will be truly interesting to see if it works tomorrow
and tomorrow is another nice day, the base for my mill is coming home!
lilalinux is now known as lilalinux_away
Gcode question: On a 3 axis mill, what is the best way to move back to a tool change position for manual tool changes? I need to have the spindle all the way up to change the tool.
I think I want a subroutine to move to my tool change position which is located a few inches from the work piece, but do I need to add the current tool length to the z axis move to retract the spindle?
Or is there an easier or best practice way to do this?
jfigie: look at g28,g28.1,g30,g30.1 http://www.linuxcnc.org/docview/html//gcode.html
OK but does g30 move all axes at once or can I move Z first then X and Y?
"G30 _axes_ will make a rapid traverse move to the position specifed by _axes_, then will make a rapid traverse move to the predefined position in parameters 5181-5186."
so i think: "G30 z zsafeheight" will first move z only, then it will rapid to tool change position
if you set TOOL_CHANGE_QUILL_UP in your ini, it will move Z all the way up when you do a tool change (M6)
you don't need to program any specific motion in your gcode then - just a safe entry move with the new tool
or, it's true, you can use G28/G30, or something like G0 G53 Z0
if you want to go all the way up, and then to a special position (like one corner) you can do that too by setting the appropriate things in the ini file
I like the TOOL_CHANGE_QUILL_UP. What would make the program more portable? I like to make the tool change position near the work so the table dosen't have to move as much. I don't want to edit the ini file for each job.
another thing you can do is use TOOL_CHANGE_AT_G30 and then when you set up the job, put the machine where you want the tool change to happen and teach the position using G30.1
The G30.1 G30 is what I was trying now. If I start with the tool in my tool change position then do G30.1 X Y and Z are stored in 5181. 5182 and 5183? The when I want to do a tool change G30 Z 5183?
no, set TOOL_CHANGE_AT_G30 in your ini file. then when you do the tool change (Tn M6) it will move to that location for you
How does it know to move Z first?
in this case, it doesn't
the only Z-only move that it knows how to do is "all the way up"
Yes but first?
let me find you the docs
[03:50:15] <cradek> http://www.linuxcnc.org/docview/html//config_ini_config.html#sub:[EMCIO]-Section
unfortunately the docs don't say in what order these moves happen, if you want several of them
if Z=0 is the top of your work, perhaps you want the move G90 G30 Z0.5 - this will move Z 0.5 above the top of work, and then continue moving Z to the G30 position
if you then want to move X,Y to the G30 position as well, program another bare G30 on the next line
cradek: The G30 seems to move all 3 axes as a linear move. But why can't I do G30 Z [#5183] since that is the position that the G30.1 stored? I dosen't seem to work. Is there a way to look at the current parameters?
Dave911_ is now known as Dave911
if you specify Zxxxx ("specified by axes") in the doc-speak, you get a move to that Z value **in the current coordinate system** and then Z moves to the **absolute (unoffset) position** stored in #5183
so the simple answer to your question is "that's not how G30 works" and the difference is the two coordinate systems
but if you read again what I said earlier, I think it's what you want
(argh, the docs are unclear about this too)
yes I will try that now - G90 G30 Z0....
I found the receipt for my spindle, but I've voided the guarantee now
lilalinux_away is now known as lilalinux
lilalinux is now known as lilalinux_away
lilalinux_away is now known as lilalinux
hmmm I need some coffee
refill mine too please
had to make it - going to be a few minutes
JUST GIMME THE GROUND COFFEE NOW!!!
Jymmm: have you ever used http://en.wikipedia.org/wiki/DRBD
NOW I SAY NOW
I"ve heard ofa few related things like that, but never played with any of them.
skunkworks_: what were you think of using it for/on?
I've looked at using it in some depth
You think a SSR could be PWM to be a pseudo variable DC voltage control?
I was just looking for cheap redudancy.
I have not played with that.. On the pc side - robocopy is similar.
wait, what???? You've never used rsynce?!
remember I am new to linux.
skunkworks_: Dude the ALL TIME EVER BEST UTILITIES IS: ssh, rsync, screen and in that order
skunkworks_: Ok, what rsync will do is sync file between hosts. Once the initial is complete, all othe runs are only the difference between the two
If you change one file, only the difference is transfered
I think robocopy isn't that smart.. it still copies the whole file if it is newer.
does it work across ssh ?
no, seriously. rsync runs across ssh
all encrypted tansfers
* skunkworks_ has not used ssh yet either... he runs and hides.
my coworker uses putty
Well, learn it, use it, love it!
