jepler: hey Jeff you about?
lerman_ is now known as lerman
dallur's thc in gEDA (unfinished) http://imagebin.ca/view/c2oCnG1.html
[08:27:55] <renesis> http://www.darkertechnologies.com/image/broken_taps_removed.jpg
i blame china.
china is todays designated victim - all blame can be dumpted on china - tomorrow spin the globe and pick a new protagonist. (like mexico, or china again)
edm's are a wonderful friend at this stage
yes - I should finish building mine, I have most all the components.
a well sized tube electrode takes out the core of the tap and you collapse the teeth in
the you can still use the part after finishing the tapping
I once broke 2 taps in the same hole - EDM was my friend - was a bad batch of O1 Drill rod.
I'm still looking for a friend like that
you seen Ben Flemings book?
and if you dont have an edm if the hole is of suitable size try dipping a welding electrode in water and it acts a bit like a thermic lance , but messy
just a bit
is that the old edm book
or something else?
last 3 years or so
has simple plans for a resitive/capacitence unit - costs about $150 to build - and is servo driven.
does that give any nice surface if you say cut a edge with it?
I bought the book, and have most the components
has a roughing and finishing setting
(dosent care for cut/edm
hm.. how well does it work/look?
better then milled?
now your comparing apples and watermelons
no seriously, I have no idea.. if a homebrew edm would be better..
I saw a picture of something edm:ed someone made with a homebrew somewhere it looked like ^ ^^^^^^^
one is green on the outside and red on the inside... the other is kinda the other way around
í lost. :p
use EDM for things to hard or dificult to machine
like to make a square drive socket
yeah.. but how does the surface turn out after you edm it?
drill the hole but use the EDM to give you the sharp square corners
looks like a really fine glass bead finish
unless you run it on rough the whole time
then its like a course sandblast finish
this unit is small so it does not have the amps to really rip apart the surface like extra heavy 3 phase units could
so it keeps you from over burning to a point
there was a good review of this unit... let me seach...
ah - found one http://www.modelenginenews.org/meng/edm/index.html
here is a pic of the whole powersuplly / servo controller in one box - from that article http://www.modelenginenews.org/meng/edm/images/edm_wiring2.jpg
big in size - small in value
they are polyproplene caps
only like 20mfd
those are used for the finishing
I am hoping at some future date to be able to use EMC with adaptive feed to to sinker EDM / and orbiting functions
such as edm'ing a thread in a full hard tool steel part.
I do this now with a commercial AGIE machine - but would love to be able to replicate the function on a homebuilt unit.
* anonimasu nods
I passed on a agie a while ago
not the most recent one..
but I guess the powersupply should have been good anyway
good machines - waste wire and or require precise settings
the sinker I use must be a good 15years old
well, im getting one someday
The Flemming R/C unit is good for small work - but soon he will have his 10amp pulsed unit out in a new book.
I want to use the Seig Super X1 base for the X/Y movements and build the tank as part of the table.
then use the regular fleming servo head for Z - or a dual Z part programable and part auto servo.
controlling depth is a paradox since eletrode erosion will play havoc with true depth measurements
how do they do it in the industry?
trial and error?
this is where cnc probing cycle might be used
I guess you can use the information from where you start sparking..
and calculate the increments you need to make over time
move to a know location and bring electrode down until you measure continuity - this resets your eletrode z offset - move back to hole location and burn away.
nice thing about the agie was it had built in lookup tables that made a good guess of what the electrode wear would be
so it would burn to a target depth and need only one reset to make proper depth.
seems like a major pain
edm is science wrapped in voodoo
how you set up you flushing plays a huge factor in electrode wear and finish
I never expect it to be simple - its a russian invention...
( not that that matters any)
major pain - yes both laser and edm are pains
but both allow you to focus power in very precise ways without being very efficient, but allowing you to do things where there are few other options.
I find milling a pain too
Milling is my favorite
I love it though..
but it's a pain feed/speeds
24,000rpm - coolant spy - zoom zoom...
and depth of cut..
bah.. I dont have that kind of machine
if I did I wouldnt be idling at irc ^_^
6x speeder head on a Bridgeport chassis
still goes zooom zoom
im still looking for a speeder head
though my next spindle will be at 8krpm :)
if the toolpost grinder works..
Ours was made by a division of Lovejoy in the UK
runs all day at 24000
rated @ 30k
but machine spindle tops out at 4K
so 24K is all we get
ER11 collet nose
speeders are not beefy for side loads
I want 5mm with a 20mm endmill at 10krpm in aluminium ;)
still - with lighter chiploads they can still remove lots of metal fast
try a 12mm dia @ 6mm depth
I dont think that'll work out
10 mm dia at 6mm depth
what kind of speed?
small tools - high feedsa
at 2krpm that's 500mm/min with the 20mm endmill..
small tools are much cheaper anyway
I use 8mm alot
I'm kind of scared of dormers feedrate recomendations :p
they are twice what I expect all the time
and 3/8" is the best value
if I think 200mm/min they say 450
true - but at what depth
and if I think "better safe then sorry they think FEED!"
more depth then I feel comfortable running
like 50% of dia..
