ahh shutup :P
what not going to defend urself ?
I called you a twig girly man
im really disapointed in a computer scicence lab assignment grade I just got
because the program i wrote was for what I thought the input file was going to be, and the grader used a different file format, resulting in my program not completing the assignment
they didn't give you a test case?
my only computer programming class, you turned in your deck of cards and they ran it
they did, and it appeared to me, that it was a header, a blank line, another header, the data to parse
if it ran the first time, you got an a, didn't run, you fixed and resubmitted
couldn't run it yourself btw
i got a 18/20
this is all supposed to be, working when submitted
and the grader said my program exited before finishing the file parse im soo disapointed
on my end, it worked perfectly, garbage in, garbage out
they are training you for real life, where you have to perform for idiots that don't care
yeah, in which case I would have added some statements to allow the user to define the input file, that was not part of the assignment, I did other things that wernt but not a user input selection set
its supposed to be basic C++ not your stuff didnt work cause it didnt do what it was supposed to do to "my file"
its the same with emc, a guy could get on here and say "ive been programing cnc machines for the past 20 years, and emc is wrong, it doesnt work the way its supposed to work when I do a Gblah blah foo foo f.005
where does the wrong apply :)
look at it from his perspective, he could be working at walmart for more money
he was given a program that worked and an input file to test your program with, and yours didn't work
I don't think doing extra is going to earn you brownie points with the average grader
the professor would probably eat it up, use your energies to brownnose the right person
yeah, ill find out tommrow, it just bothers me, cause I may have missread what i was given
my guess is that it was poorly expressed
I emailed him and said, that maybe I was to blame, and would like it shown to me
I saw the input file as 3 lines of junk at the start, so I omited them, the grader(not the professor) said my program ignored the last line of 4 lines of data. to me, and what I wrote, the only way for that to happen is a eof before the last line
there may be a cabal of frat boys that have the old assignments that you need to make friends with
heh, he is really smart, and can see things, in fact he doesnt like to be asked questions
which one, the prof or grader?
there are 2 levels of questions, level 1 is help on logic, level 2 i will point you to the manual
the grader is a grad student, Im hopping that all of us ignored the first 3 lines as well, if not, then i screwed up
usually the garbage at the beginning tells you what you need to know about the rest of the data
like how long it is
yeah, this was just a plain txt file to be read in, not parsed mind you, but read in, so it had to have structure
the code of each student had to rely upon a specific data structure
which was roughly printed out on the assignment sheet
a blank line on a piece of paper is a blank line in a file to me, am I wrong?
you could be wrong
well, if im given a database with a table entry for last name, and the grader puts the first name in the last name box, and I pull that last name out and print it as a last name, but when the printout is read its read as a first name?
just an example of course but still, it disapoints me, that I could have had the human element on my end, have a dirrect effect on the computational outcome
mid term is morrow too
better go study
no, you're whining about getting an A- on an assignment
I have to have something im doing right in, math is killing me
batman = d/dx batwings*(bruce wayne)
but can you prove that 1 = 0?
jmkasunich_ is now known as jmkasunich
guess you don't know how to use google, 260000 hits for prove 1=0
heh, no just busy reading and writting
calculus has gotten so bad, that I have resorted to using youtube for instruction
and my 160 buck calculator is about to meet the working end of my shaper
its wrong in terms of how im supposed to do things 99% of the time
so I dont have an effective learning method down yet
I don't recall needing a calculator for calculus
jmkasunich: your part from the other night wasnt wrecked was it?
eric_U: it's how TI makes money any more
eric_U: im just using it to help make sure my answers are correct
they require you to buy an overpriced Z80 with crap software
what, do they have the integral tables in there?
current price of math tables plus elementary calculus book is approximately what I paid for my first quarter of college
gezar: ever used mathematica or maxima?
no, in fact im one of a few the others I have shown how to do specific things with them, at least I didnt buy the required 230 buck monster
fenn: i use mathematica's quickmath at times
I dont own it personally, wife talked me into buying some other software that has turned out to be a complete waste of money
msg mbot for an irc interface to mathematica :)
all be damned, what channel is he in?
and #not-math apparently
Im not trying to wing the class, but I cant learn as quickly as most do, its been a long time since ive had classes
from my perspective, calc 1 should probably be learned the old fashioned way
#math is interesting
eric_U: I agree, thats where most of my time is spent, but if I do everything wrong, with no time to learn it, then I wont learn it, so I have to see whats right, so I can learn what is wrong
school is a waste of time, unless you're trying to get a job
we test every other day, with full homework assignments due each day, they are graded and returned with out much notes
fenn: I agree, and im tired of working the jobs that dont require school
that and Im learning a ton of stuff
proof 1=0: x = 1 + 1 + ....+ 1 (x times)
then multiply both sides by x, take derivative, and 1 = 0
the version involving integration by parts is much more elegant
er, dont you get x/2dx = 1?
