lerman, you kicking?
was looking for hydra
is it sunday already?
amazing isn't it
does that mean tomorrow is monday?
that's the pattern I've noticed so far
Does anyone here have access to a CNC machine?
Can you make a part for me?
maybe... can you describe the project a bit more?
a small half inch thick plastic board
with 100-200 .25in holes in it
a few cm apart
can paypal you money or something
what is it for... what tolerances?
what sort of plastic?
its an led mount
ah... so not critical
doesnt have to be that precise
cheapest plastic you got
that will work
im gona have to figure it out
one sec let me pm you
did you get the pm?
im not registered or something
[freenode-info] if you need to send private messages, please register: http://freenode.net/faq.shtml#privmsg
thats what it said on my status
what your email address then?
do you have aim
email sent... nope... dont have aim
alright out for the night
ink251: keep in touch with email
lerman_ is now known as lerman
yo ho ho
close, bottle of run ya bum!
I hate when they don't give detailed specs.... http://cgi.ebay.com/Mitsumi-FA404M-7in1-Card-Reader-with-Floppy-Drive-Black_W0QQitemZ160098397910QQihZ006QQcategoryZ44959QQrdZ1QQssPageNameZWD1VQQcmdZViewItem
one pic shows header for ribbon cable, but ya cant be sure of it
those are cool
cradek: You never saw those before?
they are a usb hub with card slots glued on top of a normal floppy drive
yes I've seen them
nice for small cases
I need the card reader and dont want to give up fdd, not sure if I have a spare internal usb header though
I'm not sure that's the standard internal usb connector
aren't they usually 3x2 not 4x1?
let me look at my mobo, 2 sec's
Hmmm,, houston we have a problem....
MY mobo has the front jackes on a special connector, damn damn damn
front panel connector is unique.
[03:28:26] <Jymmm> http://tekgems.com/images/large/SYS-VG61-unit.jpg
[03:30:50] <Jymmm> http://tekgems.com/Products/et-10972-sys-vg61.htm
I love the blower in this thing
cradek: I'm glad you said something, made me check if nothing more than the internal USB. I was planning on buying a mini usb hub and installing it internal. But I down't think I can.
you can't piggy back USB can you (passively)?
Like if I soldered and internal connector to the front panle connector just making sure sure I use one or the other
how can I link a signal to a paramter in HAL? is it possible?
hmm, big problems getting emc to shutdown... I'm runnign with a jdi.py script and a pyvcp panel, and I need to shut those down with the emc script's Cleanup() functino. Trouble is, they're both python scripts, so the only way I can get their pid is to do pidof python
Even then, things are not being shutdown correctly
anonimasu_get is now known as anonimasu
actually i think i may have fixed it
jepler or cradek: were you talking about working on a cairo-based eda program at one point in time?
wondering if anything ever came of that
circuit diagrams and pcb traces
cairo? wasnt that win98s code name
thanks for the tune
cairo is a graphics library
alex_joni: coppe is my new addiction :D
fenn: you must have me confused with someone else
fenn: nice to see you btw
Guest657 is now known as skunkworks_
fenn: you may have been thinking of this effort? http://arda.no-ip.org/crap
irc user valarq, though I haven't seen him around lately
Fenn! how are you doing? missed you at the workshop.
thats gnome canvas tho, i remember him complaining about it sucking
bbl, have to go to work..
what did i miss?
umm.. rigid tapping and a lot of junk food
hey should you be showing off the working version? :)
looked like it made jelly out of the threads
they had tapped aluminum also..
[12:21:43] <skunkworks_> http://www.youtube.com/watch?v=UF493BFh_-4\
[12:21:48] <skunkworks_> http://www.youtube.com/watch?v=UF493BFh_-4
heh love the comment on that video
"looks like the last place I worked at. A bunch of engineers talking shit and no machines making anything ROFL no wonder everything is moving to china"
fenn: nice comments here also http://www.youtube.com/watch?v=vNn-Yr7it5s
fenn: where did you see that comment?
why's the toolchange so slow?
* alex_joni feels blind
lerneaen_hydra: positioning the spindle takes a while
alex_joni: probably on the time-lapse video -- I'm just guessing
seems to take longer that it has to
lerneaen_hydra: come to the next CNC workshop and fix it for us then
another thing that makes it take longer than it needs to is that the table moves in XY needlessly before the toolchange
skunkworks_: I don't see any comments there
neither do I -- "Comments: 0"
Views: 106 | Comments: 0 | Favorited: 1 time
jepler: :p I meant the part when the spindle doesn't move (or moves imperceptably) and just sits there for a couple seconds
oh - http://www.youtube.com/watch?v=l5q1rVKYLTQ
that was the 'not compleat' video
jepler: around 0:54-0:56
huh I hadn't seen this video: http://www.youtube.com/watch?v=BPGsMDM_UPU&mode=related&search=
anyone know more about the machine?
