#linuxcnc | Logs for 2013-05-03

Back
[00:00:20] <ssi> mhaberler: so the hal driver api is four funcs, rtapi_app_main, rtapi_app_exit, write_port, read_port?
[00:00:30] <mhaberler> yes, pretty much
[00:00:34] <ssi> ok
[00:01:02] <mhaberler> when you 'loadrt <drivername>', rtapi_app_main() gets called, in a non-RT context
[00:01:14] <mhaberler> you can do file i/o there just fine
[00:01:23] <eric_unterhausen> mharberler beagleboard@googlegroups.com
[00:01:30] <mhaberler> but not in the thread functions (read_port and write_port)
[00:01:53] <mhaberler> on unloadrt <drivername>, again rtapi_app_exit gets called, again user context
[00:02:20] <ssi> and read_port, write_port are pretty generic I guess
[00:02:27] <mhaberler> ah, that gorup, thanks
[00:02:32] <ssi> so you write components to glue those to more specific things?
[00:02:53] <mhaberler> yes, there shouldnt be much of a difference except for the mmap() address and port offsets
[00:03:23] <mhaberler> the hard part is getting rtapi_app_main() and rtapi_app_exit() right
[00:03:44] <mhaberler> there needs to be a command line argument to determine which pins are under driver managment
[00:04:05] <mhaberler> probably a bit of conflict detection there helps a lot against frustrations ;)
[00:04:53] <mhaberler> then take those pins out of driver management (in theory some other process could have that pin's device file open for instance, that should be detected)
[00:05:27] <mhaberler> and while the HAL driver is active, the kernel's gpio dev files should either vanish or be deactivated
[00:06:04] <mhaberler> but you can figure this out completely outside of linuxcnc
[00:06:07] <ssi> I was looking at BBW gpio stuff earlier, and I saw reference to some magic device tat you'd write numbers into to cause the dev files for the GPIO to show up or disappear
[00:06:17] <ssi> I wonder if that's the case with BBB as well
[00:06:20] <mhaberler> with the HAL RT environment it's a tad harder to debug
[00:06:44] <mhaberler> the pinmux dev probably
[00:06:55] <pcw_home> OK you can set/clear individual bits so all you need to do is somehow tell the driver the pins are reserved
[00:07:04] <mhaberler> right, thats it
[00:07:18] <pcw_home> (for HAL/PRU etc)
[00:08:03] <mhaberler> for instance assume some other process had an in pin open with edge triggering, and then you trample over that with the hal driver - that might mean an edge triggered interrupt in the kernel and the HAL driver just ironing over the same port
[00:10:00] <mhaberler> I think we can take the non-DTB based kernels like the 3.2.21 we have now out of the equation and assume the 3.8+ DTB based one
[00:10:59] <mhaberler> a fellow over at #xenomai is already debugging the build, and I think it wont take long; days to a few weeks atm
[00:12:03] <pcw_home> The cape EEPROM can have (the driver?) setup pins but at some point you need a config setup tp say who owns which set of pins
[00:12:11] <ssi> sounds good... I've got 3.8 dtb builds working now
[00:12:16] <ssi> just waiting on the xenomai patches
[00:12:19] <eric_unterhausen> with hdmi, I would guess a lot fewer poeple will be using the lcd pins
[00:12:29] <mhaberler> ssi: I think my suggested approach would be to cook up a small C program which accesses pins as we eventually need it in hal_gpio, and make sure you get reservation/interaction right
[00:12:38] <ssi> eric_unterhausen: hdmi uses the lcd pins, sadly
[00:12:51] <mhaberler> only when that's done, actually start figuring linuxcnc hal drivers, that'd be the easy part
[00:13:09] <pcw_home> and 1 encoder :-(
[00:13:17] <ssi> mhaberler: easy part for you maybe... I'm still trying to wrap my head around it! :D
[00:13:32] <ssi> pcw_home: what about 1 encoder?
[00:13:42] <mhaberler> yeah, it's going to take a while of googling, reading and asking, and trying
[00:13:58] <eric_unterhausen> I'm still trying to read the SRM output pin chart rotated sideways
[00:14:18] <pcw_home> If you have HDMI you lose one of the three hardware encoders
[00:14:31] <mhaberler> one line I'd investigate is: how are the DTB's different from BBW and BBB and does that have a bearing on our driver, or can it be overridden at runtime
[00:14:49] <ssi> mhaberler: the hardware is very very similar between them
[00:14:51] <eric_unterhausen> pcw: so headless it is :)
[00:14:55] <mhaberler> DTB seems a boot-time thing so a bit heavy-handed
[00:14:56] <ssi> mhaberler: ram voltage is slightly different
[00:14:57] <pcw_home> if you use the on card NAND, you lose another
[00:15:19] <mhaberler> pretty sure that HDMI/NAND usage is reflected in the DTB
[00:15:47] <mhaberler> I saw a post recently about user-defined DTB manipulations, let me see if I can find that
[00:15:56] <pcw_home> Haven't looked at the PWMs vs pin muxing yet
[00:17:52] <eric_unterhausen> pin usage is encoded on compatible capes
[00:18:18] <mhaberler> a, in the eeprom?
[00:18:28] <eric_unterhausen> yes eeprom
[00:18:43] <ssi> yeah there's a table in the cape eeprom
[00:18:45] <mhaberler> ok so that'd be boot time; we need runtime
[00:19:08] <mhaberler> but basically the same mechanism I guess
[00:19:50] <mhaberler> say cape X uses pins 1,2,3 for some funny I/O - I assume those wouldnt be available for general gpio post-boot
[00:19:54] <eric_unterhausen> pad control registers readable?
[00:20:04] <mhaberler> that would defeat the whole purpose if it were
[00:22:45] <mhaberler> I think the best option for now is to post the problem description to the beagleboard group and hear what they say
[00:23:09] <mhaberler> based on that, try to reproduce results with a standalone C program and verify this works
[00:23:17] <mhaberler> then glue it into the HAL driver
[00:24:13] <mhaberler> it could well be the PRU hal driver has the same issue actually, so that solution would apply there as well
[00:27:36] <mhaberler> ok, well I'll scan the beagleboard group for related information and post the problem outline there later today
[00:27:52] <ssi> I had a dozen mini and micro usb cables here somewhere, and now that I need them I can't find them
[00:30:00] <ds3> you don't have to use the cape system
[00:30:31] <mhaberler> well eeprom emulation wouldnt be my first choice too ;)
[00:31:03] <ds3> so throw it out
[00:31:05] <ds3> I did
[00:31:11] <ds3> it is rediculous
[00:31:23] <mhaberler> you mean a bad idea?
[00:31:37] <ds3> the whole cape id with eeprom is a really bad idea, IMO
[00:31:54] <mhaberler> could be that it originated from userfriendly.com
[00:32:31] <mhaberler> thinks like that happen if you enter arduinoland from upper atmosphere
[00:32:32] <ds3> nah... there is a whole other dispute
[00:32:47] <ds3> it was thought of in the beagleboard classic era
[00:33:08] <mhaberler> what is the alternative?
[00:33:11] <ds3> peoplel couldn't figure out how to work the pinmux
[00:33:20] <ds3> writing proper code to pick what you really need for your application
[00:33:28] <ds3> not doing braindead automated things that don't suit what you need
[00:33:29] <mhaberler> well it needs digging in the TRM, but it isnt rocket science
[00:33:48] <ds3> during the beagleboard era when it was thought it, I realized it is a no go
[00:34:16] <ds3> i have a LCD board that I sold... it is wired up so that you can use the hardware PWM for backlight dimming OR just a plain GPIO for on/off mode
[00:34:40] <ds3> any eeprom setup would force everyone to use one or another.
[00:34:44] <mhaberler> what I dont understand is the relation of eeproms and DTB 'overlays'
[00:34:53] <mhaberler> any idea for me unwashed?
[00:35:04] <ds3> the overlays are an attempt to map the eeproms to the whole DTB idea
[00:35:29] <ds3> eeproms predated the DTB. before DTB on ARM happened, we used board files to identify what we want and to create the right drivers
[00:35:44] <mhaberler> map.. meaning replace by converting into dtb o format?
[00:35:49] <ds3> that code used to read the eeprom and create drivers
[00:35:56] <ds3> not quite.
[00:36:09] <ds3> DTB did away with the whole concept of a board file
[00:36:21] <mhaberler> yes, was about time
[00:36:33] <ds3> so how do you dynamically create drivers at run time? dtb fragements aka dtb overlays was born
[00:36:35] <mhaberler> the mess was getting a bit annoying
[00:36:45] <ds3> that mess wasn't that bad
[00:36:52] <ds3> DTB doesn't solve that at all.
[00:37:08] <mhaberler> am I correct assuming DTB overlays are a boot time affair?
[00:37:11] <ds3> it just renames the problem. instead of a simple .c file, you now have another data structure that the bootloader or user had to create
[00:37:18] <mhaberler> duh
[00:37:22] <ds3> no, run time
[00:37:28] <ds3> it ties in with the cape manager stuff
[00:37:37] <mhaberler> so you can apply/unapply DTB overlays at runtime?
[00:37:42] <ds3> yes
[00:38:01] <ds3> my personal preference is to revert this DTB/eeprom crap and use a simplier saner system
[00:38:02] <mhaberler> ah, then that is probably the vehicle we're looking for
[00:38:17] <ds3> but then I have interest in smaller systems like Cortex-M3 CPUs
[00:38:24] <ds3> DTBs are too heavy for that stuff
[00:39:10] <ds3> yes, you can kind of do what you want with the DTB stuff
[00:39:33] <ds3> but what is wrong with telling the kernel the gpios are inert and having the RT stuff grab it underneath it?
[00:39:49] <mhaberler> well would this problem statement make sense: based on a set of pins desired for the HAL driver, create and load a DTB overlay to reserve those pins and take them out of gpio driver management?
[00:40:14] <mhaberler> nothing, the only question is how to tell the kernel in a non-intrusive way
[00:40:32] <mhaberler> and detect possible conflicts while you pull a set of pins into HAL driver management
[00:40:35] <ds3> from what I recall, the gpio driver isn't written to allow that
[00:40:46] <ds3> as long as you don't do a gpio_request, you are fine
[00:40:59] <mhaberler> so ships in the night
[00:41:13] <ds3> and if you do, set it up as an input. then go underneath it and change the register to an output
[00:41:15] <ds3> no no
[00:41:18] <mhaberler> and hope the user dont fiddle any /dev/gpioX
[00:41:18] <ds3> this is a common thing
[00:41:28] <mhaberler> I see
[00:41:30] <ds3> you know what the pinmux is?
[00:41:45] <mhaberler> sorta kinda, I think I managed to do a few pins right, yes
[00:42:01] <ds3> let me restate it so I can get my point across
[00:42:13] <ds3> the pinmux lets you set an external pin to one of many possible functions
[00:42:20] <mhaberler> right
[00:42:35] <ds3> so if a pin is set to say... SPI, the gpio subsystem shouldn't be able to muck with it
[00:42:49] <ds3> in effect, every pin is part of the gpio driver
[00:43:02] <ds3> but unless you do a gpio_request(), the gpio driver won't do anything with it
[00:43:06] <ds3> makes sense?
[00:43:11] <mhaberler> yes
[00:43:28] <mhaberler> gpio_request() is an in-kernel thing, or a userland thing?
[00:43:31] <ds3> kernel
[00:43:34] <ds3> I don't do userland :D
[00:43:35] <mhaberler> I see
[00:43:42] <mhaberler> real men use printk
[00:43:49] <ds3> precisely
[00:44:04] <ds3> if you want to use the kernel's gpio driver, typical sequence is:
[00:45:11] <ds3> gpio_request(foo, "desc"); gpio_set_direction_output(number, value);
[00:45:46] <mhaberler> is that reflected in the sysfs or dev entries?
[00:46:15] <ds3> look at the gpio_keys.c stuff
[00:46:23] <ds3> there is no /dev entry for gpios
[00:46:36] <ds3> indirectly
[00:46:47] <mhaberler> gpio_keys.c is part of the 3.8 kernel tree I assume?
[00:46:56] <ds3> sysfs won't let you allocate a gpio unless you explicitly allow it
[00:47:06] <ds3> gpio_keys.c is part of the kernel since who knows when :D
[00:47:11] <ds3> look in drivers/input/keyboard/
[00:47:15] <mhaberler> ok
[00:47:30] <ds3> it is a driver to do buttons via a gpio pin
[00:47:49] <ds3> think of it as a non RT, limit switch ;)
[00:50:07] <mhaberler> do we get away with all userland code for the driver?
[00:51:35] <mhaberler> I see, gpio_request_one()
[00:52:32] <mhaberler> ah, there's also the code to interface to DT
[00:55:00] <mhaberler> it looks like we'd want to do something very similar like in gpio_keys.c but in userland
[00:57:25] <ds3> in general userland drivers don't work well
[00:58:04] <mhaberler> well we dont have the option, xenomai will relegate the thread to cattle-class scheduling as soon as you do syscalls say for file i/o
[00:58:33] <ds3> hmmmm
[00:58:39] <mhaberler> I can imagine an rtdm device driver, but that leaves open the question of rt-preempt
[00:58:54] <ds3> what does xenomai do with memory mapped drivers?
[00:59:03] <mhaberler> absolutely nothing
[00:59:11] <ds3> that would avoid the syscall
[00:59:37] <ds3> actually if you must do that
[00:59:38] <mhaberler> doing mmapped I/O in the rt thread is the way to do it, unfortunately
[00:59:39] <ds3> do this -
[00:59:50] <ds3> don't allocate any gpios in kernel space
[01:00:01] <ds3> i.e. make sure nothing calls gpio_request() on the pins of interest
[01:00:25] <mhaberler> (can I test for that in the setup code which is non-rt?)
[01:00:26] <ds3> use /dev/mem (yes, that is discouraged but it sounds like that is the only way for what you are describing)
[01:00:36] <mhaberler> that's the way, yes
[01:00:53] <ds3> go in there and set the pinmux register and read/write to the gpio
[01:00:58] <mhaberler> it's really all about conflict detection, yes
[01:01:04] <ds3> there is no conflict
[01:01:14] <ds3> as long as you don't have another driver requesting the same gpio lines
[01:01:22] <ssi> is there a way to check if something's gpio_request()'d a pin?
[01:01:38] <mhaberler> I know, I am referring to an accident where some other driver has requested a pin
[01:01:47] <mhaberler> yes, that is the question I'd have too
[01:01:56] <ds3> in older kernels, there is a easy way
[01:02:10] <ds3> the new kernels, I don't know other then inferring it
[01:02:14] <ssi> for that matter, once we've loaded our driver, is there a way to block drivers later from requesting one
[01:02:18] <mhaberler> and from userland, if I can have a say in it..
[01:02:29] <mhaberler> exactly
[01:02:31] <ds3> ssi: yes
[01:02:46] <mhaberler> ah: we're all ears and taking notes
[01:02:53] <ds3> ssi: use sys fs to grab the pins (it does gpio_request in the backend)
[01:03:03] <ds3> it will default to an input
[01:03:14] <ds3> then don't touch the sysfs interface for those pins
[01:03:21] <ssi> ok what about this
[01:03:21] <ds3> and use /dev/mem to override it
[01:03:24] <mhaberler> and then massage the direction register once its reserved
[01:03:30] <ssi> if another driver has gpio_request()'d a pin
[01:03:36] <ssi> and then we use sysfs to essentially do the same
[01:03:37] <ssi> will it fail?
[01:03:41] <ds3> ssi: yes
[01:03:46] <mhaberler> perfect
[01:03:48] <ssi> ok well then that seems to solve both problems
[01:03:57] <ds3> /sys/class/gpio IIRC
[01:04:17] <ds3> there is an export file. you echo the GPIO number in there. if it has been requested before, that will fail
[01:04:32] <ssi> haha does this channel have a logger somewhere?
[01:04:37] <mhaberler> am I right to assume the contents of /sys/class/gpio are different between bbb and bbw because of hdmi, nand flash?
[01:04:41] <mhaberler> yes it has
[01:04:45] <ssi> gooood
[01:05:06] <ds3> different in what sense?
[01:05:23] <ssi> mhaberler: they should be identical
[01:05:24] <mhaberler> logger[mah]: foo
[01:05:25] <logger[mah]> mhaberler: Log stored at http://linuxcnc.mah.priv.at/irc/%23linuxcnc/2013-05-03.html
[01:05:27] <ssi> I think the gpios are even pinned the same
[01:05:29] <ds3> the basics are the same for all of them cuz it is the same SoC
[01:05:37] <ds3> the difference is the drivers that are active
[01:06:10] <mhaberler> so what succeeds on a bbw would fail on a bbb then because of the hdmi/nand use?
[01:06:22] <mhaberler> if you went to reserve those pins?
[01:06:25] <ssi> there are some steps you have to take to get access to those pins
[01:06:31] <ssi> the SRM talks about it
[01:06:55] <ssi> and for my uses, I think I'd prefer to keep the HDMI active, but can scrap the nand
[01:07:10] <ds3> yes
[01:07:25] <ds3> you can force those devices to be in reset
[01:07:41] <ssi> that'll take some wiring on a cape, yes?