mbot doesnt seem to be working
at least the mathematica part
i just sent it help and info with no reply
and you have to be registered :(
no, you get X^2 = x + x .... + x (x times)
eric_U: and then when you differentiate..
take the derivitive, and you get 2x = 1 + 1 +1 ..... +1 (x times)
so 2x = x
problem is, the first line is only true if the derivative is not defined
i dont think you can take a derivative piecemeal like that?
the integration by parts doesn't go bad until the guy integrates, and then it's off by a constant that most people ignore half the time
figured it would be that the total number of pixels necessary to make a 1 are the same as those used in a o
missing the fact its a lower case oh
what if you use antialiasing?
well, if anything, I can work on being a master machinist
they require a degree now too
think about getting a Ph.D.
doctor of science in engineering
not that it ever made sense anyway
science would be a Sc.D.
art is DA right
arts, ie math
they dont give doctorates for art :)
a doctorate in math is a DA
they do too
it's PhD across the board
all the way down
all be, well, ive managed to let the compsci issue fade from memory, and I completed another assignment, now to finish studying for the compsci test
wiki sez DA is like a PhD except you just study instead of study + independent research
sounds pretty lame to me
maybe a money issue
best they don't pretend in my book
so i spent two years learning calculus and never used it, so then when i actually did need it i couldnt remember how
I know, senior year, prof pulls out an integral and I said "whaaaaaaat?"
that must have been junior year, I got a math minor
"you should have known to use green's theorem"
was it this cold this same time last year?
it was pretty cold
last december was obnoxiously warm
flowers blooming, that sort of thing, then jan hit and it was nasty cold
I dont understand weather at all.
oh well, not trying to stir anything up, it just seems extra cold this winter
here is how the weather works:
sometimes it is hot, and sometimes it is cold
nobody knows why
also, there is a chance of rain.
actually, they have gotten very good at predicting it over a short time horizon
OR HAVE THEY
"they" read the farmers almanac too?
I'm not talking about a year ahead, just the next week
they used to be pretty sure about the next day
launch all zig
i just shoveled 3 inches of partly cloudy
[11:02:15] <Dallur> https://bugs.launchpad.net/ubuntu/+source/linux-source-2.6.22/+bug/190587
Just to remind you that the updated kernel is now available and everyone should patch their Ubuntu machines and reboot
the vmsplice doesn't affect 2.6.15 which is on dapper
it has been fixed in 220.127.116.11 though.. so hardy will have it fixed
hmm.. for 2.6.25 there's 1.4M lines of diff
fenn_ is now known as fenn
for dave_1: here's m5i20 phys pins & hal logical pin list http://pastebin.ca/902396
remember kids, keep your hands out of the way! http://mitglied.lycos.de/fretless85/00002.jpg
or this will happen...
wah i have a lot of money so emc should read my mind! wah!
fretless85 what happened?
put an 3t workpiece on it
fenn, looking at that davenull email?
fenn: yeah I'm having trouble holding my tongue
I'm writing a response now
SWPadnos: try not to use too many cusses
and I haven't had coffee, so it should be "good"
fenn - you should link to your responce on cnczone about steppers.
i think gimp is pretty good btw
yeah, and if you're used to photoshop, then "gimpshop" is the thing for you
now i got fun for some weeks...
fretless85: 18.104.22.168 with rtai 3.6 work great!
and it happened on my birthday...no beer after work ive got morphium in the ambulance
maddash, on 64bit 7.10 ubuntu?
fretless85: well, i live in 32-bit land, but i don't see why it wouldn't apply to 64-bit as well
eric_U: f = g does not imply that f' = g'
uh, yeah it does
fenn: no, it doesn't
case in point: x^2 = x+4
i got time the next weeks so^^
f' + g' isnt (f + g)'
i will try to get it running
fenn: uh, yeah it is
seriously, differentiation is linear
cradek_ is now known as cradek
SWPadnos: be gentle
if f = g then they are the same thing, wtf are you on about
I'm in teacher mode at the moment, I think :)
why is it that every stepgen model i look at only does linear velocity ramping?
fenn: f = g doesn't mean that they are identical across the entire domain
fenn: and in eric_U's case, they aren't
if you're going to make up your own definitions about equality then we can't have a useful discussion
f (triple equals sign) g means that they are identical. I'm not sure he '=' conveys the same meaning
at high stepgen speeds, the adjacent freqs get closer and closer
fretless85, the boss here has a .8 module index finger
put his finger in the gear miller
it got me a morning off work taking him to hospital :), after he made the coffee
mine are "only" broken and lacerated+contused
at least morphine isnt that bad *g*
i kinda like it lol
dont get hooked
no pain killers from now on
it dosnt hurt that bad
no idea if this is good or bad ;)
but the hospital released me
i'm surprised they still use morphine
got to go there every day for changing my bandage
because it works fast i guess
shouldn't G53 be able to be used with G38.2?
whats the right word?!
nothing wrong with your english
As it is now G53 is only allowed for G0 and G1, allowing for G38.2 would be a good idea IMO, though diverging from the RS274 spec isn't good either...
but it sucks that i mashed my hand...
