i guess i could add in some gcode to have a pause for 10 seconds or so
back in a bit, gotta run into town to help my folks with something.
workaround is to press start (r shortcut) and quickly press pause (p) and then press step (t) until your safe and then press (s) to continue
even better is to use optional stop, and make sure the button is always pressed
but it's too late for that :)
I see that more and more people hit this run up to 1000 bug
first I've heard of it
(though I don't actually have a CNC at this point, so that's not much of a data point)
yeah surely someone should fix it...
lilalinux is now known as lilalinux_away
SWPadnos: You can be an Uncle to mine if it helps =)
SWPadnos: ...or an Aunt, your choice =)
SWPadnos Know of a way of getting a "time to go" to display in axis? IE how long the whole program has left to run?
or anybody else for that matter
is there a log that tells me why a desktop icon doesn't do anything in ubuntu?
Valen: you can get an estimate of total time with File/Properties. Then you can just eyeball the scrollbar to see how far along it is.
cradek, problem with that is if you have a bunch of curves and straights,
IE cutting down to a level then contouring
yeah it's not perfect
the curves have loads of code but don't take much longer than the straights
I just figured I couldn't be the only person who thought it would be a nice idea ;->
it's one of those things that you can't get right - it's irritating to write code that makes guesses/estimates
so we leave it to you to guess/estimate
in reality you're probably better at it, especially if you know the program
you generally get used to knowing how long it takes to remove metal.
and can ballpark.
if your doing the same thing many times
no, not really
i find no matter how different the part is i can generally guess how long my program is going to take just based on the volume of the area to be roughed, and the shape of it
and then the surface area to be finished
ok, perhaps if you have been doing it for more than one part ever ;->
how bad would a guess based on the estimated total run time and the time elapsed be?
depends on the program
depends on your feed override setting
depends on your machine's acceleration
you could factor feed override in
I would have thought the run time estimate would include acceleration?
or is it just feed rate * distance?
no, not currently
you'd have to preplan the entire motion to get better
it's just not worth it
but even then its not really going to matter
It'd only be out by a few % I'd imagine?
assuming you aren't running some anemic wussbag of a machine ;->
(us all being real manly men here)
a simple "fudge factor" should get it fairly close I'd imagine
I mean if its out by 5 minutes on an hour program its not really going to matter
I bet you're right and you could have a somewhat useful time-remaining estimate. but - I'm not going to do it, because it's a sucky problem and I don't think the feature is really worth the trouble
no matter how long it takes, if you need to cut a thing, you're gonna have to suffer
I'd be happy to review a patch if someone else wants to tackle it
yeah but it might let you know if you have time to go to the loo or time for dinner ;->
where would the patch look at going?
somewhere in axis I would assume?
I'm not sure how it would work
probably guess a time for each program line, then watch as the program executes and subtract the times?
loops etc make it harder of course
and various things you wait for: custom M codes, spindle at speed, machine acceleration, probing, etc etc
I would try and use the motion estimation
and just not worry about all that other stuff
loops are a big deal
you can't add up all the lines after 'where we are now' and give that as the remaining time
you have to understand the whole program flow
yeah but the motion estimate factors those in I believe, because it shows them on the plot
so not actually use the currently executing line as a base
that's all you have
My theory was to, use the estimated time, that seems to be calculable as axis does it now.
yes you have that
but you don't know how far you are in
when things are "running" count down from that time to 0, just based pureley on that clock
if you "run from line" or such it'd get pretty wonky
but any other mechanism is like you said, really hard
unless you pre-plan the entire motion, but if you have a spindle feedback and slow down while running to maintain the spindle speed your screwed again
have you come around to 'just look at the scrollbar and estimate' yet?
no, I resolutly stand by my dodgyness ;->
touchy doesn't have a scrollbar, and I miss it
see you need to add some kind of time to go support to the lower levels of emc ;-P
[04:37:31] <Valen> http://xkcd.com/612/
is good enough for 90% of the planet after all
I like the one where nasa blows up the little prince
"If English teachers teach English and Maths teachers teach Maths, then Head Teachers must teach..."
[Global Notice] Hi everybody. #freenode-newyears is now open for all your new years celebrations. Happy new years from PDPC and the freenode volunteers and thank you for another great year.
lilalinux_away is now known as lilalinux
MarkusBec_away is now known as MarkusBec
MarkusBec is now known as MarkusBec_away
MarkusBec_away is now known as MarkusBec
SWPadnos_ is now known as SWPadnos
MarkusBec is now known as MarkusBec_away
MarkusBec_away is now known as MarkusBec
For several days now, I try to move to EMC. been using Mach up until now, and basically I got it all together, what I lack is:
1. the homing - I have all limits on one parallel port pin - 15, and I try to use it for homing, but really no use
I routed the signal to pin 13 for homing, and pin 15 for limits, using all-home parport....pin-13-not and all-limit parport...pin-15, the signals come in correctly, the ini file has been scanned for errors numerous times, yet every time I home, I get the limit switches stopping the machine - any idea why???