[01:07:46] <ssi> (or airwiring)
[01:07:52] <ds3> no, in SW
[01:07:55] <ssi> ah ok
[01:07:58] <mhaberler> so it would boil down to: for each pin requested, write the number to the exports file ; if this succeeds, we can then manipulate it with /dev/mem safely?
[01:08:05] <ds3> BUT watch the electricals on those pins
[01:08:28] <ds3> mhaberler: yes. unless someone else is mucking wiht /dev/mem too in which case you are screwed
[01:08:35] <mhaberler> sure
[01:08:46] <mhaberler> idiot prevention doesnt have to be 00%
[01:08:49] <mhaberler> 100%
[01:08:54] <ds3> yep
[01:09:12] <ds3> if possible avoid the HDMI/NAND pins (GPMC and DSS)
[01:09:13] <ssi> yeah I'd say so long as we're playing nice in our own sandbox and essentially lockchecking before we poke at memory, we should be fine
[01:09:17] <ds3> they overlap a little anyways
[01:09:42] <mhaberler> is 'reserved' (ie successful write to exports) a per-process state or kernel-permanent (ie does it go away on process exit, or is it a one time thing?)
[01:10:03] <ds3> that is presistant til you reboot
[01:10:14] <ssi> so the driver needs to clean up after itself on unload too
[01:10:19] <Gamma-x> any you guys use an auto tool setter?
[01:10:19] <ds3> the write to exports creates another entry in /sys/class/gpio/
[01:10:31] <ds3> that is normally available so you can use that to set the bits via the kernel
[01:10:42] <ds3> as long as you don't touch that, you are not going to be fighting with yourself
[01:10:52] <ds3> there is an unexport too
[01:10:57] <ssi> right
[01:11:02] <ds3> main thing to watch out for is hardware fighting
[01:11:03] <ssi> I saw some writeups on that earlier
[01:11:07] <mhaberler> ah, so that would be used in rtapi_app_exit
[01:11:20] <ssi> yep
[01:11:22] <ds3> i.e. pin is reserved and set up as input cuz there is a board connected to it that is driving it
[01:11:33] <ds3> you set that to output and you are screwed
[01:11:44] <ssi> screwed how?
[01:11:55] <ds3> board potentially dying
[01:12:12] <mhaberler> electrically?
[01:12:13] <ds3> 2 signals driving each other is not a good thing
[01:12:20] <ds3> yes, electrically
[01:12:21] <mhaberler> I see
[01:12:24] <ssi> yea I could see a potential for bad stuff
[01:12:28] <ssi> input pin is high Z
[01:12:35] <ssi> output could be low Z, pulled low
[01:12:37] <ssi> could sink plenty
[01:12:44] <ds3> yes
[01:12:59] <ds3> so some people are paranoid and put a resistor in series to try to limit damage
[01:13:12] <ssi> sounds like something pcw would do :)
[01:13:49] <zultron> mhaberler, how did the hardy amd64 problem manifest itself?
[01:14:02] <zultron> Your configure logic detects the problem in el6 as well.
[01:14:02] <mhaberler> which one, the include files thing?
[01:14:05] <ds3> i should go make a board with LCD/touchscreen/DB25/level shifters on it
[01:14:09] <zultron> the 'workaronud' thing.
[01:14:10] <zultron> Yeah.
[01:14:37] <mhaberler> well look at the patched include that directory contains, I think I added a note
[01:14:39] <zultron> Compile-time or run-time error?
[01:14:44] <mhaberler> compile time
[01:14:46] <ds3> mhaberler: do you guys have the PWM output stuff working?
[01:14:49] <ssi> ds3: thanks for the brain dump
[01:14:51] <ssi> I'm having fun already
[01:14:55] <zultron> Ok, cool. Thx!
[01:15:04] <mhaberler> ds3: yes, thanks a lot for the tutorial!
[01:15:05] <ds3> ssi: np... I am hoping to make use of the result of this ;)
[01:15:13] <ssi> ds3: we all are
[01:15:23] <mhaberler> pwm… not that I know of
[01:15:25] <ssi> $45 machine control hosts would be a good thing for everyone
[01:15:40] <mhaberler> Charles has done something via the PRU, let me see
[01:15:45] <ds3> no
[01:15:49] <ds3> the PWM is seperate from the PRU
[01:16:00] <ds3> that chip has a timer unit that can generate high resolution PWM
[01:16:00] <ssi> you're talking about the PWM pinmode, yea?
[01:16:12] <ds3> I am talking about the timer module, not the pinmode
[01:16:13] <mhaberler> http://git.mah.priv.at/gitweb?p=emc2-dev.git;a=blob;f=configs/pru-examples/README.stepgen;h=3261abb80592cfb4eb964ce745394808d0061d34;hb=refs/heads/arm335x-hal-pru-module
[01:16:16] <ssi> ah
[01:16:23] <ds3> i.e. if I want to control a spindle
[01:16:31] <ds3> or modulate a laser power
[01:18:20] <ssi> mhaberler: tomorrow I'll start messing with locking/unlocking the pins with export/unexport and poking at the pinmux registers with /dev/mem
[01:18:31] <ssi> I think I actually absorbed enough of tonights conversation to be useful
[01:18:41] <ssi> my tomorrow is your this evening :)
[01:18:50] <mhaberler> excellent!
[01:19:00] <mhaberler> well yes, I'm chasing the second coffee
[01:19:18] <mhaberler> I think it's pretty much outlined?
[01:19:22] <ds3> did they yank the kernel pinmux interface?
[01:19:26] <ssi> yeah it's reasonably clear
[01:19:40] <ds3> you might be able to set up the pinmux via sysfs
[01:19:57] <ssi> ds3: where is that stuff documented? in the am335x datasheet?
[01:20:02] <ssi> I don't think they get that deep in the SRM
[01:20:17] <ssi> I mean the actual memory locations of the registers
[01:20:20] <mhaberler> I would think this applies to the rpi in a similar fashion, but then that board I leave as exercise for the determined
[01:20:33] <ssi> mhaberler: I dragged my rpi out as well... I'll mess with both
[01:20:33] <mhaberler> must be in some include file for sure
[01:20:35] <ssi> I wish I had a BBW
[01:21:08] <mhaberler> ah, having that fixed for both platforms would be great
[01:21:20] <ds3> ssi: the sysfs or the registers?
[01:21:27] <ssi> ds3: either? both? :D
[01:21:30] <ds3> that is in the TRM
[01:21:39] <ssi> that's... TI ref manual?
[01:21:43] <ds3> look in the am335x TRM... probally under the PRCM section, IIRC
[01:21:54] <ds3> yeah... that is the fat document. the datasheet is the thin document
[01:22:01] <ssi> k
[01:22:07] <ssi> got it
[01:22:10] <ds3> gpio is in a different chatper
[01:22:22] <ds3> good luck getting this info for the pi ;)
[01:22:24] <Tecan> http://tldp.org/guides.html
[01:22:34] <ssi> yea I know the pi is a black hole of closed knowledge
[01:22:36] <mhaberler> yeah if all fails, the TRM has the io adresses for the registers
[01:22:38] <ssi> that's part of the reason I'm hot for the BBB
[01:23:40] <ssi> ds3: did you see the adtrap kickstarter?
[01:23:46] <ds3> no
[01:23:48] <ssi> ok
[01:23:53] <ssi> it's an am335x board
[01:23:59] <ssi> a close friend of mine designed it
[01:24:03] <ssi> so I'm making him help me with this crap too
[01:24:07] <ds3> ah
[01:24:13] <ds3> did he bring out ALL the PRU pins?
[01:24:29] <ssi> I don't think he's using PRUs at all actually
[01:24:32] <ssi> I'll have to double check
[01:24:34] <mhaberler> nah
[01:24:37] <ds3> oh blah
[01:24:49] <ds3> I thought of doing an am335x board myself
[01:24:55] <ds3> but erratas
[01:25:02] <ssi> he has people at TI he can call and harrass
[01:25:05] <ssi> which might prove useful :)
[01:25:07] <ds3> hope he read the erratas
[01:26:05] <ssi> you weren't kidding about the TRM being fat
[01:26:11] <ds3> hehehe
[01:26:28] <ssi> pad control registers, eh
[01:26:33] <ds3> yep
[01:26:36] <ds3> that's the pinmux
[01:26:44] <ssi> yeah
[01:28:23] <ssi> ok going to bed for real
[01:28:27] <mhaberler> yes, some 3000+ pages clunker
[01:28:29] <mhaberler> cu!
[01:28:30] <ssi> talk to you all tomorrow :)
[01:30:12] <mhaberler> ssi: for completeness you might read this thread too, funny dma-based stepgen driver for the rpi: linuxcnc.org/index.php/english/forum/18-computer/20514-emc2-running-on-raspberry-pi/33082
[02:06:21] <DJ9DJ> moin
[03:02:01] <RootB_i> Guys
[03:02:07] <RootB_i> could someone tell me if my tool path is wrong?
[03:02:12] <RootB_i> My CNC went crazy milling it
[03:02:43] <archivist> define crazy
[03:03:37] <RootB_i> mm
[03:03:41] <RootB_i> My cork sheet is
[03:03:46] <RootB_i> 11 3/4 inch
[03:03:54] <RootB_i> on the circle that's suppose to pocket.
[03:03:58] <RootB_i> It began to make a crazy oval
[03:04:06] <RootB_i> that's bigger than the entire sheet
[03:12:39] <archivist> is this the first thing you have attempted to make
[03:13:07] <RootB_i> well, it's the first one i designed and took measures
[03:15:08] <archivist> have you properly measured and set your X and Y scaling factors
[03:15:58] <RootB_i> X and Y scaling factors?
[03:18:01] <RootB_i> archivist
[03:18:06] <RootB_i> here's my .dxf if you want to take a look
[03:18:07] <RootB_i> http://www.mediafire.com/?kq5t6atal94x58c
[03:18:26] <archivist> I do not want to look at a dxf
[03:18:52] <RootB_i> (._. )
[03:18:58] <archivist> that is not relevant to your machine settings
[03:20:43] <archivist> how did you set up your machine
[03:20:49] <RootB_i> mmm
[03:20:57] <RootB_i> by setup you mean..?
[03:21:08] <RootB_i> mm,step,etc?
[03:22:11] <archivist> yes
[03:38:06] <alex_joni> RootB_i: what archivist means is that you need to test if your machine travels the expected amounts
[03:38:21] <alex_joni> say g0 x0y0 - put a mark, then go x100 - measure and check
[03:38:29] <alex_joni> g0x0yo
[03:38:36] <alex_joni> then g0y100 - measure and check
[03:39:31] <archivist> also trying to ascertain his method, stepconf or hand editing or pure guesswork
[03:52:00] <alex_joni> archivist: right
[03:52:11] <alex_joni> good morrow to you, by the way
[08:04:29] <jthornton> sounds like pure guesswork to me
[08:05:07] <archivist> :)
[08:07:13] <ReadError> i just stick a dial indicator down
[08:07:58] <jthornton> you can get really close with a tape measure
[08:08:25] <ReadError> i put it in 0.01 or 0.001 steps
[08:08:27] <archivist> dti shows your crappy backlash too :)
[08:08:32] <ReadError> and just jog it
[08:08:46] <ReadError> fortunately, I have 0 backlash with rack and pinion
[08:08:51] <ReadError> its working out very well
[08:09:24] <archivist> either special were taken or you measured it wrong
[08:09:40] <archivist> special methods
[08:09:50] <skunkworks_> we used a tape measure to calibrate the k&t..
[08:09:57] <skunkworks_> ;) no not really
[08:10:25] <skunkworks_> (we used the touch probe and gauge blocks..
[08:11:38] <skunkworks_> (the lead screws are not perfect 3tpi)
[08:48:39] <carper64_lb> hi everyone hope your all having fun lol
[08:52:32] <JT-Shop> well it is raining and cold here
[08:53:53] <skunkworks> our area got a few inches of snow.. North of us got over a foot
[08:55:57] <carper64_lb> yep raining and cold hee too typical british summer lol
[08:56:46] <carper64_lb> given snow out on the hills here further north but non here
[09:07:48] <archivist> carper64_lb, another brit!
[09:33:49] <spiderdijon> warm in Bristol!
[09:34:21] <spiderdijon> well 14C :/
[09:34:55] <archivist> archivist: Forecast for Tatenhill: May 3, 2013: Chance of Rain. High 16°C (60°F). Winds 7 kph NE May 4, 2013: Scattered Clouds. High 17°C (62°F). Winds 21 kph WSW May 5, 2013: Chance of Rain. High 16°C (60°F). Winds 18 kph West May 6, 2013: Scattered Clouds. High 18°C (64°F). Winds 14 kph SW May 7, 2013: Partly Cloudy. High 21°C (69°F). Winds 10 kph South May 8, 2013: Scattered Clouds. High 21°C
[09:35:22] <archivist> must reimplement the bots weather function
[09:38:52] <archivist> I hope the northern weather remains northern so the midlands can have a fine steaming this weekend
[09:45:52] <jdh> what constitutes 'steaming' there?
[09:46:33] <archivist> http://www.middleton-leawood.org.uk/leawood/
[09:47:50] <archivist> I get to play with the levers on the engine
[09:51:47] <jdh> oh, I thought maybe the 18�C was 'steaming' for there.
[09:53:50] <L84Supper> archivist: at least you have an actual spring, 2 days ago it was 30 C, now its back to 4 C
[09:54:33] <archivist> spring is rather late this year I think
[09:56:48] <jdh> very late here also.
[09:58:05] <jdh> We had a day or two of spring a few weeks ago then it went away.
[09:59:03] <Tom_itx> jthornton, this is the flow chart for the spindle control: http://tom-itx.dyndns.org:81/~webpage/cnc/configs/sherline/Spindle.png
[09:59:47] <Tom_itx> something isn't quite right but i'm not sure just what
[10:00:37] <jdh> why so much?
[10:00:47] <spiderdijon> seasons are over rated.
[10:01:23] <Tom_itx> jdh, i was trying to allow it to be turned off during auto mode and paused but not resume until it was back on
[10:01:35] <Tom_itx> and allow it during manual mode
[10:01:47] <JT-Shop> Tom_itx: you have a spindle start and a spindle stop button?
[10:01:51] <Tom_itx> but not allow it to turn off if it was running
[10:01:52] <Tom_itx> no
[10:01:54] <Tom_itx> one button
[10:02:15] <Tom_itx> the program start button could be removed from that flow chart
[10:02:31] <Tom_itx> i was just wading thru the hal file
[10:03:02] <Tom_itx> basicall i want to allow spindle control during 'safe' conditions
[10:03:05] <JT-Shop> ok I'm reading it backwards :P
[10:03:27] <Tom_itx> i'm chinese :)
[10:03:33] <JT-Shop> lol
[10:04:15] <Tom_itx> there is some logic missing near the toggle function
[10:04:30] <Tom_itx> but i'm not sure what to put there or if it needs to be in a different place
[10:06:03] <Tom_itx> the only reason the program start button is there is because it shared the halui.mode.is-auto signal
[10:06:05] <JT-Shop> it looks like my flood toggle button on the Hardinge
[10:06:36] <Tom_itx> there is one condition it still fails in but i'll have to sit down and rethink what it was now
[10:06:52] <Tom_itx> when it fails it hangs axis
[10:07:49] <Tom_itx> i wonder if toggle2nist should be first and all the 'check' logic after it
[10:09:11] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/cnc/configs/sherline/Pause-Resume.png
[10:09:28] <Tom_itx> there's the 'pause resume' button but it's working ok
[10:09:37] <JT-Shop> put the pins in a watch window and see what is going on
[10:10:23] <Tom_itx> that's what got me that far
[10:10:28] <Tom_itx> but i ran out of time
[10:10:37] <Tom_itx> then time passed...
[10:10:42] <Tom_itx> memory lapse
[10:13:12] <carper64_lb> ia origianally from the midlands archivist
[10:14:55] <carper64_lb> but been in scotland for last 6 yrs
[10:19:39] <lando5446> Help: I've upgraded from version 2.4 to 2.5 and somehow its reverted back to using an older kinematics file. I tried recompiling and installing it, but it keeps using the older kinematics.
[10:25:36] <archivist> carper64_lb, hehe the frozen north :)
[10:29:19] <carper64_lb> yep afraid so but cant beat the salmon fishing and the gold panning up here
[10:31:01] <archivist> I used to go gold(fools) panning in Wales
[10:33:27] <carper64_lb> yep me too on th yep me too on the afon wen and mawddach
[10:34:57] <JT-Shop> Tom_itx: I watched my flood toggle in a watch window and it works as expeced and coordnated with the Axis flood button
[11:25:18] <WalterN> L84Supper: I emailed edmund optics, their f-theta lenses for 1064nm wavelength has about 5% loss with 808nm
[11:25:40] <WalterN> thats much more reasonable
[11:25:50] <JT-Shop> than?
[11:25:57] <WalterN> 12%... lol
[11:26:01] <Gamma-x> mornin aall
[11:26:04] <JT-Shop> and what the heck are you talking about
[11:26:12] <JT-Shop> run for your life....