* ETA_lokal pokes cradek jepler jmkasunich
should G53 be allowed for probing moves
as it is now it's not, and as such it's not easy to probe the tool length
now i got time but i can solder shit with one hand :(
so remember, keep your hands out of the way ;)
* ETA_lokal is lerneaen_hydra on other machines
* skunkworks figured
ETA_lokal: I've considered that too, and I agree it might be nice but I'd want to think about it some more first. Right now it's easy to have one of the coordinate systems with zero offset and use that for probing. You could submit a feature request on sourceforge to make sure the idea is not forgotten if you like
any of those that are best to use?
one that's not otherwise in use
you could just zero the coordinate system before you use it, to make sure
I like to use G30 to get above the tool sensor, and then probe straight down. This way I don't have to worry about the coordinate system when getting there
it's true the Z result is in my G54 system but I always use a difference of lengths so it doesn't matter
How far do you probe in Z, as the machine coords will change depending on the G54 offset
SWPadnos: still composing?
ETA_lokal: good question. You could use G91, I suppose
or like I suggested, use another coordinate system
but please do submit a feature request. G53 G38.2 seems useful to me too
omg PIC board with mad speed! http://www.microchipdirect.com/productsearch.aspx?Keywords=DM320001,SW006015,MA320001,AC244006
I note that NGC doesn't explicitly say what coordinate system probe results are in
it talks about the controlled point, which means TLO is used, but that's as explicit as it gets
so you could do a g30 to move above the switch then a G91 g38.2 z-1 f15?
seems pretty flexable.
it's pretty hard to probe an unknown length. You don't know how high to start, and you don't know how low to probe
oh -- pic32 is actually MIPS? I'll stop snickering quite so much at it, then.
jepler: what do you mean?
kepler: i meant, "pic32 is actually MIPS?"
instead of 4 clocks per instruction
fenn: i'm lost. isn't MIPS 'millions of instructions per second'?
imo if you need processor speed use a DSP...
[15:05:16] <jepler> http://en.wikipedia.org/wiki/MIPS_architecture
uh, nevermind then
i bet "MIPS" was pretty sci-fi stuff back in the 70's
MIPS has a time-tested RISC instruction set with existing support in the gnu toolchain
(it was also one of the very early RISC processors, which I studied in the '90s at university)
weird that the wikipedia page doesn't mention PIC's nor Microchip
maddash: last decade
whoa dude tha'ts ancient
[15:08:55] <jepler> http://www.microchip.com/stellent/idcplg?IdcService=SS_GET_PAGE&nodeId=2591
shows that the core of the PIC32 is a "MIPS M4K 32-bit core"
what's 'm4k'? is that standard terminology?
I don't care so much about the RISC vs CISC debate, but it's a huge advantage to have a well-tested, Free toolchain for the microcontroller you'll be using
[15:10:24] <alex_joni> http://www.mips.com/products/cores/32-bit-cores/mips32-m4k/
'static implementation: min. operating freq 0MHz'
why would they include that in the PIC32 user manual?
because a minimum clock rate is a requirement of some microprocessors
heh - I just disabled the 'power saving' in the bios of the compaq. now I am getting sub 10uS jitter. (before it would climb over 50us)
that is all it is called 'power saving'
DRAM uses a capacitor to store charge, but the charge is lost over time (the memory eventually floats to 1 or 0 no matter the stored value). A device which uses DRAM type storage has a refresh cycle which reads each bit of memory and writes it again. In a system like this, the minimum clock rate would be related to the maximum permitted time between refresh cycles.
SRAM, on the other hand, does not use a capacitor to store charge, so it doesn't require a refresh cycle.
