ho ho ho
A2 tool steel: feisty
milling an oversize 2x3x3 block of a2 tool steel down to 1x2x3: even fiestier
03jepler 07TRUNK * 10emc2/docs/src/hal/drivers.lyx: document petev's improvements to m5i20 encoder interface
03petev 07TRUNK * 10emc2/src/hal/drivers/ (hal_m5i20.c hal_motenc.c):
-Changed encoder reset-count to reset and changed behavior so that it
conforms to EMC canonical model.
03petev 07TRUNK * 10emc2/src/hal/drivers/m5i20.h: -Added some new registers to memory map.
03petev 07TRUNK * 10emc2/src/libnml/inifile/ (inifile.cc inifile.hh): -Added Find() method for mapping strings to double.
03petev 07TRUNK * 10emc2/src/emc/nml_intf/emc.hh: -Added typedefs for linear and angular units.
03petev 07TRUNK * 10emc2/src/emc/ini/ (emcIniFile.cc emcIniFile.hh): -Added Find() methods for linear and angular units.
03petev 07TRUNK * 10emc2/src/emc/ini/initraj.cc: -Cleaned up access to INI file.
03compile-farm 07BDI-4.51 (188.8.131.52-rtai) * 10emc2head/: build FAILED ; see http://linuxcnc.org/compile_farm/emc2head_slot6_log.txt
03compile-farm 07Ubuntu 6.06 LTS (dapper) realtime (2.6.15-magma) * 10emc2head/: build FAILED ; see http://linuxcnc.org/compile_farm/emc2head_slot7_log.txt
03petev 07TRUNK * 10emc2/src/hal/drivers/Submakefile:
-Changed to use newer FPGA image for 5i20 in m5i20/bit dir. This image
needs more testing before it is released.
xemet might try a pyvcp panel to monitor the halpin ( took too long to see what he wanted to do :(
03petev 07TRUNK * 10emc2/src/hal/drivers/m5i20/README.txt: -Updated paths to FPGA configs.
crepincdotcom: is the 'multimeter' method = watch for continuity from tool to copper layer? maybe clean the surface before probing. and check at least 3 places to see the plane is normal to Z.
03compile-farm 07Ubuntu 6.06 LTS (dapper) realtime (2.6.15-magma) * 10emc2head/: build PASSED
03compile-farm 07BDI-4.51 (184.108.40.206-rtai) * 10emc2head/: build PASSED
03petev 07TRUNK * 10emc2/src/emc/sai/saicanon.cc: -Fixed signed/unsigned warning in GET_EXTERNAL_PARAMETER_FILE_NAME().
03cradek 07TRUNK * 10emc2/src/emc/kinematics/tp.c: give queue depth to halscope
03cradek 07TRUNK * 10emc2/src/emc/motion/ (control.c motion.h): give queue depth to halscope
03jepler 07TRUNK * 10emc2/src/hal/utils/comp.g: add support for arrays and dotted HAL names
03jepler 07TRUNK * 10emc2/docs/src/hal/comp.lyx: begin to document arrays
03cradek 07TRUNK * 10emc2/scripts/emc.in:
This makes segment throughput as high while running AXIS as with xemc.
It makes a huge difference on this PII-400 and I bet people with "real"
computers won't notice the difference.
03cradek 07v2_1_branch * 10emc2/scripts/emc.in: backport simple fix for segment throughput on slow machines
what are people's opinions on cutting acme threads into one side of a spider coupling?
[03:59:27] <Jymmm> http://www.embeddedtronics.com/acmetap.html
ie. so the one end can go directly onto the end of the leadscrew without being turned down
My exp ended when I pasted the link =)
I've seen that, it's a great idea
but alas, I have no mill or lathe yet
4" or so
yeah, get a cutoff blade to make the slots,
use the grinder wheel to begin tapering the tip
hrmmm... I don't quite see myself being able to pull that off with a hand grinder and have it turn out decent
ejholmgren: is there nobody around you with a lathe you can borrow?
ejholmgren if you have a drill press, you can just chuck up the acme rod, then bolt the grinder to the dp and let er rip!
I suppose that would work for the taper
unless it's long...
I think the chuck might only be 3/8" as well
only a benchtop press
anyone ever seen a recent picture of Ed Nisley?
mine too, but 12"
tomp: the problem is that yes, the probe needs a cleaner area to test with
ejholmgren when do you need this by?
and also, the carpet tape method allows the board to move such that it is no longer normal to the Z axis. Even during milling, the sides differed by about 0.0008"
sorry to interupt guys
ejholmgren heh, well you have time to find someone with a lathe/mill
ve7it is now known as LawrenceG
crepincdotcom HUH, how rude!!! lol
no kidding ...
crepincdotcom: have you experimented with different kinds of tape? I had to try several before being happy
we didn't even get to add another power tool to the drill press/grinder mix
i have tried the two that home depot had, neither have made me happy yet
are you using the very thin stuff that looks like packing tape?
too skithy or not enough?
maybe happy isn't the right word :-)
too sticky or not enough?
eric_unter: i tried that type as well as thick padded type
lol too sticky, hard to get off
the padded is going to suck
I can never find that when I'm looking for it
cradek: beleive it or not the padded had less of a deviation from 90 degrees than the thin did, may have just been an anomoly though
I use the thin unpadded stuff that's for sticking down tile? carpet? and has nylon? threads in it
crepincdotcom lighter fluid / napatha is your friend
yes i have some of that, ill play more with it
this tape comes off too
Jymmm: all, ill try it
I have had to use lighter fluid, but not for this stuff
what do you do with the lighter fluid?
use it to get it not sticky
it eats the adhesive
ok guys bed time, just saw that someone had said my name
drill press + hand grinder + lighter fluid as coolant ...
doesn't wd-40 catch on fire?