[11:26:17] <WalterN> http://www.edmundoptics.com/lasers/laser-optics/laser-optic-assemblies/f-theta-scanning-lenses/2961
[11:26:29] <Gamma-x> lol
[11:26:32] <Gamma-x> http://www.ebay.com/itm/USED-MITSUBISHI-3-4-DIA-SHANK-FACE-MILL-END-MILL-FLY-CUTTER-MAP3500R122WA125B-/130835106487?pt=LH_DefaultDomain_0&hash=item1e76614ab7
[11:26:52] <Gamma-x> What is that used for? I have one that came in my lot of end mills...
[11:27:08] <Gamma-x> i know it says face mill but its 3/4 big....
[11:27:24] <JT-Shop> just an indexable end mill
[11:27:28] <PetefromTn> looks like a 3/4 indexable endmill
[11:27:38] <JT-Shop> he used lots of key words to get hits
[11:27:49] <JT-Shop> I never buy from people that do that
[11:29:21] <Gamma-x> JT-Shop, in what scenario would I use that?
[11:29:58] <WalterN> JT-Shop: been looking around for a while... nobody sells an f-theta lens for 808nm, they are all for 530, 1064, and 10,600nm wavelengths... so I've been asking around to see how well the 1064nm ones work with 808nm
[11:29:59] <JT-Shop> in a mill
[11:30:39] <JT-Shop> WalterN: what does the wavelength change or have to do with it?
[11:30:44] <JT-Shop> I know nothing!
[11:30:54] <WalterN> different optic materials
[11:31:13] <Gamma-x> WalterN, for laser?
[11:31:15] <JT-Shop> for different types of lasers?
[11:31:19] <WalterN> yeah
[11:31:45] <Gamma-x> well BigJohnT did go to school for astro physics...
[11:32:16] <JT-Shop> but that was when the Earth was the center of the universe...
[11:32:50] <Gamma-x> ok galileo
[11:33:01] <JT-Shop> lol
[11:34:32] <archivist> the universe moved?
[11:34:41] <JT-Shop> all the time
[11:34:46] <Gamma-x> http://www.ebay.com/itm/Lathe-Gage-tool-insert-cutter-setting-level-center-height-setter-/370592279135?pt=LH_DefaultDomain_0&hash=item5649056a5f
[11:34:50] <Gamma-x> I like that^^
[11:35:03] <mrsun> http://sphotos-e.ak.fbcdn.net/hphotos-ak-ash3/11779_10151568732563648_1851182760_n.jpg does that look "acceptable" ? :) made it in the lathe =) (toothed pulley)
[11:35:17] <WalterN> interesting
[11:35:48] <JT-Shop> could work if your material is centered up
[11:36:00] <JT-Shop> doesn't look practical to use
[11:36:19] <Gamma-x> its 10 bucks lol
[11:36:50] <mrsun> adds the guess work to the aligning of lathe tool :P
[11:37:30] <mrsun> http://www.ebay.com/itm/08-Lathe-Gage-tool-post-insert-cutter-setting-level-center-height-setter-finder-/251256171254?_trksid=p2047675.m2109&_trkparms=aid%3D555003%26algo%3DPW.CAT%26ao%3D1%26asc%3D244%26meid%3D7400327559917707873%26pid%3D100010%26prg%3D1127%26rk%3D3%26sd%3D370592279135%26 that one looked alot better
[11:37:46] <JT-Shop> #52 http://www.hardinge.com/usr/pdf/tooling/1318.pdf
[11:38:07] <JT-Shop> on page 4
[11:40:44] <L84Supper> WalterN: kewl
[11:41:54] <WalterN> L84Supper: alright... now.. what do some of these f-theta specs mean? lol
[11:42:19] <WalterN> like scan length and scan field
[11:42:35] <WalterN> http://www.edmundoptics.com/lasers/laser-optics/laser-optic-assemblies/f-theta-scanning-lenses/59961
[11:43:55] <JT-Shop> http://www.vertikatrykescanada.com/boutique.shtml?cat=ALL&search=Card_id41
[11:44:23] <JT-Shop> http://www.youtube.com/watch?v=-l6Y8PA_neY
[11:44:30] <Tecan> (-l6Y8PA_neY) "one - in bluegrass style - iron horse" by "oddis123" is "Music" - Length: 0:04:18
[11:46:04] <WalterN> meh... here.. http://www.youtube.com/watch?v=bBwa8nqtris
[11:46:07] <Tecan> (bBwa8nqtris) "eScala - Palladio (HQ) Album Version" by "XxNemoAintEmoxX" is "Music" - Length: 0:03:57
[11:48:04] <tjb1> JT-Shop: June 10th I will be working there :)
[11:49:24] <mrsun> hmm looks like the teeth became a little bit wide :/
[11:49:34] <mrsun> backlash eminent
[11:52:16] <GammaX> JT-Shop you sure are a redneck! also... That was your sign! lol
[11:52:49] <jdh> tjb1: what will you be dong?
[11:53:03] <tjb1> ME
[11:53:18] <GammaX> Your gunna be doin yourself?
[11:53:40] <tjb1> no
[11:53:45] <tjb1> manufacturing engineer
[11:54:01] <GammaX> ahhhh lol
[11:54:25] <jdh> my title is manufacturing engineer
[11:55:46] <jthornton> tjb1, make them successful
[11:55:57] <tjb1> their stock is going up ;)
[12:02:59] <PetefromTn> Coolant trough progress...http://s150.photobucket.com/user/matospeter/library/Cincinatti%20Arrow%20500%20Retrofit?sort=3&page=1
[12:03:42] <GammaX> PetefromTn just covered it in epoxy paint?
[12:03:56] <PetefromTn> GammaX: Nope...
[12:04:31] <GammaX> Loooks great
[12:05:41] <PetefromTn> they have been ground down completely on the interior, painted with some special paint that converts rust to primer, sanded down, re-roughed up again, then painted entirely on the inside with Epoxy resin from West System, finally they were scuff sanded and THEN I painted them with three coats of the industrial Epoxy white paint which is chemical,acid, and immersion proof according to the manufacturer...
[12:06:34] <GammaX> self etching primer... that stuff is amazing!
[12:06:53] <Connor> PetefromTn: Looking good!
[12:07:25] <GammaX> Im going to have to make a new lube container for my lubematic...
[12:08:02] <GammaX> cracked on all for sides and it also seems to have started to erode on the inside...
[12:08:13] <PetefromTn> Connor: thanks man, been a PIA.
[12:08:38] <Connor> Sounds like it. but, I think the epoxy stuff was the right way to go..
[12:08:56] <PetefromTn> GammaX: that sucks,,, is there a container you can repurpose for it?
[12:09:26] <PetefromTn> Connor: yeah I think so too... expensive tho... Hopefully will be making some testing chips this weekend with full flood coolant now...
[12:09:41] <Connor> Cool! :)
[12:10:07] <jthornton> I had to buy a new oil container for the C
[12:10:07] <GammaX> PetefromTn I was thinkin about just taking a block of something cheap and making one... but realistichly I dont have anything other than tupper wear....
[12:10:18] <jthornton> HNC it cost $500
[12:10:23] <PetefromTn> Connor: Yeah it is cool... I cannot wait.
[12:10:32] <jthornton> PetefromTn, looking good
[12:10:33] <PetefromTn> GammaX: doesn't it need to be metal?
[12:10:54] <PetefromTn> jthornton: Thanks man. I am pretty pleased with the results.
[12:11:02] <GammaX> PetefromTn no, the container on it is plastic.
[12:11:48] <jthornton> what brand oiler is it?
[12:11:48] <PetefromTn> jthornton: Was it stainless?
[12:11:48] <GammaX> PetefromTn Im thinkin about buying either a cheap block or alum, delron or plexi glass... w/e I can find thats about 6x6x6.
[12:11:48] <GammaX> lubematic
[12:11:48] <jthornton> no plastic
[12:11:58] <PetefromTn> oh okay...
[12:12:15] <PetefromTn> GammaX: is it a pressure cylinder or just a resorvoir?
[12:12:20] <GammaX> res
[12:12:39] <PetefromTn> well then about anything should work really.
[12:13:10] <jthornton> like this one? http://www.ebay.com/itm/Lube-Matic-Automatic-Oil-Lubrication-Pump-Reservoir-Unit-100V-Used-Warranty-/200861518562?pt=LH_DefaultDomain_0&hash=item2ec4478ae2
[12:13:13] <ssi> PetefromTn: looks great
[12:13:16] <ssi> PetefromTn: what a pile of work, eh?
[12:13:46] <GammaX> PetefromTn agreed.,.... Just wanna have it look good in the end, gotta find a cheap block of something rather than my food storage containers.
[12:13:53] <PetefromTn> ssi: yeah man it was a pain in the ass for sure. I just hope this will be a real long lasting solution...
[12:13:55] <GammaX> jthornton yup that is just about her.
[12:14:31] <IchGuckLive> hi ALL B)
[12:15:03] <PetefromTn> IchGuckLive: Hi dude..
[12:15:49] <PetefromTn> ssi: done anything on your plane today?
[12:21:27] <Gamma-x> off to the shop I go!
[12:21:28] <Gamma-x> brb
[12:21:52] <PetefromTn> well have fun LOL
[12:22:49] <IchGuckLive> dont brek to many bits
[12:24:26] <ssi> PetefromTn: haha no, not today
[12:24:35] <ssi> I probably won't go down there this weekend either
[12:24:46] <ssi> it's about 65 miles from my house, so I tend to go down and spend all weekend working on stuff
[12:24:53] <PetefromTn> well dude gotta have priorities...
[12:25:05] <ssi> but this weekend I have obligations at home tomorrow night, so I think I'm gonna stay home and rebuild my plasma table
[12:25:24] <PetefromTn> watch some youtube videos of those rv7's and you might change your mind LOL...
[12:25:36] <ssi> yeah man, it helps with the motivation for sure
[12:25:43] <PetefromTn> I'd like to have a plasma..
[12:25:46] <ssi> I have one more major purchase to make, the propeller
[12:25:59] <ssi> I should be getting paid from doing that mill conversion soon, and I think all that money is gonna go toward my prop
[12:26:11] <IchGuckLive> pete its cheep so do it
[12:26:21] <PetefromTn> do what?
[12:26:26] <IchGuckLive> lasma
[12:26:29] <IchGuckLive> p
[12:26:38] <PetefromTn> oh yeah REAL CHEAP!!
[12:26:56] <ssi> heh quit whining:
[12:26:56] <ssi> Compact hub Constant-speed prop for (I)O-360 (180/200hp) 72"
[12:26:57] <ssi> Part Number = PROP C2YR-1BFP/F74972
[12:26:57] <ssi> Price = $7160.00
[12:27:01] <PetefromTn> Honestly tho I am quite sure I could build one, I have build a lot more complex stuff...
[12:27:02] <ssi> :(
[12:27:42] <cpresser> ssi: build the prop yourself :D
[12:27:42] <PetefromTn> HOLY SMOKES those things are expensive...whats the completed plane usually cost building it yourself?
[12:27:43] <IchGuckLive> PetefromTn: just 3 axis and lots of speed
[12:27:46] <ssi> cpresser: no thanks
[12:27:58] <ssi> PetefromTn: depends on what you put in it, but $60-100k
[12:28:09] <ssi> i haven't kept a detailed accounting of all the money I've spent, but it's a lot
[12:28:19] <PetefromTn> WOW!! I was thinking maybe $40-50 or so...
[12:28:41] <ssi> in the past it could be done that cheap
[12:28:50] <ssi> it's getting hard to do it that cheap these days
[12:28:55] <ssi> metals prices are through the roof
[12:29:02] <PetefromTn> I hear ya...
[12:29:16] <PetefromTn> how much of it is in kit form?
[12:29:21] <ssi> I think the cost on an RV7 kit, just the airframe kit, is around $24k these days
[12:29:35] <PetefromTn> and is that including what exactly?
[12:29:51] <ssi> http://n999za.files.wordpress.com/2009/08/rv-7_standard_kit_lg.jpg
[12:29:54] <ssi> all that :)
[12:30:05] <ssi> it's a pretty complete kit
[12:30:20] <ssi> all the sheetmetal, weldments, canopy, fiberglass parts
[12:30:34] <PetefromTn> yeah it appears to be actually. is that all plasitc coated sheet aluminum laser cut?
[12:30:36] <ssi> just add engine, interior, paint, avionics
[12:30:44] <ssi> no its cnc punched on a trumpf punch
[12:30:52] <ssi> pretty impressive operation really
[12:30:52] <PetefromTn> yeah and a $7500.00 prop LOL
[12:30:57] <ssi> and all the holes are prepunched
[12:31:01] <ssi> so everything self-jigs
[12:31:06] <ssi> cleco it together and match drill
[12:31:19] <ssi> they've got it figured out
[12:31:32] <PetefromTn> Is a lot of it stitched together with those aircraft solid rivets?
[12:31:32] <ssi> you don't have to run a $7500 prop... fixed pitch props are $2200 brand new
[12:31:37] <ssi> yeah all of it
[12:31:42] <ssi> 10k rivets or so :)
[12:32:05] <PetefromTn> Yeah man but when you're done you have a badass fully acrobatic sweetass plane right...
[12:32:12] <ssi> yep
[12:32:16] <ssi> one of the best
[12:32:19] <PetefromTn> SICK...
[12:32:27] <ssi> it's not the top performer in any category, but it's extremely good at a lot of things
[12:32:37] <ssi> 180kt cruise easily (some people are getting as high as 188kt)
[12:32:45] <ssi> 35kt stall speed
[12:32:46] <PetefromTn> It also looks quite comfortable and can take two folks easily..
[12:32:50] <ssi> 300ft takeoff distance, 500ft landing
[12:33:00] <PetefromTn> 300'!!!!
[12:33:01] <ssi> 2500fpm climb rate
[12:33:10] <ssi> yep
[12:33:11] * cpresser just bought a few props. 2$ for 4 pieces :)
[12:33:11] <PetefromTn> what is the most common engine?
[12:33:18] <ssi> 180hp O-360
[12:33:22] <ssi> but I'm doing a 200hp IO-360
[12:33:25] <PetefromTn> who makes that?
[12:33:28] <ssi> lycoming
[12:33:34] <ssi> the 200hp is an angle valve engine
[12:33:35] <PetefromTn> $20k?
[12:33:37] <ssi> basically a hemi head
[12:33:41] <ssi> haha more like $45k new
[12:33:47] <ssi> I bought a used one
[12:33:49] <PetefromTn> JEEZ...
[12:34:02] <PetefromTn> not for the faint of wallet like me LOL...
[12:34:11] <ssi> IO-360 200HP
[12:34:11] <ssi> Part Number = EA IO-360-A1B6 RT
[12:34:12] <ssi> Price = $42085.00
[12:34:24] <ssi> that's vans's OEM discounted price
[12:34:34] <ssi> An OEM agreement with Lycoming allows Van's to sell to RV builders factory new engines at prices far below list. This agreement specifies that engines may be sold only to those who have purchased a complete RV kit. The engine must be installed in that aircraft. Only one engine may be sold per RV kit.
[12:34:56] <PetefromTn> always something right...
[12:34:59] <ssi> :)
[12:35:05] <ssi> that's ok... a "kit" consists of the tail
[12:35:08] <ssi> which is sub $2k
[12:35:16] <ssi> that gets you a builder number, and eligibility for the oem discounts
[12:35:30] <PetefromTn> that the same $24k kit?
[12:35:40] <ssi> there's four kits in an RV-7
[12:35:44] <ssi> tail, wing, fuselage, finish
[12:35:58] <ssi> the prices go up a bit every year
[12:36:07] <ssi> looks like right now it's about $21,5, or $22,5 for a nosewheel kit
[12:36:24] <ssi> $1745 tail, $6960 wing, $6170 fuse, $6635 finish for the taildragger kit
[12:36:27] <ssi> http://www.vansaircraft.com/public/kit-prices.htm
[12:36:29] <PetefromTn> whatabout the pic you posted with the complete airframe kit?
[12:36:43] <ssi> that's just all four kits spread out on the floor
[12:36:50] <PetefromTn> aah..
[12:36:56] <ssi> it's good that they break it up
[12:37:02] <ssi> I wouldn't have ever been able to buy a $20k kit
[12:37:05] <ssi> but I just bought what I could when I could
[12:37:10] <ssi> $6k at a time is easy enough to come up with
[12:37:12] <PetefromTn> yeah easier to swallow in chunks LOL
[12:37:16] <ssi> in fact the whole project has been that way
[12:37:30] <PetefromTn> How long is the average build time?
[12:37:32] <ssi> $2k tail, $6k wing, $6k fuse, $6k finish, $7k EFIS, $7k engine, $8k prop
[12:37:37] <ssi> eh it's all over the place
[12:37:45] <ssi> there's some pro guys that built one in three months
[12:37:49] <ssi> but you really gotta bust ass to do that
[12:37:54] <ssi> most people take 3-5 years I think
[12:38:09] <PetefromTn> does the kit come with all the control mechanicals?
[12:38:11] <ssi> I started in 2007
[12:38:13] <ssi> yep
[12:38:29] <ssi> the controls are simple... stick weldments, pushrods, bellcranks
[12:38:34] <ssi> rudder is cabled directly to the pedals
[12:38:48] <PetefromTn> even with the cost of metal that sounds pretty expensive for 2d parts?