When the whole microprocessor is based around SRAM and SRAM-like storage, then the clock can be permitted to go down to 0Hz. If it has DRAM-like storage, then the clock cannot be permitted to go down to 0Zh.
fenn: not sure.. It is a micro mini case similar to the deskpro
* maddash has yet to see a DRAM-based uc
doesn't say on the front
733 pentium III
take that back - I swapped proccessors in it - 866mhz III
so far - no change. Before it would just walk up in jitter
maddash: I doubt those exist any longer, in fact despite my advanced age I've never used one either
with these inexpensive pic's floating around, i see no reason to keep the rabbits around
now, if only I could return them and get my money back...
ok - up to 12us (after it went into screen saver)
good idea, but i was actually thinking of calling up rabbitsemi and showing them their shittiness
somehow i dont think they would be easily convinced
The advantage of DRAM is that it uses fewer transistors (less space) per bit stored AND less power.
for instance the MCS-48 (very old) specifies a minimum clock of 1MHz (see pages 12 and 26 of http://home.mnet-online.de/al/mcs-48/mcs-48.pdf
lerman: thanks for your measured response
Well, I figured it was my turn. :-)
I hope his response is to READ more, not TYPE more, but I'm not holding my breath
he has had a lot of good helpful advice from several clueful people
sorry, talking about stuff going on in the mailing list
got a link?
it's all in the last couple days, just check the recent archives
I have enough pent-up anger to launch a reasonably lengthed diatribe
er, what search term am I looking for? 'davenull'?
Are you sure that isn't /dev/null ?
more diatribes are not needed!
how is it possible that the cpus listed on http://www.microchipdirect.com/Chart.aspx?branchId=8165&mid=1&treeid=8
have internal oscs at 7.37MHz but have cpu freqs of 80MHz?
you can use an external oscillator?
or maybe PLLs
jepler: entering the jitter from the latency-test in step_conf.. which one do you use? servo or base? highest?
Ok - I think the manual implies base thread jitter
skunkworks: if the servo is much larger than base I would be concerned that there's something funky in the system; otherwise, I'd enter the base thread value
(if the servo thread is a bit higher than the base, it reflects the fact that servo thread has to wait for base thread to finish sometimes)
is this http://sourceforge.net/mailarchive/forum.php?thread_name=2422.214.171.124.5.1202856126.TlMUQ2BZT0x2RA%3D%3D.squirrel%4077.103.68.5&forum_name=emc-users
ah - they are within .4uS
SWPadnos++, good replies
jmkasunich: i find we already use LEM, but a version from 'nanalem' , the HNC-300P, which (obviously) i am not so happy with. I will look at those other models, esp since LEM USA seems to be Milwaukee :)
so, has anyone noticed that xkcd comics have tooltips that add to the joke?
here's the text from that basilisk one: "U+FDD0 is actually Unicode for the eye of the basilisk, though for safety reasons no font actually renders it."
Rapid traverse override feature request #1536176 ? why couldn't the G0 output be routed through the jog speed slider?
am i reading this correctly,is"davenull" the guy thats having the concerns about emc?
he seems to be
glass scale guy?
"5Hz should be enough for anybody!"
send him to baking class with Gamma
k,im gonna send him an email and my number and offer some assistance
have you gotten things working with glass scales only (and steppers)?
on one of those web bbses, some guy is saying that, like a wiggler, a coax indicator doesn't have to chuck up accurately to work right. This doesn't seem right to me - does anyone see it?
jlmjvm: did you ever get the adaptive feed working?
the small amount of angular error will affect the "scale" of the eccentricity measurement but not the "offset"
no,only tried it that 1 day when alex was helping me
it's still rotating around the true center of rotation
cradek probably true, I have center finding indicators
ok I think I see it
it's just like swinging a DTI
very useful toys
that's what it is, a dti with a rotating indicator
so it works the same no matter how it's mounted
mine are morse taper mounted
archivist: to stick in a tailstock?
one is not accuratly mounted and seems ok
the only reason to use one is to put it in a spindle
is R8 really any better than, say, MT4?
i just dont get it
methinks depth accuracy will depend on taper
neither one will self-release will it?
the collet pulls up as it tightens regardless of the taper
dunno which is best
SWPadnos:i dont have glass scales,but i thought alex told me emc was capable of using either encoders or scales for ferror
amount will depend on steel elasticirty and taper and force
and i think thats about all you can ask for using steppers
it can certainly use either device for feedback, but a control system that has feedback not tightly coupled to the actuator will be hard (or impossible) to tune
so you could certainly fault out with an ferror, but you can't control the motors using PID
and in general, using pid for steppers is doomed to failure
(at least I don't know how to do that, and neither does anyone I know)
correct on all statements
seems he wants something that doesnt exist,be it open source or commercial software
* archivist was contemplating a method of over driving steppers last night (but keep average current within limits ish)
jlmjvm, I think you're right, but he seems to have incorrectly learned that it's all EMCs fault
jlmjvm: he just started buying parts before reading/designing: it's something we've all done I bet.
it's my favorite hobby
this is why we go to CNC workshop - to get rid of the evidence
the more money i spend, the less of a hobby it becomes. then it's an addiction :(
yes,thats why i was gonna email him
im planning on attending the fest this year
the more the merrier
some of us do it mistakenly and hide the evidence, others of us do it overtly and with great gusto :-)
i look forward to meeting yall
* archivist build build his machine with "evidence"
btw,thats how we talk down here
and the plural of "y'all" ??