I would assume so
there's a youtube video where someone shoots a can of it
resulting in a very large firey explosion
people use that as a coolant all the time, but they are the ones that believe it's mostly water
[04:27:16] <ejholmgren> http://www.youtube.com/watch?v=_dHUfvRsCb4
I take that back, he hit it with a sledgehammer
[04:28:41] <ejholmgren> http://www.youtube.com/watch?v=HJX-ty0nhpU&NR
wd-40 + supersoaker
[04:29:08] <eric_unter> http://www.youtube.com/watch?v=z-1ZoxWDU7o
I like the title on that one better
I'm gonna say that the supersoaker guy was really lucky
[04:31:10] <Jymmm> http://www.youtube.com/watch?v=AV6WSx9sj4M&NR
you guys convinced me to never even be tempted to use wd40 as a coolant again
they charged some guys around here under the patriot act for blowing up a microwave in a local abandoned quarry
judge threw out the charges though
the DA's office charged under patriot act?
the partiot act should never ever of happened.
why would you use wd40 as a coolant
also people use flammable coolants all the time
kerosene and gasoline, notably
kerosene is amazing for plastics
wd40 works pretty well
a lot more people have a can of wd40 than have a can of kerosene
any time i've seen wd40 used by anyone the improvement has been minimal
at least in the lubricants i've seen used
it's been a long time since I've used it at all, I grew to dislike it over many years as a bicycle mechanic
it generally smokes and is completely useless
not that a good drip cutting fluid won't smoke like the dickens
toastyde1th is now known as toastydeath
I want to use flood, but a bridgeport is not made for it
you can use spray
or, low pressure flood
is spray different from mist?
just wanted to make sure there wasn't an alternative I wasn't aware of
at work, the toolmakers all milled coolant grooves in the table
so they can flood
mine has that, I suppose I could cut down the volume
they also made all sorts of ridiculous splash guards, though
i mean a low pressure flood just drips off unless you're cutting aluminum or using a tiny endmill
the Haas at work has all sorts of extra sheet metal stuff tacked on to the stock sheet metal, and it was built for flood
the floor used to get pretty wet
it can, if you crank the rpm up
usually there's not much use for flood coolant on a manual mill anyway
misting systems are all that's necessary
mine was built for cnc
do you run it with carbide tooling and at high speeds and feeds
only handle on the whole machine was the knee handle
I've never cut anything with it, bare bones when I bought it, and I'm really slow
yeah i was wondering how you were cutting on a manual mill with ballscrews
motors were gone, pulleys were gone, cabinets stripped
tends to not work too well
so are you converting to manual
or looking to do CNC on it
no, converting to emc
it came with a mist system, I'll probably try that
the only time you really should care about flood coolant is running at high end carbide speeds
otherwise the there's enough heat dissapation in the chips
and the mist helps the chip flow and vibration
I've heard mist is annoying in the shop
i like more of a drip
so just cut back the air pressure?
thats what i would do
because you're not trying to coat the whole damn area with a fine mist
reminds me, I have to get the air balance system in the knee to stop leaking air
acme screws and 1000 lbs of knee/table don't go together too well
air balance system?
i've never heard of that
what's wrong with your acme screws?
all of the bp cncs have a pneumatic piston in the pedestal of the base that pushes up on the bottom of the table
oh, sort of backlash eliminator
acme screw is ok, it just is a bear against the immense load of the knee/table
i dunno dude
acme screws work just fine on the 3500 lb knees on the big horizontals
in fact, I hear one of the failure modes is that the table will start floating
and on the 5000 lb HBM heads
I wonder if they have more robust acme screws
is it a bridgeport specific issue?
and lower gearing
i mean the knee screw isn't exactly big on a bridgeport
it was enough of an issue that they put a pneumatic lift in there
as cheap as they were, I don't think they did it for fun
I have no problem with the series I at work, but mine is pretty much unusable the way it is
what exactly is the float
i'm not seeing what problem they're fixing except maybe backlash
they may have started with the mills that had the steppers on the knee, and kept doing it out of habit.