that fan sure does oscillate quickly ;)
is kinematics flexible enough to be able to program that as an XYZ machine?
how will it behave if you command a feed through the middle, the spindle can't spin inf fast, will the feed be lowered gracefully?
depends on the kinematics - that's a case it would have to handle
I bet you'll get Xpos=-nan Ypos=-inf
and the steppers will go tits up
or maybe +inf
'tits up'? ;)
or maybe down?
depends on the funeral ceremony
but mostly up, I think
yeah, I mostly seen it that way
next year we need to leave the camera running and make sure the last frame is of a hand reaching to take it down :)
this is fun to watch http://www.youtube.com/watch?v=Wk2VYwCIcZE
the live center seems a bit underdimensioned for the cutting forces, wouldn't there be lots of deflection?
alex_joni: I just saw that before..
that's a better exampl
that doesn't sound that bad does it?
heh nice jackhammer
yeah.. it sounds like shot
right now the stage of the planner that determines the maximum (cartesian) velocity and acceleration for a move has no knowledge of the kinematics at all
there was some discussion about having derivative and limit functions in kinematics
but I think it lasted about 5 minutes
it's easy to identify that some parts are missing that prevent emc being useful on nontrivial machines, but harder to decide how to fix it
even if you could adaptively slow the feed on a move that passes near to or through the center of rotation on a turntable kins machine, that will have consequences for part finish ..
that kinbd of slowing is already done for trivial machines
yes naturally, however that's up to the code generator to take into account
figure it out in cam, use inverse time mode
I think there's no general solution; we can't make emc run any shape of machine and get everything right
IMO the ideal behavior would be if you specify a feed A, and it would attempt to run the axes to acheive that speed, without exceeding any of the maxvels for an axis, slowing down feed to compensate
so in the case of the ornamental engraver and a line that goes through the middle it would have to pause and wait for it to spin half a turn
(don't know what to do about joint limits though)
s/maxvel for axis/maxvel for joint
SWPadnos: yes but for trivial machines either you'll hit the desired speed on a particular straight move (if it's long enough) or not -- so you can just look at the F-numbers in the program and see if your machine is that fast
and a particular move will have one cruise velocity
in the case of a kinematics machine, there might be an acceleration profile (in cartesian space) through the whole move
sure, but a G2/G3 that uses axes with different limits will exhibit the "finish problem"
since dtheta/dt is not constant
well I can see a couple of janky solutions: an internal adaptive-feed-from-kinematics, or an external one that has to be hooked up for proper functioning
isn't each 'axis' in the ini limited.. Wouldn't that limit the motion? or is there a need for limits on top of that?
yeah - programmed feed rate ...
the trouble comes in with nontrivial kinematics - what does "10 degrees/second" mean in cartesian coordinates?
the answer: it depends on R
even worse is the question "how fast do I need to rotate theta for a linear move in X"
the answer: it depends on how close to the center of rotation you get
right - but you have that problem with a normal xyza (4axis machine)
no you don't, because combined linear/rotary moves calculate the time to complete the XYZ portion and use that for the time to complete the rotary portion
ie, the roptary part is effectively done in inverse time mode
as cradek said - inverse time seems like it the way to go.
who cares about time, i'd want constant chip loading and have it do it as fast as it can (to a limit)
feed rate is just a speed limit anyway
[14:25:55] <lerneaen_hydra> http://www.snotr.com/video/83
fenn: how are your projects coming?
slower than you can shake a stick at
today i am playing with opengl in python
[14:48:45] <fenn> http://fenn.dyndns.org/pub/irc/circle.py
wondering why the mouse wheel binding doesnt seem to do anything
Cool.. Slow here also. although I have been playing some.. http://www.electronicsam.com/images/KandT/servostart/ampmess.JPG
did you see that jepler found a $60 fpga and put 4 axis servo into it + i/o?
he also has a step generator setup but untested.
(that is what is running that h-bridge)
skunkworks__: you sure like to sing my praises
:) damn cool.
that's a big capacitor..
it is 1900uf at 350v I think
what voltage do you run those servos at?
I am needing about 150v for the monster servos
the one in the picture is running about 50v 10a
i finally got a laser printer, so i should be making pcb's like mad (but i'm not)
:) I have been milling them http://www.electronicsam.com/images/KandT/servostart/top.JPG
never tried the laser transfer
that's very clean
fenn: did you see the big thread on cnczone where someone hacked a inkjet to print directly to copper clad?
the big deal was cooking the ink iirc
how do you make the clear areas so smooth? usually there's machining marks left over
That circuit board took a couple of hours to mill... The machine will only do about 7ipm before it starts to shake too bad.