("all y'all" :) )
y'all is plural
it's kinda both, but
all y'all means "all of you"
similar to "you", which may be singular or plural
the reason im willing to help this guy is i know a guy that struggled for quite some time to get emc to work before he found the group,the difference is he didnt give up or resort to taking a jab at the software
I've seen several of yall recently :)
you all uns
you uns would be if your from tennessee
uns is german
A friend of mine from South Carolina insists that "y'all" is singular and "all'y'all" is the plural...
fretless85: short for 'you and yours' i think
jepler, yep, I have Texan friends who say the same thing
its a german word
no, "uns" is a suffix - weuns, you-uns, young'uns ...
or something like that
and also that "you guys" (the local regional spelling of "all'y'all") is sexist
ok but its a german word tho
young'uns I've heard, but not you-uns or you'uns
we'uns is goin' to th' store
lots of apostrophes when you spell that stuff out ;)
it's hard to spell weird pronunciations and poor grammar ;)
err - regional variations in the language, I mean
fretless85: compared to the English you learned in school, the language spoken by people in the southern and south-eastern part of the USA will be pretty different
jepler, i guess its like "plattdeusch" here
"Me name be Ralph Nader. Me gonna make an appeal. Homies, save the rainforest. Aiii? Keep it real."
spoken in some locations in germany
shee-it, man. you down on da 'Nade?
don't be messin' wit ma homey
i can understand it a bit but.... not very well
SWPadnos it's his most famous speech!
cradek: submitted to sf.net
fretless85: a friend from frankfurt has done her impression of some local dialect/accent, but I don't know enough german to appreciate it ..
someone needs to find a video of the jive talk on the airplane movie..
what all will actually be happening at the fest
Oh stewardess! I speak jive!
that was funny
bro was On! "Mr Stryker was in control"
jlmjvm: at fest I spend most of my time hanging out with the other emc folks who are there, and often we work together on enhancements to the software. We also evangelize emc to people who pass by.
jepler, plattdeutsch is german mixed with english and netherlandish...pretty confusing
jlmjvm, the CNC workshop has a lot of classes on machining, machine building, some other stuff like casting and anodizing were done a couple of years ago, demos of peoples machines, etc.
jlmjvm: you'll also find tutorial sessions on software packages (particularly windows CAM software, so I stay away) and people selling software and hardware (such as servo and stepper motors)
what we do is work on a machine (the Mazak), and actually talk about EMC with our mouths while writing code :)
jlmjvm: there's also a swap meet, raffle, and dinner
they actually do talk
yes, we have test-to-speech devices ;)
err - text
cool,just wanted to make sure an intermediate guy like me wouldnt be in the way
heh - a new product line: "Test to speech devices"
"that didn't work"
"the signal is wrong"
us over the water watch the webcam
im thinking about going to the mach3 convention in Knoxville,Tn and taking some emc discs to hand out
and offer assistance to people that havent been able to get it going
we joked about giving some free emc licenses for the raffle at workshop last year :-)
since i have a true dual boot machine
unlimited tech support package
free irc support
two years ago Ray Henry did some "emc classes" -- I think he covered some setup stuff and also basic hand-writing of gcode. unfortunately, Ray didn't attend last year and nobody jumped in to fill his shoes.
i would like to meet ray,hope he can attend this year
heh - too funny
this is what it goes with http://youtube.com/watch?v=d28IFaRvkJ8
texan dictionary http://homepages.nildram.co.uk/~jimella/texan.htm
, but they missed 'offen' as in, i gonna knock that clean offen yo haid
and don't forget the automatic translator: http://www.rinkworks.com/dialect/
tomp2: Often so as to be no longer supported or attached Example: "Now stan still so ah can shoot that apple often yore had"
tomp2, they spelled it "Often" - it's the sixth word
he also missed "Nukular" = "nuclear"
us brits iz gud at speelin
jepler: I take it - halscope cannot handle double_freq?
do you mean doublestep?
skunkworks: it does if you know how to interpret what you're seeing
it seems to invert..
the fastest step rate will show a constant high signal on the step line
it also depends on where in the thread you put the scope.sample function
a high signal for a period represents a step during that period, a low represents no step
oh - true, the stepgen output isn't changed, only the port pin
getting 40K out of this poor thing.