I don't know if they ever made manual series II, and if those had air assist or not
and i have no idea
but that does explain what that one CNC is always leaking at school
apparently, there is a leather seal on the piston that drys up and leaks air
it's a bridgeport CNC
everytime I walk by
tells you if you have air pressure in the building though
there is an air spindle speed adjust motor and air spindle lock too
but the most likely leak is in the knee
i have no idea, so much stuff on that thing doesn't work
that's why they are cheap
I think there may be more working now in hobby shops than there used to be when bp sold them to machine shops
it was pretty common for people to buy them and never use them
production shops don't have any use for a cnc bridgeport
not any more, there was a time when the shops were all manual
yes, and even then
bp shows up with a cnc, people bought them because of the name
and they didn't work
the cnc bridgeport didn't fill any particular niche when NC was first coming out
not when nc came out
but when the bridgeport cnc came out
I don't know what was competing back then
air >re<sist. used on vertical axis. xy doesnt get affected by gravity, but Z does. so down is easier than up. the servo notices this difference and is difficult to tune correctly in both directions. the air piston ( or weight ) makes the response simlilar.
bridgeports are for quick modifications to parts in production shops
tomp -- that would be a good point, but my knee is manual, always was
or onesies and twosies with the right person at the wheels :)
yeah, that too
we do up to tensies
but we employ toolmakers to turn the wheels
but they have the accu-rite conversational CNC thing
with manual handwheels
we have one of those
so they can do roughing patterns and arcs, but still keeps the versatility
those are so cool
the machinist just took the quick change spindle out of it
it stopped being reliable
should say quick change drawbar
i've got no direct experience with this, but i hear homemade quick change drawbars are more reliable
than most commercial drawbars
personally i'd like a bridgeport with a CAT-20 spindle
the accu-rite thing was really neat, but it could be a pain
never used it
what was wrong
I have a NMBT-30 spindle in mine
i've just seen it used
never saw a custom made pneumatic drawbar
Joe Vicars used to sell one
there are a number of designs that use a pneumatic torque wrench
people make them out of impact wrenches
and other air nonsense
that's what I meant
OR DID I MEAN TORQUE WRENCH
we wil never know.
same thing, almost.
I guess getting the socket to advance and retract would be the only tricky part
the coolest thing is Joe's toolchanger for the Sherline
almost makes me want to get a sherline so I can make Joe's toolchanger
[05:39:36] <eric_unter> http://home.insightbb.com/~joevicar3/Automatic_Tool_Changer_Plans.htm
the video is great
anyone got those plans?
I've seen it's predicessor years ago.
[05:41:57] <eric_unter> http://home.insightbb.com/~joevicar3/Customer.htm
power drawbar pics uses impact wrench
cool tool changer, pretty fast too!
i'm not sure i'll ever buy a bridgeport
getting a big cinci seems a better choice
for the same price ballpark
ablit higher shipping
bp has its advantages, the basic design is not one of them
i'd buy a bridgeport with a conversational cnc
there seem to be some nice tree mills
my problem with that is twofold: first, don't want to send any royalties to Hurco for that bogus patent
and second, commercial controls cost too much
yeah, but if I bought a conversational bridgeport i would do so as a small business
for a hobbyist, I'd pick up a 30 hp vertical Cinci, loaded with tooling, and call it a day
I used to use the accu-rite a lot, and I'd rather be able to dump my program easily over the network
and have good software on my desktop
at that point, it makes more sense to just write it for a vmc
our accu-rite programs are never more than a couple little commands
If I was starting a small business, that's what I would get
accu-rite retrofit is over 10k
to me the accu-rite and a vmc are totally seperate tools for totally seperate tasks
and in that vein, I would probably have at least one accu-rite bridgeport
but, that would be my strategy
not necessarily what someone else would use it for
I know they were selling them, so there must be shop owners that agree with you
and maybe I get that view from the shop I work at now
which is how they use them
and centroid is still in business, they pretty much do the same thing
each toolmaker has one, and uses it to do arcs and other things that would require extensive setup on rotary tables and indexing fixtures
they get a conversation cnc bridgeport, and a normal bridgeport
this is at a school?
no, a production ship
we make air bearings for CMMs, semiconductor fabs, and other metrology industries
the part is more technologically advanced than our processes are, admittedly
but it's not a high volume part, so it makes sense
custom bearings get made in the manual department
the lab next to mine at school deals with air bearing spindles
university == research business
mechanical engineering building has exactly one classroom, and 4 floors of labs
FWIW I found 4" x 36" sanding belts up to 400 grit
cradek_ is now known as cradek
hey on a certain program (lots of moves .003 long) I found that my mill was stuttering too - I found a simple fix for it - will be in 2.1.3
the nice ?
seems crazy, but it works
while it was stuttering along (I was watching the queue size in halscope) I reniced it, and it loaded the rest of the program (2000 lines) into the queue in about 2 seconds
cradek: cool. (I think I could really notice it on the gantry - because it is only driven on 1 side - when it would studder - it would really shake the gantry.
cradek: stuttering = starving for data?
tomp: yes, my machine is a PII-400 and it couldn't keep motion "fed" with segments before the fix
now it feeds the whole program in 3 seconds (program runs for 30-60 seconds) on that machine
have to get ready for work... bbl
cardek: thanks, bye
If I used the *.sh method of installing emc2 - is that the full version?
