toastyde1th is now known as toastydeath
anyone know how to make a big F1 button for a keyboard?
why not a proper estop mushroom?
and suporting hardware
it's for my dad he is 82 and plays age of empires and the F1 is the pause button
guess halui is out of the question then :)
yea I guess so
and it's not really E-stop ;)
If I could make him a big red push button to do the F1 key it would make it easier for hime
just a shot in the dark to see if anyone here knew how
I'd get an USB keyb, and hack that one
build a big button, and connect it to the proper wires
one possibility is to get one of those key-to-keyboard things (like they use for MAME cabinets) and wire up a big button to that
he has a usb keyboard
the idea is to use another one so you can hack it up
BigJohnT: one you can cut apart
ok I have plenty laying around
can you plug two keyboards in?
ok, I'll give that a try
try it first.. before hacking up the keyb
BTW your up real late alex_joni
I know.. had an afternoon sleep
I love a nap on the weekend
well.. I do too.. except when I can't sleep afterwards
I can sleep at the drop of a hat
ok it's not real pretty but the picture quality is better. http://i47.photobucket.com/albums/f163/johnplctech/Plasma%20Cutter/Plasma011.jpg
Oh Boy I get to hack up a keyboard!
I need to fit a ball screw to my lathe cross slide...
* BigJohnT looks around for an old keyboard and a 9" angle grinder
a screwdriver might be more appropriate
LOL, but not as much fun!
well, that goes without saying
* BigJohnT flipps keyboard over, there is some screws there after all
ok, I'll put the angle grinder back up...
cool, I found a hack on instructables for the keyboard
* BigJohnT hears the dinner bell so ttugl
SWPadnos: Looky what I found... http://areca.us/products/ide_to_sata_ll_sub.htm
strange that they have SATA2 for the drives but SATA1 for the host connection
if anything, I'd want it the other way around
[02:03:55] <skunkworks_> http://www.jhchoppers.com/ebay/cnc/ATC%20Testing%201%20to%202.wmv
seems like a lot of 'stuff' to do what it does
[02:04:28] <skunkworks_> http://www.cnczone.com/forums/showthread.php?t=53998
man, that's a lot of work
certainly doesn't look like the most efficient option - separate motors per tool?
no power drawbar (so it's not complete), and R8 needs spindle orient for toolchange
i dont get it. why not just lower the quill or whatever
lowering the quill brings the drawbar out of reach of the air wrench he's using for a power drawbar
and without spindle orient (which I'm assuming he doesn't have), he's pounding the hell out of the R8 taper (around the slot in the toolholder / the peg in the taper)
by spinning it in the taper socket?
yeah - I'm betting that the drawbar grabs the collet, then when it binds on the peg, it ends up spinning the collet until it's correctly oriented, at which time it pulls up into the taper and tightens
scraping all the way
I'm not sure if that can happen though - you may need a longer drawbar to even reach that far
(and who knows, he may have an orient function)
something has to keep the spindle from turning when the air wrench is banging on it. probably the motor dynamic braking
you could make a frankenstein impact socket to act like a long spline (so you can tighten the drawbar with the quill down)
there's a brake for the spindle, you don't need a drive on the motor to prevent motion
oh i figured it had some universal motor that you could just short the terminals
easier than adding a solenoid to pull the brake, and orienting it so the brake works, etc
I think there is enough mass in the spindle for it not to be a problem..
not really - it's pretty easy to turn by hand
you could have a snap-fit catcher that goes on the end of the quill to catch the toolholder as it drops out
so you loosen/tighten the taper when the quill is up, but the catcher holds it in approximately the right place as you move up/down
the mazak used ball detents, but you could also use some kind of elastic claw shape
lower quill onto tool rack. "clunk" tool is held in catcher. raise quill. run air wrench to tighten taper
ah - then you need a claw to pull it out (like the mazak)
nah you can just drop it (need something to loosen the claw though)
and you still need some spindle orient I think
might as well have a spindle brake then, huh?
[02:30:47] <skunkworks_> http://www.cnczone.com/forums/showthread.php?t=29060&highlight=impact+wrench
why spindle orient?
because there are keys on the toolholders
these are guys with cnc equipment so why do we see so many mediocre solutions?
what I don't understand is why someone thinks that 6 (or 12) stepper motors and associated electronics is a good idea for a toolchanger :)
haas does it with one motor?