866 pent III
whats the odds of making emc work on a 6 axis fanuc robot?
jlmjvm, excellent, depending on your skill at writing the kinematics module for it
at least a few people are doing that already
I think there's even a youtube video of someone's homemade robot moving marbles or somesuch
the local trade school has one and the instructor wants to convert it
[17:28:47] <fenn> http://www.youtube.com/watch?v=AoOWi1r6-Eo
skunkworks: 40 kHz step pulses?
i saw that vid earlier
from germany i think
so am i correct,6 axis kins would need to be written?
alex said he runs it in teleop mode, records the joint positions with streamer? and then runs it again in uncoordinated mode so that joint limits arent violated :\
or something like that
basicly he has 2 configs
one with kins, one with trivkins
alex_joni: how does he play back the recording but still use emc?
loads the kins one, jogs in teleop, uses teach-in.py to record position
he saves points
then converts those to g-code
ok so they're not coordinated moves
they are G1 moves
but the arm wont move linear
the tooltip won't move linear
if the points are near enough, the approximation is good enough
hmm.. I've never seen http://www.youtube.com/watch?v=PVuPHD1LpL8&feature=related
I want to attach some .deb packages to a wiki page (debian_lenny_compile_rtai page) but the files are big, app. 80 MB. is there a free space for this in EMC server?
yes, but getting them there isn't trivial without making an account for you
acemi: can you put them somewhere else online?
I (or someone else) can put them on linuxcnc.org then
there are share sites which let you do that, if you don't have online space
I think to install them into esnip.com but this site is blocked now from my contry.
alex_joni: after upload them in a site, I'll send you a message
acemi: sounds good
alex_joni:Would it be possible to use emc on a 6 axis fanuc robot,the local trade school has 1.
jlmjvm: possible.. yes
would it be straightforward? .. probably not so much :)
does anyone here have any clue why my geckodrive stepper drives make a loud annoying beeeeeep sound?
quiteBIGeye: more info..
well, i and Vq^ have a machine, i think he has posted som pic's of it...
do the kins exist for something like this?
and when we turn the power on to the steppers there is a quite loud high pitch sound
is it a hobbycnc board?
partly from the steppers themselfes but mostly from the engines...
no home buildt from scratch
with a 24volt pwm power suply
i would guess either noise in the drive or not high enough pwm/chopper frequency
got a scope?
yea, Vq^ is bringing it over this weekend
pwm freqency in the suply?
are you using a switching psu?
that's probably it then
aah okay... well some huge capacitors might fixt it right?
put a load on it (light bulb) and see if there's a ripple on the voltage
no, but some smallish high-esr caps might
equivalent series resistance
okay, well now you lost me...
generall speaking, the bigger a capacitor gets, the slower it is to charge/discharge
they're made of metal, which has an inherent resistance (and inductance)
or any material really
yea we had some 4700u that we threw on there... it helped but didnt really do the trick...
should we try with something like 470u in parallel?
try a non-electrolytic capacitor
actually, a better test would be to use a different PSU and see if that fixes it
i think we had that soud when we ran from a car bettery to...
ok, then it's not the PSU
maby at a difrent freq but some noise...
if we dn't find it using the scope i'll get back to you guys...
btw what components are you using?
jlmjvm: yes and no
there are some kins for a 6dof robot called puma
which is very similar to a fanuc
but they might not fit
k,was just trying to get some info before i just dive in on such a large project
seems like it would be quite a showboat if it could be done
kinda hate not trying to do it,open source controlled full size robot in a post secondary educational environment
jlmjvm: do you know anything about the robot?
what servos/motors/feedback, etc?
emc demonstrating its superiorty over all others
factory fanuc servos
it certainly will be way worse than a fanuc robot control
g-code is not for robots :)
so if its a working fanuc robot leave it alone?
I'd say so
jlmjvm: "if it's not broken, don't fix it" :O)
wasnt my idea,the school wants it
i will relay the info
jlmjvm: depends what they use the robot for
if they use it to teach students how to work with it, how to program it, etc
then don't touch it :)
if they want to go deeper into the software, and try different approaches for trajectory planning & such
then emc2 will be what they want
but it will take some work before emc2 will be equivalent to the control they have now
i really dont know what their reason is to convert it,obviously they are impressed with the mill i did for them
impressed with emc that is
if you can't reach something with them.. drag them in here :P
emc needs a bit of an overhaul to work *really* well with non-trivial kinematics
then we can disect it more in detail
fenn: yes, that too
and it needs a *real* overhaul to be similar to a robot control
which is compliment to all that have developed it
eh, not so much
* alex_joni thinks about another interpreter, another GUI possibly, etc
that's an addition, not an overhaul
I meant lots of work
writing new code is easier
fenn: should be.. :)
if you ever get something that you think would be a comparable let me know,ive got the hardware available to test it on
alex_joni, new interpreter?! you know my ideas...
fretless85: I catched lerman the other day
what he said?
he asked if there's a standard somewhere
standard for what?
for the language
wasn't it that what you wondered about?
the klartext heidenhain hasnt changed since the first heidenhain cnc
fretless85: that's not what I said..
is it defined somewhere?
