[00:18:14] <dmess> http://www.dataq.com/products/software/linux/index.html
check this out
that sounds sweet Tomp2
looks good, will take some time to read all that, i got the 'LabJack' http://www.labjack.com/labjack_u3.php?prodId=53
i need to buckle down and make a driver for it
the big block is that theres no generic usb driver for emc ( i have nothing to hack up), and if the device isnt already seen by linux ( like uhci stuff ), then it's pretty much new territory.
but unfinished business # 378... i didnt finish the vismach machine i started last nite :-[
hey, but thx for reminding me, i see theres an OpenLabJack project :)
i may order 1 to try just for shits and giggles
theres a sweet lookin o'scope for 1500$
have you ever dabbled in X10 stuff??
i got the x10 barking dog alarm ( prox switch sets off doberman recordings :)
runing thru heyu??
no, old w98 stuff
oh i see... im wired into the building on THIS emc box... no machine attatched... ; ) thk god
i want the heyu idea tho, my eeepc responds to 'computer internet' and 'computer email' ( the app is not open src :(
you talk to it
devices listen.. if thats what you mean
[00:43:19] <dmess> http://misterhouse.sourceforge.net/
too late i think
[00:46:18] <dmess> http://www.aartech.ca/
everything X10 for you my friend
? "We have some of the best pricing in Canada and pricing on par with the USA on most of our items." what, they offer to charge you more? USA is more expensive ;)
now it is..... we used to bend over for 30%.... i am in canada
im more in line to push it a little harder if there's 30 more point in it for me
i'm trying to get my sons to sell and install a starter system door to door 2 lamps 2 appliences a cheap pc for 300 bucks
print up a pdf that show how the buyer will save money ( see 'living like ed' )
I was just looking up links to my company ablap
and i found a chat link to here
OK, I've hit it again....
I did not know this exhisted
what do you do for work skunkworks?
I though one was free to jog manually wherever, but (new Y axis setup ) seems to be stopping at what EMC thinks are the limits.
Which is a LONG way from where I want it...
And with the table where it is, it will require some jerking around to get it physically apart to manually move the screw...
Why do they have the "Override Limits" checkbox when you check it, it immediately resets itself to unchecked?!?!
are they soft limits or over travel switches
I guess I just run the thing until it strips it out of the coupler...
K`zan, on my machine, I just set the limits out of the way. Using axis, I can verify where its going before I turn a program loose.... on tight setups, I even jog around the perimeter to make sure cutter clears clamps(watching both axis and tool path)
LawrenceG: I just rebuilt the Y axis and when it came back together it was a long way from where axis thought it was.
I guess I could have gone back in and reworked the config file to limits that would be useful.
do you have a home routine?
But, running it the only way it would go nicely stripped the screw out of the coupler :) and took care of the problem.
Yes, EMC thought it was at home...
Lemme go build another coupler, bbiab.
understand.... manual cranking is required or open up the soft limits
was it close to the home position when it thought it was there?
Nope, major change and the screw got repositioned somewhere in the middle of nowhere (according to EMC :) in the process.
the limit override only works when you are using limit switches.... it then lets you jog off the switch
Only way I could think of toget around it otherwise was to go back in and re-do the configuration file, I guess to reset the limits to something useful.'
Getting single shaft steppers might have been cheap, but I now question the wisdom of it...
I kind of like linear motors
maybe you can add a crank to the far end of the screws
they never break
So would I with a bigger budget :).
trust automation makes some nice amps
THis thing is errr, afro-engineered as a learning process (maybe a bit of engraving or PCB etching) and it sure as hell is being that :)
Iam not much help this is my first time on the site
i am use to delta tau
Off to make up another coupler....
DASVT: I am IT/MIS. cnc is a hobby
DASVT: Take a look at emc2. pretty damn powerful.
well i am all ears
I have one of the highest precision machines in my shop
I work in the machine tool research center a GE global research
should not be for 350 k
hydrostatic bearing for the slides
but i am interested in any open cnc
I want to build a precision cnc grinder to make linear bearing in my shop
I might consider the EMC
i had a Toshiba F-Mach for a while..
I have not seen a toshiba
i read some papers on them in the past
are you a member of aspe
how dos the emc compare to fanuc controls in performance?
aspe? The American Society of Plumbing Engineers? http://www.aspe.org/new/index.php
american society of precision engineers
I figured that might be the one :}
google finds a lot of aspe groups!
[01:46:50] <DASVT> http://www.aspe.net/index.html
DASVT, welcome to emc
they are a good bunch of guys
all the national lab guys are members
yadda yadda yadda
how accurate have you seen an emc machine
I am a retired software engineer.... used to do industrial scanning and optimization controls for the forest industry....I love to watch computers run machinery
CNC chainsaws? ;)
precision finger removal
actually... I had an idea for a firewood product that was a computerized chainsaw that climbed trees limbing on the way up and bucking the tree to 16" lengths on the way down.... raining firewood
this is a desperate geek chat
sounds ok... but i hate to see your umbrella
hardhats may not be effective :}
my machining skills start to expire when approaching 1 thou or so
have you used emc
DASVT, where are you located?