I agree that's the amazing part
as far as the vertical motion goes, you could just have the pockets on the table, and raise the knee
that would be much less expensive, and you'd get a powered knee out of the deal
yeah if there's a knee
would be a lot simpler than the catcher stuff
I want a powered knee, but I'm too lazy to fix the air assist
according to some people, air assist isn't needed
with enough gear down
my knee weighs about the same as a series I bridgeport
ok, well the people I've heard that from are using BPs, so YMMV
not sure how long the leadscrew is gonna last
doesn't really matter much - there's no backlash due to the weight
as long as you map errors from time to time, you're fine
I suppose, but it's a lot of weight on an acme thread, that's gonna cause some galling
what kind of machine do you have?
just cover it with a huge black glob of moly grease :)
series II bp
people seem to forget that the knee was actually designed as a motion axis just like saddle and table
so you should be able to move it around
yes, but it was also designed to have an air assist
built into the pedestal
well in that case, I'd fix that ;)
I know, but it weighs about the same as a bp series I
get a car jack..
I can raise it up, block it and take out the pedestal
but I'm too lazy
the earlier series II actually used a ballscrew on the knee
I want to retrofit mine
i predict that if you get it working you'll forget all about retrofitting it
it's so hard to move that knee, I dunno
but you're probably right
conditions must be bad, township sent the plows out
its pretty bad here
I ran the snowblower this morning - cutting thru knee high drifts
we got rain all week
this evening I can't even tell I did it
chimney finally gave in and started dripping
well, I got lights and camera, no action yet
trying to video turning a fusee (tapered thread)
fscking piece of crap camera driver / video capture utility
and the camera can barely see the threads - its quite low resolution
I need to put my frame grabber in the camera
I want a program that recognizes if my daughter misses the bus
do you have a digital camera that will do video.. I have a cheap digital camera that will do decent video
well, the good news is I have a video
the bad news is its over a gigabyte
gotta increase the feedrates ;)
gotta improve the compression
turning your Shoptask into a HSM would also be more impressive anyway :)
3:26 duration, 640x480 res, 1.7G
compression should definitely take care of that then
its on my windows box, cause I didn't want to fight with drivers
that's closer to the size of a full-length movie on DVD
anybody know something about video post-processing
didn't jepler or someone figure out a magic incantation at Fest last year?
(which might still be in your user dir)
that incantation turned single frames into a movie
this is already a movie, in some dialect of AVI
oh, you need to transcode
You can reduce using divx
also, its on the dose box
how small you want it?
small enough to upload to youtube
I think they have a limit of 100M
shit, all my video stuf ins in the other machine, hang on...
I have 768K upload bandwidth, so 100M will take a about a half hour
uploading the original is out of the question - whatever processing is done needs done here
you pay by the mg or what?
I don't want to spend all night
it takes a day to upload a gig via DSL
and this is closer to 2 gig than 1
it's gonna take longer to transcode it - about 12 to 20 hours
2gb to under 100MB.... heh
I suspect it's not actually compressed at all at the moment
its only 3 minutes and 26 seconds....
640x480x3 = ~1M/frame
and I think it was only doing about 10 frames/second
oh, in that case it's not compressed
jmkasunich: Then whatcha waiting for... get to it! LOL
do you have mencoder?
~200 seconds * 10 or so * 921.6k ~= 1.7G :)
cradek: I think so - I'm pretty sure we used it for the time-lapse
the internet says the camera output is using an SP54 codec (I think)
mencoder in.avi -o out.avi -oac copy -ovc lavc -lavcopts vbitrate=250
what he said
(what did he say?)
now can you tell me how to copy the file to my linux box from the doze lappy?
do you have networking or USB on the windows machine?
Places-> network -> entire network -> windows network -> <name of windows network> -> <name of laptop> -> share :)
On the winbox, right click the fold and select SHARE, then browse from your nix box to that share name
hmmm. work laptop - could be locked down somewhat
(then again, there's VMWare on it, so who knos)
my largest usb stick is 1G, so that won't work
messing with network stuff now
Then again, xp there is Movie Maker that will reduce the size down too
I can fax you a 2G if you like
movie maker is probably the way to go
compress on the doze box, and upload from there
all the click and drool "how to youtube" instructions are aimed at that environment anyway
jmkasunich: If you haven't already done so, install the SP54 Codec on that box first.
In regards to the r8 retrofits, they take out the alignment pin so no orient problem.