JT-Work: Hey, where's my jerky!
my other brother John just ate some
he just said is was good and passed the QC but more research is needed
He is LHAF
Lazy Half Assed Fscker?
Laughing His A$$ Off
I like mine better =)
he said no mater what you call him he still has the jerky and he wins LOL
that is until his daughter finds out he had some and then he is without
* JT-Work wanders off to Bald Knob, AR to visit a customer
MarkusBec_away is now known as MarkusBec
[15:56:47] <frallzor> http://pici.se/gal/oddname/Mill/
todays work =D
first 5 pics
last 5 even
er you need a larger garage
no I need less crap :P
there is room we just thought we cleared enough
so the second round have to come a bit earlier :P
I dont see much space to walk around it :)
I know the problem well
you can, but not as good as it will be
when the last round of cleaning is done
just cleared enough for it to come home
Hey All, during ./configure if emc2 I get checking /usr/include/python2.5/Python.h usability... python is clearly installed (cat /usr/include/python2.5/Python.h shows me the file) why does emc2 think it's not usable?
frallzor: great work!
you are building fast...
bit wankish atm when its not cleaned out enough :P
but at least its home
ries: could you put the whole configure output on pastebin?
mine is 45 minutes away from here :s
cradek: I think so, let me see
mine is allmost the same, since i will hesitate that long to go outside now
I was planning to have it done by x-mass... but I failed, I couldn't get old of a good gauge plate yet
cncbasher in the forum can fix if you want from uk
* ries wonders if it's possible to pipe straight to pastebin :D
you can possibly do it with wget using --post-file
haven't tried, though
would be nice....
is that the last line?
the error isn't show (I forgot how to pipe that)
But it basicly states that the requires header Python.h missing
./configure --flags >logfile 2>&1
could you also pastebin config.log?
[16:14:04] <celeron55_> http://www.knoppix.net/forum/viewtopic.php?t=28739
I think I see what's going on...
Dave911_ is now known as Dave911
lilalinux is now known as lilalinux_away
lilalinux_away is now known as lilalinux
cradek: the problem was my fault, I fooled around with ubuntu's method if building a new kernel, but that went wrong... so I just used a vanially kernel and compiled 'the good old way' rather then the ubuntu way, by doing this 'something' went wrong and teh asm directory got lost
The error I got about Python not usable I don't understand, but it was solved by putting that asm directory in place again
ries: sorry I got distracted. You can see in the config.log the Python.h error is really a broken kernel header problem:
/usr/include/linux/errno.h:4:23: error: asm/errno.h: No such file or directory
line 782 in your http://pastebin.com/d16d5e89
you are right that it has *something* to do with your kernel or header packages
cradek: no problem... I am rusty on kernel complications and complications on Linux in general, last time I did it was a 2.4 kernel (my first one was a 2.0 kernel) I think it happened when I was trying to do teh ubuntu method of compilations, which I didn't quite follow.. I did a wrong delete action I guess
rtai installed on a new kernel...running /usr/realtime/testsuite/kern/latency$ ./run works fine, but when using latency-test then teh computers hangs, any suggestion to check?
Dave911_ is now known as Dave911
jepler: the belt driven zenbot has a full step resolution of 139.4 steps/inch.
or about .007"
skunkworks_: Yeah, sounds about right. I got information to the effect of microstepping giving about .001" resolution (but presumably not accuracy)
* jepler wonders whether he should go add to all his old zenbot blog posts: the current zenbot is totally different and you shouldn't pay much attention to this
I was wondering what he was doing for a drive system initally - thought he had some sort of right angle gear box... Until I looked at the manual. ;)
it's just toothed belts with a pulley on the stepper motor shaft, right?
as soon as I feel like doing it I have a couple of boards to mill
hopefully the holiday weekend furnishes the time
I am hoping time this weekend to work on a few of my projects also. time will tell.