what do you mean 'full version'? it doesn't have the source iirc.
will it run a real machine if installed this way?
where does it install the drivers - path?
drivers? do you mean machine configs and such?
duerz: use dpkg -L to get a list of installed files
dpkg -L emc2 will show you where all the files are
no, the drivers for interface cards
duerz: dpkg -L emc2 will show you where all the files are installed
ok - looking
alex_joni: what is nice? (que optimization)
I got an e-mail from eric johnson and he was talking about looking : If you view the
header file /src/hal/drivers/hal_vti.h, you will see four structs defined.
after doing this dpkg i have not seen hal_vti.h?
i found a hal_vti.ko - but cannot open it
you don't have the source. I would build a 'run in place' so you can see it. follow the directions here http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Installing_EMC2
- section 3
then if he wants you to change the code - you can and rebuild
what is a struct?
duerz: a struct is a datatype in the "C" programming language
duerz: you can see the sources online at http://cvs.linuxcnc.org/
[13:49:45] <alex_joni> http://cvs.linuxcnc.org/cvs/emc2/src/hal/drivers/hal_vti.c?rev=220.127.116.11;content-type=text%2Fplain
[13:50:00] <alex_joni> http://cvs.linuxcnc.org/cvs/emc2/src/hal/drivers/hal_vti.h?rev=18.104.22.168;content-type=text%2Fplain
duerz: .ko is a binary format, like .exe or .dll on windows
I'v egot some problems..
I found a program that causes ferror always.
I'll post it in a few secs..
with a stepper machine
with ferror cranked up to 2mm
and min ferror to 0.5
anonimasu: what emc2 version?
the problem you are seeing is mostl likely because of an incorrect max_vel, max_accel, INPUT_SCALE and BASE_PERIOD combination
alex_joni: known issue?
why didnt I see it before..?
anonimasu: 2.1.0+ will report an actual erro about what you did wrong
I've made corners at higher speed..
anonimasu: probably changed something
"joint 1 following error"~
I posted a new program with a straight entry..
that's what I changed..
but the corners in cut are much sharper then the entry..
iirc anonimasu has one of jon elson's boards, not using software stepgen
this is another machine
anonimasu: please upgrade and see what the error is with emc2.1.x
or pastebin your ini
need to run out and carry the machine back into the office..
upgrading isnt a problem other then that
I might aswell upgrade emc first..
be back in a few minutes..
alex_joni: want to have a look at the program?
sure, pastebin the program too
[14:20:17] <anonimasu> http://www.pastebin.ca/394659
N0170 is where it happens
cradek: 0 means no tool, right? So what does this mean? G43 H0
It is OK for the H number to be zero; an offset value of zero will be used.
G49 is the usual way to cancel length offset
cradek: it's posted off the sheetcam emc post..
it's for tool length offsets..
but as i dont want any..
it cancels them..
anonimasu: it ran here, on the example 'configs/stepper/stepper_inch.ini' configuration
I understand - it's a little odd, but the spec allows it to be canceled that way
cradek: that's weird..
I'll have the machine up in a few minutes..
what do you want to know?
the periods are the stock ones..
anonimasu: pastebin the ini file if you can
cradek: I just tried the method you suggested on the mailing list.. but I can't find any servo
anonimasu: it's most likely that (A) you did not increase STEPGEN_MAXxxx when you increased MAX_ACCELEREATION or MAX_VELOCITY or (B) that you set STEPGEN_MAXVEL above the number of steps per second that you can get with your BASE_PERIOD and step timings
anonimasu: that's why alex suggests putting the inifile on pastebin
and now the printer doesn't work anymore.. it makes a funny noice, and some smoke comes out of it when I try to connect it to the toaster
jepler: I can do that now..
but I should upgrade first..
anonimasu: the problem will probably be still there after the upgrade, but you'll have a message explaining it
if you can pastebin the ini, we'll start analyzing.. while you upgrade :P
yeah, but i can make a new config file..
I'll do that
alex_joni: should I just let apt do the work?
run the update function?
anonimasu, is this ubuntu?
anonimasu: if you have an emc2 installed, yeah
but the one im running is from cvs..
has emc2 ever been installed with apt?
ok, then you should be able to upgrade to 2.1.x via the GUIs or apt
doing it now
if you want 2.1.x, you need to change the emc2 repository
or it'll only tell you about 2.0.5
what version is your suggestion that I run?
err what version do you suggest that I run
my suggestion is 2.1.x
whats the address to the repository?
if you dont got it in your head I'll look it up :)
are you using synaptic?
same address but use emc2.1 instead of emc2
[14:41:27] <jepler> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?UpdatingTo2.1
^^ says how to change to the 2.1 repository using the GUI
hmm.. jepler was faster
jepler: I know how to edit a sources.list..
that wiki page tells you what to edit it to :)
anonimasu: same line, just the last part changed from emc2 to emc2.1
already did that :)
[14:44:03] <anonimasu> http://www.pastebin.ca/394679
note: the 2.0.x configs need some tweaking to work
the pre-2.1 probably works without modifications (if it's not too old)
anonimasu: what PC?
a PIII 500
MINle _LIMIT = -1000.0
is that a pastebin typo?