JymmmEMC: it must already be there - media player can play the original movie, and movie maker has opened it
I'm at the RTFM stage now
jmkasunich: ok, cool.
jmkasunich: turn off and close any programs... including virus scanner
there's a limit to what I can do
eric_U: you around?
it says it will take a minute or so to compress to 20M
(there must have been next to no compression originally)
it did some wizard thing, broke the movie into clips and did bits and pieces
No, those are where it was paused originally
paused or stopped
no, there were no pauses or stops
I think I got it figured out, resaving now
I ran a "automovie" thing since I didn't know how to use moviemaker
the initial load split the file into 16 shots, even tho there are no transitions or anything in the original movie
the wizard took a few seconds from each block, and did fades, etc, between them
If you drag each of those "thumbnails" to the timeline it'll make the movie
but after seeing how the wizard worked, I figured out the user interface enough to do what you just said
its compressing now
uploading - kinda lame, there is no progress bar
If you have recently uploaded this video, you may need to wait a few minutes for the video to process."
hurry up hurry up....
that means "please wait while our amazing technology degrades your video quality even further!"
I watched the compressed version (not youtubified yet) and it looked pretty good - wound up at 39 meg
it should be interesting to see the difference :)
[04:33:32] <jmkasunich> http://youtube.com/watch?v=ACvRilmIKDQ
you need mist coolant ;)
I'm quite pleased with the roughing
yeah - it's pretty cool
the only optimization I'd want to see (which I don't know about doing in G-code subroutines) would be to skip any moves that move outside the stock radius
maybe some parameter could tell the sub that
I rapid them
they did look faster :)
ah, right, you still execute them just in case a later move ges back inside the diameter
if you know the diameter is monotonically increasing, it would be easy to cut it short
make and install the mating nut
cast it out of epoxy
still uninstallable or extractable
the pitch varies from 40 tpi (the skinny part) to about 22 tpi at the end
nice work... totally useless... but nice programming
its not useless
that was the "transmission" for a rat-trap powered vehicle
ever used it on a real part??
[04:47:53] <jmkasunich> http://jmkasunich.com/cgi-bin/blosxom/shoptask/eng-week-02-21-08.html
click on the 2nd photo
compressor scrolls is all i see it working for.. and the machine would have to haul ass
uh he just showed what it's used for
and clock fusees: http://www.villagetimekeepers.com/clockproject.htm
I admit there aren't many parts that need both variable diameter and variable pitch
plastic extrusion screws are the only thing I can think of, and I bet they're milled, not turned
creep grind + finish grind
also a lot of mixing equipment requires variable pitch/diameter
some extrusion equipment
for other productws
any good tutorials for using opengl and python?
I think lewing is trying to send us a message in morse code
anyone know if there is a quick test for checking a thrust bearing (small, used on the X axis of a lathe)
other than sound and temperature, no
spindle speed max, it should not be too loud
let it run for a bit, and feel the bearing housing
it should be warm, possibly hot, but not to the point of burning you
if it burns you, it's bad
let me give you the history -
trying to fix up an old CNC lathe for use with EMC; the X axis had about 0.010-0.020 of lash; turns out there is play on the screw
i am assuming there is more to it?
so today I finally got around to disassembling it; the mount is: nut, star washer, regular washer (thrust washer?), roller (?) thrust washer, bronze bushing, thrust bearing of the same kind, washing/thrustwasher, a "flange" on the ball screw
this is where i am at - can't find anything wrong so I can either reassemble and hope or ?????
what actually has the backlash
it's something in the screw itself?
like in that stack
the screw "mount"
and not the nut?
man that sucks
not the nut, the nut is tight
yep. don't want to blindly replace the thrust washers
is it a roller bearing?
i'd hold it gently in a vise, put an indicator on one side of the bushing
and grab the ID, and give it a gentle push in and out
and see if your .020 show up
or in, whatever
it depends on what kind of loads the bearing is designed to handle
not sure if that would help as he bearing has a permanent thrust surface on one side
it is not open on bothsides like the ones i have seen before
huh. i don't know what to tell you man
guess I should just reassemble and re-measure... maybe the thrust nut worked itself loose over time (it is 20+ years old and was stored outside for a while
it's worth a shot
but if that doesn't fix it, i don't have any suggestions =(
let me try that
i am going to bed, i wish you luck sir
anybody in here use EMC with a PlasmaCAM system?
anyone alive in here?
Can anybody help me get my limit/home switches working?