[12:39:10] <PetefromTn> anyone ever built one from scratch?
[12:39:17] <ssi> the older kits were scratchbuildable
[12:39:29] <ssi> the value of the prepunching can't be overstated
[12:39:31] <PetefromTn> so you can buy plans/templates?
[12:39:32] <ssi> it makes it SO much easier and faster
[12:39:36] <ssi> yea
[12:39:42] <ssi> there's tons of scratchbuildable airplane designs
[12:39:47] <ssi> no point scratchbuilding an RV
[12:40:01] <PetefromTn> can't save money?
[12:40:09] <ssi> only if you don't value your time
[12:40:27] <ssi> scratchbuilding sheetmetal planes is very time consuming
[12:40:33] <ssi> you're better off scratchbuilding tube and fabric
[12:40:35] <mrsun> anyone got any good idea how i can radius the top corners of http://sphotos-e.ak.fbcdn.net/hphotos-ak-ash3/11779_10151568732563648_1851182760_n.jpg easily ? :)
[12:40:50] <PetefromTn> Honestly even IF I could afford to build something like that It would take so long for me to buid it making it all myself would not matter that much really LOL
[12:41:02] <ssi> mrsun: which corner?
[12:41:11] <mrsun> ssi, top corners of the teeth of the pulley
[12:41:14] <PetefromTn> sure throw it in a late and radius it..
[12:41:29] <ssi> mrsun: you mean the corners of the teeth where they fit into the belt valley?
[12:41:30] <mrsun> when i cut them they get very sharp and T5 profile says 0.4 radius on corners
[12:41:35] <PetefromTn> or do you mean each tooth?
[12:41:49] <mrsun> the top of the teeths that goes into the belts valleys
[12:41:56] <ssi> should have done that in your cnc toolpath
[12:42:00] <ssi> depending on how you're cutting them
[12:42:02] <mrsun> ssi, no cnc
[12:42:03] <PetefromTn> yeah agreed...
[12:42:08] <ssi> yea that's gonna be a bitch by hand
[12:42:10] <mrsun> doing it by hand in the lathe :P
[12:42:18] <ssi> file :)
[12:42:46] <PetefromTn> you might be able to use a triangular file between teeth pretty easily...
[12:43:00] <ssi> PetefromTn: the problem with scratchbuilding sheetmetal is you have to painstakingly lay out your rivet lines with tape measure and rivet fans
[12:43:03] <mrsun> nah it goes to deep :/
[12:43:13] <ssi> PetefromTn: and build elaborate fixtures to make sure everything is drilled and riveted straight and true and not twisted
[12:43:17] <mrsun> thinking of grinding me another tool to cut a chamfer on the teeth :P
[12:43:21] <ssi> and even with a lot of care, it's tough to build a straight airplane
[12:43:25] <ssi> and a straight airplane is a fast airplane
[12:43:28] <mrsun> i guess the radius isnt very critical
[12:43:35] <mrsun> but dont want to dig intot he belt with sharp corners
[12:43:47] <ssi> plus I think the raw materials are likely at least half the cost of the kit
[12:43:49] <ssi> probably more than half
[12:43:57] <PetefromTn> ssi: having been a professional cabinet builder for many years I am quite used to having to make fixtures and doing things the long way...
[12:43:58] <ssi> letting them kit everything for you is worth it
[12:44:11] <ssi> hell, just the hardware in the kit is probably gonna cost you $3k
[12:44:16] <ssi> the canpy is $1k
[12:44:19] <ssi> the cowl is $1k
[12:44:19] <PetefromTn> agreed...
[12:44:30] <IchGuckLive> arent ther good T5 belts in the usa available and puleys
[12:44:46] <ssi> IchGuckLive: mrsun is suffering from the same problem with pulleys that pete is suffering from with airplanes
[12:44:50] <PetefromTn> actually it is the canopy and the nose nacelles that are the parts that would be really tough to make...
[12:44:54] <ssi> the folley that it's cheaper to make it yourself :)
[12:45:09] <PetefromTn> LOL...
[12:45:12] <IchGuckLive> ist 5USD here
[12:45:46] <PetefromTn> If I had a fourth axis and CNC working properly I would make pulleys myself despite the manufactured costs..
[12:45:58] <IchGuckLive> http://maedler.de/product/1643/1616/zahnriemenraeder-t-profil
[12:46:03] <PetefromTn> Especially since you can make them any size and style you need...
[12:46:31] <ssi> PetefromTn: yea it'd be a nice thing to have the ability to do with a fourht axis
[12:46:32] <IchGuckLive> in B 10,16,25mm 1inch wide
[12:46:34] <ssi> but making them by hand is silly
[12:46:50] <ssi> considering you can buy them for ten bucks apiece max
[12:46:51] <PetefromTn> probably...
[12:47:07] <PetefromTn> actually some are quite expensive when they get larger..
[12:47:43] <ssi> reprap people obsess about printing timing pulleys
[12:47:45] <PetefromTn> besides most guys get into this to learn machining and enjoy making stuff so its all good...
[12:47:50] <ssi> and the pulleys are junky
[12:47:57] <IchGuckLive> the biggest T5 Z60 Wide 25mm is 15USD
[12:47:59] <ssi> buying $20 worth of pulleys makes those machines work so much better
[12:48:04] <PetefromTn> repraps seem to be junky anyways...
[12:48:13] <ssi> PetefromTn: like anything else, it depends on the build quality
[12:48:35] <ssi> there's a weird fundamental philosophy rampant in that particular community
[12:48:42] <PetefromTn> yeah but I have yet to see a homebuilt plastic squirter that made anything I'd want to have really....
[12:49:08] <PetefromTn> much rather have a plasma table or CNC router table...
[12:49:09] <ssi> they've been pretty useful to me
[12:49:26] <PetefromTn> Oh I am sure they have their uses..
[12:49:27] <ssi> mostly for toy stuff, but still useful
[12:49:27] <andypugh> I made my own pulleys because I needed something particularly long, but with teeth only on one end.
[12:49:42] <IchGuckLive> ballscrews need a 20/25 pulley drive to fit the steppers
[12:49:47] <ssi> http://www.youtube.com/watch?v=VsHNA3oLXC8
[12:49:50] <Tecan> (VsHNA3oLXC8) "3dof quadruped first gait" by "imcmahon" is "Tech" - Length: 0:00:30
[12:49:57] <IchGuckLive> 400 steps meets 5mm at 20/25
[12:50:34] <PetefromTn> that yours?
[12:50:45] <ssi> ya
[12:50:59] <PetefromTn> okay I stand corrected LOL...
[12:51:08] <PetefromTn> get some rubber on those damn feet...
[12:51:10] <ssi> plus I've built machines using printed parts
[12:51:32] <ssi> http://sphotos-c.ak.fbcdn.net/hphotos-ak-frc1/577596_10100133843429202_1708616203_n.jpg
[12:51:49] <ssi> naturally you're not gonna do any milling with such machines
[12:51:54] <ssi> but it's awesome for design validation
[12:51:54] <IchGuckLive> why is ther 2min of advertising
[12:52:12] <ssi> and I could easily build a laser cutter with printed parts and extrusion
[12:52:17] <ssi> IchGuckLive: because youtube blows :P
[12:52:38] <PetefromTn> WOAH wait a minute did you build that five axis?
[12:52:48] <ssi> it's not five axis, it's constrained to three
[12:52:49] <ssi> but yea I built that
[12:53:06] <ssi> I took a design someone else did called the rostock delta and designed a better frame for it
[12:53:10] <PetefromTn> those are amazing.... I love watching that move...
[12:53:27] <PetefromTn> why is it only three axes?
[12:53:43] <ssi> the arms are parallel, and they move on the carraiges together
[12:53:45] <PetefromTn> cannot the head tilt on multiple plains?
[12:53:47] <ssi> that constrains the motion to keep the platform level
[12:54:03] <ssi> in order to be able to tilt the platform, you have to be able to move the arms of one axis independently
[12:54:16] <ssi> there's no free lunch... you still need six joints to get six DOF
[12:54:47] <PetefromTn> huh, I thought if say you moved the far two axes up and the front one down the table will tilt?
[12:54:54] <ssi> no, it will stay level
[12:54:59] <PetefromTn> HOW?
[12:55:05] <ssi> http://www.youtube.com/watch?v=AYs6jASd_Ww
[12:55:12] <Tecan> (AYs6jASd_Ww) "Rostock delta robot 3D printer prototype" by "Johann Rocholl" is "Tech" - Length: 0:01:35
[12:56:07] <PetefromTn> yeah but still freakin' cool...
[12:56:23] <ssi> you can get six axis out of it, you just need to be able to move the arms on each side independently
[12:56:28] <L84Supper> ssi: what's the application for the delta?
[12:56:34] <ssi> like this:
[12:56:35] <ssi> http://www.youtube.com/watch?v=G_UmhUjZhNo
[12:56:39] <Tecan> (G_UmhUjZhNo) "LinuxCNC (EMC2) hexapod parallel robot machine tool" by "PKM077" is "Tech" - Length: 0:03:33
[12:56:52] <ssi> L84Supper: mine? nothing... it sits in my office at work and looks cool
[12:56:59] <ssi> I had planned on putting an extruder head on it
[12:57:00] <ssi> but I haven't yet
[12:57:41] <PetefromTn> what's the difference? http://www.youtube.com/watch?v=nebJ59TcYlQ
[12:57:43] <Tecan> (nebJ59TcYlQ) "CNC hexapod parallel robot 5 axis milling" by "PKM077" is "Tech" - Length: 0:00:43
[12:58:04] <ssi> the difference is six motors instead of three
[12:58:30] <PetefromTn> aah nevermind I just watched that first movie... awesome stuff...
[12:58:37] <PetefromTn> How does it avoid binding?
[12:58:52] <ssi> mine? it's got u-joints at the end of each arm
[12:58:56] <ssi> 3d printed u-joints :)
[12:59:09] <PetefromTn> smartass...
[12:59:18] <ssi> eh?
[12:59:43] <PetefromTn> that six axis one has some sweet movement, love the joints on the table as you are calling it...
[13:00:26] <PetefromTn> just kiddin man...
[13:00:40] <ssi> it's a slick machine, I'd love to build something like that
[13:00:43] <ssi> but one damn project at a time :D
[13:00:46] <ssi> well
[13:00:50] <ssi> one dozen at a time
[13:00:53] <ssi> no new ones
[13:01:01] <L84Supper> http://www.youtube.com/watch?v=W_gxLKSsSIE
[13:01:08] <Tecan> (W_gxLKSsSIE) "Robot learns to flip pancakes" by "Petar Kormushev" is "Tech" - Length: 0:01:39
[13:01:25] <ssi> haha I saw that
[13:02:19] <ssi> I love machine learning
[13:02:48] <PetefromTn> thats really cool... sensors all over the room?
[13:03:22] <ssi> yea motion capture
[13:03:52] <PetefromTn> check this one out...http://www.youtube.com/watch?NR=1&feature=endscreen&v=0-Kpv-ZOcKY
[13:03:56] <Tecan> (0-Kpv-ZOcKY) "High Speed Adept Quattro Robot handling steel balls on conveyor" by "AdeptRobots" is "Tech" - Length: 0:03:16
[13:04:20] <L84Supper> yeah the Adept and the ABB deltas
[13:05:09] <PetefromTn> wonder what the positioning accuracy is? looks pretty damn precise.
[13:05:15] <L84Supper> they usually have fun demos at most of the automation and manufacturing shows
[13:05:34] <L84Supper> they are fast but not extremely accurate
[13:05:44] <L84Supper> check the specs on the Adept
[13:06:00] <PetefromTn> I bet i could do that faster by hand.......meh maybe not LOL
[13:06:09] <L84Supper> http://www.adept.com/products/robots/parallel/quattro-s650h/technical-data
[13:06:23] <L84Supper> ±0.1 mm (Uni-directional)
[13:07:23] <PetefromTn> not bad really... better than a lot of homebuilt CNC mills LOL
[13:07:26] <L84Supper> https://www.youtube.com/watch?v=v5eR0eHknZk
[13:08:13] <PetefromTn> I see that term scara a lot what does it mean?
[13:08:19] <L84Supper> compare the delta to the scara http://www.adept.com/products/robots/scara/cobra-s600/technical-data
[13:09:16] <L84Supper> http://en.wikipedia.org/wiki/SCARA
[13:09:17] <IchGuckLive> L84Supper: get a drill bit onto it !
[13:09:41] <PetefromTn> better yet an endmill LOL
[13:10:49] <IchGuckLive> https://www.youtube.com/watch?v=v9oeOYMRvuQ&feature=endscreen&NR=1
[13:11:37] <IchGuckLive> Why a picker on pankakes let them drive thrue get it up and pack it into
[13:12:09] <PetefromTn> all I wanna know is who the hell is gonna eat all of those damn pancakes LOL
[13:12:46] <ssi> ahahaha
[13:12:52] <IchGuckLive> what a fake
[13:12:55] <IchGuckLive> https://www.youtube.com/watch?feature=endscreen&v=bbdQbyff_Sk&NR=1
[13:13:09] <L84Supper> http://www.staubli.com/en/robotics/6-axis-scara-industrial-robot/low-payload-6-axis-scara-robot/6-axis-industrial-robot-tx40/ Repeatability ±0.02 mm
[13:14:13] <IchGuckLive> in off thunderstorm around O.O
[13:14:40] <L84Supper> the same reach Scara is half that res. ±0.01 mm http://www.staubli.com/en/robotics/6-axis-scara-industrial-robot/low-payload-6-axis-scara-robot/ts40/
[13:14:42] <PetefromTn> so can LinuxCNC run both the three axis delta as well as that six axis?
[13:14:52] <L84Supper> yes
[13:15:02] <PetefromTn> linuxCNC ROCKS!!
[13:15:09] <ssi> yep
[13:15:12] <ssi> I haven't tried it yet
[13:15:12] <L84Supper> IIRC the limit is ~9 axis for the kinematics
[13:15:20] <ssi> that's part of the reason I'm working on this beagleboard project
[13:15:29] <mazafaka> there is a dubstep all around on Youtube... And it sounds like someone simply doesn't know what to play next. I am probably not alone who thinks so...
[13:15:31] <PetefromTn> whazza beagleboard?
[13:15:32] <ssi> I'd like to have something better than RAMPS/marlin for controlling machines like this
[13:15:49] <ssi> http://beagleboard.org/Products/BeagleBone%20Black
[13:16:10] <PetefromTn> likea arduino?
[13:16:18] <ssi> no, much much much more hardware than arduino
[13:16:28] <ssi> it runs linuxcnc fine
[13:16:28] <PetefromTn> minipc?
[13:16:31] <ssi> yeah
[13:16:44] <ssi> I'm working on hal drivers for running the GPIOs on the expansion ports
[13:16:49] <ssi> and there's folks working on stepgens and such
[13:16:56] <PetefromTn> what an absolutely amazing world we live in today....
[13:17:01] <ssi> our goal is to have boards that can plug into the beagle and do full machine control
[13:17:11] <PetefromTn> can it rus mesa?
[13:17:14] <ssi> ain't that the truth
[13:17:15] <PetefromTn> run
[13:17:18] <L84Supper> maximum number of joints is 9
[13:17:29] <PetefromTn> hell thats plenty no?
[13:17:31] <ssi> well there's no mesa hardware directly for it now, but PCW is already noodling on a mesa board for it
[13:17:42] <PetefromTn> PCW is DA MAN...
[13:17:43] <ssi> L84Supper: is that a kinematics limit?
[13:17:56] <ssi> ie can you write kins that have more joints?
[13:18:14] <PetefromTn> dunno shit about kinematics...
[13:18:15] <ssi> I thought I'd heard of people doing 18DOF hexapod walking robots in linuxcnc
[13:18:21] <ssi> that's another thing I'd like to explore
[13:19:03] <PetefromTn> kinda chuckle when I read about all the kins on linuxCNCspeak...
[13:19:06] <L84Supper> ssi: I never looked deeply into the code enough to see what the limits are
[13:19:54] <PetefromTn> so ya gott be a rocket scientist with 1/4 inch cokebottles to understand it?
[13:20:08] <PetefromTn> I might be able to get the cokebottles...
[13:20:46] <ssi> PetefromTn: there's a learning curve
[13:20:56] <ssi> I'm no rocket scientist
[13:20:56] <PetefromTn> that beagle is completely USB then?
[13:20:56] <L84Supper> http://www.linuxcnc.org/docs/devel/html/config/ini_config.html#_axis_lt_num_gt_section_a_id_sub_axis_section_a
[13:21:02] <skunkworks> the maximum amount of joints that the trajectory planner can deal with is 9
[13:21:05] <ssi> PetefromTn: not sure what you mean by that exactly, but no
[13:21:11] <skunkworks> In hal you can have many more...
[13:21:30] <PetefromTn> yeah but not actual axis of movement right?
[13:21:38] <ssi> an axis is a joint
[13:21:44] <PetefromTn> understand that...
[13:21:57] <skunkworks> a joint is a joint - an axis can be made up of multible joints..