I used a 1/16 mill to clear
cradek: speaking of milled circuit boards, does the CVS version of gcode.ulp produce files that do your tool-length-computation thing?
and then the pointed mechanical etching bit to do the fine work.
lets see a hi-res scan of that board :)
jepler: I think so
03jepler 07TRUNK * 10emc2/src/emc/usr_intf/axis/extensions/minigl.c: additional opengl functions and constants
skunkworks__: I don't think your rounds are round... around the cap pads it seems like the circular cutouts are very squarish
SWPadnos: can you open a .sat file?
err - I dunno
IGES is the official name
I can - what do you want me to open it with?
in that case, maybe ;)
cradek: anything, I was wondering if the geometry is right
it's ACIS, not IGES
[15:12:28] <alex_joni> http://188.8.131.52/~juve/bariera%20optica%20-%20senzor.sat
SWPadnos: sorry.. yeah
(I wouldn't have known if not for filext.com)
cradek: it should be something roughly a meter long
speaking of directly printing on copper clad.. http://www.warrantyvoidifremoved.com/taxonomy_menu/18/17/21
how's the board fed through the rollers?
lerneaen_hydra: not sure exactly; i think he made some kind of custom drive roller
cradek: I think the pads are 8 sided
shows the toner cartridge sitting on the desk still hooked up to the printer: http://www.flickr.com/photos/warrantyvoid/430512524/in/set-72157600042844344/
That is the fuser.
are they different things?
I suppose he runs it thru the toner - then thru the fuser to 'fuse' the toner to the board
so the blue dust inside the printer has something to do with that
this method is clearly superior to press-n-peel
dont let the boss see the sink in that state
have you tried laser-printer-iron-on method?
I did it once, it worked suprisingly well.
did a smurf explode?
did you use the special paper? I used transparencies and was pretty unhappy with how it worked (I also tried several other things)
glossy inkjet paper was second best I think
I used inkjet photo paper
cradek: managed to open that?
maybe it's me :-)
alex_joni: sorry, I was joking, I don't have that kind of software
oh :P ok
I thought swp was helping you - I didn't realize you were waiting on me
it's not that important.. somehow it's off by a factor of 10
not sure if it's the program I'm exporting it with, or the next one I'm importing it into
I'll have the proby scale it all by 10 ;)
it is a 'proby scale' uses an external probe to decide on the scale amount. ;)
oh -- "the new guy"?
the new pro
it's actually a student working a couple weeks durign summer
college around here requires you to do that
both to gain some experience and to .. (something else I forget atm :P)
nope, they don't get paid
I'd probably say "intern" then
[16:29:12] <cradek> http://www.cornwellauction.com/flod17.jpg
what is this?
it's a yak shaving machine
it's an offset-platform multibore dinglehoozit
a copy engraver/miller
If any of you are using firefox and the save password feature please read this asap -> http://www.heise-security.co.uk/services/browsercheck/demos/moz/pass1.shtml
cradek that looks a useful toy
hmm.. internet at home must have gone down
skunkworks__ is now known as skunkworks
seems something foobared at my regular irc client's place
JymmmEMC: did you get your machine straitened out?
skunkworks: No, still have the leadscrew soaking in charcoal fluid to get rid of all the grease.
what axis acceleration values do people here have for small stepper based mills?
skunkworks: I'm suspecting that the lead end-mount of the leadscrew is a little off compared to the right, and the ballnut mount is putting too much upward pressure on the leadscrew.
lerneaen_hydra: I think chris's maxnc is 20 inch/s^2, max velocity .5667 inch/s
so about 500 mm/s^2 and 15 mm/s
(looking at configs/max/max.ini)
hmm accel that high?
oh, how small was it?
like 10-20kg X/Y assembly?
yeah certainly under 20kg
you did say "small"
yeah, guess I'm used to mazak sized and anything that's not room sized is small to me ;)
lerneaen_hydra: Don't you mean "anything that you don't build a room around for it...."
you basically don't notice an acceleration phase -- if my math is right, it takes about 30ms to accelerate, during which time it travels about 1/4mm
machining usually has soft curves too so its usually irrelevant
this thing cuts a fair number of nominally square (45 degree and 90 degree) curves when milling circuit boards
I guess "square" isn't quite the word I want to refer to a 45 degree bend
but not rounded at all
oh, I see
yeah PCBs are another matter
the mill I'm running now seems to need an accel of about 200mm/s^s to not miss steps
how do you find good values for accel? start low, increase, see where it misses steps, back down a bit?
lerneaen_hydra: trial and error..