I mean a pdf to download
* Vq^ is also interested in heidenhein specs
you mean like the gcode?
fretless85: yes, like the rs274ngc spec
g00 rapid linear movenent...
let me search...
g00 = rapid move, parameters: ..., it's forbidden to ...
im a bit disabled
puma acronym...a programmable universal machine accessory, not a brand. several manufactures. it describes the mechanism
[18:29:17] <fretless85> http://mitglied.lycos.de/fretless85/00002.jpg
tomp2: the puma560 is a product
only one hand to write
alex_joni: if i were to create a non-tcl frontend for emc2, should i try to replace emcsh or build something on top of it?
if it's non-tcl.. why bother with emcsh ?
look at halui.. that should be one of the simplest GUIs
fretless85: been beating up someone? :o)
tomp2: the kins we have in emc2 are pretty tighly linked to puma560
the guy from germany who built the robot on youtube, said it doesn't fit his robot
Vq^, no workpiece mashed my hand
there are some assumptions in the 560 kins (not really 6*R kins)
alex_joni, im searching there should be something on the heidenhain hp
fretless85: I didn't find anything close to a full spec
did you search on the hdh site?
for a couple of minutes
if not i can do one?
what would full spec look like? i have operator manuals, user manuals, programming manuals for a limited series of Heid controls.
I doubt you can make a complete spec
there's little things i learned about what it cant do, but only by using it
things to avoid
i work with hdh since 8 years now
actually 1993 began coding dialogs
does FANUC publish a spec? i bet thats secret
a spec is usually much more than simply explaining each code
it's also things like: what codes can be combined how, what is not permitted on a line, etc
the book with all the diagrams showing 'fishtail' errors etc?
not sure what fishtail errors are
sometimes I think you can construct an interpreter from a good users manual
overcompensate a sqr corner, and the tool moves past the corner, generates a quarter circle arc, back onto an extended line, back onto the true path
tail of a fish
the docs on the Heidenhain site are all they'll give to you, at any rate
it's as complete and as accepted as RS274NGC ;)
tomp2: that's good to know :)
(ngc always sounds so startrek )
hdh isnt the best in sup. ;)
but when you got into that control you wont go back,,,
tomp2: according to paul's latest rant the original EMC was called VGER
i like it, customers with large shops fulla control like it, it's great for operators who 'dont wanna learn', just answer the questions
fenn: ? gotta read !
yeah, that's an interesting document he linked to
especially with all the stuff he left out
sorry i been deleteing those mails pre-supposing its some pissing contest
it is ;)
[18:44:58] <SWPadnos> http://www.isd.mel.nist.gov/personnel/kramer/pubs/RS274NGC_3.web/RS274NGC_3TOC.html
well, there's nothing new in computing , right? sutherland & knupf did it all way back
[18:45:41] <fenn> http://www.isd.mel.nist.gov/personnel/kramer/pubs/RS274NGC_3.web/RS274NGC_3.html#1001730
[18:48:25] <alex_joni> http://www.isd.mel.nist.gov/documents/kramer/RS274VGER_11.ps
may be of use http://faculty.etsu.edu/hemphill/entc3710/heid-op/h-toc.htm#top-o-page
tomp2, dont think so
not to you ;), to whoever wanted the 'spec'
every needed info is in my head :P
maddash: only theoretically
given physical limitations you won'te get even spaced microsteps on a stepper
and if the RS274/VGER spec is what is wanted, there is no such Heidenhain document. BUT there are very few primitives, and a simple macro system with 'fill in the field'
alex_joni: you mean the stepper shaft positions are inaccurate?
fretless85: me too, i've already tried what doesnt work, and know how to get what i want
tomp2: fretless85's question was if users would be interested in a hdh interp for emc2
got huge amount of user cycles made by me and working buddys +know how to do my own cycles
my head weighs ten pounds
fretless85: i would like to know how to make my own cycles in heidenhain :)
got skills in q programing?
yea i use them for everything...
nice :) i love em
maddash very inaccurate for microstepping
you will need cycledesign from hdh
we don't have F AUTO in all our machines so we made a standard to use Q1 instead...
afaik microstepping is only useful for giving a smoother movement
to implant it into the control
got a 408 that cant do f auto ;)
and an 355
okay, i'm not that familiar with models... :)
but our 406 or 408 cant take auto eitghter...
410 got auto
i made some trocoidials using only Q's and "cp ipa"
in short q's are the shit! :)
thats an heavy q program http://mitglied.lycos.de/fretless85/test/Farben.jpg
[19:14:45] <fretless85> http://mitglied.lycos.de/fretless85/00003.jpg
radius done with q parameters
you are mad! did you ramp that?!? how many passes?
[19:16:28] <fretless85> http://mitglied.lycos.de/fretless85/test/RIMG1372.JPG
aaah you cheater :) you don't have to manually turn your a and b axis...
i wish i could program them...
well... i have to register so we don't spam this channel... i'll be back tomorrow...
oy vey. goyim!