[13:22:03] <skunkworks> :)
[13:22:06] <ssi> well yea
[13:22:15] <PetefromTn> don't understand that really...LOL
[13:22:21] <ssi> well think about the walking robots
[13:22:27] <ssi> they are still only 6 axis
[13:22:33] <ssi> X, Y, Z, A, B, C
[13:22:41] <ssi> but they have 12 or 18 joints
[13:22:54] <PetefromTn> so you mean PHYSICAL joints that are not intelligent?
[13:23:01] <ssi> physical things that move
[13:23:08] <skunkworks> or a hexapod.. multible joints need to move to make a strait line in the 'X' direction
[13:23:13] <ssi> you have rotational joints and translational joints
[13:23:28] <ssi> linear and rotary
[13:23:47] <PetefromTn> rotational being the u joints and translational being the linear actuators...
[13:24:18] <PetefromTn> but the rotationals are only slaves to the linears right they don't have power and feedback correct?
[13:24:19] <ssi> no in the case of my delta, there's only linear joints
[13:24:22] <ssi> the motors move things in a line
[13:24:24] <ssi> that's it
[13:24:37] <ssi> the u joints aren't considered joints
[13:24:49] <ssi> from a kins perspective anyway
[13:25:18] <PetefromTn> but in a walking hexapod the KNEES for instance are generally just dumb physical joints no?
[13:25:26] <skunkworks> Again - a hexapod - it can do motion in XYZABC with 6 linear actuators..
[13:25:56] <PetefromTn> forgive my dumbassness.....
[13:26:09] <ssi> PetefromTn: in a walker, the knee is a joint, because a motor can move it
[13:26:19] <skunkworks> I don't consider a walking robot the same as a cnc machine.. - I am talking about a machining style hexapod
[13:26:24] <PetefromTn> oh so the knee is powered and feedbacked then....
[13:26:30] <ssi> yes
[13:26:49] <PetefromTn> agreed two different animals ;p
[13:26:50] <ssi> my walker design has three degrees of freedom per leg
[13:27:08] <ssi> one is the coxa, it's rotational in the XY plane (looking top down)
[13:27:10] <ssi> that's the hip joint
[13:27:27] <ssi> one is the knee, it's rotational in the XZ plane OF THE LEG (coxa rotation will move the leg out of the XZ plane)
[13:27:42] <ssi> sorry
[13:27:49] <ssi> hip is two joints
[13:27:51] <ssi> coxa and femur
[13:28:01] <ssi> coxa rotates about Z in the xy plane
[13:28:11] <ssi> femur joint rotates about Y in the XZ plane (assuming leg is in the XZ plane)
[13:28:23] <ssi> tibia joint is the knee, it rotates about Y in the XZ plane
[13:28:40] <ssi> with those three joints, you can position the foot in any point within its envelope
[13:28:47] <PetefromTn> so how long have you been a nerd exactly LOL...
[13:28:55] <ssi> my whole damn life
[13:29:04] <PetefromTn> long live the nerds!!
[13:29:14] <ssi> I had to study long and hard to get that far with the walkers
[13:29:18] <ssi> it's not simple stuff
[13:29:24] <PetefromTn> no doubt...
[13:29:44] <ssi> also it's doubly hard for me because I'm a high-school dropout wannabe engineer with no educatino
[13:29:49] <ssi> so I have to go everywhere the long way
[13:29:51] <PetefromTn> it's a shame you don't live nextdoor to me we would get along famously LOL...
[13:30:13] <ssi> haha in terms of folks in this channel, we practically do live next door
[13:30:23] <PetefromTn> I'm the same in that regard, school of hard knocks...
[13:30:39] <PetefromTn> agreed. might have to meet up sometime..
[13:31:04] <PetefromTn> I've got my pal Connor here so it's all good he is in K-town...
[13:31:16] <PetefromTn> guess what he's a nerd too LOL...
[13:31:20] <jdh> I'll be driving by one of you next w/e
[13:31:27] <ssi> yea I'm familiar with connor :)
[13:31:29] <ssi> he's got a g0704, as do I
[13:31:32] <PetefromTn> met in person?
[13:31:33] <ssi> but he actually made progress on his
[13:31:35] <ssi> no
[13:31:35] <jdh> me too!
[13:31:38] <ssi> we've talked about it
[13:31:55] <PetefromTn> jdh: hey man who ya talkin bout?
[13:32:03] <jdh> you, or ssi
[13:32:12] <ssi> jdh: feel free to drop in and fix my HNC!
[13:32:21] <jdh> going to nashville for a wedding... via atlanta or knoxville
[13:32:22] <PetefromTn> We need to have a Pizza summit at my house LOL..
[13:32:54] <ssi> haha
[13:32:58] <PetefromTn> right after he fixed my VMC LOL...
[13:33:09] <ssi> PetefromTn: man I'm jealous of that machine
[13:33:13] <ssi> I need to find something like that
[13:33:30] <ssi> I thought about using the money from this mill conversion to buy a mill of my own
[13:33:32] <PetefromTn> they're out there that is for sure. If I run across a deal I will shout at ya...
[13:33:36] <ssi> but I need to get that prop
[13:33:40] <jdh> I40 all the way is boring, and I hate drivign through the mountains at night... the other way involves driving through atlanta
[13:33:41] <ssi> yeah please let me know if you see one
[13:33:48] <PetefromTn> that atrump is cool...
[13:33:56] <ssi> it's cool, but less automated than I'd like
[13:33:57] <ssi> and not mine
[13:34:01] <ssi> which is an important distinction :)
[13:34:16] <PetefromTn> jdh: you can come up 75 and be near me in about twenty..
[13:34:45] <PetefromTn> LOL I know that feeling. When I worked in that CNC shop running all those brand new Haas I felt like I was cheating on my wife LOL....
[13:35:04] <ssi> well I mean, I'll never even make parts on it
[13:35:08] <ssi> I'm just converting it for money
[13:35:16] <ssi> I have one more thing to do and then I get paid and go do something else :)
[13:35:31] <ssi> I wrote them a program for drilling pipe, and that's it
[13:35:39] <PetefromTn> Oh really you must know this stuff pretty well then. Perhaps you need to come up here and look over my mess LOL...
[13:35:54] <ssi> no, not really
[13:35:58] <ssi> but I'm getting pretty good at faking it
[13:36:00] <PetefromTn> How good are ya at PID tuning?
[13:36:13] <ssi> good enough that I can get andy and pcw to help me with it :D
[13:36:37] <ssi> I've tuned two servo systems now
[13:36:43] <PetefromTn> I don't have a scope other than HAl but it is working still need to get a resistor on my Z drive...
[13:36:44] <ssi> one velocity mode, one torque mode
[13:37:07] <ssi> doesn't exactly make me an expert
[13:37:07] <PetefromTn> mine's all velocity but PCW said it might be easier to tune the Z in torque since it is not counterweighted.
[13:37:22] <PetefromTn> well I have NEVER tuned one so you got one up on me LOL...
[13:37:43] <PetefromTn> Even tho it is working I honestly STILL have never tuned one LOL...
[13:37:49] <ssi> so here's everything I know about tuning
[13:37:59] <PetefromTn> Watching Magnum PI...
[13:38:24] <ssi> velocity mode: raise P til it oscillates, back it off and add some D to stabilize. Scope it, and you'll observe an error spike while moving the axis... add FF1 to neutralize
[13:38:41] <PetefromTn> Some hot babes on there....
[13:38:48] <ssi> torque mode: raise D till it oscillates, back off and add P until it oscillates, back off
[13:39:02] <ssi> scope it and observe error while accellerating, add FF2 to neutralize
[13:39:03] <PetefromTn> yup that is kinda what we did but found that messing with FF1 etc worked better than D....
[13:39:04] <ssi> that's about it
[13:39:19] <ssi> D should allow you to run more P stably
[13:40:22] <PetefromTn> only issue is when you are moving towards oscillation on the Z axis with that HUGE millhead it is pretty nerve wracking and downright scary when it does it. Sounds like the machine will tear itself apart ROFL...
[13:40:40] <ssi> yes, it's scary
[13:40:50] <PetefromTn> Oh and no lower limit switch makes it even more fun...
[13:40:54] <ssi> haha
[13:42:07] <PetefromTn> I think I still need to play with it because the Z never really quiets down ever. It is not whining bad but it is always doing something kinda reminds me of the servos on my RF45 DC with geckos, kinda dithering constantly when at rest...
[13:42:49] <PetefromTn> I'd love it to be whisper quiet when the head is not in motion but that may actually be impossible with no counterweight...
[13:43:05] <ssi> you have an RF45 too?
[13:43:56] <tjtr33> yow! mouser delivered BBB in < 1 week
[13:44:07] <ssi> yea I got mine quick
[13:44:15] <PetefromTn> HAD... Sold it to purchase this monster...
[13:44:16] <ssi> I have a second one coming from adafruit too... they got a restock
[13:44:16] <jdh> what do you do with them?
[13:44:30] <PetefromTn> whazza BBB?
[13:44:35] <ssi> PetefromTn: daww I was gonna say, maybe I'll buy your RF45 for a song now that it's taking up space :)
[13:44:38] <ssi> PetefromTn: the beaglebone black
[13:44:48] <PetefromTn> aah...
[13:45:47] <PetefromTn> bought my RF45 for $1500 new over a decade ago from Lathemaster.com. Retrofitted it with DC servos and did one of the first belt drive conversions ever, used it to learn on and then sold it for almost $6k about a year ago....
[13:46:03] <PetefromTn> Not a bad trip really...
[13:46:08] <ssi> yeah that's pretty good
[13:46:17] <PetefromTn> RF45 is a decent machine once CNC'd
[13:46:21] <ssi> yeah
[13:46:30] <ssi> the g0704 ought to be an ok machine cncd
[13:46:32] <ssi> I have DC servos for it
[13:46:37] <ssi> and AMC drives
[13:46:38] <PetefromTn> Nothing like this bad boy tho...
[13:46:55] <PetefromTn> yeah the 0704 is decent Connor has a nice one...
[13:47:04] <jdh> send me the drives and servos, I'll try them out on mine.
[13:47:26] <ssi> I've already got ballscrews for it... and it's disassembled
[13:47:30] <ssi> I just need to get off my ass and finish it
[13:47:32] <PetefromTn> I'd look seriously at the DMM servo systems for any of these smaller benchtop mills...excellent stuff..
[13:47:34] <jdh> that's the hard part
[13:48:09] <ssi> PetefromTn: what's the cost?
[13:48:13] <PetefromTn> If I were to do it again I would find an IH clone and do DMM servos...
[13:49:05] <ssi> the AMC drives were cheap.. I paid about $124 for three of them
[13:49:13] <ssi> the servos were around $60 apiece, with encoders
[13:49:18] <PetefromTn> reasonable and honeslty when I added it all up on my geckos, motors, cabling, encoders, BOB, power supply, etc etc etc I really would have saved some cash and been done a lot sooner with less headache had the DMM been available when I stared...
[13:49:29] <ssi> mesa 7i77 kit is $240
[13:49:43] <jdh> ssi: where did you get the motors?
[13:49:44] <ssi> all I need is a decent 60v power supply and I'm good to go
[13:49:46] <ssi> jdh: ebay :D
[13:50:11] <PetefromTn> bought mine brand new from a guy on the zone..
[13:50:28] <ssi> I also wouldn't screw with gecko for servos
[13:50:37] <ssi> I'm not crazy about step/dir servo drives
[13:50:42] <PetefromTn> well I get to go pick up my kids here in a bit...
[13:51:32] <PetefromTn> yeah thats what I am saying.... AC servos are awesome. You don't have to buy thier complete kit either prolly could just buy the motors and drives and use mesa...
[13:52:00] <ssi> yea but what's the price?
[13:53:07] <PetefromTn> http://dmm-tech.com/products_main.html
[13:54:05] <ssi> not terrible, but more spendy than I'd like
[13:54:40] <PetefromTn> oops sorry http://dmm-tech.com/pricing.html
[13:54:46] <jdh> how big of a servo do you need for a g0704 ?
[13:54:47] <ssi> yea I found it
[13:54:57] <ssi> jdh: probably not all that big
[13:55:14] <PetefromTn> they have package deals that include everything you need for discounts. Nice people too. might call em...
[13:55:16] <ssi> one thing I'm not real clear about is the relationship between stepper torque and servo torque
[13:55:26] <ssi> the servos I have are much less rated torque than I'd expect
[13:56:20] <ssi> I guess cause the speed
[13:56:21] <PetefromTn> cool my wife is home and wants to pickup the kids LOL
[13:56:23] <ssi> yea that makes sense
[13:56:29] <ssi> http://www.ebay.com/itm/MCG-Servo-Motor-12384246NC-Z1XX-170VDC-6000-RPM-83OZ-IN-2-93AMP-/150831468008?pt=LH_DefaultDomain_0&hash=item231e418de8
[13:56:30] <skunkworks> the steppers torque rating is at 0 rpm
[13:56:33] <ssi> that for instance
[13:56:47] <ssi> 170V, 3.93A is a 500W motor
[13:56:52] <ssi> that's like 2/3hp
[13:57:01] <ssi> but it's 83oz
[13:57:04] <andypugh> The 9-axis limit is mainly because there are no G-code letters left to use for axes. There is no reason to have any limit on the number of _joints_
[13:57:04] <ssi> probably at 6000rpm
[13:57:08] <ssi> so I guess you need to gear it
[13:57:13] <ssi> or potentially need to gear it
[13:57:23] <ssi> andypugh: ahh ok
[13:57:29] <PetefromTn> thats a 6k RPM motor...
[13:57:31] <ssi> andypugh: so controlling an 18dof robot might be possible
[13:57:34] <ssi> PetefromTn: right
[13:57:39] <ssi> PetefromTn: it's also a high voltage servo
[13:57:40] <andypugh> I don't know how many joints the joints_axes3 branch allows.
[13:57:54] <PetefromTn> did you already buy these?
[13:58:00] <ssi> no those arent' exactly what I have
[13:58:06] <ssi> mine are the same brand, but they're 60v servos
[13:58:14] <PetefromTn> same other specs?
[13:58:18] <ssi> no, not really
[13:58:23] <ssi> I was just trying to find an example :)
[13:58:31] <ssi> these are kind anice though cause they're nema23
[13:58:34] <ssi> the ones I have aren't
[13:58:35] <PetefromTn> already bought em?
[13:58:41] <ssi> yea a year ago
[13:59:11] <ssi> these 170v ones would actually be pretty good with the AMC 200V brushless drives
[13:59:15] <ssi> like the ones I just used on the atrump
[13:59:21] <ssi> and I like that they come with CPCs on them
[13:59:37] <ssi> maybe I'll buy some
[13:59:51] <PetefromTn> you need nema23?
[13:59:59] <ssi> nah that'd mean I'd need more drives and a harder to get psu
[14:00:06] <ssi> don't need nema23, but it's convenient
[14:00:15] <PetefromTn> whatya got now?
[14:00:23] <ssi> one of my servos is exactly like this:
[14:00:23] <ssi> http://www.ebay.com/itm/MCG-60VDC-5300RPM-Servo-Motor-2234-ME3563-2-with-Encoder-/251257349029?pt=LH_DefaultDomain_0&hash=item3a801ae3a5
[14:00:28] <PetefromTn> 34's?
[14:00:32] <ssi> and that isn't a flange, it's some kind of custom block that's screwed to the end
[14:00:52] <ssi> 50oz-in, 60VDC, 3.5A
[14:01:01] <ssi> 5300rpm
[14:01:51] <mrsun> Lin Engineering 4118 High Super Torque Stepper Motor with CUI Encoder <-- cant go wrong with a high super torque stepper! :P
[14:01:54] <PetefromTn> Honestly man I know you already bought them but I would SERIOUSLY consider buying a package from someone like DMM that comes with brand new motors, plug and play everything, saves you so much damn time is not funny...
[14:02:15] <ssi> PetefromTn: I've got enough experience with these systems at this point that it won't be an issue
[14:02:36] <ssi> I don't have $1k+ to throw at some company to prebake me a package
[14:02:42] <PetefromTn> agreed but no matter your experience soldering, shielding, shrink wrapping takes a bunch of time...
[14:03:01] <ssi> doesn't matter
[14:03:02] <PetefromTn> and the components are designed to work together...
[14:03:04] <ssi> i have all the tools
[14:03:14] <PetefromTn> okay man just saying...
[14:03:18] <ssi> http://sphotos-b.xx.fbcdn.net/hphotos-ash3/944683_10100131793367542_829866394_n.jpg
[14:03:21] <ssi> you saw that right?
[14:03:24] <PetefromTn> I have all the tools too and it was still a beotch...
[14:03:30] <ssi> I like wiring and lacing and such :)
[14:04:02] <PetefromTn> very nice man, you need to come over and tidy up my junk...
[14:04:14] <jdh> you might re-phrase that.
[14:04:16] <PetefromTn> CNC junk of course...LLOL
[14:04:17] <ssi> hahah
[14:04:30] <ssi> MCG SERVO MOTOR WITH ORIENTAL GEAR HEAD 4GN180KA
[14:04:31] <ssi> hahah
[14:04:46] <PetefromTn> those digital stepper drives?
[14:04:55] <ssi> servo drives
[14:05:10] <PetefromTn> keling?