It is a lathe, and the x-axis seems to work ok, but the z just errors
The error is "joint 1 on limit switch"
have you tried to ad debounce?
how do you do that?
i dunno, but some one will comeby soon who has a clue :)
I'm reading about it on the net.
can you see the joint 1 limit switch working correctly with halmeter?
Yes, it seems to be working ok, It is yellow when off the switch, and red when it is on.
and have you checked that this is the correct polarity that EMC wants? (e.g. check that the z-limit works the same as your x-limit)
I guess I should add that is errors when homing.
I used stepconf, and set both axis the same.
max + home
i don't know much about homing, esp. usign the limit switch for homing, sorry...
Thats ok, hopefully the homing guru will come along. ;)
StepDi1: it would be helpful if you coudl pastebin your ini file
did you create it yourself? or using stepconf?
do you home using Home All? or do you try to home each axis at a time?
after you homed X, and try to home Z, does X move off the limit switch?
do you have a shared pin for the limits?
I just stepped in alex_joni, what is the question?
Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2008-03-09.txt
BigJohnT: roughly 40 minutes ago StepDi1 asked for help setting up homing/limits on a lathe
ok, there are some bugs with homing in stepconf at the moment
there is a slight issue with stepconf and missing axes (Y in his case)
I used stepconf
good morning alex
BigJohnT: I know.. that's why I wanted the file on pastebin
yep the file on pastebin is the best way
I tried it both ways, home all, and each axis
StepDi1: open an editor (for example gedit), open the file, select all, copy, open a browser, go to http://www.pastebin.ca/
paste, give us the number
ok, hold on.
[13:06:52] <StepDi1> http://www.pastebin.ca/935212
The x-axis seems to be working ok
does the z axis home first?
It stops as soon as it goes active low with "on limit" error
when do you get the error?
StepDil: I just logged on can you repete the problem for me?
while homing X?
Never makes it to x
is z going the wrong way after touching the switch
If I home x by it self ... X works
Does not go anyway stop as soon as active low
just like it ran into the limit
StepDi1: can you pastebin the hal file too?
StepDi1: I notice you have SEARCH_VEL and LATCH_VEL
does it move on the switch, then back off, and on the switch again?
[13:14:36] <StepDi1> http://www.pastebin.ca/935220
No just on
[13:15:00] <alex_joni> http://www.linuxcnc.org/docview/html//config_ini_homing.html
that's how it should move..
btw.. LATCH_VEL should usually be way lower than SEARCH_VEL
It does that on the x, but not the z
that could be the problem
StepDi1: you need to change the hal file
net max-home-z => axis.1.home-sw-in
net max-home-z => axis.1.pos-lim-sw-in
those need to be *.2.*
as it is now, emc2 think it's homing Z, but the error is on axis 1 (which is Y)
by george you have found it
and therefore it trips with an error, as it didn't expect any limits activated on Y
is that a bug in stepconf
I thought about that, but wondered why the rest of the axis used a *.1*
BigJohnT: yes, it's a bug in stepconf
StepDi1: only axis.* needs to have a 2
stepgen is ok with .1
as you only have 2 stepgens
ok, let me give it a whirl
so you need to fix _only_ lines 57 and 58 from that pastebin
It works, must be a bug in stepconf.
does jepler know about this one?
That is the first hand edit I have done
if you run stepconf again you will have to change it back again
yup, any rerun of stepconf will overwrite your *fix*
StepDi1: also, take a moment to read about SEARCH_VEL and LATCH_VEL
SEARCH_VEL is for a rough localisation of the switch (so it doesn't take too much time to find it)
and LATCH_VEL is the one for a precise localisation (slow speed, high precision)
I set LATCH_VEL to half SEARCH_VEL
i have another question. Can i do lathe threading with a 1 pulse per rev encoder on the spindle?
in theory yes
I saw some chatter on the mail list but forgot the specifics
if the results are good enough, that's for you to figure out
StepDi1: if your spindle speed is fairly constant, then it might be enough
that means taking cuts which don't influence the spindle speed much
I could change my trigger to more pulses. I guess the real question is X pulses on 1 input pin.
you need a index pulse
meaning 2 input pin on a parl. port?
that is absolutely necessary (so multiple passes start in the same location)
alex_joni: does jepler know about the homing bug?
I was afraid you were going to say that :(
StepDi1: as I said.. you might get away using only the index pulse
BigJohnT: he might know, but I'll bug him about it
ok, then I won't
Could you send me in a direction on the hal/ini setup?
