:26:34] <toastydeath> i'm saying with a g1-style rapid
[20:26:39] <anonimasu> ah
[20:26:52] <anonimasu> it changes the KV factor when you do FMAX moves..
[20:26:54] <toastydeath> like, most controls open up the positional control
[20:26:57] <toastydeath> oh
[20:26:58] <toastydeath> k.
[20:27:05] <dmess> technically yes... and no hockey stick moves
[20:27:10] <anonimasu> hehe
[20:27:16] <toastydeath> lol hockey stick moves
[20:27:19] <anonimasu> heh
[20:27:30] <toastydeath> i never understood the advantage to doing that
[20:27:34] <toastydeath> but all our machines do it
[20:27:36] <anonimasu> nope
[20:27:42] <dmess> what you never bit off a bolt on the 2nd part
[20:27:49] <Skullworks-PGAB> logger_emc bookmark
[20:27:53] <anonimasu> ^_^
[20:27:56] <anonimasu> no, I machined a clamp though
[20:27:59] <skunkworks__> logger_emc: bookmark
[20:27:59] <skunkworks__> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2007-12-07.txt
[20:28:09] <dmess> there is NO advantage
[20:28:12] <toastydeath> dmess: i'm talking about the style of g0 move where all axis feed at max until one gets where it is going
[20:28:14] <anonimasu> but I slowed the feed down until it cut ok
[20:28:15] <anonimasu> :p
[20:28:16] <toastydeath> are we on the sape page?
[20:28:17] <anonimasu> before it hit
[20:28:25] <Skullworks-PGAB> thanks - still on 1st cup of coffee
[20:28:28] <toastydeath> *same
[20:28:32] <anonimasu> ok
[20:28:40] <anonimasu> nothing like that on the heid's..
[20:28:52] <toastydeath> yeah, it produces a literal hockey stick type movement
[20:29:01] <anonimasu> -------------------/
[20:29:07] <anonimasu> I've seen them
[20:29:09] <toastydeath> i generally g0 z+ until i'm up above
[20:29:14] <toastydeath> any sort of fixture
[20:29:17] <toastydeath> then tell it to move
[20:29:23] <dmess> we've all been there... Tosnuc lets you choose.. interpolated g00 or original g00
[20:29:28] <toastydeath> to avoid that kind of crash
[20:29:47] <anonimasu> yep
[20:30:15] <anonimasu> im still watching that clip(the one with the steel)
[20:30:25] <anonimasu> not the facemilling one
[20:30:52] <toastydeath> although yesterday my co-worker tricked me into thinking i'd scrapped a $2000 block of aluminum with several more thousand dollars of machine/personell time into it
[20:31:04] <archivist> hehe
[20:31:12] <toastydeath> i cussed for about fifteen minutes after i'd figured out what was going on
[20:31:14] <anonimasu> gah :)
[20:31:59] <toastydeath> yeah it's funny hindsight
[20:32:15] <toastydeath> it also lets me know the experienced guys now feel i'm "part of the team"
[20:32:28] <toastydeath> they wouldn't ever do that with some of the other newer guys
[20:32:40] <toastydeath> nor would they let the newer guys cut on a block that expensive
[20:33:06] <toastydeath> bbl
[20:33:38] <anonimasu> laters :)
[20:33:42] <anonimasu> im going to code some
[20:42:01] <dmess> we had a mandelli lose its mind this week on a 70,000 dollar piece of Aermet100... luckilly its in an area i think we cut all out... its gonna be miserable im sure but we should save the forging
[20:42:25] <Skullworks-PGAB> been reading the emails about tool carosel - Fixed pot vrs floating pot.
[20:42:58] <dmess> fixed pleas... slower but safer and no recovery required
[20:43:30] <dmess> just make the caroselle go faster...
[20:43:39] <Skullworks-PGAB> almost all machines since the early 80's that use a swing arm tool changer are floating pot - Except Kitamura...
[20:44:26] <dmess> okk's can usually select either or..