I don't see any constraint violation running anonimasu's program either (running check_constraints.hal in sim)
that's a pastebin error..
it looks ok in the config..
however I have to unpause with the mouse instead of the keyboard, for some reason, after the G4P7/M1 sequence
* anonimasu nods
well, if you turn off the optional pause stuff..
it's a pause per part.. im making 2m of them
so I can re-light the torch is something fails..
I don't think it's pushing hard on constraints
INPUT_SCALE = 33.3
cool, optional pause
and MAX_VEL = 30
BASE_PERIOD = 50
I have emc 2.1.2 now :)
MAX_VELOCITY = 29 / STEPGEN_MAXVEL = 29.2
why is AXES=2, but there are Z moves in the file?
is this too little headroom?
finding the needed headroom seems to be an inexact science, but 5% is more typical than .5%
yes that's less than 1%
try 5% (but I'd expect troubles on jogging, not coordinated motion)
I ran it at 30.
it dosent make a difference..
I tried tweaking them closer to see what happened..
cradek: because my post spits out Z moves..
cradek: for pierceing..
but I've yet to build a Z axis..
try increasing it -- make STEPGEN_MAXVEL = round up(MAX_VELOICITY * 1.1) == maybe try 32
jepler: just a sec..
i can try it right now
anonimasu: I suggest you define the Z axis
that N0170 move has a Z component. I wonder if that's the problem (Z vel/accel values are not loaded because it's an invalid axis)
if jepler's advice doesn't help, try enabling your Z in the ini, even though you don't use it
or try removing all Z words from your file
yes, or that
I can try, but I dont think that matters..
joint 1 following error
emc/task/taskintf.cc 592: Error on axis 1, command number 62
now let's see with Z enabled
why didnt this happen on my other program?
that fixes it?
I didn't see it like this.. but usually it would wait (hang) forever on Z moves
thanks for the help..
I guess I should move to 2.1 as soon as I've made the parts I need for now
anonimasu: at least you have a fresh update now :P
alex_joni: but not runinng :)
T = testing
why does openoffice come with out ubuntu dist?
werent we low on space before?
it's on ubuntu, and the easiest thing to do is the fewest changes to the base distribution
anonimasu: we're not eager to make an OS release ourselves (there is way too less manpower to support that)
that's why we chose to use a pretty good distro, with a GREAT support base behind it
keeping the changes we make (adding a handfull of packages) at a minimum allows people to use that great support base
it might be nice to try making an xubuntu-based EMC, but not nice enough to actually sit down and do the work myself ;)
(or ask others to do it)
SWPadnos: you mean install xubuntu and then install emc? that's easy
no - I mean basing the emc2 CD on xubuntu
it may allow enough room for compilers and such
SWPadnos: I don't think the xubuntu disk is way smaller than the regular ubuntu one
I don't know - checking now :)
in theory, we could remove openoffice and add development tools, but I don't think that would serve most of our users who want to use a live cd to install
vs. 698MB regular ubuntu
40M is a lot when you're talking about tools (vs. guis)
yeah, but gcc & devel libs probably is a bit more than that
sure - though a CD has a max of ~730M
I don't know what's missing from xubuntu that would be needed by emc though - gtk and that kind of thing
stuff that's probably included with normal ubuntu
SWPadnos: demo'ed emc off of my keychain drive yesterday.. (just happened to have it with me.) He thought it was cool.
skunkworks: to who?
A local machinist that has some old equipment.
it worked on his computer thankfully :) just new enough but not too new.
heh - boots from USB but still has a parallel port? :)
hmm. didn't look ;)
new enought to boot from usb - old enough for it to recognise the video
skunkworks: you're a great salesman except for the making-no-money part of it
03jepler 07v2_1_branch * 10emc2/src/hal/drivers/pluto_servo.comp: correct digital input pin names in documentation
thats ok - I get enough out of it ;)
we use ubuntu & kernel 2.6, if 2.4 was ok, then damnsmalllinux is 50 meg (word pro ffox) or it's more full featured version dsl-n at less than 100meg. lotsa of room for dev tools. also matt shaver did some work with puppy. both good for low mips older hardware. it's kde/gnome that suck up the cd space.
I used the full dapper install on my PII-400/256 meg, and it works well - are you using even older hardware than that?
* skunkworks also ran dapper for w while on a PII-400/256
I don't remember if I had more ram in it when I installed - I may have stolen some since then
150 mhz Pentium, 48 meg ram, 2.1 gig hd.... was very snappy with fluxbox , included synaptic and a few tools
yikes, that's underpowered even by my standards
nice that you can still make it go
* alex_joni runs home
underpowered by my stds too, but it shows what can be done.