25 secs to 5m/min :/
that's acceleration per second ;)
so .5 seconds to 5m/min
real units here :p
that's more sane
Z is only 75mm/s^2 though, as it's so heavy
lots of mass
03alex_joni 07TRUNK * 10emc2/src/emc/rs274ngc/rs274ngc_pre.cc: fix bug #1734309, on abort the already opened % condition was not properly reset
skunkworks: you still at the keyboard?
03jepler 07TRUNK * 10emc2/src/emc/usr_intf/axis/extensions/seticon.c:
new window manager functions (not used in AXIS):
* get X window ID of wrapper window
* set the NET_WM_STATE of a window
* set the WM_TRANSIENT_FOR of a window by window ID
alex_joni: lerman is around, now.
lerman: thanks, but I nailed it
I was under the impression the error I was seeing was because of o-words
but it was a red herring
03alex_joni 07v2_1_branch * 10emc2/src/emc/rs274ngc/rs274ngc_pre.cc: backport fix for #1734309, on abort the already opened %-condition was not properly reset
is there a way to link a hal pin to a parameter?
03jepler 07TRUNK * 10emc2/configs/demo_sim_cl/demo_sim_cl.ini: get rid of mentions of bridgeporttask and minimilltask
03jepler 07TRUNK * 10emc2/configs/etch-servo/etch.ini: get rid of mentions of bridgeporttask and minimilltask
03jepler 07TRUNK * 10emc2/configs/demo_mazak/demo_mazak.ini: get rid of mentions of bridgeporttask and minimilltask
03jepler 07TRUNK * 10emc2/configs/dallur-thc/dallur-advanced.ini: get rid of mentions of bridgeporttask and minimilltask
03jepler 07TRUNK * 10emc2/configs/plasma-thc-sim/plasma_mm.ini: get rid of mentions of bridgeporttask and minimilltask
03jepler 07TRUNK * 10emc2/configs/demo_step_cl/demo_step_cl.ini: get rid of mentions of bridgeporttask and minimilltask
03jepler 07TRUNK * 10emc2/configs/puma/puma.ini: get rid of mentions of bridgeporttask and minimilltask
03jepler 07TRUNK * 10emc2/configs/halui_halvcp/halui.ini: get rid of mentions of bridgeporttask and minimilltask
xemet: no, parameters cannot be linked.
03jepler 07TRUNK * 10emc2/configs/plasma-thc/plasma_mm.ini: get rid of mentions of bridgeporttask and minimilltask
03jepler 07TRUNK * 10emc2/configs/hexapod-sim/minitetra.ini: get rid of mentions of bridgeporttask and minimilltask
03jepler 07TRUNK * 10emc2/configs/sim/ (10 files): get rid of mentions of bridgeporttask and minimilltask
xemet: that is the main difference between a pin and a parameter.
03jepler 07TRUNK * 10emc2/configs/max/max.ini: get rid of mentions of bridgeporttask and minimilltask
03jepler 07TRUNK * 10emc2/configs/nist-lathe/inch.ini: get rid of mentions of bridgeporttask and minimilltask
03jepler 07TRUNK * 10emc2/configs/scara/scara.ini: get rid of mentions of bridgeporttask and minimilltask
03jepler 07TRUNK * 10emc2/configs/stepper/ (pluto_inch.ini sim_inch.ini stepper_inch.ini stepper_mm.ini): get rid of mentions of bridgeporttask and minimilltask
so there is no way to write a parameter except from halmd
03jepler 07TRUNK * 10emc2/configs/ppmc/ppmc.ini: get rid of mentions of bridgeporttask and minimilltask
xemet: in a .comp it is very easy to turn a parameter into a pin; in a .c file it is a bit harder.
03jepler 07TRUNK * 10emc2/configs/stepper-gantry/stepper_mm.ini: get rid of mentions of bridgeporttask and minimilltask
03jepler 07TRUNK * 10emc2/configs/stg/stg.ini: get rid of mentions of bridgeporttask and minimilltask
03jepler 07TRUNK * 10emc2/configs/vti/vti.ini: get rid of mentions of bridgeporttask and minimilltask
03jepler 07TRUNK * 10emc2/configs/stepper-xyza/inch.ini: get rid of mentions of bridgeporttask and minimilltask
03jepler 07TRUNK * 10emc2/configs/univstep/univstep.ini: get rid of mentions of bridgeporttask and minimilltask
03jepler 07TRUNK * 10emc2/configs/univpwm/univpwm.ini: get rid of mentions of bridgeporttask and minimilltask
I have a pid and I would like to change P-I-Dgain paramters using three pyvcp bars...but I think it is not possible
xemet: no, you would have to change the source code to make them pins, or change pyvcp so that it can invoke "halcmd setp"
uhm...I don't think I would be able to change it
it would be helpful to have a way to esaly change the pid P I and D values
have you tried the calibration screen? it works nicely
I think some of those numbers are way too touchy to use sliders
anonimasu: no, it is not automatic
but it lets you enter parameters, then test, then enter new parameters until you are satisfied
if you have a servo based system (or any closed loop system) can the commanded/actual position on one axis affect the position of another axis to keep toolpath as close to ideal?
anonimasu: there is an automatic tuning HAL module in development but I have not personally used it successfully
so...:) is there a calibration screen?
never seen it!