[20:44:31] <Skullworks-PGAB> Sliding carosels - like older FADAL and Haas are always fixed pot
[20:44:48] <dmess> they have to be
[20:45:38] <Skullworks-PGAB> true
[20:46:56] <Skullworks-PGAB> Kitamura is different cause they had an intermediate pot to hold tools during tool changes and then put the tool back into its own fixed numbered pocket
[20:47:44] <Skullworks-PGAB> They also have some of the worst toolchanger foulups I've ever seen
[20:54:11] <Skullworks-PGAB> I would agree fixed pot methods are best for "Home builders" but for people trying to salvage Good machine tool iron which has been surplused or otherwised removed from service do to the old age death of the OEM control and spares are not available anymore for that original control... Those "builders" will need a floating pot solution.
[20:56:27] <anonimasu> "pot"
[20:56:30] <anonimasu> ?
[20:57:15] <Skullworks-PGAB> The Mazak used at the workshop is a hybred machine - a Quill machine with a swing arm ATC - rather unique. I'm surprised they are able to rotate the ATC magazine while the swing arm is out of home position.
[20:58:13] <Skullworks-PGAB> POT is what most machine tool makers refer to the pocket which holds the tool in the ATC magazine.
[20:58:27] <anonimasu> ah ok
[20:58:43] <Skullworks-PGAB> Don't have any idea where the term came from.
[21:00:23] <skunkworks__> http://www.electronicsam.com/images/KandT/conversion/toolchangerspindle.JPG
[21:00:42] <Skullworks-PGAB> I love the old OKUMA MC4 mills - you could single step thru each move of the tool changer forward or backward
[21:02:39] <Skullworks-PGAB> Sam - what is that in the spindle?
[21:03:21] <SWPadnos> looks like a dremel-type tool
[21:03:22] <skunkworks__> a welded up mount for a die grinder..
[21:03:22] <cradek> even this could index while the arm is horizontal couldn't it?
[21:03:51] <Skullworks-PGAB> yeah
[21:04:11] <cradek> it would be holding the old and new tools in the horizontal position while it indexes to where the old tool goes
[21:04:32] <cradek> it would not be very fast.
[21:04:54] <skunkworks__> are you talking about mine?
[21:04:57] <cradek> yes
[21:05:01] <skunkworks__> ah
[21:05:08] <skunkworks__> yes - I suppose..
[21:05:10] <Skullworks-PGAB> but remember that the swing arm would have to stay extended if there was a tool held in the arm
[21:05:21] <skunkworks__> right now it is random access. It reads the barcode on the tool
[21:05:25] <cradek> yes
[21:05:45] <skunkworks__> you call the tool befor hand - it spins the chain reading each tool.
[21:05:52] <cradek> cool
[21:05:57] <cradek> you could do that too!
[21:06:03] <skunkworks__> that is my hope :)
[21:06:15] <cradek> it would just look for the tool every time
[21:06:26] <anonimasu> you can make a index table..
[21:06:27] <cradek> that's neat because it doesn't need to remember anything
[21:06:29] <anonimasu> for faster lookup also..
[21:06:42] <anonimasu> so that once you've used a tool one time it wont search for it..
[21:06:49] <skunkworks__> you just call the next tool way before you need it ;)
[21:06:56] <anonimasu> yeah.. that works too
[21:07:08] <cradek> does it scan through them pretty fast?
[21:07:14] <skunkworks__> like - right after the previous tool chain.
[21:07:17] <skunkworks__> change
[21:07:32] <skunkworks__> I think it takes about 60 seconds to go all the way around
[21:07:54] <cradek> how many pockets is it?
[21:08:00] <anonimasu> lots :)
[21:08:05] <cradek> no kidding
[21:08:10] <Skullworks-PGAB> I always call the next tool on the next line after M6 - the Okuma I use has a large ATC like sam's but it moves very slow
[21:09:03] <cradek> even with fixed pockets sam's could be pretty fast if it can move both directions (and you aren't using 30 tools)
[21:09:43] <cradek> I notice you have approximately 1 tool holder for it :-)
[21:10:12] <cradek> it looks like about 50 pockets, wow
[21:10:19] <SWPadnos> 64?