[14:05:12] <ssi> AMC
[14:05:30] <ssi> http://www.a-m-c.com/download/datasheet/bx15a20.pdf
[14:05:30] <jdh> doesn't look like enough wires?
[14:05:39] <PetefromTn> still man it's a PIA...
[14:05:43] <ssi> enough for what? :)
[14:05:54] <jdh> for the drives
[14:05:56] <ssi> PetefromTn: compared to wiring an airplane, this stuff is cake
[14:06:15] <ssi> jdh: +10V/GND, ENA+/ENA-, motor +/-, HV +/-
[14:06:19] <ssi> what more do you need?
[14:06:27] <PetefromTn> Then I guess I won't be building no damned airplanes then....
[14:06:48] <jdh> oh, we have lots of AMC drives, but all with hall switches and <something else>
[14:07:01] <ssi> they're running brushed dc motors in current mode
[14:07:07] <ssi> current feedback is internal to the drive
[14:07:27] <ssi> motor interface is just motor power and 8 pin encoder
[14:07:31] <ssi> encoder is going straight to the 7i77
[14:07:55] <PetefromTn> woah just noticed they are 190v...NICE
[14:08:28] <ssi> yep
[14:08:36] <ssi> I had to get those because the servos on the atrump were 140V
[14:08:50] <PetefromTn> so they can run what 1500 watts or so?
[14:08:51] <ssi> the AMC drives I have at home for the 704 are 30A8T
[14:08:55] <ssi> 30A, 80V
[14:09:06] <ssi> those little ones are 15A 200V
[14:09:06] <ssi> so 3kw
[14:09:20] <ssi> and the 30A8T are 30A, 80V, 2.4kw
[14:10:13] <PetefromTn> well I guess if you are the master wiring and looming guru and it floats your boat you are better off building it yourself, Me I hate that shit...
[14:10:25] <ssi> yeah I actually enjoy it
[14:10:41] <ssi> it's a meticulous fussy sort of work, but it's satisfying to me
[14:10:50] <tjtr33> very nice lacing ( lost art since tiewraps ) but my goto tool for debugging busted machine w/o prints is: tear the loom apart :)
[14:10:54] <PetefromTn> honestly these commercial style AC servos were kinda fun to install
[14:11:10] <ssi> tjtr33: I am nice so I made them a wirebook for it :)
[14:11:35] <ssi> and I do lacing because of the aviation work
[14:11:35] <tjtr33> numbered wires! how modern
[14:11:37] <ssi> tie wraps are too bulky and heavy
[14:11:54] <ssi> they don't pull through bulkheads
[14:11:55] <ssi> etc
[14:12:04] <PetefromTn> that picture looks like a geeks wet dream....
[14:12:35] <ssi> i'm obviously proud of it :P
[14:12:50] <PetefromTn> you should be LOL...
[14:13:10] <PetefromTn> none of those wires look shielded no?
[14:13:15] <ssi> no, none of them are
[14:13:25] <PetefromTn> whazzabout noise?
[14:14:08] <ssi> I would like to go back and do shielded 2twisted for the +/-10v lines
[14:14:08] <ssi> but so far there haven't been any issues
[14:14:08] <ssi> encoder lines are differential, and I think they're ok as is
[14:14:08] <ssi> although they might ought to have been twisted in the loom
[14:14:08] <PetefromTn> diff enc's rock...
[14:15:05] <PetefromTn> I need to wire up my new encoder for my spindle motor and dunno how best to do it yet. Connor was working on some kinda board to mask the 180 out sensor for my 2-1 belt drive.
[14:51:03] <andypugh> Does anyone have an IKEA Janso lamp, and a multimeter?
[14:51:48] <andypugh> I was surprised (in fact literally shocked) to find that the 4V outputs have 100V AC relative to earth.
[14:51:57] <Tom_itx> what's special about that lamp?
[14:52:20] <andypugh> The lamp is small, bright, and very cheap. It makes a very good machine light.
[14:52:24] <Tom_itx> ahh
[14:52:26] <Tom_itx> nice
[14:52:31] <Tom_itx> led?
[14:52:40] <andypugh> It also seems to have a very "special" sort of 4V wall-wart.
[14:52:53] <Tom_itx> floating gnd?
[14:53:05] <andypugh> Bear in mind that 100V to earth is pretty strange in a 240V AC country.
[14:53:13] <Tom_itx> maybe it just has 1 diode in it
[14:53:50] <Tom_itx> so it's not the lamp you got for the curly cord?
[14:53:52] <skunkworks> capasitively coupled oddness?
[14:53:55] <andypugh> It has to be floating gnd, as there is no earth terminal on the PSU. But I wouldn't be expecting an electric shock from it.
[14:53:58] <cradek> is it smooth DC-ish?
[14:54:11] <andypugh> 100V AC
[14:54:23] <Gamma-X> hey guys, im homing all my steppers and when the x starts to home it hits the limit switch and just faults, I have all limit switches setup to be hom and limit. THoughts?
[14:54:27] <andypugh> Smooth 4V DC
[14:54:38] <Tom_itx> take it apart and see
[14:55:16] <andypugh> I have three. Two are already hacksawed out of their cases, rewired, put in a project-box and mounted inside my machine frame.
[14:55:24] <cradek> I predict something like http://www.datasheetcatalog.org/datasheet/stmicroelectronics/6941.pdf
[14:55:27] <skunkworks> Gamma-X, I think you need to do some hal magic to make it so the other axis don't see the limit switch trip.
[14:55:41] <cradek> (these are an awesome idea if you don't need isolation)
[14:55:47] <andypugh> One is totally unmolested and on my bedside table. All three show the same "feature"
[14:56:35] <Tom_itx> cradek i doubt they'd put something so 'expensive' in a lamp walwart
[14:56:45] <andypugh> cradek: I think they are a little SMPS, there are caps and ICs and a small transformer in there.
[14:57:07] <Tom_itx> but then who can say what china can put out for a buck
[14:57:41] <andypugh> Gamma-X: HOME_IGNORE_LIMITS http://www.linuxcnc.org/docs/2.4/html/config_ini_homing.html
[14:59:00] <Tom_itx> andy, what model lamp is it?
[14:59:10] <andypugh> IKEA Jansjö
[14:59:41] <andypugh> http://www.ikea.com/us/en/catalog/categories/series/18294/
[15:00:29] <Tom_itx> http://www.ikea.com/us/en/catalog/products/20169658/
[15:00:33] <Tom_itx> different base
[15:00:47] <andypugh> That's the exact one I am using.
[15:01:10] <Gamma-X> andypugh, skunkworks well I have the same setup on all axis, first it homes the z and that works great, then it homes the x and it faults when it hits the switch..
[15:01:28] <skunkworks> this makes me cry http://evworld.com/journal.cfm?jid=69
[15:01:37] <andypugh> Gamma-X: You need the flag in every axis.
[15:01:53] <Gamma-X> I do
[15:02:00] <Gamma-X> There all setup the same from step config.
[15:02:23] <Gamma-X> home+limit
[15:04:32] <andypugh> Are the already-homed axes off the limit before next one homes?
[15:04:42] <Gamma-X> yes it backs off of it
[15:05:39] <Gamma-X> wait i may be mistaken
[15:06:31] <Gamma-X> no i was right.
[15:07:53] <andypugh> Does me saying "How strange" help at all? (I thought not)
[15:08:35] <andypugh> When if faults, which axis is homing, and which axis faults?
[15:08:43] <Gamma-X> yeah i dk lol
[15:08:57] <Gamma-X> the x is homing
[15:08:57] <Gamma-X> and it faults on 1
[15:09:01] <skunkworks> Gamma-X, how are they setup? how many pins used and how many wires?
[15:09:06] <Gamma-X> joint 1
[15:09:25] <Gamma-X> closes config
[15:09:28] <Gamma-X> 2 wires directly into a bob
[15:09:34] <andypugh> OK, so it is moving X and seeing the Y limit. You may have your limit switches crossed.
[15:10:53] <Gamma-X> joint 1 is y?
[15:11:06] <andypugh> Yes
[15:11:37] <andypugh> Joint 0 is X. (typically, not absolutely written in stone)
[15:14:03] <Gamma-X> andypugh, your the man
[15:14:11] <Gamma-X> somehow gpot x and y confgused...
[15:14:18] <cradek> I don't think that's right
[15:14:29] <cradek> you said you've wired all your limit switches together
[15:14:50] <cradek> so X bonking into the switch makes axis.0.whatever-limit, axis.1.whatever-limit, and axis.2.whatever-limit ALL go true
[15:14:55] <Gamma-X> I need some better freaking couplers
[15:15:26] <Gamma-X> I bought delrin couplers hoping they would be legit and there deff not.... If I over tighten them they break and if there too loose then they slip as they are now on my y
[15:15:30] <cradek> you've presumably got it set for ignoring limits during homing
[15:15:45] <cradek> but that means when X is homing only X's limits (axis.0.whatever-limit) is ignored
[15:15:58] <Gamma-X> its fixed now.
[15:17:01] <Gamma-X> semi...
[15:17:01] <cradek> oh they're not wired together?
[15:17:03] <cradek> can you say again clearly what your hardware setup is
[15:17:17] <Gamma-X> rf45
[15:17:21] <cradek> and what exact error you're getting in what exact situation
[15:17:26] <Gamma-X> well its fixed now
[15:17:34] <Gamma-X> rf45 with steppers.
[15:17:59] <Gamma-X> directly mounted, had some wiring messed up and had to set opposit up for homing off the limit switches.
[15:19:05] <Gamma-X> nowe I just need to put in the correct parameters for the table length and such
[15:20:17] <Gamma-X> and should be good to go, im also getting some studdering in my y which im not sure exactly what to put in for that values... x and y are both kelingers 570 ounce hooked up to ma860h and the z is a 1200oz hooked to the same
[15:37:17] <Gamma-X> linear move on line 14 would exceed joint 2's negative limit.....
[15:51:37] <DJ9DJ> gn8
[15:58:36] <JT-Shop> this should be easy for me but I'm lost. I do a G0 Z0.375 and it moves to Z0.210...
[16:00:47] <JT-Shop> dang Z tool offset
[16:02:52] <JT-Shop> the preview tab shows Z0.210 and the DRO tab shows Z-0.70
[16:03:08] <cradek> that doesn't smell right
[16:03:24] <cradek> this isn't 2.5 is it?
[16:03:38] <JT-Shop> yes
[16:03:48] <JT-Shop> 2.5.2
[16:04:22] <cradek> ugh
[16:04:46] <cradek> is it in relative mode in the menu?
[16:04:54] <JT-Shop> each time I start the scrip it increments the Z axis by -0.010 from the commanded position
[16:05:02] <ds3> g92 active?
[16:05:36] <JT-Shop> show relative is checked
[16:06:12] <JT-Shop> g92 on the DRO page both show 0.0000
[16:09:18] <ds3> blah sorry not g92
[16:09:28] <ds3> g91 I mean
[16:11:48] <ssi> I recall reading some rumblings that there were significant fixes in 2.5.1 for 5i25 or perhaps 7i77...
[16:12:02] <ssi> I'm wondering if I need to get that machine updated
[16:12:06] <ssi> it has no network and is in a shop :/
[16:12:07] <JT-Shop> ds3: no it is in G90
[16:17:26] <Gamma-X> I got bad connection, dont know if this went through, but when running the default linux cnc engraving program I get program exceeds machine minimum on axis z. WHat is the probab;ly cause of this?
[16:18:16] <andypugh> ssi: I don't see anything big here; http://wiki.linuxcnc.org/cgi-bin/wiki.pl?Released_2.5.X
[16:18:54] <ssi> maybe i misconstrued a reference to this:
[16:18:54] <ssi> Hostmot2: Add support for 6i25
[16:18:54] <JT-Shop> Gamma-X: did you read the header in that file? I bet the answer is no
[16:19:22] <PCW> only updates for 5i25/7i77 are for pncconf
[16:19:22] <andypugh> I think that just involved spotting that it was a 6i25 not a 5i25.
[16:19:31] <ssi> ah ok
[16:19:40] <PCW> and most 6I25s lie anyway
[16:19:46] <ssi> I downloaded pncconf suppotr stuff directly from the forum
[16:19:49] <ssi> cmorley's stuff I think
[16:19:59] <PCW> (say they are 5I25s)
[16:19:59] <ssi> ok good I won't stress about getting it updated then
[16:20:18] <ssi> I'll have my second 5i25 tomorrow hopefully :D
[16:20:21] <andypugh> Gamma-X: You need to touch-off the Z or it won't run. I seem to be in the minority in not thinking that this is a good feature.
[16:20:39] <ssi> I'm gonna run it directly with a G540 and see how that works
[16:21:17] <Gamma-X> andypugh, ahhhh ok
[16:21:44] <JT-Shop> andypugh: not really
[16:21:55] <JT-Shop> in the minority I mean
[16:21:57] <PCW> theres someone on the forum that has a config file that does the magic to get the G540 charge pump working
[16:22:34] <PCW> I don't think pncconf has the bolierplate for that
[16:22:43] <ssi> good to know
[16:23:33] <andypugh> OK, lets see what the lathe did after I dived indoors to save the quiche from immolation.
[16:23:57] <ssi> lol
[16:24:04] <PCW> the 5I25 G540 config has a spare stepgen that's used to generate the chargepump squarewave
[16:24:17] <ssi> oh neat
[16:24:20] <ds3> so you failed to make food grade charcoal?
[16:24:37] <ssi> I don't fully understand how the chargepump works
[16:24:42] <ssi> I guess it's like a capacitor based voltage doubler?
[16:24:56] <ds3> THE charge pump? where?
[16:25:26] <ssi> well, in a G540 for instance
[16:25:26] <ssi> or elsewhere :)
[16:25:45] <Gamma-X> for the z axis config, the table travel I have 0 to 10. should the 10 be negative?
[16:25:46] <Gamma-X> or should it be 10 to 0?
[16:25:48] <ssi> I guess it's more accurate to say that a voltage doubler can be made from a charge pump
[16:27:18] <ssi> does the G540 actually have a charge pump on it, or is the charge pump some holdover from parallel port specs that the G540 uses as a watchdog?
[16:27:24] <ssi> the docs dont really make that clear
[16:27:39] <JT-Shop> Gecko docs?
[16:27:42] <ssi> ya
[16:27:46] <PCW> Dont know how its implemented
[16:27:49] <ssi> NOTE: If you are not using Mach3 and do not have access to a charge pump, you can disable this option. If you remove the cover of the REV1 G540 there is a switch located on the underside near the 12- position header; this switch is accessible through a cutout on REV3 and above G540s. Move the slider towards the edge of the board or where is says “OFF” to disable the charge pump. Push the slider in or slide it to where it says “ON” to enable this
[16:27:54] <JT-Shop> it is on the G540
[16:27:59] <ssi> The G540 will not come out of FAULT mode unless there is a charge pump (watchdog timer) signal present. The charge pump sends a 10kHz signal to the G540 on pin 16, and must be enabled for the indicator LED to come out of FAULT mode.
[16:28:10] <ssi> that makes it sound like it just uses that as a watchdog
[16:29:00] <PCW> Yes its a watchdog (which may or may not be implemented as a charge pump)
[16:29:24] <ssi> oh ok
[16:29:30] <ssi> I thought maybe it needed the higher voltages for someting
[16:29:38] <ssi> but why would it... it's got +50v on the main power pins
[16:31:47] <JT-Shop> http://www.youtube.com/watch?v=wYwb-O27qv0
[16:31:50] <Tecan> (wYwb-O27qv0) "I'll Never Smoke Weed With Willie Again" by "Trevor Vacon" is "Music" - Length: 0:02:42
[16:32:36] <ds3> you can also invert voltages with a charge pump
[16:33:13] <PCW> A 2 diode 'voltage doubler' will not double a unipolar pulse train (like the signal from the parallel port) but it does
[16:33:14] <PCW> require the input to toggle to generate an output voltage
[16:33:27] <ssi> gotcha
[16:39:54] <JT-Shop> it seems to be something with named variables...
[16:42:35] <jdh> I've read that the g540 watchdog doesn't normally work with linuxcnc (but does with mach)
[16:43:23] <PCW> That's just because you need to set the parallel in EPP mode for it to work
[16:43:33] <PCW> no bigge
[16:47:09] <PCW> It probably arguable the LInuxCNC should always set the parallel ports into EPP mode
[16:47:57] <jdh> I've also read that EPP mode wasn't always enough.
[16:48:17] <jdh> but, I read some really weird things on the interwebbes
[16:49:14] <PCW> Nope its just that Gecko chose a pin that open collector in normal PP mode
[16:49:21] <jdh> http://www.mail-archive.com/emc-users@lists.sourceforge.net/msg35125.html
[16:51:14] <PCW> Yep but if EPP mode doesnt do it mach wont either
[16:52:44] <PCW> I don't think its that badly designed, but it does need the driver to be push-pull no open collector with a weak pullup (std mode)
[16:53:20] <PCW> and with PCI cards it may be a puzzle to set them into EPP mode
[16:53:49] <jdh> I ran across some linux utility that made it simple.
[16:54:41] <PCW> on some cards...