I'm using the pid to control the spindle speed of my micro milling machine
ie if one axis is 0.1mm from ideal, could another axis move to compensate and try to keep the tool along the toolpath
xemet: in AXIS, choose from the menu: Machine > Calibration
but that will probably not show you anything about your spindle, because it looks at the [AXIS] sections of the inifile
nothing to do with that
xemet: if you wish to understand how to convert a param to a pin in a .c file, you may find this helpful: http://emergent.unpy.net/files/sandbox/pid.c-make-pgain-a-pin.patch
it converts "Pgain" (only) into a pin
skunkworks: have a look at TI's uc3638 datasheet.... It looks like it would be a great thing to marry to the h bridge (see page 6 of data sheet)
skunkworks: I am working on a controller board at the moment to see how to use it effectively
thanks jepler...I could try that
there are 3 main parts to this conversion: First, change the HAL shared memory item into a pointer. second, change the creation step to use hal_pin_xxx and HAL_IN instead of hal_param_xxx and HAL_RW. third, use the "dereference" operator everywhere the value is used (e.g., '*(pid->pgain)' instead of 'pid->pgain')
you would have to do the same work, but for I, D, and the ff012 if you use them
well...I don't know if it is convenietn for me to do that work...maybe is better if I change the values from terminal
I've reached a good configuration I would like to fine tune it
Do you have a particular method to tune pid?
Today it was the first time in my life I've seen a PID
and I found it a little hard to tune :)
what are for example FF0 FF1 and FF2? I've them setted to 0. Have I to change them?
not for this application
I bet you can use P only
increase it until it oscillates around the set point, then reduce it until it stops oscillating
at that point you might be done
I tried P only, but it was a little oscillating on load
maybe reduce P, or add a little D
it was ok when there was no load applied
the aim is to mantain constant (or about costant ) the speed when load is applied
using only P the speed was always above the commanded value
I've read this is normal using only P
and read I've to use I to reduce the offset
maybe you need some I then if the exact value is important to you
I compensates for steady state error
now is not bad, only at low speed (in my micro milling low means about 5000 rpm) it tends to oscillate when load is applied
above 7000 it works fine
try a little less P or more D to fix the oscillation
D is very touchy - change it a little at a time
tuning is terrible"
yeah...it was fun...but a for me it was a process based on try
I don't know a method
is your feedback an encoder?
my feedback is an hall sensor
there are two magnets on a plate attached to the spindle
about how many pulses per pid run?
and the halla sensor read the passage
let me think
I've the pid function attached to a thread with 1/25 s period
I need a high period
vecause I only have two magnet
so at low speed I've about 83 pulse per rotation
pulse per second
3.3 / period
that's way too few
you will not be able to tune it
that only gives around 4 pulses /period at 7000rpm...
think what happens when it goes from 3 to 4 per period
try a .5 second period
so you say to increase the period of the pid
either that, or add more resolution
in order to have more pulses per period
(more magnets, or encoder)
you need more somehow
no...to add more resolution I've to change everything
that was only an experiment...
or, if there's a low-pass filter component for HAL. but that's roughly the same as lengthening the period
then you have to increase period
tell me another thing
I've the counter update function attached to the same thread of the pid
is that right?
that should run fast enough to catch every pulse
if it's not a hardware counter?
there are two functions in the counter
xemet: oh I remember, sorry
no I'm using the hal counter component
capture-position should be in a fast thread
update-counters can be in the slow thread I think
it is in the base-thread
update is in the same thread of the pid
just run update-counters before the pid that reads its output
yes I think that's fine
well the best thing would be to have more resolution I think
because a long period means a long response time
that would give you faster response because you can run pid faster
according to the manpage, update-counters is for fast thread and capture-position is for slow thread
yes, update is in the fast and capture is in the slow now I remember
ùthe first is the function that read pulses and increase counter
in TRUNK, the velocity estimate is better at low speeds because it is based on the period when the count was seen
the second read the counter and calc velocity position etc.
(in 'encoder' module only, not in 'counter')
however, add more resolution is hared because the space in my Proxxon MF70 is very small
try longer period, it might be fine
I think I'm the first in the world to have a sensor placed in a Proxxon MF70 and read the speed
How do you like the MF70?
I was considering one.
and increasing it by pwm signal
the MF70 is good for very small parts
did you DIY the cnc conversion, or use a kit?