[21:10:19] <toastydeath> skullworks-pgab: we do the same, cat-50 toolchangers just move too slow when you have more than 25 tools
[21:11:13] <Skullworks-PGAB> best to store tools on a rack outside the ATC long term - or the get so gummed up they sometimes stick in the spindle. :(
[21:11:22] <toastydeath> we do
[21:11:54] <toastydeath> all our horizontal mills have an interesting toolchanger though
[21:12:04] <toastydeath> 25x cat-40, on what is essentially a lathe turret
[21:12:11] <toastydeath> pretty quick tool to tool
[21:12:25] <skunkworks__> I think it is 60 tools
[21:12:58] <Skullworks-PGAB> you want to see weird - watch a Hitachi Hi-Cell run
[21:13:29] <Skullworks-PGAB> its a large lathe with a 8 tool turret with live tooling
[21:14:02] <archivist> turning squares etc
[21:14:21] <archivist> and hex from round
[21:14:30] <Skullworks-PGAB> while its turning a part the back of the machine opens and an ATC swaps out a tool on the far side of the turret.
[21:14:43] <toastydeath> haha that's so cool
[21:15:17] <Skullworks-PGAB> 8 tools in the turret + 24 in the magaine
[21:15:23] <toastydeath> doesn't mazak make a version of the integrex that has a lower turret and an upper 5-axis spindle
[21:16:15] <Skullworks-PGAB> most the big machine tool builders have show boat machine they build very few of
[21:16:40] <Skullworks-PGAB> but the HiCell was a very common machine
[21:17:02] <Skullworks-PGAB> used by Custom Wheel makers
[21:17:24] <toastydeath> nice
[21:18:13] <Skullworks-PGAB> yeah - did the lug bolt holes in the same chuking as turning the rim
[21:18:49] <Skullworks-PGAB> then a robot flips the wheel for doing the back side
[21:28:09] <SWPadnos___> SWPadnos___ is now known as SWPadnos
[21:38:42] <dmess> wierd... check out a chiron dz16.... 2.0 seconds chip to chip
[22:13:32] <dmess> im back and hi'er than last time...
[22:20:08] <anonimasu> nice
[22:21:34] <dmess> what???
[22:23:10] <anonimasu> I dont know
[22:26:11] <alex_joni> http://home.arcor.de/hans.plumanns/L-Test.jpg
[22:31:41] <hmtr> Ok, when I left, we were talking about using encoders for feedrate pots on Mesa gear.
[22:31:57] <alex_joni> hmtr: feedrate override
[22:32:06] <anonimasu> well, that's kind of bad..
[22:32:12] <alex_joni> the actual feedrate is coded in the files
[22:32:14] <alex_joni> anonimasu: why?
[22:33:13] <anonimasu> because when you turn your feedrate down, you want to know that it's at zero..
[22:33:20] <anonimasu> regardless of machine on/fooff
[22:33:22] <anonimasu> off..
[22:33:24] <hmtr> so if I'm understanding it correctly, I need 5 axes, 1 spindle, 2 override encoders, and 1 mpg = 9 axes
[22:33:37] <hmtr> or 9 channels
[22:33:55] <anonimasu> hm
[22:33:58] <anonimasu> for feedrate it'd work
[22:34:02] <anonimasu> not feed-override
[22:34:09] <dmess> find this tune and your understand.... anon... 8-) ..... Toby Keith&Willy Nelson I'll Never Smoke Weed With Willie Again.mp3
[22:34:24] <anonimasu> feed override should be a pot with a fixed limit
[22:34:32] <alex_joni> anonimasu: nope
[22:34:33] <dmess> still gigglin'
[22:34:33] <anonimasu> so when you turn it to zero your machine stops.
[22:34:39] <anonimasu> alex_joni: that's retarded..
[22:34:40] <alex_joni> how can you set it from the GUI then?
[22:34:48] <anonimasu> alex_joni: should you set it from the gui really?
[22:34:54] <alex_joni> feed-override?
[22:34:55] <alex_joni> sure
[22:34:59] <alex_joni> from any GUI imo
[22:35:19] <alex_joni> what if you don't have a knob? then you don't use it at all?