[19:24:08] <fretless85> http://mitglied.lycos.de/fretless85/DSC00080.JPG http://mitglied.lycos.de/fretless85/DSC00079.JPG
programmed by hand on heidenhain
[19:24:32] <fretless85> http://mitglied.lycos.de/fretless85/test/Wanne-Q.jpg
fretless85: i never use cycle design, all by hand ( not so smart, but the cycle design tool is not neccesary )
you can call your own cycles by program call
and i dont think a heidenhain interpreter is needed/wanted, but a conversational tool is a good idea ( tied to g code is a restriction that wasnt part of the original emc design , it was meant to be flexible command structure )
oh, and a heidenhain interp isnt a bad idea, just not neccesary
Dallur: please ping me when around..
no not neccesary
that's conversational (tm), buddy! :)
if you only want to program in an cam
cp1 is nice http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Cp1
i hacked heidenhain output with it
anything on heidenhain is just substitution ( code wise) just like gcode macros, o word macros
you compare an steam powered car with an viper :P
now, if emc could handle cycle 19 plane tiliting :-)
and polar programming
+q parameter logics
alex_joni: files are ready. download link: http://www.4shared.com/dir/5689661/f0d409d0/sharing.html
acemi: do you want wiki.linuxcnc.org specifically ?
acemi: those are kernel debs for lenny .. right?
acemi: switch over to #emc-devel please
fretless85: ever used APT?
[19:45:30] <fenn> http://aptos.sourceforge.net/WhatisApt.html
let me take a look
ah no never used such
you might like it
i like to program by hand ;)
it is very much by hand
ah this might be more relevant http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?AptProgrammingForEMC
* archivist experiments live webcam from emc box
why is it pink?
crappy torch light
just testing to make sure it handles the cpu load vnc emc webcam and browser
and machine vision possibilities
I need to get far better lighting and better mounting for tha camera (bluetack to machine atm)
hmm no nasty dmesg, seems to be ok :)
may change to camera in/on a microscope for better vision of cut depth
okay, im home
and check this out
fenn you remember my bitching about my lab assignment right?
it's white now
should www.linuxcnc.org have "hacked" all over it?
well, anyway, I go and I meet with the grader, sure enough it doesnt run, so we try it on another compiler, same issue, so I go into the classroom, and load it onto that computer, works correctly, then I get home and the grader has emaild me "I re ran your program and I dont know why but it work now"
awallin: grr.. no
alex_joni: I see it too
oh this sucks
sorry to bring this news to your attention...
making some waffles...
that's what it says on piratebay whenever they are offline :)
alex_joni, changed camera better colour and focus, will work on it later
archivist: sorry.. been working on something else
yup I noticed
archivist: do you ever stack up lots of gear blanks to cut them all at once?
yes, but not that often
more often if hobbing gears
btw lighting/camera much better
this one is from an educational microscope
shows promise, will add crosshairs and make a mount
what size thread is that?
heh using php to overlay crosshairs?
you can print crosshairs on clear paper and cover the lense?
or etch some into another lense, same result :)
and no, is the answer
fenn i got my 20/20 :)
it would be out of focus
have to put the crosshairs behind the lens
or at a focal range with another lens and make it like a scops sight
i dont know..
are AC servo motors suitable for scara robot?
jymm, some very dead net somewhere
Ok cool... using a new irc client, not used ot it yet =)
oh, which client?
thanks for sending me unsolicited commercial email, cnczone!
if you think that is fun try downloading a trial from BobCad...
they wants an evil profit
how much is bobcad now?
and its full 3d now right?
$300 for an upgrade
what does 2 1/2 d cad/cam mean?
2D motion at several depths
like a pocket of some shape at some depth, plusa hole, plus some other pockets at some other depths
vs. 3D continuous contouring, like a mouse shape, for instance
and 3d isn't 5 axis either
perhaps they should demote "3D cam" to 2.75D cam :)
very nearly 3D
and 5 axis would be 5/6 = 0.83333D cam
er, wait a minnit
2 1/2 means it will move to a location Z down then follow a 2 axis profile...
i have bobcrap 21 at my machine shop but I don't use it
well, ive been a turbocad owner for 10 years if not more, and so I get uber deals on stuff, they have a cam plugin and i was wondering it if was any good
don't know about turbocad
but i guess it doesnt do the hills
TurboCADCAM seems pretty reasonable.
I can get the newest version for 200 and the cam for 300
I was able to customize the POST for EMC pretty easily, and I hadn't used TurboCAD since the DOS days
I'm a Solidworks guy but it don't do cam :-(
I upgrade every 2 years or so, I dont use it much but its more affordable for me