StepDi1: there are sample configs included with emc2
probably nist-lathe is close
there is also a lathe in sim
lathe.hal is what you want
I'm digging for it now
[13:40:35] <alex_joni> https://sourceforge.net/tracker/index.php?func=detail&aid=1910452&group_id=6744&atid=106744
now you have done it you bugged jepler
fenn_ is now known as fenn
Hello ALL !
I installed the new EMC2.2, WOW ! I am using the outdated 2.05 because i couldn´t reconfigure my machine
can anyone help?
Sorry overtyping! My drives are PHASE-DRIVES
[14:56:04] <SWPadnos> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?UpdatingTo2.1
[14:56:07] <SWPadnos> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?UpdatingTo2.2
./update-config.sh line77: sintax error
md5sum: emc.nml was modified; yuo must update it by hand
Ito_Brazil: Hang in there someone can answer your question, I'm not the one...
jmkasunich: very nice video
The link you have on the youtube page doesn't seem to exist.. http://jmkasunich.com/cgi-bin/blosxom/shoptask/fusee-1.html
[16:25:16] <skunkworks> http://youtube.com/watch?v=ACvRilmIKDQ
it would be neat to see close-up photos of the finished part
those roughing passes do look great
darn, I guess there's no sound
it took multiple passes to thread; how is it not ruining the previous thread pass?
too bad there's no sound
JymmmEMC: the encoder has an index impulse
it always syncs on that before running the pass
it's not very obvious but it sits and waits a bit before the run
cutting threads always takes multiple passes that have to line up
those were tiny threads, what if you were cutting someing like 1/4-20?
I think he said the left side is about 20 tpi and the right is about 40
sorry, I meant some really course thread where you have to remove a lot of material
you should come to cnc workshop... I cut many many 1/4-20 threads there :-)
JymmmEMC: you just use more passes. the idea is the same
JymmmEMC: look at the link I just sent
like acme thread or something like
ok, so multiple passes.
(i do wish that these video paused the end result though)
yeah, a photo would be great
M DV has a Snapshot button, where it takes a still on video for about 5 seconds.
My DV has a Snapshot button, where it takes a still on video for about 5 seconds.
that would be perfect here
I haven't written the blog post that goes with the video yet - I'll put a close-up of the finished part there
jmkasunich: cool beans
did we switch to DST today?
jmkasunich: you should tag your video with text overlay too
cool beans is right!
JymmmEMC: yes that's what you hav eto do if you don't have cnc :-)
I thought that was cnc
well he has to reverse the motor to back up - a real lathe lets you disengage the leadscrew
no, this is had threading
if your cutting metric threads on some lathes you have to back up...
yeah, what a pain
the monarch makes manual thread cutting pretty easy.. You can set stops to turn off the spindle..
why turn it off?
for us slow people..
that are not coordinated ;)
I like the sensible cut depths on jmkasunich's vid
I stomp on the brake and disengage the feed and back out the tool all at once
skunkworks: I've seen a "ramp" that presses the halfnut lever to disengage
mine has a ramp to disengage the feed but not the half nut
jmkasunich: what RPM are you threading at?
cnc is sure nice for lathes.
has anyone thought about really nasty threads like buttress?
I have to thread at 60 with my manual lathe so I can see when to jump up and down on the brake and levers
prior to CNC, I never threaded under power on the shoptask
this has got to be better than a pull tab for a beer can then
I can easily thread at 1000rpm on my sherline
Gawd, this is so fscking weird... not used to actually having daytime weekends off yet.
[16:48:57] <alex_joni> http://www.peters-cnc-ecke.de/forumupload/uploadFiles/2142_122869528792_Mach3-Keyboard_000.png
threading on a manual would be easier if one could rig a good way to back out the x rapidly. ;-)
* BigJohnT has gotta get/build a cnc lathe now
usually on a manual you need a relief cut at the end. with cnc you can just back out reliably
dave I've thought of that a few times
alex_joni: oh god
cradek: some guy trying to build a special keyboard to control a machine
something to back out the X feed knob at a certian spot on the Z travel
[Global Notice] Hi all! I am about to take down services for a tiny while to update it to the latest revision, while it's gone there will be NO services (nickserv, chanserv etc) -- downtime should be, er, hopefully a few minutes only, I will let you know when services are back up! Thank you for using freenode and have a great day!
he says he wants to make plans and interface open
alex_joni: onscreen, or a real keyboard?