[22:35:28] <dmess> override is required to prove out..
[22:35:40] <alex_joni> hmtr: you said you don't want encoder on the spindle?
[22:35:48] <anonimasu> dmess: prove out?
[22:36:44] <hmtr> no encoder is required on the spindle. too much gear train to try to reverse for rigid tapping
[22:36:56] <alex_joni> hmtr: then a short recalc :)
[22:37:07] <dmess> tape try out... 1st off... whatever you call it... a complex part will NEVER be perfect 1 time out.. who are you kidding...
[22:37:08] <hmtr> :)
[22:37:12] <alex_joni> 5 axes = 5 DACs, 5 ADCs, 5 encoder counters
[22:37:32] <alex_joni> 1 mpg = 1 encoder counter (low speed)
[22:37:48] <alex_joni> 2 override mpgs = 2 encoder counters (low speed)
[22:38:06] <anonimasu> dmess: Yes, but the debate is about having a encoder instead of a pot for feedrate override
[22:38:07] <alex_joni> 1 spindle = 1 DAC, 1 ADC (measuring current speed)
[22:38:18] <alex_joni> anonimasu: the idea is having multiple points of control
[22:38:23] <dmess> some of our programs run 40 hrs for 1 operation of 40...
[22:38:33] <alex_joni> and that's the whole idea behind the NIST control principle
[22:38:53] <dmess> gotta have a KNOB..
[22:38:54] <alex_joni> we'll stick with that, so if you think you have something better.. I suggest you try to convince them :)
[22:39:00] <anonimasu> alex_joni: I sure as hell wouldnt want anyone else poking around with my machine when i run it.
[22:39:08] <hmtr> I think I'm starting to grasp it a little
[22:39:08] <anonimasu> alex_joni: that's outright dangerous
[22:39:15] <alex_joni> anonimasu: someone else?
[22:39:21] <anonimasu> alex_joni: the idea is really nice, but "multiple points of control"
[22:39:27] <anonimasu> how is that applicable to a vmc?
[22:39:29] <alex_joni> I am talking about a GUI slider and a phisical knob
[22:39:37] <alex_joni> physical even
[22:39:47] <anonimasu> Yeah, but why would you ever touch the gui slider/even want it when you have a physical knob?
[22:40:02] <alex_joni> because I want to watch the screen for the actual feedrate
[22:40:05] <hmtr> I'm assuming the spindle can be run as completely open loop? without the analog speed input?
[22:40:10] <alex_joni> hmtr: sure
[22:40:19] <dmess> we have multi pt. controls... siemens 840D... on each end of a 4 spindle monster...
[22:40:21] <alex_joni> but it's nice to have though..
[22:40:45] <alex_joni> dmess: then I'm *sure* you don't have pots for the knobs
[22:40:49] <dmess> 1 operator 1 key.. though.. ; )
[22:41:02] <alex_joni> because no matter what.. you can't make the pot turn from somewhere else :)
[22:41:13] <dmess> keys... alll work on bothe sides
[22:41:19] <alex_joni> and disregarding it (while it reads 50%) would be even worse imo
[22:42:01] <alex_joni> but then again, emc2 is configurable enough for anyone to do as he pleases
[22:42:18] <dmess> but is an 840 D.. X 2..... that is the SMARTEST and most flex control currrently on the planet...
[22:42:33] <alex_joni> dmess: I don't argue that it's not..
[22:42:47] <dmess> 27 axes to 1 control
[22:42:52] <hmtr> For my first EMC, the least amount of configuration will be better of course!!
[22:43:11] <alex_joni> hmtr: right.. that's why we said 5 axes might be a bit of a bite
[22:43:51] <hmtr> well I was planning on paying you to come finish it after I failed miserably :)
[22:44:12] <dmess> ive installed a DME evolution 5 axis at a college and had to train them on it.... fun WOWOOW
[22:44:28] <hmtr> I assumed that the amount of axes was going to be the simple part
[22:44:50] <alex_joni> hmtr: you'll face a couple of hurdles
[22:44:54] <alex_joni> get the wiring done
[22:45:03] <alex_joni> testing, fixing to get movement
[22:45:12] <alex_joni> tuning
[22:45:13] <alex_joni> etc
[22:45:22] <dmess> it is for them.. everything has an encoder.. the spindle.. t/c...part loader...