I've built my conversion
I give you a link
[19:50:43] <xemet> http://www.emc2cnc.altervista.org/index.php?option=com_content&task=view&id=17&Itemid=31
it is my page
14:42:15 < cradek> think what happens when it goes from 3 to 4 per period
but it is in italian
at page 2 you can see some pictures
[19:51:43] <xemet> http://www.emc2cnc.altervista.org/index.php?option=com_content&task=view&id=17&Itemid=31&limit=1&limitstart=1
I will write a page about the spindle speed regulator using the PWM signal generated by EMC2 and the sensor added
* bill2or3 looks
that sure is a small mill
"Così ho realizzato un sistema senza Hard Disk e senza lettori CD, Ubuntu con EMC2 gira in modalità Live Persistent da una Flash Card di 1 Gb montata in un adattatore IDE."
03alex_joni 07v2_1_branch * 10emc2/debian/changelog: note bugfix for #1734309
isn't running the mini-itx board naked behind tempting fate to throw some metal chips in it?
xemet: so you are using a 1GB flash disk to boot and to store the OS?
jepler: that's what I'm reading too
it is a live system
I wish I was as smart as you guys who have been able to get that working
xemet, it looks nice.
the ITX board now is covered :)
ah nice :)
xemet: USB, or the kind that plug into an IDE port?
very neat - I like the pictures of the bearings on fingertip
I've ran also the system from a USB pen, but it was slower
xemet: will you write a page in English on wiki.linuxcnc.org about installing emc this way?
jepler: petev fixed it - now the pid autotune seems to work for em.
I could do that
but I've to remember all the things I've done
I has having the same issues you where at the workshop
I've read some links on the web
skunkworks: oh -- I guess I missed that
skunkworks: that's good to hear
because I was not able to do that
xemet: if I understand what I read, it is installed on this flash so that it still runs like the "live CD", with another area storing your files?
* jepler <-- trying to read italian using his mostly-forgotten french from university
I turn on the other PC and send you some link
xemet: OK I'll read them
xemet: if you need to install a newer version of emc, how do you do it?
remount rw? then update/upgrade ?
no I think the root fs is already rw, it's a copy of what's on the disk
ok I'm here
it's compressed on the disk
the base is that:
[20:01:36] <xemet2> https://wiki.ubuntu.com/LiveUsbPendrivePersistent
you could update it in ram (would have to do it every time)
it is for the USB pen
^ that is what I used also
I had to change some things
except using the emc live cd
xemet2: thanks -- it's good to know that this procedure works with the emc live cd as well
had to format the first partition to FAT16 instead of FAT32
and had to create the partitions with gparted and to format them using mkfs to give them the right name
I had to use lilo iirc for mine to boot
however, to use the Compact flash on ide
the difference was that:
[20:04:00] <xemet2> http://ubuntuforums.org/showthread.php?t=71567&page=18&highlight=initrd
change the script casper in the initrd
because if not it will hang at start on the "mounting live CD" if I remember well...
if you need to upgrade the system, it will be stored in the persistent part of the Compact Flash
oh is the persistent part some kind of overlay?
I've updated EMC2 two times...I don't know how much space I've now in my persistent partition
what's an overlay?
I think I can't update it forever...
something like a transparent filesystem
which can get mounted over another one
replacing things which are already there
because the casper-rw had a limited space
ues...I think so
the files on the original show through if nothing on the persistent is newer
maybe it is not a very good system...because the best method is to remake the live partition with the changes
it is compressed so that is very hard
there is the procedure
I've used it to make a lot of live cds
if it is IDE you could just install to the device normally right?
then there would be no problem like this (assuming it is large enough)
yes, if it is big enough
I've read that Compact flash suffer if it is written a lot of time
cradek: The only issue using a Cf as a HDD is the constant write cycles to the CF will degrade it big time
so I thought it was better if it is read only
like the live CD
at least the live partition
[20:10:00] <xemet2> https://help.ubuntu.com/community/LiveCDCustomization?action=show&redirect=LiveCDCustomization%2F6.06
just use that to upgrade the live content
and recopy it in the card...
"Keep in mind that not everybody we want to convert to Ubuntu is a power user. I can pretty much guarantee that anything that looks like typing Scary Strings is going to turn people off, not on, no matter how "powerful" it is behind the scenes."
* alex_joni heads to bed
good night all
xemet2: Can you image the CF so it can be emailed and burned to another CF ?
LawrenceG: I will look
xemet2: including partitions?
JymmmEMC: there's a good app/OS called G4U that does it
lerneaen_hydra: partitions too?