two daughtercards for arduino: GPIB (software to be written) and 10-pin programming header so the arduino can program other AVRs
I'm tempted to try amforth http://amforth.sourceforge.net/
which sounds cool because the forth environment on the avr can actually compile new words (no PC-side development environment required) .. on the other hand, you're stuck programming in forth
and you have to replace the oh so convenient arduino bootloader to do it, too
GPIB ? /me may/am interested
(no, that was not a note of sarcasm; it really is nice for a fast edit/compile/test cycle
archivist_emc: well, I have a scope with GPIB, but no PC GPIB interface
archivist_emc: GPIB hardware interface is pretty simple and there seems to be documentation
I have loads of toys with interfaces but no PC card either
I'd be satisfied if I could capture a (HGPL, presuambly) plot from the GPIB interface and send it to the PC over USB serial
I do have a Racal GPIB analyser somewhere for testing the protocol
that's the general idea
jepler: sounds like you'll ned to make a few of those pcb's - make it an even thosuand just to be safe
the GPIB usb thingies on ebay cost too much for what they are
I wasn't kidding about the 1K.... ppl will buy a few of them at a time
as usual for my projects, the source code and board designs will probably end up on my blog
true, but many dont have a mill =)
I want http://cgi.ebay.co.uk/Systron-Donner-Microwave-Frequency-Counter-GPIB-26-5GHz_W0QQitemZ280430645561QQcmdZViewItemQQptZUK_BOI_Electrical_Test_Measurement_Equipment_ET?hash=item414af81139
but selling fabricated boards for $10 each and then enduring the users who assume I owe them endless support is not something that is worth it to me
that matches the synthesizer
jepler: toss up a forum and state "Any support is user based, see forum"
archivist_emc nice fcounter, i have a similar here, anritsu, only goes up to 23GHz though
I have an EIP to 18ghz but the synth goes to 26
alex_joni: "..it works OK"
err archivist_emc that is
archivist_emc: a EIP ???
archivist_emc: why do you need => 18GHz?
we do have a band up there somewhere
archivist_emc: It's use is restricted though
testing the gear is most the fun though
archivist_emc: commercial or amateur ?
my playing has generally been amateur but im game to be paid :)
But > 10.5 GHz is restricted, even for testing.
Jymmm: more than likely any eventual release will be GPL software and dual GPL/CC-BY-SA board design. If it really is a good business idea to sell the boards (and you could be right about that) you're welcome to the opportunity.
but don't count this chicken; the egg isn't even laid yet
jepler: I have just been seeing a alot of nice kits lately. I'm really surprised some PCB house hasn't doesnt it as a sister company.
I wonder what this stuff is like http://www.prologix.biz/
Jymmm: sparkfun.com's boardmaking side lets individuals put their board designs up for sale, but it doesn't (A) help them get a cart with the required compoenents or (B) give a cut to the board designer
jepler: those bastards! Should be like cafepress.com
archivist_emc: This is the digital output, someone else has made a analog one... http://www.cliftonlaboratories.com/si570_kit_from_k5bcq.htm#Frequency_Spectrum
Analog here and rs-232 iface too http://www.pongrance.com/dds.html
Desc of 1st one http://www.qsl.net/k5bcq/Kits/Kits.html
jymm this is what I have 100 meg thought to 26ghz http://www.collection.archivist.info/showresult.php?srcdata=title&srcprog=searchv10.php&searchv4page=1&errlev=0&searchstr=systron&srcdata=title&srcprog=searchv10.php&Type=EQ&Accn_no=71&dir=2009_01_01_a&file=P3072786_1625.jpg&subject=15545
archivist_emc: Cool, but you really haven't said what you're paying with =)
Ive used it for testing counters and filters and used it for stability testing of a crystal standard I have
and missed another on ebay this weekend
archivist_emc: Ah, so mostly TE.
[20:18:14] <Jymmm> http://www.gather.com/viewArticle.action?articleId=281474977306204
Other than voltage/amerage ratings, are all triacs basically the same?
I can't see how this pkg can really handle 16A http://www.taydaelectronics.com/servlet/the-39/BTA16-dsh-600B-BTA16-dsh-600-Triac-SGS-dsh-THOMSON/Detail
well you can bet it needs good cooling =)
they are at saturation when on so wont get that hot
archivist_emc: and at 50% ?
its a swiche device...on or off
like pwm ?
archivist_emc: bt that's unique to triacs alone?
except that triacs don't shut off until there's no current going through them
so PWM isn't all that useful for those, and they can't be used to switch DC loads
unless you only want to turn the load on
SWPadnos: Not on/off but control 1-100%
they turn on for a cycle or part of usually
Wait, are they a AC or DC device?