[22:45:22] <alex_joni> hmtr: got a timeschedule?
[22:45:48] <hmtr> is a month of machine down time asking to much?
[22:46:15] <dmess> WAY...
[22:46:15] <alex_joni> depends how long your workdays are :)
[22:46:38] <alex_joni> hmtr: this is after you have *all* the components?
[22:46:41] <hmtr> I've done quite a few Heidenhain/Siemens/Acramatic retros, but a lot of the control loops are already done for you
[22:46:53] <alex_joni> hmtr: ok, that helps a lot :)
[22:46:58] <alex_joni> where are you located?
[22:47:05] <hmtr> alex_joni: I plan to have everything ahead of time
[22:47:16] <dmess> and how many GOOD ppl you throw at the job... 1000 hrs a work 100 ppl 10 hrs
[22:47:23] <hmtr> and a heck of a plan when I attack it
[22:47:30] <alex_joni> dmess: that doesn't work :D
[22:47:48] <dmess> it does SO... worked it for ALOTTA yrs
[22:48:12] <alex_joni> wanna bet that you can't do a 1000h job with 100 ppl in 10 hrs?
[22:48:21] <alex_joni> (using the exact numbers you said..)
[22:48:23] <dmess> 7:30 meeting... q#1??? whos in trouble on a job...
[22:48:32] <alex_joni> I agree the concept works.. but not with those numbers
[22:48:58] <dmess> okk 10 ppl in 100 hours is a GUARANTEE...to me to
[22:49:06] <alex_joni> I agree that works
[22:49:15] <dmess> if i pick em... 1st part right
[22:49:21] <alex_joni> hmtr: where are you located?
[22:49:28] <hmtr> wichita, ks
[22:49:44] <dmess> i may be visiting...
[22:49:46] <alex_joni> hmtr: I suggest you subscribe to emc-users
[22:49:49] <toastydeath> it should work
[22:49:56] <toastydeath> 1000 hrs x 100 people on a job that took 10 hours
[22:49:57] <alex_joni> that's a mailing list we have for emc
[22:50:10] <toastydeath> because the cycle time
[22:50:12] <toastydeath> will drop with quantity
[22:50:15] <alex_joni> toastydeath: not in 10 hours.. you lose a lot of time to get them into gear
[22:50:28] <alex_joni> toastydeath: it wasn't a machining job
[22:50:30] <toastydeath> oh
[22:50:37] <dmess> ive done the 300 hrs in 48 before... 10 good guys...
[22:50:37] <alex_joni> it was a "thing" in general.. as a concept
[22:51:02] <alex_joni> 1000h, with 100ppl in ~20h I can believe :D
[22:51:09] <alex_joni> maybe 15 :)
[22:51:23] <dmess> thats total progect times... TO the shipping dorrs on time..
[22:51:28] <alex_joni> hmtr: you can find people in your area on the emc-users list
[22:51:32] <toastydeath> it depends on what you are doing
[22:51:40] <toastydeath> sometimes it takes longer as it scales, sometimes it drops off
[22:51:42] <hmtr> I also have a few months to get my plan together...maybe even practice on a servo knee mill I have here first
[22:51:46] <alex_joni> toastydeath: handcarving flutes
[22:51:52] <toastydeath> jesus christ what.
[22:51:57] <alex_joni> hmtr: that's a great plan
[22:52:07] <toastydeath> hand carving flutes.
[22:52:07] <alex_joni> toastydeath: first thing that popped into my mind :)
[22:52:17] <alex_joni> software development
[22:52:19] <toastydeath> haha, glad i'm not the only one
[22:52:19] <alex_joni> better?
[22:52:25] <toastydeath> sure, that's better
[22:52:30] <archivist> all hand errors copied
[22:52:33] <dmess> i sent the 1st steel mill ladles and runouts to China.... we did it in under 1 yr
[22:52:35] <hmtr> where is this list at?