I don't know...I'm not so expert :(
BSD based, copies the entire disk or partition
lerneaen_hydra: like ghost?
if you tell me how to do I can try
[20:12:54] <lerneaen_hydra> http://www.feyrer.de/g4u/
uploads the disk to an ftp server
lerneaen_hydra: excellent, getting tired of ghost
it's really nice actually
uhm I've to read
lerneaen_hydra: can if work from a live OS?
it boots from a CD/other device
so it's its own OS
lerneaen_hydra: ie... cp a live hdd to another hdd?
and therefore completely independant on whatever is installed on the computer
lerneaen_hydra: No, I mean live disk-to-disk backup
oh this isn't for live backup
but I'll still check it out, thanks.
"5.1 Supported filesystems
One of the questions arising a lot is "what filesystems does g4u support". The answer is: "all of them". g4u reads the disk bit by bit, starting from byte #0 to the end. It includes any MBR, boot record, partition table and the partitions themselves without further investigating the structure of the data stored in these partitions. "
returning to the MF70, if some one is interested, here is a picture of the spindle sensor: http://imagebin.org/9470
one of the 4.7 nF capacitor was replaced with a 100nF
xemet2: made by an individual
as opposed to a product made by a company
usually a "homebrew" item is also unique
and here the sensor mounted: http://imagebin.org/9471
yes homebrew :)
how much did the mill cost?
the mill cost about 250 EUR
xemet2: what is the sensor for?
to read the spindle speed
ah - no tapping.. (index)
xemet2: 250€? that's almost free
it is very small...
the proxxon mills are very small
that's... that's.. very cheap
is the quality any good?
it probably weighs about 5 to 8 kg and has a travel of 50mm on Y
[20:21:38] <xemet2> http://www.emc2cnc.altervista.org/index.php?option=com_content&task=view&id=17&Itemid=31&limit=1&limitstart=1
is it made out of aluminum?
we had another user with one of those -- <crepincdotcom> or something like that
he managed to mill some circuit boards with it, I think
less than 50 mm on Y :)
looks like cast iron though
used the MF70 to make most of the parts
[20:23:54] <xemet2> http://video.google.it/videoplay?docid=-6418470384160839689
with a mill that costs 250€ you could probably make a complete cnc solution for €500
[20:24:17] <xemet2> http://video.google.it/videoplay?docid=-8248420887418059619
yes...I think I'vespent maybe less than 500
I use it more to experiment
you're milling very very slowly in the video, why?
because I cannot go too fast because I would break the mill...
I so want a micro mill to play with.
still, isn't that mill at least 2mm in diameter?
the best thing is that you can place it on your desk
the mill is one flute 2 mm
ideally I'd want to find one a bit bigger, maybe twice that size
20-30kg in weight
* skunkworks would use servos and the pluto.
I can mount max 3.125 mm diameter
I've the pluto
I will try a day
still have to find theservos
it looked like you were machining at something like 20-40mm/min
maybe a bit more, say 60
shouldn't it be going at at least 300-400?
Z -1 mm
that's what I've always done and it's worked for me
yeah sounds about right
I've tried 100
the mill broken
maybe those mills are bad...
anyone else here have any input on that?
a 2mm endmill
1mm cutting depth
sounds about right..
you will need lube..
oh, one flute
yeah ethanol would be quite good IMO
I have no lube
that could be why
maybe it clogs up
wd-40 works good :)
try with ethanol maybe?
no mess afterwards
wd40 works very very well..
I don't know if ethanol would work - kerosene will
anonimasu: 5-56 is the same/similar to wd40, right?
yes the mess is the problem
lerneaen_hydra: not very..
skunkworks: all the literature I've looked at has reomended ethanol as a cutting lube, though *only* for alu
lerneaen_hydra: I use wd40 for milling alu.
I think I tried 5-56 but it werent as nice
[20:33:47] <skunkworks> http://www.cnczone.com/forums/showpost.php?p=321680&postcount=167
lerneaen_hydra: never run across that... doesn't mean anything though.
IMO it works really well, and there's no mess to clean up
you need to ventilate the fumes somewhat though
lerneaen_hydra: did you try wd40?
not in alu, only in steel
[20:35:28] <skunkworks> http://s12.photobucket.com/albums/a220/ArieKabaalstra/Beagle-1/?action=view¤t=acee4cfc.flv
also tried thick thread cutting oil, didn't work as well
wd40 is cheap on jula..
they have it at jula?
2 cans for 99
skunkworks: the alu has lots of burrs :/
lerneaen_hydra: use better tooling..
in my experience if you have ethanol there aren't any burrs at all
and use lube..
oh, that's not my mill
most of my alu machining woes has been due to crap tooling.. and no lube..
that spindle sounds a bit underpowered
taking a 1mm cut in alu with a 3mm endmill and sagging a lot in speed
[20:38:00] <anonimasu> http://en.wikipedia.org/wiki/WD-40#Formulation
ah, the napthna (heptane?) is probably a good cutting compound
though not at all similar molecularly to ethanol
lol "WD-40 dissolves cocaine."