I have a mill/drill/lathe running emc since the late 90's
looks like Albany ISP, but you have VT in your name, so I get suspicious that you're in Vermont.
Vancouver Island off the west coast of Canada
opps.. wrong answer
I'm thinking I may have been in the shop where DASVT works recently :)
or one like it :)
tenths are still magic for me
I am just amazed at some of the precision stuff in hydraulic valves
it's kind of amazing that hydraulics work at all
hell of a mess when they dont
heh - I've noticed that, with a Greenlee hydraulic punch :)
Ok Got that mess fixed!
Break time. put up tools and get coffee.
any hints for how to color a collection in vismach ? ( i gots zombie windows with disbox = Color([disbox],"green") )
is vismach in the emc source or distributed separately?
its in the bin dir, its pythin, so yes you got the src
came with 8.04/emc live cd
ah ok - lib/python/vismach.py
duh, yeh the bin dir they call lib :-[
I think it gets copied to emc/bin when you make, but in CVS it's lib/python
k, and color seem to be rgb like (0.5, 1, .377) not tcltk names
looks that way
ah, the zombie window dies! drill it with xkill muhahaha
sudo kill --silver_bullet zombie_task
wooden stake crucifix holy water die die die
I think we're both mixing vampire remedies with zombie remedies
I hope I don't meet any zombies (unless they're also vampires), because I don't remember how to kill them
can you imagine me with a zombie wife ( kingston trio 1960 )
I don't think that was on the album we used to have
i was searching the web today... and came across all the usual items. has there been any milestones in something like mastercam functionality in linux?
so emc2 is working great on my Ubuntu Studio 8.04 64 bit box although i have yet to plug the engraver into the lpt port and power it up.
no more 333 mhz gateway control boxes!
BigJohnT: thanks, I've got things working with stepconf's test
SWPadnos_ is now known as SWPadnos
Anyone had problem with electrically weak PC parallel port, that the communication exist only into PC, and no out
we develop IO card, with stepper generators, ADC converters and so on, and It works only with parrallel port on HP computers
how isolated is your card?
if you're sourcing a lot of voltage from the paraport, that might be a problem
we not sourcing any voltage from parallel
but card is double opto-isloated
I dunno. I guess you could check the voltage and amperage outputs of all the computers you're testing with, and see where the difference is
realy, it shouldn't matter
I've never heard of a parallel port device that only worked with ertain brands of computers
we develop and run that card on HP PC
and we dont have HP to connect to laser and we choose thoshiba
and communication is only into PC not from PC
but we check volatage and amp that you saying, thanks
micges: do you have the schematic online somewhere?
only in EAGLE program
eagle is fine for me :-)
tomorrow in work :)
If you can provide the schematic it's easier to have a look at the error
But the most commonly mistake is tou source current from the parallel port instead of sinking int into the port
must note to check tomorrow
for the outputs, do you have the anode of the opto's connected to the port and the cathodes to the ground?
if so, turn it around, supply 5V to the anode side and reverse the logic in the software
Should the USA go metric? (1 2 3 ... LastPage) Replies: 205 Views: 7,620
gotta love these web-bbses
(yes that's the most popular thread)
cradek: where was that again?
huh. I guess the owner of that blog wasn't alive in the '70s
hmmm. or maybethey were :)
sometimes people post pictures of projects. that can be neat. but generally, the signal to noise is much lower than our lists
that just seems like a question that anyone over 30 would know the answer to
micges: in my opinion, it's better to design parport interface boards so that only tiny (logic-level) currents are ever sourced or sunk from the parport pins. Use single-transistor inverters or standard logic drivers with specified source/sink currents between parport and optoisolation. otherwise, there's just too much variation among parports to create a reliable product.
SWPadnos: ok you win, when will you have it ready by
I am sure you guys can bang it out during the emc fest.. :)
as long as SWPadnos can find a gigbit hub
giga is bigger thatn deca
you're talking small potatoes - decagigabit :)
I could cara lessa
hey - when did you start channeling Giacus?
careful there -- "FAST is a Registered Trademark of Fairchild Semiconductor Corp."
07:18 < cradek> Should the USA go metric? (1 2 3 ... LastPage) Replies: 205 Views: 7,620
engineering is base 1000
mils/inches is base 1000
and a mil is like the smallest shit i can see
renesis_: dont hold back, say what you really mean.
no im done i already won the argument
as I said, anyone over 30 would know the answer
aint gonna happen
anyone over 5 [years of experience on the internet] knows not to read or write in that thread
morning mr radek
i gotta solder in more scsi wire into like 20 more .012 vias
ima break down the taigthing and clean it up see what kinda room i got for an anti backlash nut
maybe you can work only in inches/mils, but in my part of inchland, people write measurements like 19' 8 3/16"
heh, i started referencing after flips approaching from the same direction, now i only get 1x backlash error instead of 2x yay
whats 8 3/16
3/16 is .375
so what thats 27.375"?