* JymmmEMC braces for impact!!!
serial port connected
oh that's not so bad
We're going for a ride kids!
from this photo, I thought he was making a (terrible) gui
for gods sake man strap yourself down!
so far he's deciding on the layout
BigJohnT: you too!
BigJohn .. I've considered a servo motor with stop for x+ and a micro switch limit for the backout
* BigJohnT is going to take a nap now
dave_1 I was thinking a lot simpler method of just having a strap or something to "unscrew" the X as it travels down the Z
One can pretty fancy if doing a lot of threading.
I'm headed for cnc ... it is just not happening quickly
other things keep getting in the way.
I don't do much threading on the lathe
My mazak did something wierd yesterday and I need to check it out.
at this time my priorities are: mazak as primary machine, then get the surface grinder going, the the small vertical mill and finally the lathe
gotta go get something done ... see ya
I'm just trying to get my portable plasma table going
that should keep you out of trouble for awhile
[Global Notice] Hi all! Services upgrade went swimmingly, all should be back to normal! Thanks for the patience and have a great day!
Hi all... has anyone used VariCAD to any extent???
[Global Notice] Hi all! I know, you're all sick of hearing from me today -- we are however experiencing some most peculiar routing issues and I shall attempt to fix them by trying some magic, expect turbulence as I try bring Europe back to us!
[Global Notice] Hi all! Re-routing appears to have fixed the problem, we may however be experiencing a slight bit of lag while the affected side catches up! Again, thank you for your patience and for using freenode.
Hello ALL, i need some help. i´m a 2.05 EMC user and need a help tu try the 2.2 aproach
well, my machine drivers are a kind of diferent, i use PHASE DRIVES
Anyone can show me where is the "new" stepgen ?
I mean, i´m configuring the core_stepper.hal but not sure thats only those files to change on the directory
Hello there! I need some help about stepgen configuration
what do you need?
how is the laser coming along?
I'm getting real close with my plasma cutter
what left ?
table and XY axes
and building the frame
and buying the plasma machine
finish the Z axis
and buying the fume extraction + table
alex must be just starting his
BigJohnT: was talking about yours
I have the table and the xy and the plasma cutter and the fume extraction etc.
alex_joni: you also building something ?
there is no frame as it is portable
he's making fun of me...
Hello dudes! I was using the 2.05 version of EMC2, now i´m trying the 2.2 but my configurations not worked
can anyone help?
stepper or servo?
but I saw that everybody serious on EMC have/will/are making some machine :)
alex made his tripod machine a while ago
that wasn't a serious machine.. but it worked
alex_joni: please help me to solve my upgrade?
most important that is working
* BigJohnT goes back out to my shop for a while
alex_joni: Thanks! but my machine is a little different
Ito_Brazil: did you do EVERYTHING on these pages:
[20:01:45] <jmkasunich> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?UpdatingTo2.1
[20:01:49] <jmkasunich> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?UpdatingTo2.2
I use phasedrives, i cant find the stepgen types on new manual
Ito_Brazil: the most changes are between 2.0.x and 2.1.x
Ito_Brazil: stepgen types haven't changed
the sintaxe changed right?
on stepper.hal its a few different , but OK i will try some more hours
[20:03:33] <jmkasunich> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?UPDATING
that page is linked from BOTH of the previous pages, and shows the config file differences
read section 1.3, stepgen and freqgen changes
Some other (offtopic?) stuff happened
alex_joni: can you check emc_devel ?
i plugged an ISA LPT and torned off the onboard LPT1, on DOS it worked nice, on LINUX, no LPT seen
methinks alex is just jealous
ISA LPT is not plug&play.. right?
alex_joni: weren't you originally going to build a plasma table?
fenn: that sure is a right machine
fenn: not conventional table
I'm still dreaming of a cable tripod for plasma cutting
but big.. 10x10m at least
i think it will swing around when you accelerate
unless you get near-zero jerk
but then it's hard to get the cutting speed constant
fenn: probably a tripod is not viable, and I'll have to do a hexapod
that's what i think
I'm also concearned with cable elongation at those lengths
maybe preload will help with stiffness (you can do preload since it's redundant kinematics)
but then you get no Z control (do you really need Z?)
for torch height control
what's a typical range of movement?