[22:52:50] <hmtr> I think I found it
[22:52:52] <alex_joni> http://www.linuxcnc.org/content/view/4/8/lang,en/
[22:53:20] <alex_joni> hmtr: beeing a new user.. it would be nice for us to hear some feedback about how *hard* it was to find/discover things
[22:53:31] <alex_joni> or maybe suggestions where we can make improvements
[22:53:42] <hmtr> whats crazy is the amount of aerospace work where I'm located vs the amount of retrofitters/integrators
[22:54:00] <hmtr> not many retro companies in the area
[22:54:10] <alex_joni> sounds like a good thing :)
[22:55:05] <hmtr> so is this more for the developers or integrators?
[22:55:20] <alex_joni> emc-users is for endusers & integrators
[22:55:21] <Ziegler> Hey gents... OT: anyone ever graph large amounts of XY data (50+ MB of x,y delineated data)
[22:55:23] <alex_joni> most are both
[22:55:33] <alex_joni> hmtr: there's emc-devel for developers
[22:55:48] <hmtr> I meant this channel
[22:56:06] <alex_joni> users, integrators, machining, off-topic stuff
[22:56:16] <hmtr> regardless, I do appreciate the time and info :)
[22:56:23] <alex_joni> hmtr: there's another one for developers :)
[22:56:35] <Ziegler> (heh... off topic)
[22:56:53] <alex_joni> Ziegler: yeah, that's quite common :)
[22:56:59] <alex_joni> Ziegler: btw, graph how?
[22:57:03] <hmtr> I'm frying my brain right now tryin to grasp this stuff, along with keeping the normal customer's happy
[22:57:34] <Ziegler> I dont care how... as long as I can deliver graphically... an xy scatter chart (points connected)
[22:58:18] <alex_joni> Ziegler: obvious question.. did you try gnuplot?
[22:58:24] <Ziegler> NOPE!
[22:58:28] <Ziegler> ill look into it
[22:58:32] <Skullworks-PGAB> these logs - help shape the program package
[23:00:14] <alex_joni> Skullworks-PGAB: sorry?
[23:00:21] <Ziegler> installing
[23:00:38] <alex_joni> Ziegler: I'm not sure how well it works with large amount of data..
[23:00:55] <Ziegler> Ill find out pretty quick here
[23:00:55] <Skullworks-PGAB> people asking hre helps shape the developement
[23:01:32] <alex_joni> Ziegler: I also just found http://en.wikipedia.org/wiki/Grace_(plotting_tool)
[23:01:39] <alex_joni> Skullworks-PGAB: sure does
[23:01:48] <Ziegler> my only other thought was doing something in perl or python... but thats last resort
[23:02:04] <Ziegler> looking at wiki
[23:02:15] <alex_joni> Ziegler: I wouldn't chose python/perl if I had another choice..
[23:02:28] <alex_joni> had a python project lately with 30MB+ vertexes
[23:02:37] <Ziegler> and?
[23:02:42] <alex_joni> and it easily grew out of memory on 392MB RAM
[23:02:53] <Ziegler> yuck
[23:02:58] <alex_joni> I had some transformations to do though..
[23:03:30] <alex_joni> Ziegler: if you end up with python, numpy saves you a *LOT* of issues and code
[23:03:40] <Skullworks-PGAB> * Skullworks-PGAB was up all nite trying to "de-Vista" a HP Pavilion 9335 laptop. Drivers just don't exist for some of the hardware...
[23:03:46] <alex_joni> http://numpy.scipy.org/
[23:08:00] <jepler> *
[23:08:00] <jepler> Turrican 1 was also released for the Atari ST, CDTV, Mega Drive/Genesis, Game Boy, PC Engine, ZX Spectrum and Amstrad CPC. The console ports were handled by The Code Monkeys and published by Accolade.