236.375 inches or 6003.937mm
no, that's nineteen feet plus eight and three sixteenths inches (5.999 meters)
oh right the '
wtf feet ha
like im an achitect or something =\
all thru highschool science everything was metric. I really did like it better. but after highschool...
well - and college.
science and engineering is base 1000, mm/cm/m isnt
if I think the original question on the webbbs was a troll, does it make me a troll too if I mention it here and unwittingly set everyone off?
cradek: no, it means you're foolish
so really, i dont care if you can convert cubic centimeters into mass of water
jepler: which is worse?
cradek: one is 12 times worse than the other. or maybe 16.
whats wrong with beeing a troll
debate is good for engineering
can anyone tell me what size drill you use to make a hole you can fit a #4 screw through?
biased arguments help to flesh out possibilities for implimentation and failures
like .13 i think
yeah wtf at drill naming, can we troll abou that?
but yeah i been poking those holes in PCB alot for last project =\
close, the answer is #32
i use endmills to do circles when im doing my cutout pass, shrug
okay so whats #32?
if I'm foolish enough to participate I'd make fun of these american schemes, not inches (there's nothing wrong with inches as long as you avoid using fractions).
fractions is okay because it matches peoples rulers
renesis_: they need new rulers then - mine has 0.10 ... 0.90 in each inch
okay well not everyone has engineering rulers with like 6 scales
you know what's worse? when people use BOTH
heh this is fun
liek if im laying out some shit that needs to be mounted, i try and stick to a .0625 grid
* cradek 's IQ drops 20 points
wood here in peru is dimensioned by inches, but lengthed in metric
and of course they quote 10% tolerance
well here does this piss you off more...
Sweeper: don't worry; a 2x4 isn't actually two inches (5.08mm) on the short measurement
jepler: here it IS
er, 50.8mm I may mean
in layout for pcb, standard unit is really like 0.1"
Sweeper: oh that's crazy talk
that threw me too
and people will fraction off that
for their sub grids
12.5, 6.125, 3.0625 mils, heheh
i try and do like a .025 big grid and .005 subgrids
sometimes its frustratuing not to be able to hit the exact middle of your grid
well, for me at least i dont think anyone else cares or notices, heathens
considering that most SMT ICs have metric pin spacing (like 1mm, 0.75mm, or 0.5mm), you also need a metric grid
unless you only do clinky huge through hole stuff
you do that in lib editors, use whatever grids and sub grids work
naw, qfp lately
if you want traces to connect to the pins, then you need a routing grid that matches the ICs
soic is sane tho, i <3 it, .05 pitch
my app does that automatically
actually eda for kids (diptrace) does too
of course, copper connects should be detected
diptrace has some fucked up algorithm that corrects for you at low grid levels
with fine pitch, having an exact grid is very useful - it's a lot harder to route between pads if you're off grid, for example
which is total ass for precision
um, wtf is a trace grid, heh
ill drop to .001 for traces if its an issue (never is really, maybe final DRC pass)
yeah pls to explain your acronym or i cant respond
oh, google knows
ir2 is now known as ravennb
i dont even know how that makes sense =) finer grid isnt gonna chance space/trace settings or anything
anyway i have to go relabel a ton of schema parts =(
which switch do you user for your cnc machines as limiting switch?
mine are a little small (If the machine runs with to high speed in it they're broken)
heh, i was gonna test some cheapo microswitches to see if they are consistent
yes they were cheap
Nice job ravennb: normally you use some sort of ramp so that the machine can never phycally break the switch.
usually people will use switches that have rollers, and not put them "in the way"
its the same stuff they use for arcade machine buttons, comes in all prices =\
like this http://timeguy.com/cradek-files/01198731403/limitswitches.jpg
I think mounting limit and home switches is a bigger challenge than picking one that will give acceptable repeatability
ha yeh those
well yeah but i havge no clue what the range is, so i want an idea before i even try
the roller/lever ones i would think kinda bend over time with use (they do in joysticks)
[15:12:06] <ravennb> http://cgi.ebay.de/Taster-Drucktaster-Offner-schwarz-TSO-11_W0QQitemZ120189378974QQihZ002QQcategoryZ31109QQrdZ1QQssPageNameZWD1VQQcmdZViewItem
this switches do I have^^
they shouldn't - you shouldn't be running into limits all the time
yeah I doubt cradek has tested those switches over 10000 activations
yeh thats true
you'll run into them once per session if you use them for homing as well
ravennb, those switches are crap :)
very much so.
that those are use interface stuff
yes I´m not running all the time into the switches but If it happens one time with to high speed they're broken
or just dumb sensing, not tolerance critical
You would have to mount them so the hard limit would be hit before the switch would get crushed.