[16:55:28] <Jymmm> dd if=/dev/zero 0f=/dev/sda bs=512
[16:55:29] <jdh> I think it made one of my pci pport cards with with my 7i43, but not the other
[16:57:50] <PCW> jymmm is spreading a unix virus
[16:59:05] <Jymmm> PCW: Nuh uh... If I were, I would not have injected the "typo injected for the safety of the truly stupid that randomly run nix commands seen on irc" =)
[17:06:25] <JT-Shop> network is back up now to figure out why a mouse won't work on the ubuntu computer... neither USB or PS2 mice work
[17:17:04] <PCW> needs /dev/cheese
[17:20:42] <JT-Shop> I pulled out the LiveCD and the mouse started to work
[17:21:04] <JT-Shop> now I can't reproduce the problem I'm having on the Hardinge with the ngcgui sim :(
[17:26:24] <Jymmm> sudo rm -rf /dev/StinkingRatUser
[17:28:08] <JT-Shop> it's messing up now :)(
[17:38:39] <JT-Shop> sometimes
[17:38:57] <Jymmm> PEBKAC
[17:39:08] <JT-Shop> two differnent chairs
[17:39:14] <Jymmm> lol
[17:39:56] <JT-Shop> what makes my eye phone cause noise on my speakers
[17:40:06] <Jymmm> RFI
[17:40:26] <Jymmm> your speakers are not shielded well enough
[17:40:28] <JT-Shop> it is annoying, do I need a tin foil shield
[17:40:42] <Jymmm> that might help
[17:53:53] <JT-Shop> somehow when the file is loaded with a G10 L20 it messes up and goes too far on Z but if the G10 L20 is allowed to complete everything is ok
[17:57:25] <andypugh> I put something on eBay, 1p start, no reserve. Someone asked what the least I would accept was. Then it started to get silly: http://contact.ebay.co.uk/ws/eBayISAPI.dll?ShowAllQuestions&iid=111061250684&requested=andypugh&redirect=0&frm=2047675&ShowASQAlways=1&ssPageName=PageAskSellerQuestion_VI
[17:58:59] <JT-Shop> that is too funny
[18:06:07] <JT-Shop> R1 is a bike?
[18:10:27] <JT-Shop> nevermind
[18:10:51] <JT-Shop> andypugh: how did you bend the frame?
[18:15:54] <andypugh> Track day crash
[18:16:06] <JT-Shop> ouch
[18:17:14] <JT-Shop> http://www.youtube.com/watch?v=mieS87TpNS0
[18:17:16] <Tecan> (mieS87TpNS0) "motorcycle crash May 2nd 2013" by "kyle anderson" is "Autos" - Length: 0:01:51
[18:17:49] <JT-Shop> I guess he had no idea how fast the GL1800 will stop
[18:19:13] <JT-Shop> amazing balance for a motorcycle with no driver
[18:21:34] <r00t4rd3d> excavator: http://i.imgur.com/amPSmUb.jpg
[18:22:36] <JT-Shop> cool, can't tell the brand by the lack of color :)
[18:23:02] <JT-Shop> a mono track?
[18:28:00] <r00t4rd3d> its going to sit on top of a 3/4 board
[18:28:46] <r00t4rd3d> for construction sign
[18:29:10] <JT-Shop> pretty cool
[18:29:45] <r00t4rd3d> its just a mini rough draft to see how my design turned ou t
[18:30:12] <r00t4rd3d> i made it from scratch
[18:36:46] * JT-Shop goes to make pizza from scratch
[18:41:20] <PetefromTn> ...
[18:44:14] <PetefromTn> anyone watching the solar impulse flight progress?
[18:49:09] <PCW> Wasted 1/2 a day chasing a FPGA bug that turned out to be a single missing 'a'
[18:50:03] <PetefromTn> man that sucks LOL... however after watching Connor play with those wordings and seeing how a simple typo can halt the works I am not surprised to hear it...
[18:52:33] <PCW> If it halted things instead of causing obscure bugs when operating, It would have been a lot better
[18:53:53] <PetefromTn> agreed, I guess that one slipped by huh LOL Im sorry for you man its no fun.
[18:55:50] <PCW> Luckily I don't think there many copies in the wild except for maybe a few custom made bitfiles
[18:56:40] <Connor> PCW Which cards would be affected ?
[18:57:00] <PetefromTn> THE WILD...love that. is it true that the beagle board stuff is something you are working on mesa gear for?
[18:57:01] <PCW> All sserial V41
[18:57:18] <PCW> only made a few for specials
[18:58:13] <PCW> I am working on a cape
[18:59:33] <PetefromTn> you gonna be a superman?
[19:00:16] <PCW> Well it will take a while (getiin old...)
[19:02:20] <PetefromTn> Aw cmon man you're only as old as you feel......until you fall and break a hip.
[19:03:18] <PetefromTn> Looks like the Solar impulse is around 150 miles west of LA... at 21640 ft.
[19:03:28] <PCW> I feel fine, just cant see worth a darn
[19:03:41] <jdh> bionic eyes!
[19:04:57] <PetefromTn> linuxCNC controlled robotic eyes...we have the technology, we can make you see better faster farther....
[19:06:10] <PetefromTn> Dayum that Epoxy industrial paint smells strong. I have to wear a respirator just to go out in the shop and feel the tackiness LOL
[19:07:33] <PCW> Does it need baking?
[19:10:10] <PetefromTn> apparently not the first two coats dried nice and even in about six hours. the third coat just went on about five hours ago so I am checking tackiness so I can start removing the tape. I am getting overly anxious about testing them with water...
[19:11:46] <PCW> I saw the pictures. looked really nice
[19:12:06] <PetefromTn> It is supposed to rail ALL WEEKEND so it should be a GREAT weekend to mess around in the shop. If this works I should be able to finally fill the troughs and test the coolant pump and adjust the VFD etc etc
[19:12:20] <PetefromTn> Thanks man it has been a bunch of work but I am glad I did it.
[19:13:07] <PetefromTn> Listening to the conversations with Solar Impulse you can hear the garbled HF radio talk sounds cool and funny.
[19:14:02] <PetefromTn> Can't go in the shop right now because it stinks of epoxy paint so I am stuck doing something else LOL...
[19:20:22] <PetefromTn> Connor: Hey man wazzyup?
[19:20:38] <Connor> Working on a website for a client.
[19:21:37] <PetefromTn> Did you get the rigid tap sorted on your CNC?
[19:22:08] <Connor> It works.. But I need some spiral Flute taps..
[19:22:50] <PetefromTn> I got a couple LOL...
[19:23:28] <Connor> Yea.. I need to get some to test with.. wouldn't want to break yours.
[19:24:58] <PetefromTn> I still have the 3/8-16 spiral flute I used in the RF45 CNC floating tap head videos I made LOL...
[19:25:14] <Connor> how does backlash affect rigid tapping ?
[19:26:39] <PetefromTn> It doesn't like it LOL....
[19:26:50] <archivist> Connor, I would not contemplate rigid tapping with backlash
[19:27:11] <PetefromTn> How much backlash ya got?
[19:27:23] <PCW> probably a big tap would pull the work
[19:27:28] <Connor> I haven't checked it in a whie on the Z
[19:27:28] <archivist> a floating holder would help there
[19:28:06] <PetefromTn> yup floater...
[19:28:17] <Connor> Didn't seam to cause any issues when i tested 1/4-20 on MDF.. but.. then again.. mdf isn't that strong
[19:28:25] <PCW> I mean push on Z when retracting (backlash is sort of floating)
[19:28:45] <PetefromTn> I originally tapped in some 1/4 inch plastic that was not held too rigidly...
[19:30:16] <Connor> I think I'm going to finish up the PDB mount.. may even try to do it this weekend.
[19:30:28] <Connor> going to use Alum for the lift plate..
[19:31:09] <Connor> I need to do some calculations.. because the design is slightly different than Hoss's.. I'm not going to bother with a top plate since my air cylinder already has one basicly..
[19:31:38] <PetefromTn> forget Hoss....make a CONNOR SPECIAL!!
[19:31:56] <Connor> I use his plans and stuff as "guide lines"
[19:33:01] <PetefromTn> gotta blaze your own trail man...
[19:34:06] <Connor> I need to cut my draw bar down.. so I don't have 4" sticking up..
[19:34:36] <PetefromTn> yup 4" whippin around can be distracting......and noisy.....and dangerous LOL
[19:34:40] <Connor> I'm not sure I like the idea of using just lock tight to hold the lug nut on it..
[19:34:56] <PetefromTn> I can Tig it on fer ya if ya want....
[19:35:25] <Connor> I was thinking of cross drilling it.. and pinning it.. but,, not sure how hard that would be..
[19:35:59] <PetefromTn> Dunno man never seen the setup. Tig is permanent,obviously
[19:37:15] <Connor> basicly a lug nut on the 7/16 Threaded Rod.. I think it's 7/16.. what ever R8 thread size is..
[19:37:33] <Connor> It's a high grade threaded rod.. not your home depot grade.
[19:37:40] <PetefromTn> yeah I think it is 7/16 as I recall...
[19:37:49] <PetefromTn> well I hope so...
[19:38:13] <Connor> screw the lug nut so that it's flush with the top of the rod.. or just slightly below and call it good.
[19:38:35] <PetefromTn> You can drill it thru the nut and I can button weld it kinda
[19:38:39] <Connor> PetefromTn: What I *DO* need is the pulleys re-done.
[19:38:57] <PetefromTn> How so?
[19:39:08] <Connor> The run-out sucks.
[19:39:31] <Connor> I purchased two new ones so we can turn them down later on on your lathe.. vs my dinky little crap one.
[19:39:56] <PetefromTn> cando...
[19:39:59] <Connor> not to mention, I didn't chuck them right.. I didn't realize the flanges were press fit..
[19:41:11] <PetefromTn> It's gonna be so cool to see the flood coolant running tomorrow on my monster...
[19:41:31] <Connor> also need to buy a 4" chuck and adapter plate so I can use my little 4' rotary table.. I made one from flat alum kinda for my 3" but could never get it 100% true.
[19:41:37] <Connor> I bet.
[19:41:54] <generic_nick|2> get ready for the mess! lol
[19:42:30] <PetefromTn> Hell I'm used to the mess.... This machine is actually quite easy to clean out thoo..
[19:42:49] <generic_nick|2> nice. what is it?
[19:42:51] <generic_nick|2> http://farm9.staticflickr.com/8395/8704736685_63028c81dd_c.jpg
[19:43:40] <PetefromTn> skateboard truck?
[19:47:04] <generic_nick|2> yes
[19:47:22] <generic_nick|2> i meant what is your mill
[19:47:32] <Jymmm> coffee
[19:47:43] <PetefromTn> mine?
[19:48:02] <generic_nick|2> yes
[19:48:27] <PetefromTn> It is a Cincinatti Arrow 500 VMC 1994
[19:48:49] <generic_nick|2> cool
[19:49:06] <PetefromTn> Yeah I hope it will be soon LOL... so far it has been a PIA.
[19:49:16] <generic_nick|2> lol they all are
[19:50:28] <generic_nick|2> mine is a PITA right now.... just swapped out my 7i33 for a 7i48 and at least one axis is running away.
[19:50:48] <generic_nick|2> hoping i just reversed the analog out wires
[19:50:53] <generic_nick|2> or encoder wires
[19:50:54] <Connor> generic_nick|2: What do you have ?
[19:51:03] <generic_nick|2> shizuoka ans
[19:51:28] <PetefromTn> love the Shiz..
[19:51:37] <PetefromTn> toolchanger?
[19:51:45] <generic_nick|2> cant complain too much for a knee mill
[19:51:46] <generic_nick|2> yes
[19:51:53] <PetefromTn> SWEET
[19:51:56] <generic_nick|2> quickdraw II
[19:52:12] <generic_nick|2> a little flakey but works most of the time
[19:52:28] <Connor> any pictures ?
[19:53:13] <generic_nick|2> closest i have on this computer to pictures. http://www.youtube.com/watch?v=h_iJe436esA
[19:53:17] <Tecan> (h_iJe436esA) "Shizuoka running on emc2" by "jeepinSD" is "Comedy" - Length: 0:00:51
[19:53:54] <PetefromTn> is that the infamous "tool chucker" LOL
[19:54:00] <generic_nick|2> lol yea
[19:54:10] <generic_nick|2> doesnt chuck tools much anymore
[19:54:10] <Connor> tool chucker ?
[19:54:36] <PetefromTn> Connor: Yup they were known to sometimes throw toolholders under the original control LOL
[19:54:47] <generic_nick|2> i gutted all the electronics and fitted it with SSR's and did the logic in classicladder
[19:54:55] <Connor> Eep. That's not good
[19:55:11] <PetefromTn> You da man..... Now I know who to talk to about my toolchanger LOL
[19:55:18] <generic_nick|2> replaced the big pneumatic/hydo cylinder with an electric linear actuator
[19:55:46] <generic_nick|2> haha, sure, but i really dont know what i'm doing lol
[19:56:04] <generic_nick|2> i need to upgrade to a vmc.
[19:56:18] <generic_nick|2> moving into a new shop so ill have room for it soon
[19:56:46] <PetefromTn> ya making bicycle parts?
[19:56:53] <generic_nick|2> was
[19:57:09] <PetefromTn> sweet.
[19:57:43] <PetefromTn> I have a SWB USS custom built recumbent bike I ride LOL...
[19:57:58] <generic_nick|2> nice
[19:59:34] <PetefromTn> yup it screams downhill LOL...
[19:59:49] <PetefromTn> that thing cat30?
[19:59:58] <generic_nick|2> nmbt40
[20:00:16] <PetefromTn> huh....does it use retention knobs?
[20:00:27] <generic_nick|2> nope, drawbar
[20:00:35] <PetefromTn> wow...
[20:00:57] <PetefromTn> How is the drawbar actuated, with your linear actuator?
[20:01:05] <generic_nick|2> fairly common
[20:01:25] <generic_nick|2> no, it's like a power drawbar for a bridgeport.
[20:01:33] <generic_nick|2> impact gun and an air cyl
[20:01:34] <PetefromTn> impact?
[20:01:38] <PetefromTn> LOL
[20:01:56] <generic_nick|2> it litterally is an electric impact gun with the handle cut off and hard wired
[20:02:07] <PetefromTn> dja make that?
[20:02:07] <generic_nick|2> i get replacements at harbor freight
[20:02:16] <generic_nick|2> nope, factory lol
[20:02:23] <PetefromTn> cool...
[20:02:41] <generic_nick|2> it works pretty good
[20:02:44] <PetefromTn> Glad my machine came with all that LOL... just gotta figure out how to make it work with LinuxCNC/Mesa
[20:03:08] <generic_nick|2> so did mine, i just changed it some as the years went on and stuff wore out
[20:03:19] <PetefromTn> Need to find a suitable 25 ohm braking resistor for the Z axis on mine...
[20:03:40] <generic_nick|2> stove burner?
[20:03:41] <PetefromTn> Im sure it did, nice machines from most accounts...
[20:04:08] <PetefromTn> Actually I used a water heater element for my Spindle drive brake and it worked great...
[20:04:23] <PetefromTn> Stops right now LOL...
[20:04:23] <generic_nick|2> nice
[20:04:38] <generic_nick|2> i got some huge resistors at the surplus yard down the road
[20:04:49] <generic_nick|2> they're like 2.2kw
[20:05:45] <PetefromTn> Solar Impulse is just southwest of Lake Havasu City...
[20:06:12] <PetefromTn> What's the resistance?
[20:07:20] <PetefromTn> How many holders ya got ?
[20:08:17] <Connor> PetefromTn: You still getting the sserial errors ?
[20:08:45] <PetefromTn> Honestly I dunno have not even turned it on in a week or two... Been futzing with the troughs...
[20:09:12] <PetefromTn> I don't think so tho, only errors are from the spindle noise issues..
[20:14:50] <PetefromTn> Gettin' bored... wish I could mill something LOL
[20:18:49] <generic_nick|2> 24 holders
[20:19:41] <generic_nick|2> hmmm pcw_home: i just swapped out my 7i33 for a 7i48 and now i'm getting heavy occilation
[20:19:52] <generic_nick|2> do i need to completely retune this thing?
[20:22:39] <PetefromTn> damn I hope not ...
[20:25:43] <generic_nick|2> its weird, y seems ok, but x and z dont
[20:26:07] <generic_nick|2> they were fine before
[20:30:39] <generic_nick|2> i have the pwm frequency set at 24000
[20:34:39] <generic_nick|2> it's getting better with tuning. i had to cut the p and i in half basically
[20:35:24] <PetefromTn> man I suck at PID tuning...
[20:35:50] <generic_nick|2> me too
[20:35:52] <r00t4rd3d> http://i.imgur.com/amPSmUb.jpg
[20:35:58] <generic_nick|2> mainly just guess work for me
[20:36:18] <skunkworks> Me too - me too
[20:36:25] <generic_nick|2> cool!
[20:36:59] <skunkworks> But it doesn't have to be perfect....
[20:37:10] <generic_nick|2> got my 4th axis all ready to go. just gotta plug it in and wire the home switch
[20:37:24] <generic_nick|2> was hoping to test it tonight
[20:37:36] <r00t4rd3d> how i use my cnc
[20:38:04] <r00t4rd3d> https://www.youtube.com/watch?v=07Y0cy-nvAg
[20:38:08] <r00t4rd3d> tecan!