B.T.W: Merry Christmas and happy new year ::::)))
when the switch closes, emc sees a limit signal on all 3 axes. HOME_IGNORE_LIMIT ignores a limit only on the current axis. You can't use a single input as "all home and all limit" (that's why there's no such option in stepconf). If you can't have separate inputs, then the best you can do is "all home" and no limit switches, or limit switches and no home.
er, I mean HOME_IGNORE_LIMITS, got the spelling slightly wrong
orotemo: look at section 0.1.3.4: HOME_IGNORE_LIMITS
orotemo: read this and follow jepler' instructions
Yeah, I did it all, and I thank you, but to no avail.
I'll go try again now, because I can see that my latest ini hasn't the HOME_IGNORE_LIMITS flag, so I'll double check. but you basically say it is totally possible to have one switch for each axis for both homing and limits?
sorry, it is only now that I read jelpers' first relpy
so you say I can't do that in emc
isn't it possible to use some ladder file and make the machine ignore all limits while homing?
that's probably possible. The pins axis.#.homing are TRUE when the respective axis is homing
so the short version of the story is that you would use axis.0.homing OR axis.1.homing OR axis.2.homing to mask the limit signals to all the axes
I don't know whether the long version of the story works out OK in the end
Sorry mate, new here, bbl?
he's be back later
Cheers, you guys helped me a lot!
heres' an idea for a startup, normally, I would get you guys a beer, but since we can't, a startup should make it possible to get a guy/girl a beer voucher at his favorite bar... how do you like that? :)
happy new year to all !
happy new year
* ries just woke up :)
HNY even :P
lilalinux is now known as lilalinux_away
alex_jon1 is now known as alex_joni
Hey all, new to emc architecture can anyone explain to me what EMC clips are in a centera environment?
wozer: I don't even know what your asking...
wozer: emc is a machine tool control. we dont understand 'clips' or 'centera' perhaps you are thinking of some other emc
yes, centera emc data management,,, this is the wrong plac efor yopu, happy new year
cradek: those solid state relays in the HNC... i was planning on triggering them with my 7i37, do you need to use diodes with those like with regular relays?
jt-plasma: thanks for the email the other day, that's a big help
happy new year everybody!
danimal_garage, no, there's no inductive spike with a solid state relay
since there's no inductor
SWPadnos: thats what i was thinking, just wanted to be sure
man, this stainless steel braided conduit they used on this machine must be pricey
it is, especially if you can find sealed stainless-sheathed stuff
which I have looked for but not really been able to find
hmm it might not be stainless, i think it's braided copper
maybe i should take it to the recycling yard :)
might get a 12pk out of it
or re use it and do a quality job
i dont need it, half the stuff is going in the cabinet where the conduit comes out of
plus it's all frayed
i'd rather just get some new conduit that i can easily cut to length, and doesnt look 32 years old
plus i'm out of beer :)
jimbo7555_ is now known as jimbo7555
good next year at all.in Italy are 22:45 .another 1,30 hours and go in new 2010.
same here =)
end of a decade.... oh well... talk to you guys next year
its only 13:50 local time here
we go start for fire
it's 3:53 here
i'm in 2010 now
how is it?
fire? no fireworks in Italy but throwing burning sticks in the air? =P
Athlocatle, good year
motioncontrol: thanks a lot!
cradek: i'm a little confused about the power supply in the left cabinet... there's a transformer, cap, and a plate with some rectifiers and what appears to be transistors, and a fuse. There's a series of blue wires coming from it, approx 14gage, and there's a multi contuctor cable, approx 25 gage coming from it as well. whats going on there?
not sure what the multi conductor wire is for
i know it went to the logic board, but i dont have schematics
Athlocatle: QUICK! What are the lottery numbers!!!
danimal_garage: does it have four big transistors? where do/did the wires all go?
yes, 4 big transistors
I remember seeing (removing) something like that
i'm not sure exactly where everything went, i'd have to look, but the wires were blue
the rest were red
well some went to the contact block below the ssr's
i think some went to the caps that are in there
the thinner gage wires that were in the multi conductor cable went to the logic board in the case... a few of the wires went to a switch for the turret
to raise or lower it
the only thing it could be for as far as i know is for the vari speed drive, the clutches for hi/lo gear, or the turret for something
greetings from 2010
happy new year alex_joni
four big transistors might be the varispeed stepper driver
(if it's a stepper)
it's not for the clutches
i think you're right
you know the wires are all numbered and they match the numbers on those terminal blocks, right?
are the clutches 110v ac?
yes I think so
Happy New Year danimal_garage
lemme look real quick brb
yep, looks to go to the varispeed unit
i guess i can just remove all that
is it just some sort of stepper drive of sorts?