03jepler 07TRUNK * 10emc2/src/hal/utils/comp.g: support parameter arrays
03jepler 07TRUNK * 10emc2/src/hal/drivers/pluto_servo.comp: use arrays. improve documentation
03jepler 07TRUNK * 10emc2/docs/src/hal/comp.lyx: note new features; other improvements and markup fixes
03jepler 07TRUNK * 10emc2/src/hal/drivers/pluto_servo.comp: remove traces of epp_soft, which was never implemented
03jepler 07TRUNK * 10emc2/src/hal/drivers/pluto_servo.comp: whitespace fix
03jepler 07TRUNK * 10emc2/src/hal/drivers/pluto_servo.comp: doc fixes and layout tweaks
cnczone.com seems to be down. what am I to do? work?
that was strange
I am sorry.. I wasn't quick enough
me too :)
shame on you! you had 20 seconds
I love poutines.
SWPadnos, how do you like poutines?
I don't know what a poutine is
SWPadnos, don't you have family in Canada???
I mean, no
just the thing to introduce to the US
poutine... fries, gravey, cheese.
hmmm. I see that
I think I don't like them
but I haven't tested the theory
cradek, anerica's starting to get them??
[18:09:42] <jepler> http://en.wikipedia.org/wiki/Poutine
I used to live in Montreal, they fed me that stuff during my first week, I will never recover ...
no, but americans are fat enough already ...
a-l-p-h-a: not that I've seen
A dish made with French fries and melted curd cheese and topped with gravy ?
"a dish consisting of french fries topped with fresh cheese curds and covered with hot gravy (usually brown gravy) and sometimes other additional ingredients."
Dallur, but you sure as hell enjoyed it didn't you? Other than it took 5 years off your life.
I must say that I do like "Pizza Skins" at Pizzeria Uno
what's a pizza skin?
sounds like putain to me :D
I will eat curly fries with "nacho cheese sauce", so how bad can poutine be?
nice flaky crust with mashed potato, cheese, and bacon bits in it
the only thing I can use to describe it is : soggy fries
soggy fries with cheese
jepler, curly fries, the spicy ones?
those are yumm...
I sometimes like cheese-covered fries, and there are a couple of local places that make very nice fries/gravy
the ones from Arby's are yummy... but aren't those just McCain's curly fries?
I'm just not so sure about cheese + gravy
SWPadnos, so you have had one.
sounds kinda like the kfc bowls. just dump a bunch of junk in the bowl
no, mashed ~= fries
ooooooooh... know what else is good.
sweet potatoe fries.
is that agreeing with me? :)
or nice mashed sweet potatoes with lots of butter, maple syrup, and/or brown suger
that's a lot of sugar.
I hate corn syrup.
I <3 maple syrup.
I didn't know until recently that corn syrup was sweet.. for some reason I thought it was an oil. Odd
should have the 'syrup' given it away? ;)
skunkworks, you're americant?
corn syrup is a staple in you foods...
my word for anything american is americant. :)
I need to make a macro that parses my for the word "american" to append a t.
just must remember to disable it when typing real letters.
as long as it also replaces Canada with "Canuckia", it should be marketable
I think the problem resently is that america can.. And does.
SWPadnos: nahh just remember to pronounce it Kanada (like Nevada)
SWPadnos (can nada)
skunkworks, can and does... but not very well. :)
skunkworks, iraq war.
skunkworks, keeping jobs in America.
I'll stop... it may hurt too much.
a little update on this thread for jepler. nothing really new
[18:21:27] <skunkworks> http://www.cnczone.com/forums/showthread.php?t=33853
this is interesting http://www.cnczone.com/forums/showpost.php?p=271373&postcount=29
for most excellent web hosting: http://www.nouptime.com/
200kb of storage???
10kB of bandwidth
try the support page :)
"Your first name is wrong".
they should have the submit button move when you try to click on it.
hahah... it's a random number when you refresh the home page... "currently hosting over X websites".
same with the happy customer quotes
03jepler 07TRUNK * 10emc2/src/hal/utils/comp.g: fix stripping of dash before hash in c identifier names
jepler: short question .. :)
why doesn't the hal_input manpage appear on www.linuxcnc.org ?
err.. sorry, was looking at 2.1
I see that devel has it
alex_joni: yeah, it hasn't been backported to 2.1 yet
I've only talked about doing that
jack seems to be digging himself into a hole.
* alex_joni swallows a reply
rtFm regarding /dev/rtf :)
guess I should start packing :/
look at the bright side: at least Germany is nicer than Manchester, New Hampshire
sure hope so
I'm sure it'll be true this weekend: http://www.wunderground.com/cgi-bin/findweather/getForecast?query=Manchester%2C+NH
oh -oh something else messed up now.
jack@rex:/$ sudo mknod /devrtf3 c 150 3
for gosh sakes, read the screen before you post
heh - but where was the typo?
email, command line ...
are you going to ask the question or shall I?
seems paul is helping him (off list) so he can handle it, it's nice of him
hmmm. didn't see that (in this mail)
I did notice the word "fedora" in there though
did everyone have some pi today?
no pi here
here, have some of mine
thanks - I'm not too hungry, so I'll just have .14
you can have the 3.0015926535897932384626433832795 back
[20:23:26] <alex_joni> http://3.141592653589793238462643383279502884197169399375105820974944592.com/index1.html
alex_joni: your pi's bigger then mine
it's longer not sure about bigger
longer = bigger
oh.. yes it is
happy pi day
yours ended in 795, mine is 7950288
it's steak and a BJ day too :)
you have to call my GF and tell her what today is!