if you're looking at controlling a DC load, then a TRIAC is not useful, unless you have a separate means of shutting off the load
is the junction voltage of a triac .6? that would only be 9.6w
or miss cycles
triacs are DC control, but they can only be usefully used to switch AC loads
SWPadnos: I have a seperate power switch
here's the rule for a TRIAC: it will turn on when the control input is on. It will stay on if either the control input is on, or there is current flowing through its output
Ok, added extra rubberband to door
it's extra windy today
anyone wanna take a wild guess why my VFD triggers my RCD but not the persons that sold me it? =)
same speces on the RCD
cant really say what we have much that shoooould leak current
because the person who sold it took you for a complete fool
are you saying that he lied? :P
well the facts are that it shouldnt work afaik with a normal RCD =)
so its a bit odd :P
do you have a three phase rcd?
if you mean a 300mA Type B no
30mA Type A
no i mean a three phase rcd
as opposed to a single phase rcd
never heard of =)
the documentation doesnt mention that either
what model is it?
our RCD does the whole net
one to three
ok, should be 3ph then
since oven and stuff works it should be
but I read in a nice pdf about delay when powering on
it shouldnt be instant when going via RCD
at least 40ms or something
[21:08:34] <frallzor> http://hvacity.danfoss.com/pdf_files/rcd.pdf
page 3 left hand column
seems to have to do with the three phases not being conected exactly simoultaneously
know of a simple solution to that?
the pdf seems to suggest using a switch to connect the vfd that is very fast and accurate
perhaps even a contactor
yeah atm im just plugging it "to the wall" :P
are you using a cee plug and socket?
standard 5 pole round red connector
Well, my 7i43 (or whatever the number is) is a curiously partial success. I have one axis working out of 4. As the amps are identical and the wiring is too, I am a bit stumped. I guess I will need to get a scope on the output pins.
yay, an axis working!
But I used to have 4 :-)
your glass is 1/4 full!
There is a ribbon cable that I need to buzz-out very pedantically, I will try inverting the signals, and then it is time for oscilloscopes, entrails and incantations.
On Emc2 wiki there is a path to g code generators. In it there is a program to generate counterbore's but I am not able to get it to work when Kim and I installed it. can any one help point us in the right direction?
is there an error?
Yes but I am not at the shop at the moment. the program comes up with a python error and wont load the counterbore program
pretty hard to guess without knowing what the error says...
i thought you guys read minds
not on tuesdays, sorry
I will have to start it up tomorrow and let you know what the error is
kim is up at the shop and is going to let us know what the error is shortly
is Kim like Rex from Secret life of machines?
I used to vaguely know Rex :-)
or what was the guy's name who always had to wrestle with the animals on Wild Kingdom?
"My assistant Jim will wrestle the lion to the ground"
I'd rather be Rex but only barely
I've never seen "The secret life of machines"
"Now Rex is going to drive this car that we've rigged up to flip over and roll. Ready Rex?"
SWPadnos: oh, you must
(bonus: the flipping device was some kind of explosive thing if I remember right)
sounds like mythbusters with additional danger
"Then they had safety elevators. Now Rex will stand up on this elevator we made, and I'll cut the cable! You'll see it latches in place instantly."
sounds like the road runner and the acme explosive company
SWPadnos u can see most of 'secret life of machines' on youtube, it was a classic series
Search Youtube. The FAX machine one is very good
andypugh yes the fax machine is my favorite
tim (?) has a website now, and he has the whole series available on purpose
andypugh: the giant sewing machine was neat too
[22:01:21] <pjm__> http://www.timhunkin.com/
but yeah, the semaphore fax demo!
he recently did a new machine 'whack a banker'
Hi, this is Kim at the Bridgeport II (BP2 unit "a" that is, unit "b" is still in pieces, LOL) Anyway, screenshot of error is now at http://imagebin.ca/view/S9Q9P4Sx.html
hm, bizarre, that looks like fine syntax to me
(just guessing:) did you fill in the spindle rpm?
Don't know, where do I do that?
oh don't you get a screen with things to fill out when you start the program?
maybe I don't understand what you did to get here
I may have, haven't seen it in a long time.
I just picked counterbore.py from the nc_files directory.
I get a different error: AttributeError: Application instance has no attribute 'NumberOfSpirals'
Is there something I should do to somehow reset or re-initialize?
the screen to enter data never comes up just the error message
oh ok, that's a different problem than what I see. I get the screen, fill it out, and then get this other error
when I look at the code, it looks like the relevant part is commented out
I opened counterbore.py in gedit just now, but I don't know much about python. Wait, is this one of those python versioning things?
3.4 vs. 3.5 or whatever it is?
I guess I don't know
shall I "python --version" or something?