[23:08:04] <jepler> "Subsong 2" from Turrican 1 is actually the song "Escape" from The Transformers: The Movie soundtrack.[2] Ramiro Vaca is responsible for the complete C64 score, including the cover of the song Escape in "Subsong 2".[citation needed] Chris Hülsbeck composed for the Amiga conversion of Turrican I, II, III as well as Mega Turrican (Mega Drive/Genesis) & Super Turrican (SNES). Subsongs 4 and 7 of the music in the Amiga version of world 4 were clearly inspire
[23:08:11] <jepler> argh!
[23:08:14] <jepler> sorry
[23:08:16] <jepler> http://matplotlib.sourceforge.net/screenshots.html
[23:10:06] <jepler> (oooh they have something in python that typesets many TeX math equations!)
[23:10:34] <hmtr> well, gotta run. Thanks for the info.
[23:13:29] <dmess> Jepler.. pass the pipe.. what have you found??
[23:14:02] <dmess> im an old c64 kid
[23:14:44] <archivist> c64! that makes you a youngster
[23:14:52] <alex_joni> good night guys
[23:14:54] <dmess> my monitor from 1984 is still playing ps3
[23:15:33] <dmess> i toothed on the PET 's wit internal tape drives
[23:15:35] <archivist> my commodore PET has died (ram error)
[23:15:55] <dmess> too bad...
[23:15:59] <archivist> comes up 0 bytes free
[23:16:18] <dmess> we had a SUPER-PET at the h/s i went to...
[23:16:18] <SWPadnos> those were the days - the Commodore Personal Electronic Transactor
[23:16:44] <archivist> if you saw the hacks I did to it! S100 bus etc
[23:17:06] <dmess> is that what it meant.. i NEVER knew.. or could find someone to tell me...
[23:17:49] <archivist> it was in all the mags and the commodore adds
[23:18:20] <dmess> even Butterfield.. the commadore GURU on the servers at the time
[23:20:08] <dmess> i bought a system for $1600..in a computer store in sept... at x-mas it was in canadian tire for $599 without the 399 monitor....
[23:20:49] <dmess> but the monitor LIVES on...
[23:20:53] <archivist> we had a cpm box on IEEE488 of the back of the PET at work, 45 mins cycle time eprom to eprom during development
[23:21:33] <dmess> great in the van.. with a dvd pluged into it and it in the ELIMINATOR powerpack....
[23:22:06] <dmess> you sound like you made IT
[23:22:45] <jepler> dmess: that was just some text from a wikipedia page
[23:23:41] <dmess> ok ... was wondering.... you should share ehhh....
[23:24:07] <jepler> this is what I was reading in the first place: http://www.metafilter.com/67268/RunStopRestore
[23:24:56] <dmess> there are many OLD console game available..
[23:26:30] <dmess> we never believed we could write a 32k program without a syntax error.. LOL
[23:27:12] <dmess> thats 32000 correct keystrokes.... LOL
[23:27:22] <dmess> i still cant type..
[23:27:47] <archivist> now we haf sepeeling chkers
[23:27:49] <SWPadnos> good thing IRC doesn't use a typewriter
[23:27:56] <SWPadnos> TTYIRC
[23:28:06] <dmess> nEt
[23:29:04] <archivist> Im thinking of the slow fixed speed tty we used on baudot code
[23:29:35] <dmess> tty... how abot ttd??
[23:29:44] <dmess> 300 baud..
[23:29:46] <toastydeath> morse code
[23:29:46] <SWPadnos> or TTL
[23:31:03] <archivist> I think my first attempt at rtty was 300 or slower (such a long time ago)
[23:31:38] <dmess> yes.. they are regular /STD device for the deaf..... that WHY blacberry has a 5+% deaf consumer rate
[23:33:06] <dmess> you hav to put the phone in ti.. and type... transmitting at 300 baud to the other end.... then he can type...
[23:34:00] <dmess> RIM.. doesnt even KNOW what they have on their hands...
[23:34:21] <archivist> some of my Open University was on a telytype with paper tape
[23:36:37] <dmess> my college was still on paper tape to the machine
[23:37:03] <dmess> apt 360 in the back
[23:39:48] <archivist> hmm that must be 30 ish years ago