60-100mm should be enough
that's more than i expected
to accomodate material thickness? or just for cutting some distance above the surface?
well.. imagine thick sheets warping
metal sheets are never straight
maybe 50mm might be enough
but that's really tight
"fun" weather here: http://jmkasunich.com/cgi-bin/blosxom/index.html
ok, well it might be easier to make a flying Z axis instead of going full-on hexapod
sounds like some dude ranch
the motor moves
that won't fix the swinging problem
preload fixes the swinging (maybe) at a cost of not being able to loosen the cables and give you some Z movement
you will always have Z, unless you pull the cables infinitely tight
or, you could have a pneumatic cylinder pushing on the moving head (regulated to produce a constant preload force)
like the tricept robot
what part of 10m x 10m cable tripod did you miss? that would be one huge cylinder, and a very tall roof to hold it
a cable tetrahedron with an artificial 'gravity' component
i'm just spewing ideas :)
I can envision a regular cable hexapod, which would require significant height, and a largish moving platform
i guess you could do a hexapod with 3 cables going up and 3 going down, holding a rod vertically (or at whatever angle)
or a pair of tripods, one controlling the bottom of a rod and the other the top
what you just said
the regular hexapod constrains rotation around the tool axis (which doesn't really matter for plasma)
you can't do 3 going down
the dual tripod doesn't, but it also doesn't rely on gravity for preload
because you have a big cutting table there
the motors would be suspended some distance above the cutting surface
alex_joni: the ones going down are the same as the ones you had on your toy tripod
oh, so you'd have a long tool ?
hmm.. that's interesting
it's more like a spar on a crane
or a piece of pipe sticking off the back of the torch head
the bottom ones can be fully tight
that's where the torch attaches
then using the top ones one can even do beveling
if work table is Z=0, top of work is Z=50mm or so, torch head is Z=100, winches at Z=500, and top of the "spar" would be Z=1500 or so
all winches at Z=500?
or maybe 3 at 500, and other 3 at 2000
if all are at 500, you aren't relying on gravity
i think someone will have to build one to find out what's wrong with the idea :)
another trick that would make the kins a little less critical would be to spring load the spar, so cable tension is nearly constant even if you have small errors in the kins
i'm wondering where the preload comes from, since you have six degrees of freedom and six actuators
rotation around the spar is uncontrolled
so you are really only controlling 5 DOF with six actuators, and the sixth gives you preload
(this assumes that the top three cables join at a single point, and the bottom three cables join at a single point - the tool can rotate around the line between those points
if you had two discs on the ends of the spar, and they were out of phase by 60 degrees, you could control the rotation (i think)
I don't think so
I don't see how
it would interact with Z
you can only control tilt
if you tighten the bottom three cables and loosen the top three, the tool would turn to align the bottom disk with the winches, but it would also move up
if the cables weren't joining to the spar radially (like bike spokes)
if you loosen the bottom and tighten the top, it would turn to align the top disk with the winches, but it would also go down
(then there are those new fancy radial spokes)
I'd use a disk on the bottom, simply to keep it from rotating around the tool due to things like cable/hose drag
there is no need to _control_ that rotation, just prevent it
if you can mount the cutting head centered
right - head shoots thru center of disk
height sensors, etc, around the disk
the spar could actually be a tripod of lightweight tubing, coming up from the edges of the disk where the lower cables are attached
that would free up the center of the disk for the torch
right, or a triangular prism
the three tubes would meet at the top, to form a "point" where the three top cables would attach
light but rigid
hmm, no not a prism, an octahedron
a triangular pyramid
I know.. let's put the cutting head consumables in the weight center of the pyrammid
then they get influenced by the pyramid power
and you don't have to change them so often :P
* alex_joni can't help it sometimes
and pyramids will sharpen endmills too :)
fenn: something to consider for your hexapod
[20:44:40] <fenn> http://www.instructables.com/id/EEUFGGNF5Y3YROZ/
fenn: mythbusters did something on that.. I think it was busted.. Imagine that..
* OoBIGeye is surprised!!! :O)
[21:11:47] <Vq^> http://xkcd.org/373/
plus or minus some bs
i should put that on a print one day
"5.125 +/- some crap"
welcome mr White
toastydeath, I know someone who puts +-0
are they like, employed
or is this just a hobby
no tolerance, zilch eff all
he actually means as close as you can get
hey. I was wondering what the current practice is for integrating pyvcp into AXIS is still as outlined on http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?PyVCP
there is actually a standard for that
used on old school jig borers.
hmm he is old
i believe it's orthoganal dimensioning, everything comes from one 0,0
I was wondering if current practice for integrating pyvcp into AXIS is still as described in http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?PyVCP
and there are no tolerances on the print
jepler: any comments on that?