[20:38:10] <skunkworks> this was a first try... http://youtu.be/FgOqEz5Tk-Y
[20:38:13] <Tecan> (FgOqEz5Tk-Y) "Kearney and Trecker first closed loop movement with EMC2." by "samcoinc" is "Education" - Length: 0:00:14
[20:38:36] <PetefromTn> sweet.....make me one LL
[20:44:30] <generic_nick|2> i cant put any ff1 into this thing
[20:44:59] <generic_nick|2> even of i have .01 ff1 it over travels
[20:49:32] <PetefromTn> whats your setup for testing the overtravel?
[20:50:00] <generic_nick|2> watching the motors lol
[20:50:13] <generic_nick|2> i got it somewhat decent, i think i'm going to run with it
[20:50:35] <generic_nick|2> time to set the hal and ini up for the 4th axis
[20:51:15] <PetefromTn> get it done man....and post vids.
[20:55:52] <generic_nick|2> spindle isnt changing rpm's
[20:55:57] <generic_nick|2> hmmm
[21:04:17] <andypugh> 99% of the controllers in the world have no FF1
[21:04:53] <andypugh> But, depending on the units, 0.000001 might be correct.
[21:05:32] <andypugh> It's just numbers, and sometimes they come out little.
[21:10:03] <generic_nick|2> yea it ran fine without it, i was just told my pcw that it needed it so i thought id try it
[21:10:57] <andypugh> Reading Back. Connor: I wouldn't take Hoss as Gospel. He is certainly self-confident, but there are better engineers.
[21:13:43] <andypugh> Impact gun-drawbars work well, and probably clamp tighter than any of the spring ones. However is it possible to make your own spring-style drawbar that fits down a standard drawbar bore. http://www.youtube.com/watch?v=pxrzJ_KfcQ0
[21:13:46] <Tecan> (pxrzJ_KfcQ0) "Harrison retrofit automatic drawbar" by "andy pugh" is "Tech" - Length: 0:01:27
[21:15:39] <Connor> andypugh: Agreed.. Which is why I said I use his stuff as guide lines.. and sometimes not even then.
[21:16:12] <PetefromTn> That looks amazing and difficult to build at best LOL
[21:16:55] <PetefromTn> Actually looking at it again not so much... sweet setup.
[21:17:49] <PetefromTn> Just peeled the tape off the troughs, almost passed out from the fumes LOL. Gonna take the tape off the other one in the AM I think.
[21:20:08] <andypugh> PetefromTn: The "clever" bit was making the finger-collet. I clamped 4 pieces of square tool-steel together in a 4-jaw chuck, and machined as one (there was a pressed-on collar for some operartions). So, the collet was split before machining, not after: http://www.youtube.com/watch?v=IJ9qVexXirk
[21:20:11] <Tecan> (IJ9qVexXirk) "Machining a split collet" by "andy pugh" is "Tech" - Length: 0:01:01
[21:21:31] <ssi> evening folks
[21:21:42] <andypugh> Morning
[21:23:27] <Tom_itx> g'day
[21:24:39] <PetefromTn> sup..
[21:25:02] <PetefromTn> andypugh: Hey man that is just poetry, Nice work.
[21:25:40] <PetefromTn> I think I have a head rush from paint fumes fellas... seeing pink hippos in dancing tutus.
[21:26:18] <andypugh> That paint has to have a "street price"
[21:26:41] <PetefromTn> Yeah man I can hook you up, got about 3/4 gallon left...
[21:27:02] <PetefromTn> Should be enough to lunch your bean for at least a few days if not forever.
[21:27:27] <andypugh> Did I mention that one of my eBay auctions went a bit wierd in the questions?
[21:27:36] <Tom_itx> saw that
[21:28:20] <andypugh> Also, the evidence is that nobody on eBay buys HSK taper tooling
[21:29:38] <PetefromTn> huh?
[21:31:06] <andypugh> I am selling BT40, BT50, DIN40, Din50, 40INT, Flash-change40 and HSK tools that I bought in a job-lot. Only the HSK have no watchers.
[21:32:49] <ssi> it has not been a productive day :(
[21:33:27] <toastydeath> anyone who runs an hsk machine probably doesn't trust used holders
[21:33:45] <toastydeath> i know i wouldn't
[21:34:00] <PetefromTn> ssi: sorry dude..
[21:34:29] <ssi> I just now got home
[21:34:42] <ssi> I want to get the files backed up off my win/mach3 plasma box
[21:34:47] <ssi> and install linuxcnc on it
[21:34:54] <ssi> but I can't get the monitor to wake up on it :(
[21:38:07] <Gamma-x> hey guys
[21:38:26] <ssi> hey gamma
[21:38:30] <Gamma-x> finally got my machine runnin!
[21:38:32] <ssi> woooo
[21:38:45] <Gamma-x> starting to wish I went with the servo route!
[21:38:52] <ssi> why's that
[21:39:19] <Gamma-x> cant seem to get all the settings rite for the stepper configs....
[21:39:32] <Gamma-x> speed/ velocity, length etc.
[21:40:10] <ssi> you think servos are easier? :D
[21:40:13] <jdh> heh
[21:40:45] <Gamma-x> lol
[21:41:11] <ssi> I think a person's first machine being steppers is not a bad thing :)
[21:41:17] <Gamma-x> well im getting a lot of studdering and my delron couplers are not standing up!
[21:41:31] <Tom_itx> use steel couplers
[21:41:44] <ssi> lovejoy couplers
[21:41:47] <Gamma-x> need to find some
[21:41:53] <Gamma-x> Tom_itx, have a source?
[21:41:55] <ssi> mcmaster.ceom
[21:41:58] <Tom_itx> ^^
[21:42:05] <Tom_itx> or sdi
[21:42:10] <Gamma-x> sdi?
[21:42:12] <jdh> lovejoys always seem to have lash in the spider
[21:42:13] <ssi> mcmaster.com should be bookmarked in your browser
[21:42:16] <Tom_itx> stock drive products
[21:42:17] <generic_nick|2> i had lovejoy couplers on my mill for a while
[21:42:19] <ssi> under "buy all your machine parts here plz"
[21:42:20] <generic_nick|2> they worked fine
[21:42:37] <jdh> mcmaster has oldham couplers
[21:42:45] <ssi> the lovejoys I've used have tight spiders that don't have noticable lash
[21:42:47] <Tom_itx> sdp-si sry
[21:42:53] <ssi> certainly less than the screws they're driving in most cases :)
[21:42:58] <ssi> oldham works too
[21:43:04] <ssi> but I think lovejoys are probably more affordable
[21:43:40] <Tom_itx> https://sdp-si.com/eStore/CoverPg/Couplings.htm
[21:44:11] <jdh> saving $20 on couplings isn't really worthwhile if you are spending a couple $k
[21:44:53] <ssi> oldhams also don't tolerate angular misalignment quite as well
[21:45:01] <ssi> but they do tolerate lateral misalignment well
[21:47:07] <Gamma-x> wow sdp-si has ALOT of couplings...
[21:47:20] <Gamma-x> Tom_itx, which one would have the most grip?
[21:48:46] <Tom_itx> depends what you need
[21:49:03] <generic_nick|2> hmmm looks like i just have to figure out my enable for the a axis.
[21:49:44] <Tom_itx> mine are hard couplers and i haven't had any problems with them
[21:49:49] <Tom_itx> you need to get them aligned
[22:06:44] <ssi> hooray
[22:06:48] <ssi> got the stupid winblows box to boot
[22:07:54] <generic_nick|2> cant believe the only thing holding me up on my 4th axis is the enable pin lol
[22:09:34] <generic_nick|2> one other issue, the encoder counts per degree are not a round number. it's like 600.666666666666666666666666666666666666666666666
[22:09:56] <generic_nick|2> anyone have a trick to make it not be off after x amount of revs?
[22:15:17] <Gamma-x> jesus
[22:15:25] <pcw_home> generic_nick|2 on the 7I48 did you set PWM mode to 2?
[22:15:25] <Gamma-x> mcmaster has sooooo many types of couplings.
[22:16:15] <ssi> Gamma-x: yep
[22:16:21] <ssi> Gamma-x: what part of the world do you live in?
[22:16:30] <Gamma-x> ssi, denver co
[22:20:39] <ssi> you should be able to get stuff from mcmaster 2 day by default, if not overnight
[22:21:25] <Gamma-x> ssi, just gotta figure out what kind of coupler i need!
[22:21:58] <Gamma-x> i want\\ them to fit directly onto partly flattened shaft....
[22:22:06] <Gamma-x> http://www.automationtechnologiesinc.com/products-page/nema-23/nema-23-three_eighths-inch-dual-shaft-with-a-flat-570-oz-in
[22:22:42] <Gamma-x> and http://www.automationtechnologiesinc.com/products-page/nema-34/nema34-stepper-motor%E2%80%931200ozin-6amp-single-shaftkl34h2120-60-4a-key
[22:23:04] <ssi> anything with a 3/8" bore and a setscrew should be fine
[22:24:58] <Gamma-x> my ball screw is 16mm....
[22:25:55] <ssi> 16mm where the couplers attach?
[22:25:58] <ssi> are the ends turned?
[22:26:07] <ssi> or are you trying to couple directly to ballscrew races?
[22:26:54] <Gamma-x> the ends are turned...
[22:27:00] <ssi> to 16mm?
[22:27:01] <Gamma-x> to 16mm
[22:27:17] <ssi> so the trick is finding a two piece coupler
[22:27:27] <ssi> that offers both 16mm and 3/8 bore sizes
[22:27:48] <ssi> 16mm is pretty close to 5/8"
[22:27:59] <ssi> http://www.mcmaster.com/#standard-shaft-couplings/=mli3gq
[22:28:02] <ssi> olhams
[22:28:08] <ssi> get a 3/8 and a 5/8
[22:28:14] <ssi> you might need to ream the 5/8 out to 16mm
[22:28:18] <jdh> dont' they have metric?
[22:28:19] <ssi> but it's only .004" difference
[22:28:21] <ssi> so should be close
[22:29:10] <ssi> looks like you want the P size
[22:29:24] <generic_nick|2> yes pcw_home
[22:29:31] <generic_nick|2> i just had to retune it
[22:29:54] <generic_nick|2> pcw_home: stumped on enabling my servo drive
[22:30:06] <ssi> generic_nick|2: what's the drive?
[22:31:00] <generic_nick|2> pac sci, but it isnt the drive, it's the enable pin on the 7i48
[22:31:10] <generic_nick|2> it isnt turning on
[22:31:17] <generic_nick|2> it did on the 7i33
[22:31:46] <pcw_home> The 7I48 has 6 uncommitted OPTO outputs for drive enables
[22:32:33] <pcw_home> basically switches though polarized and limited to I think 15 mA
[22:32:47] <generic_nick|2> pcw_home: when i talked to you on the phone a while back you mentioned something about the first and last pins having to be true when i checked them in dmesg
[22:33:00] <ssi> looks like you have to wire a pin to the enable inputs and pull them down?
[22:33:00] <tjtr33> for special sizes, ask the guys at mcmaster, they can get stuff not listed in catalog.
[22:33:17] <generic_nick|2> they are false with emc turned on
[22:33:17] <ssi> oh from the fpga never mind
[22:33:38] <generic_nick|2> it's a config issue, not wiring
[22:33:45] <pcw_home> If you have analog outputs, the enables _are_ on
[22:34:12] <pcw_home> it is most likely a wiring issue
[22:34:16] <generic_nick|2> hmmm
[22:34:52] <generic_nick|2> oh so the enable pins on the 7i48 do not put out 5v like the 7i33?
[22:35:31] <pcw_home> No they are OPTO outputs (polarized switches)
[22:35:37] <generic_nick|2> ah
[22:35:46] <generic_nick|2> thats the problem then
[22:35:50] <generic_nick|2> thanks
[22:38:56] <ssi> startin to make some progress :D
[22:39:16] <ssi> jthornton: around perchance?
[22:43:02] <ssi> pcw_home: I really appreciate the fact that you're happy that your customers hack on your products
[22:43:28] <ssi> I'm looking at this THC product I have, and it'd be trivially easy to make it possible to have linuxcnc control and display the arc voltage settings
[22:43:42] <ssi> but I doubt they'll be hacker friendly enough to make it practical
[22:43:55] <ssi> likely because I think they have more expensive products they would rather you buy with those features
[22:44:13] <Gamma-x> too many options...
[22:44:23] <Gamma-x> and som e are toooo expensive!
[22:47:04] <ssi> sweet, this machine has reasonable latency numbers
[22:47:06] <Gamma-x> anyone wanna make some cash by turning me some couplers?
[22:47:08] <pcw_home> Thats why we made a very simple THC for LInuxCNC, so LinuxCNC can do the control and then you can
[22:47:10] <pcw_home> add features as needed, not being locked into what the external hardware can do
[22:47:21] <ssi> pcw_home: I may end up buying that if this thing gives me any trouble
[22:47:45] <ssi> I bought this one because it was inexpensive and I was in mach3 land back then
[22:47:55] <ssi> a land I'm happy I expatriated from :D
[22:48:12] <ssi> oh jeez maybe it doesnt' have decent latency :P
[22:48:26] <ssi> couple of 225us spikes
[22:48:35] <pcw_home> oops
[22:48:58] <ssi> although that shouldn't be the end of the world with mesa doing the stepgen
[22:49:20] <ssi> mornin mhaberler :D
[22:51:22] <ssi> ugh I should just order another cheapy dell
[22:52:03] <ReadError> ssi: microcenter has refurbs cheap all the time
[22:52:17] <ssi> last one I ordered from tigerdirect, off lease opti 755 for $199
[22:52:27] <ssi> and it has great latency numbers
[22:52:28] <ssi> 17us or so
[22:52:49] <ssi> network doesn't seem to work on this stupid little lenovo box
[22:53:47] <generic_nick|2> yay, the 4th axis works!
[22:53:52] <ssi> sweet :D
[22:54:55] <generic_nick|2> there are a few issues.... first is the pid values are extremely low in order to get rid of occilation. not sure it's going to have the holding torque i need with the values that low.
[22:57:28] <generic_nick|2> secondly, i gotta figure out the velocity and acceleration rates. guess i gotta do some homework on that. the max velocity is set at 300, which is fine for the xy and z axis, but that seems to equate to 300 degrees a minute as well, which is slow as hell
[22:57:57] <ssi> generic_nick|2: there's a master max velocity, and a per axis max velocity
[22:58:30] <ssi> so the network on this thing, the link is flapping up and down
[22:58:43] <ssi> it's a tg3, and I get a link up at 1000mbps, link down over and over in dmseg
[22:58:49] <ssi> extra lame
[22:59:06] <generic_nick|2> yea, i have that set, but the sliders are in ipm so not sure how that will work with the 4th axis
[22:59:51] <ssi> http://www.linuxcnc.org/docs/devel/html/config/ini_config.html
[22:59:59] <ssi> MAX_LINEAR_VELOCITY = 1.0 - The maximum velocity for linear jogs, in machine units per second.
[23:00:02] <ssi> MAX_ANGULAR_VELOCITY = 1.0 - The maximum velocity for angular jogs, in machine units per second.
[23:00:36] <ssi> dunno what the machine units are for angle
[23:00:38] <ssi> but I guess degrees
[23:05:12] <Gamma-x> I think im going to have to make my own couplings...
[23:07:49] <ssi> why
[23:09:17] <generic_nick|2> thanks ssi, thats what i was looking for
[23:09:36] <generic_nick|2> i got the 4th axis spinning at 8500 degrees per minute
[23:09:58] <generic_nick|2> still a bunch of tuning to do, but it's a start
[23:10:20] <ssi> sweet
[23:10:33] <ssi> 140 degrees per second
[23:10:52] <ssi> shit my RV should roll twice that fast ;)
[23:10:56] <Gamma-x> ssi, just no ones with right size....
[23:11:11] <ssi> I told you what to get
[23:11:47] <Gamma-x> the oldhams?
[23:11:51] <Gamma-x> I dont trust em...
[23:12:15] <Gamma-x> and I need one that I can buy 1/2 x 16mm with keyway
[23:12:32] <Gamma-x> I can buy that one ^^^
[23:14:13] <mhaberler> good morning Ian
[23:16:20] <generic_nick|2> that worked like a charm ssi. now the sliders are right.
[23:17:32] <generic_nick|2> just gotta do some tuning over the weekend and it's good to go. still gotta get it on the mill and wire up the home switch, but that's easy.
[23:25:29] <Gamma-x> ssi, for 30 bucks I can get machinable couplings.
[23:25:49] <Gamma-x> http://www.mcmaster.com/#standard-shaft-couplings/=mlitnh
[23:27:40] <ssi> have at it
[23:27:49] <ssi> I've done all the helping I care to on that one :)
[23:28:09] <Gamma-x> lol I just want something that WONT slip at all!
[23:28:25] <Gamma-x> I appreciate the assistance
[23:28:42] <PetefromTn> nitey nite dudes....
[23:41:41] <ssi> build faster, kernel!