% python -V
you oughta ask jthornton if he's got a newer version - I'm pretty sure the one on the wiki today doesn't work
Can I revert to 2.4 using Synaptic?
better to fix the script
yeah, don't do that kind of destruction...
it's got other problems anyway - I bet it's a simple fix to update it (if in fact the thing you're seeing is a python version problem)
OK. Is that something I can fix now myself, or should I wait to chat with jthornton?
only you can know that :-)
use The Source, Luke!
Well, I mean if jthornton has the only copy of what I need then I must wait, but if you can tell me where to go to get what I need then maybe I can fix it now with your help?
I vote we talk to jthorton
it runs on my box and I have 2.5.2
not running it from emc though
the version from the wiki just now?
OK, well, Bridgeport_II wins with 51% of the vote, LOL (it's his machine). So I guess we'll wait to hear from jthornton.
you ever coming back to NE, Kim?
is "nfm" the not version of "wfm"?
Yes, before the end of the year, at least briefly. After that don't know. Still looking for work here, there, everywhere.
arg. hope you find what you want soon.
Yeah, thanks. John and I both.
kim - you have mad skills for not finding work. Scary.
that sounded funny.
skunkworks_: I'm still tryig to figure out what you said
so am I
Kim try turning off EMC and then run counterbore
for how smart kim is - and not finding work.. Scary.
Nobody wants to make anything anymore. Everyone wants to have it made in China and brought over in big container ships. So should I learn Mandarin or Schezwan?
skunkworks_: And I didn't even see him comment on anything
mah screenshot lection.archivist.info/bore.png
[22:29:53] <archivist_emc> http://www.collection.archivist.info/bore.png
* archivist_emc needs a job in UK
archivist_emc: that's what I get too - notice the code is cut off and there's an error in your terminal
archivist_emc: Hey, you don't have enough shells opened, try 20 more at least
Jymmm: met him at the cnc workshop.
cradek, I was talking in terms of not crashing
counterbore.py gives command not found. I said python and got >>> but now I don't know what to do. (I hope ^c or Q or esc exits, LOL)
I ran it by python counterbore.py
$ python counterbore.py gave me File "counterbore.py", line 422 if len(self.SpindleRPMVar.get()) > 0: SyntaxError: invalid syntax
the ^ pointer is pointing at the colon after the last zero.
re copy it from the wiki, have you added a syntax error accidently
Don't know, but recopying is a good idea, will do...
We did not edit it at all (should we have?) don't know what to do with it.
what are you pasting it into? I have had good luck using gedit..
heh - those should be files though..
OK, I guess this rates a "Doh!" Thanks for the recopy suggestion, it seems to be working fine now. (Although I've never seen it before, LOL)
We never pasted any of them into anything, just did a "save link as..."
and saved them to the nc_files directory.
dont trust its output though as we see an error when running it
Oh, OK, thanks for the tip, what kind of error(s) are you seeing, and is it (are they) repeatable?
this error AttributeError: Application instance has no attribute 'NumberOfSpirals' and short gcode
seems to be not calculating spiral count correctly
As I recall John T mentioned something about spirals needing to be fixed on the wiki page re that program
What version of python are you running? (2.5.2 here) Shall I enter your values to see what mine does on the same data?
2.5.2 here and yes he mentions At this time there is a bug if you have a path that does not require a spiral... working on it
From wiki page ...
* At this time there is a bug if you have a path that does not require a spiral... working on it
I added more variation and still get the error
* BridgeportIIa must close in about 10 minutes in order to arrive on time for a delicious meal of roast beast
I have time to try entering your values, what capscrew, etc did you pick?
dont be late no food left
Thanks for the warning, LOL
* archivist_emc slaps jt for the code not matching the picture http://www.collection.archivist.info/bore2.png
his pic on the wiki has an extra box to fill
I just discovered that when I went to fill in the posted values, the two images are with and without a "Spiral De(covered up)" box.
I was trying to copy the wiki values to test :)
Also, is stepover % 0-1.00 or 0-100?
OK, John and I will play with it later, it's at least starting now. I don't know what the error was. Bad download??
* BridgeportIIa can smell the roast beast cooking all the way across town
Thanks very much for your help gents, I've got to go now. John can stay and chat awhile though, I'll bet.
KimK is now known as KimK_afk
weird roast beast...
ries are you here? =)
frallzor: YEs sir, I am
If I get disconnected, wout a couple of mins so I can start my generaotr... I will nto have elctrocity this evening (checking what time....)
how did you mount the v-groove bearings?
did you make the bearing support bases with a lathe?