I've found that out by myself... now I search for new ones which I could mount like these
those don't have enough travel for that though
hehe, thatd be like an adventure every time you home
and if you try to use a cam, the actuator is likely to bind (since the cam would be going across the intended activation direction)
i wanna buy a few types of cherry switches
because i like their logo
and then test to see if price is proportional to precision
SWPadnos: true.. there is not spring to those switches.
hehe, their keyboards are awesome, each key revieted into a pretty impressive steel plate
(after they make contact)
makes them impossible to fix if you spill into them =(
actually, there's a tiny bit of overtravel (the contacts are on the side of the barrel, and the plunger sticks a contact between them)
but they're noisy as hell, and don't last long
[15:16:40] <SWPadnos> http://cgi.ebay.com/Microswitch-BZ-2RL2-enclosed-limit-switch-roller-NNB_W0QQitemZ250032974470QQihZ015QQcategoryZ55834QQrdZ1QQssPageNameZWD1VQQcmdZViewItem
hey that Hardy install problem I have http://ubuntuforums.org/showthread.php?t=600126
looks as though not just Hardy
even better: http://cgi.ebay.com/SWITCH-CHERRY-E13-SPDT-roller-limit-snap-action-NEW_W0QQitemZ310052793634QQihZ021QQcategoryZ111606QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
[15:20:17] <renesis_> http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=CH139-ND
oh those are nice (kinda ugly, but whatevers)
doesn't everyone have about 100 of them in their junk inventory?
wonder if they are used
i have arcade button switches =)
ok thanks for the tips
[15:24:56] <renesis_> http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=CH301-ND
7779K52 McMaster-Carr $3
i wonder what brand
mcmaster is so sneaky about that
hmmm, let me see
they are Omron
K`zan, good morning.... running out the door to coffee with the local hams
oh omron is good shit
LawrenceG: Looking at the real estate listings. Hope you had fun at the coffee :)!
is there anyway to dispaly in realtime either in Axis or some other method the "error" in the PID loop? I'm trying to tune in my PID loop and i just think it would be nice to see the difference of the set point to the real point, well the difference so I can see if i can understand more about what's wrong with the PID setup
pid.0.error, I think
look at it in halscope
oh cool, its graphical too
well i'm not downstairs, but i'll have to check that later, thanks SWP
sure. pid.xx.error is the hard part though
there's a good start to a tuning page on the wiki - I think it tells how to set up a good test environment and what to look at in halscope
then it stops, just before telling you how to actually tune :)
[19:40:37] <SWPadnos> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Tuning_EMC2/HAL_PID_Loops
oh, maybe it doesn't tell you what to look at. oh well (you can update it when you figure it out :) )
some new milling videos on youtube. mostly the same part and the same program as my previous video though...
semi finish: http://www.youtube.com/watch?v=Ulcbgo1tO8s
schrupen vorslichten schlicten? sounds like edm :)
lewin1 is now known as lewing
awallin: am I seening about 5000mm/min rapids? (196ipm)
awallin: very nice videos
skunkworks: hi, yes I believe the rapids are set to 4000mm/min or 5000mm/min. that would be 157 ipm or 197 ipm
rapids are mostly for show :)
awalli1: looks great
alex_joni: thanks, I couldn't have done it without you ;) next stop is fitting a spindle encoder and trying rigid tapping.
me? you must confuse me :)
I'm the regular loudmouth.. trying to put in my 2 cents usually ;)
ok so without the regular #emc crew which seems to offer knowledgeable 24h support on all issues
awalli1: ok, that's more spot-on
on that note.. I'm off to bed :D
ha. good night.
awalli1: you need to add those videos to your post on cnczone
skunkworks: hm, I did a video post quite recently, but there's a much older post about the machine somewhere
don't have time to follow the forums really, all I read is email and rss feeds nowadays
ah - ok
I was referencing this on your youtube video http://www.cnczone.com/forums/showthread.php?t=40018&page=2
thanks for digging up that post! I can add the youtube videos there. Those old things were done with stepper motors. Although I think we had ballscrews and linear rails then also
the machine was really nothing to show off when it had acme screws and dovetail ways :)
any suggestions for webcam streaming server for a local lan ? ( to watch the machine using vid4l device from different sites around the office)
i'd like to stream to a port so even the W$ boxes can see it, the only app i found so far is 'webcam_server' and it uses a java i/f on a web page, and the w$ boxes cant handle java5 or 6.
tom1, u could try 'motion' under linux, it will stream video as a mjpeg stream that u can watch in a normal browser