WD-40 dissolves cocaine - it has been used by a pub landlord to prevent drug-taking in his pub's toilets.
I like #11 http://news.bbc.co.uk/1/hi/magazine/4566526.stm
#23 ?!?!?!? wth
toast is now known as toastydeath
are we well?
tomorrow is going to be a fun day!
why is that?
got a potential new client coming round
they asked us to quote for a bunch of work ...
never got anywhere with it
cool - no pressure ;)
now, they want us to "just be a back stop, a fallback position should our machine break, we can use you till its fixed"
basically, they simply can;t afford to be 'down' for a week
and there are very VERY few people who can do this particular job
ever gone into a game of cards holding all the aces?
more than once.. I know what you mean.
could someone test minfest.com and see what's there?
is it the same as lerneaenhydra,net?
is this something where you charge them the same amount as if you did it all? ;)
well, that sort of thing ... frankly I'd rather have the regular income
sounds like they are asking for an insurance policy
lerneaen_hydra: works here
they should go see an insurance salesman then, we are engineers ;)
skunkworks: hmm, wtf, that's not my domain O_o
I'm tempted to play the "regular work, or no deal" card,
so ask them to pay for the policy up front...premiums are the same as if they did the entire run with you guys ;)
why the fsck is it pointed to my site
no need to do the work, just pay the same ;)
after all, they may never need it .. so I'm not *actually* losing anyting if I dont get it
robin_sz: oh btw, the Orac manual officially claims it is a 300lb lathe and it sure feels like 300lb
lerneaen_hydra, do you mean minfest.com, or www. minfest.com?
;; ANSWER SECTION:
lerneaenhydra.net. 600 IN A 184.108.40.206
wtf is up with that
minfest.com. 3600 IN A 220.127.116.11
why do I and skunkworks get directed to my page
minfest.com is a framed redirect to www.minfest,com
;; ANSWER SECTION:
www.minfest.com. 557 IN CNAME ppnu.mine.nu.
ppnu.mine.nu. 14357 IN A 18.104.22.168
which is a dynamic DNS thing
it's a swedish registrar, and both my domain and minfest share the same DNS servers
I haven't changed IP in a long time though
ppnu.mine.nu. 14357 IN A 22.214.171.124
is the entry for www.minfest.com
whois minfest.com gives me the same registrar that I have
its a dyndns thing
NOTHING TO DO WITH REGISTRAR
whoever owns minfest.com uses dynamic dns
so they probably last ha dyour IP, and havent logged in for a few while, even months
[23:06:11] <lerneaen_hydra> http://www.whois.net/whois_new.cgi?d=minfest&tld=com
<-- I don't get that then
NOTHING TO DO WITH WHOIS
unless it's minfest -> ppny.mine.nu -> my IP
whois does NOT affect eouting
and look at the NS entries for ppny.mine.nu
hmm, dyndns kills your routing if you don't touch it after a month
whois is purely a databse of registrant info, it contains references to your nameservers, but, it is not involved inrouting in any way
I suspect they must have been allocated your IP at some point?
or they mistyped their IP?
could be that too
I assume you are just getting loads of crap traffic on your connection now?
no, funnily enough I haven't
maybe their site was never popular :)
6 hits so far over a week
you run vhosts by name?
including the hits now from #emc
right, well apache can split an IP address to hundreds of sites
oh, right, yeah
it looks at the actual name that was used, and serves up a differnet site etc
so, time to put goatse pictures up for www.minfest.com
he he he :)
or worse ... ;)
easy to set up?
its debian based?
ubuntu or whatever?
put a file for the domain, I think there is a sample somewhere
I'll leave that for another day, later here now
if i was very rich
i would collect old rolling mills
from steel mills
i think that would be an amazing thing to collect
robin_sz: AH HA! You posed for goatse!
robin_sz: I mean, I've called you an asshole, but I never meant it literally!
JymmmEMC, whatever turns you on
robin_sz: Nah, none of your goatse pictorials... now tubgirl, oh baby!
JymmmEMC, I thought lemonparty was more your sort of thing?
is that like golden showers?
it is hostile in here
toastydeath: It's all good CLEAN fun =)
* robin_sz grosses JymmmEMC out
robin_sz: Eh, not easy thing to do.
true, but .. possible
robin_sz: I work at a datacenter, I see all kinds of crap all the time.
we have a lot of "colorful" client content
I think I can come up with worse :)
robin_sz: After your goatse pics, I bet you could =)
alrighty folks, I gots to get outta here and pickup a 20X dvd burner before heading to work... laters!