[20:25:10] <SWPadnos> http://www.steakandbjday.com/
/topic < a-l-p-h-a> blow job?
why, that's for chip clearing, of course :)
I need to setup my lathe CNC power supply again... and pump out some lathed parts...
did all the prep work... just dunno how this is gonna go.. I have a 72VDC power supply right... but the motors are rated 2.8VDC 2.8Ohms.
is that too much voltage with geckos/motor setup?
it's a little high, but should work
if the motors get very hot at 0 speed, you should use lower voltage
err - was that right?
42VDC is my other transformer.
72V is only a little over the 25x ratio they recommend, so it should be fine
I'm looking for an email on the geckodrive list that tells what to look for regarding reducing supply voltage
but 72V is a bit close to 80VDC
I thought the ratio was 20x.
shouldn't be a problem, assuming proper supply / safety design
a-l-p-h-a: what kind of current on the motors?
the range is 20-25x
a-l-p-h-a: 5-30x is OK
ok - recommended range is 20-25x :)
the higher the voltage the more you push'em
2.8VDC/2.8Ohms... V=IR, there for I = 1Amp.
umm... no clue what amps I'm pushing.
actually U=IR, but I get the point :P
increased voltage increases top speed and the corner frequency
only in Romania ;)
V = Voltage. I = amps, R = resistance.
actually it's E
[20:40:36] <a-l-p-h-a> http://en.wikipedia.org/wiki/Ohms_Law
I = V/R
or conversly V = IR
oooh... sometimes E or U, instead of V
V is potential, yes
in my deranged mind are "V=IR" and "P=IE" which don't even use the same letters
E is actually electric field in those equations
but I learned that difference of potential is called U
right - P=IE, E=IR
potential is always a difference
so I guess I learned it both ways
SWPadnos: and it wasn't a romanian school
U = Va - Vb
I've never seen U used, but often E/V are interchanged
defensive. :) lol
I was kidding about P.
oh - that looks like an algebraic/calculus variable substitution
[20:44:15] <alex_joni> http://de.wikipedia.org/wiki/Spannungsabfall
it's always U in german
SWPadnos, any luck with that email?
what to look for.
geckodrive list that tells what to look for regarding reducing supply voltage
oh - that one. I got sidetracked ;)
happy Pi day!
happy steak & bj day
she's just so loose, I wouldn't do her...
well... if I was naked already, and she was... probably.
but still... she's been in so many pornos, it's just scary...
a-l-p-h-a, try looking at the message from Mariss titled 'The "20 times rule"', dated 10/30/04
I think the deal is that if the motor gets too hot at high speed (no load), reduce voltage. if it gets too hot at rest, decrease current setting
03jepler 07TRUNK * 10emc2/src/hal/utils/comp.g: fix stripping of dash before hash in c identifier names
that looked familiar
that message should have been: add experimental 'personality' support, which can control which pins exist and how large pin arrays are
cvs --oops --newmessage="DWIM!"
there is 'cvs admin' to replace log messages, which I did right away
but it can never fix the commit message shown here or in the commit e-mails
oh - I noticed something in the emails. where is the colorization of the path/filename done?
is that at navi.cx or on the CVS server?
e-mail doesn't come in colors
no but the IRC script does some coloring
sorry - the CIA messages
yes I think that's done by navi.cx
<tap> <tap> <TAP> ... is this thing on?
ok. it looks like there's a problem with names that have underscores in them
jepler, good evening
any that don't have underscores, the path is one color and hte name is different. in those that have underscores, the path and filename are the same (uncolored, I think)
I turned off color in my irc long ago
SWPadnos: not here..
here they are always the same (path coloured, filename not)
hmmm. you can see it in the logs on linmuxcnc as well (though the colorizations are stored as chars 00-xx)
if I would bother.. prolly
alex_joni, its not a good evening?
* robin_sz thinks it is
im home from work, and just got changed into something comfortable
I feel all happy and relaxed now
ah I'll go home in just a few minutes
03jepler 07TRUNK * 10emc2/src/hal/components/logic.comp: component to demo new, experimental 'personality' feature
it makes a nice change at the end of a long day covered in dirt from metal, to get changed into a nice dress
ok - that one wasn't colorized like the previous one
today, I am hating powder coaters.
SWPadnos: just the same colours as always here
actually, no its not just today, its an ongoing thing
robin_sz: electrostatic ones?
not the equipment, the companies that do the work
we bought 25kg of yellow powder to do a job that SHOULD have taken about 1kg MAXIMUM ...
they used ~18kg
at £140 a box, thats expensive ...
this was 200 small parts, 200mm x 100mm
what that 4 sq m?
hmm.. pretty small
60-70 um ?