no idea - I tore it out
yea, i'm doing that now, i just wanted to know what i was removing
for no other reason than curiosity
did you keep the air pressure sensors in there and working?
unfortunately I got bored by the time I got down to those details
there's a lube level sensor too I think
i cant see myself needing them, i'll be using a chuck or fixture most of the time
i guess good air pressure is still needed for the turret though, huh?
if I do a tool change without air, it sits and waits, and then I turn on the air
thats true, i guess emc wont get a tool changed signal unless it actually finishes the cycle, so it has a saftey mechanism by default
i wonder if radio shack is open today, i need some standoffs
so there's 2 prox switches for homing per axis, right? the fine one being the one that's on the cog. are those just 12v prox switches?
standoffs!? too bad you don't have a lathe to use
i have another lathe
but for what they cost, it isnt worth my time
i want ones with male threads so i can just tap holes in the pannel. doing threads on my manual lathe will take a while
i ran out of scrap computers to take them out of :)
I just tap the panel and put shcs through the board and regular nylon cylinder type standoffs
not a bad idea
probably stronger than those little brass ones i get from my computer cases
much simpler and less precision required, but it's a little finicky to hold all the parts at the same time
I drill/tap one hole, install the board with one screw, level/straighten it, transfer punch the other three holes, take board out, drill/tap
sounds tedious but it's pretty fast
spiral point tap in the cordless drill...
yea, i'll have to do that as well. on my mill, i mounted the boards in the removable trays that were in my empty disk drive slots in my pc case
so i just took them out and drilled/tapped them on my manual mill
so they were spot on
yeah that would be nicer
but i cant really do that with the lathe the way i'm doing it
the pico and mesa boards are going in the left cabinet
welli guess i can make a smaller board that they all mount to and just mount the board in the cabinet
the panel comes out doesn't it?
or too much stuff on it already?
yea, and i guess it'll be half empty by the time i get to that point
wouldnt hurt to pull it off
except i'll have to pull the vice off my mill
but thats just me being lazy
gives you a chance to get all that goop out from under it!
manual mill is clean :)
cnc on the other hand... flood coolant=mess
i gotta wire up the servo amp PS... what do you think that thing draws?
too much for one of those 10A SSR's to turn on and off?
HNY to you too frallzor
you can't even put 240v through those can you?
didnt know it was 240v till now :)
surge current on a transformer with a dead short (big empty caps) across it is going to be pretty darn high briefly
you better check - don't trust me :-)
hmmm i wanted a way to enable/disable it via EMC
hey hey, no cnc now! =P
only thing i have that can handle that is one of my magnetic motor contacts
doint spoil my moment!
but i dunno, seems kinda hack
danimal_garage: ding ding
cradek: good idea?
frallzor: hny! is it 2010 where you are?
i got a billion contacts
does emc still work or do we have a Y2.01k problem?
ok, well maybe like 6
I dont have emc at the moment, so I cant tell :P
is it standard practice to use a motor contact as a heavy duty relay?
or at least acceptable practice?
cant see it hurting much
I'd think so...
wonder why ppl celebrate ny so much
its not like its going to end :P
"oh noes, what if it doesnt occur next year" :
adios, and thanks
cradek: no problem with emc2 here
at least not until 2038
thanks for the email, big help
the wire colors are exactly the same as my resolvers
I just found it by accident
working on the lathe now
yea, even blind mice find the cheese every now and then :)
getting the power side of it squared away
I just dragged home another 10hp motor for a phase converter, I'm going to use the first one I found
this one needs to be rebuilt too...
what do you have that needs a 10hp?
the HCNC is 7.5 hp so I need at least a 10 to manage
your lathe and mill are under 3hp, right?
did you get the drive wiring sorted out
drive? as in the servo drive?
the lathe is 7.5hp 6000 RPM
cool... mine is only 2hp
my mill is 1.5hp LOL
not entirely, i did look into it a little, but i wanted to get the power side of the lathe done and the pic/mesa boards mounted first
4200RPM but makes too much noise above 3k RPM
lol mine are 3hp and 2hp
is your spindle just an AC motor?
that makes it easier
you have a servo, right?
0-10v velocity input
i want one :)
come an get it:)
on my way
i'd feel more comfortable rigid tapping on the mill with a servo as a spindle motor
if you have an encoder on the spindle you don't need a servo motor
I'm never comfortable rigid tapping LOL
yea, i will, but a servo would change direction faster
didn't think about that
well i gotta get ready to go to my buddy's for NY
do you need a phase converter for your setup?
have a good new years
well, i'm using my vfd to convert it
but i do have a rotary converter
I hope the neighbors don't make much noise and wake me up :)
same here.. i doubt i'll make it to midnight
* jt-plasma listens to Pink Bach
Dark Side of the Moon