owhite: it should be about correctly described on that page
i have one or two books where the practice is described
hm. there's another method described on http://linuxcnc.org/docs/2.2/html/hal_vcp.html
how are they different?
oh, don't read the wiki text labelled 1jan2007, that's historical by now
you should just make your pyvcp xml file, then a HAL file that hooks stuff up, and then refer to these in the correct way in your ini file
I like that appraoch more :-)
the manual "1.4 Using pyVCP with AXIS" should be a pretty straightforward description of how it's done. A recent version of HAL might suggest you use 'net' instead of linksp in the HAL file
If you make an ineresting pyvcp panel do post it to the wiki or a link to the mailinglist or something!
yeah I've made a new circuit board that hangs off of a M5i20 card. It receives data from a laser, and also sends PWM signals to same from the M5i20 card. EMC/pyvcp is sending PWM controls to that board. There were some interesting tricks, at least to me, to make that happen.
I was planning on documenting it all and seeing if i could put it up on the wiki.
what are you doing with the laser?
sounds cool, definitely worth documenting in the wiki or if you have your own site
just as soon as I get my pyvcp panel into axis . :-)
note that AXIS will probably re-name your pyvcp pins like so: pyvcp.mybutton so you might need to change your HAL file too
that's something I thing jepler/cradek could reasonably be asked to change in AXIS
yeah. the pyvcp panel is showing in axis but the buttons dont appear to be connected to anything. I'm doing some 'halcmd show stuff'
did you have the pyvcp all hooked up when it ran stand-alone?
if you put the same HAL file you used into postgui_hal= then the pin-names might be wrong due to the reason I mentioned above
HAL/EMC should complain about missing pin names at startup
not getting any complaints at start up.
not getting any complaints at start up
oh sorry. xchat screwed up.
check with halmeter or halcmd show what the pyvcp pins are called now that they are integrated into AXIS
On this page, click on FAQ, then on DEVICE, and read item #2 http://www.wireless.att.com/cell-phone-service/tutorials/deviceTutorial.jsp?q_prodId=prod1050430
hm. looks like my .hal file that is passed to POSTGUI_HALFILE of my .ini does not want to load_rt hal_m5i20
it's a bit late to loadrt hal_m5i20 in the postgui
so I should split out the call to hal_m5i20 from the .hal stuff that handles gui-related pins, me thinks.
I bet you already have it there
was that a suggestion?
yeah, reread the hal files :P
there is one called motion something
in there there's the loadrt hal_m5i20
you only need that once
(that's why the call to loadrt hal_m5i20 in m5i20_io.hal is commented out)
okay let me poke around. I dont use a m5i20_motion.hal. but I can take a look at that example.
I think I can load it pre-GUI, then do the addf m5i20.0.digital-out-write servo-thread -1 in the post-GUI?
mib_0ft8l9sg is now known as [ALS]
cradek: did you get my message yesterday about the diesel pump?
[ALS] is now known as ALS
good night all
jmkasunich: you're strictly using the printer port for everything right now correct? reading encoder back in.. What count encoder is it?
ALS: no I guess I didn't - do you mean irc or email?
sorry I looked back and still don't spot it
ALS: still no luck. what did you say?
cradek: deisel? ck the delivery valve is not stuck open on the pump, after sitting lease get gui and get stuck open and piston will pull the fuel back down and not reload
sometimes you can take a small alluminum rod and with a small hammer tap on the top of the valve seat through the hole under the line coming out of the pump, you ck and see if its stuck up by measuring each hight of the valve
so poke a rod in the output hole of the injector pump and see if it's stuck up or down?
measuring each hight of the valve
ok interesting, I'll try carefully doing that
if you tap on it and it opens you will here the spring pop it shut
they are pushed up but have a spring to pull them back against the cam?
(seems like a very small hole)
I might have brass that size
the fuel pushes the valve open and closes
ok maybe the top block comes off?
don't think so...
try and measure the high with a small pin
clearances are in the microns, dont let any grit in
who makes the pump
not sure, says yanmar on it
thats is what the problem sounds like to me though stuck open valve
the top of the valve may unscrew but usally special sockets required
I have had the element stick
bbl, thanks ALS
pump has a number of parts inner piston and outer element thats rotated in some types to set cutoff
mib_ot78xcf1 is now known as ALS
fuel meter helix
usally is control by a rack
cradek: look around the net for some technical drawings of the injection pump