[21:48:47] <SWPadnos> http://www.cncgear.com/Files/colors1.png http://www.cncgear.com/Files/colors2.png
SWPadnos: the first one had the same colours (even if there was a _ in there=
actually, there wasn't an _ in either of those, but the colorization changed anyway
your client is borked
I've noticed odd things that I'd expect would be taken care of by TCP, but who knows
that looks like a byte is missing from my receive stream
if it's repetitive.. then I doubt it
it may be more random, but there seemed to be a pattern yesterday
I've noticed the problem before, but not a pattern
alex_joni, basically, they must have either lost or stolen about 15kg+ ... I hate it when that happens
how did they justify that?
not spoken to them yet
"hey, it hurts when I do this" ...
what an idiot
I mena yeah it was funny ...
but he was lucky not to ge tsmacked on the head by the hose end when he fell off
and to not fall off onto his head ...
bike helmet would have been an idea
robin_sz: staying off would have been a better idea
I suspect hes not THAT bright
it's a real bummer that 17" LCD laptops are so heavy
still waiting for mine
SWPadnos: 21" :P
those are even heavier :P
I'm always torn between light weight and computing power
see.. then the 17" are lighter
VAIO VAIO VAIO
for MMTB people
heh - I went with power last time, and it's still a viable machine (6 years later)
but it weighs ~10 pounds, and gets heavy after a while
this is an 833mhz ... not too bad
it gets heavier?
mine is a P3 1GHz / 512M / 40G / 1400x1050 15.1" display
go schumi :)
so it's stillusable, but the video is lacking, and it's not great fopr FPGA synthesis / place+route
mschuhmacher: too bad you retired :D
I´m not the Formula 1 Schumacher im a Schuhmacher :-)
hmmm. then there's the problem of legacy ports ...
mschuhmacher: so you're making shoes? CNC manufactured?
I dont make anything, but maybe in the Future...
I´m a newbie
in cnc retrofitting
learning is half the fun
or all of it in some cases :)
not that people always make racing jokes when you're around :)
who's Michael's little brother that rides his coat tails?
ALEX is that YOU????
robin_sz: nope, I'm not that crazy
thats awesome ...
big FANUC bot
heh - the reaction to that video is probably very different here than in your average IRC channel
I'd rather ride the Robocoaster
robin_sz: done properly
[22:28:33] <alex_joni> http://www.robocoaster.com/video/video1.html
[22:29:28] <alex_joni> http://www.robocoaster.com/video/video4.html
O M G
way way way too crazy...
his head was sooo close to the ground...
and his feet drag on the ground too!!!
CRAZY!... I would NEVER ride that thing.
as it slaps him repeatedly into the floor ... ooops, programming error ...
his head was square
still, thats top fun ...
they guy just "did it" ...
China is awesome... love tea.
nice with milk and sugar :)
seriously... I love coffee and tea... in HK, there's called a "ying-yang", it's half coffee, half milk, with condensed milk. It takes some getting used to, but I like it.
best of both worlds.
there's a drink* called
that G204 mentioned, is this the G203 now?
I don't think so
the G204 is the "vampire" drive, which hasn't been released yet,AFAIK
the G203 is an "in between" version I believe
but I could be all wrong on that
fired an email off to Mariss... rather not call...
he's very friendly, and I pay phone bills.
alex_joni is now known as alex_joni_away
03jepler 07TRUNK * 10emc2/src/hal/hal_lib.c: CONFIG_PROC_FS makes no sense on SIM but at least one oddball system (debian stable) defines it in /usr/include/linux/autoconf.h
03jepler 07TRUNK * 10emc2/src/hal/classicladder/classicladder_gtk.c: make 'load' button work
03jepler 07v2_1_branch * 10emc2/src/hal/classicladder/classicladder_gtk.c: merge rev 1.19: make 'load' button work
cradek: I want to note that fix in debian/changelog on the branch, but I'm not sure what to do. Do I want to use 'dch -i' and increment the version to 2.1.3?
sure go ahead
I have some to add there too
03jepler 07v2_1_branch * 10emc2/debian/changelog: note classicladder fix
/* init all structure members */
hal_float_t maxcmd_dd;/* param: limit for 2nd derivative of cmd */
hal_float_t cmd_dd;/* opt. param: 2nd derivative of command */
hal_float_t ff2gain; /* param: feedforward 2nd derivative */
is this a bug?
you mean, because those items are not initialized?
are you having trouble with ff2, or did you see a warning when compiling, or something else?
I dont know if it works correct
I agree that the others are set to zero, but ff2 stuff isn't
it would probably be best to initialize those fields, but I think it turns out those bytes are "zero" anyway
I use ff2 on my lathe successfully
(however, I do not see any guarantee that hal_malloc() returns a buffer which has been memset() to 0, so it may be a bug)
want to explain the problem you're having? maybe I can help
jmkasunich: is the intent that hal_malloc returns a pointer to zero'd memory?
I want to use emc2 and therefore I read the code
mschuhmacher: that's great, code always benefits when others review it
mschuhmacher: once I find out what jmkasunich thinks (he is the creator of emc's hal) I'll either improve the documentation, or add initialization for those fields, or both
mschuhmacher: ff2 was added later, so it was probably an oversight that those fields weren't explicitly set to zero