I seen where they want a .5mm chipload per tooth at only 1mm depth
is that sane?
I usually think of it as "they have more clue then I'll ever get" :p
and start 25% off where they want it and use the feedrate override to get there..
uses fast/shallow moves to remove more material in Z level roughing operations
I'm going to put on a german copy of Andromeda and fall asleep with my brandy... nite all
I did a cvs update and then ran 'make', but I still get EMC2 - pre-2.2 CVS HEAD. shouln't it be pre2.3?
where did you run cvs up ?
was it in the topdir?
hummm, maybe it was in /src
the VERSION file is in the topdir
right, I'm running make now, do I need to run make again after I do cvs up in the topdir
yes, I think so
actually it's configure which gets the version from that file
hopefully not 'make clean' though...
not for VERSION..
now it's linking, so there is hope..
but it's wise to make clean if it's been a really long time since your last update :)
just abour 7 days
then it's safe
I was thinking more like a couple monthgs
alex_joni: do you think there is space in cvs and in the standard emc for a specialized m5i20 driver?
cvs is already 30MB how big could your driver be?
about the size of the current m5i20 driver, but specialized for working with the pico-systems amps
yep, now I'm running pre-2.3!
awallin_emc: worst case you can stick it in a branch
(you can branch only one folder.. )
did you see my comment on your blog :)
yep, I do need to moderate them before they come online ;)
r u guys running 2.2 yet
It's nice if you can look at the opengl stuff, I have lot's of other things going on but lately I've been reading about voronoi diagrams and offset generation
* anonimasu nods
I think we should leave any cad parts for later
and work soley on making toolpaths.. for a start
and get some python/octave/random bindings
i like the 'emergent spiral
is there a way to do that with arbitrary outlines?
what is the voronoi diagram used for in doing offsets?
is it just for finding the center of arcs?
i dont have access to online journals so m. held's page is worthless to me
fenn: send me your email address and I can spam you with all the pdf you can read ;) anders dot wallin at helsinki dot fi
but it'll be later in the evening
hmm. scripts/realtime stop gives me error messages eventhough I've done halcmd unloadrt all ??
sometimes hal modules don't actually unload when you tell them to unload
halrun -U gets 'em though
etla@etla-desktop:~/emc2.dev$ scripts/halrun -U
ERROR: Module hal_lib is in use by pid,threads
damn. need to reboot.
hi Martzis, what's going on?
I have not had much time for cnc
oh, too bad...
I currently re-routing the cables of my mill
Machine configuration directory is '/home/etla/emc2.dev/configs/sim'
Machine configuration file is 'servo_sim.ini'
(time=1194790284.262529,pid=5405): Registering server on TCP port 5005.
what does servo-sim have to do with TCP??
awallin_emc: the NML channels are all owned by emcserv
it listens on TCP for connections..
oh, anyway that crashes later on
servo_sim.hal:223: pin 'axis.0.neg-lim-sw-in' was already linked
Shutting down and cleaning up EMC2...
I have been designing EMC and Mach compatible HID input device
[14:16:31] <Martzis> http://martzis.wippiespace.com/images/image/USB%20HID%20input%20device/64button_8-axis.jpg
the demo config has a slight problem, it refers only to axis 0 around that line...
[14:16:39] <Martzis> http://martzis.wippiespace.com/images/image/USB%20HID%20input%20device/inputkortti_th_ttl.jpg
The current design supports 64 buttons, 4 potentiometers and 4 mpgs
it should be 1,2, etc
alex_joni: there's also wcomp.9 which should be wcomp.8 I now got it running.
seems there's been a bit of a copy/paste errors
Martzis: anything in particular you want to do with the HID? or just normal things like jogwheels, buttons etc
Just for normal buttons and wheels
I thought that this would allow easy modernization for some old UI panel
I think it could be useful for something like this:
[14:30:36] <Martzis> http://i215.photobucket.com/albums/cc21/jonne_photo/HPIM0166.jpg
Current design is input only, it does not support output for ligths.
what kind of update rates do you get over USB to the PIC?
I have not calculated exact numbers...
Enough for user interface
But not enough for PID loop
for E-stop and other potentially dangerous stuff I would not rely on USB...
I would not either
I would like to know if EMC can control HID LED items. Usage Page 0508 (LEDs) contains definition for them. Those seem to be also used in usb keyboards (numlock etc.).
for halscope, is there a 'hard' limit to 16k samples? I do have RAM to spare if that's it...
Martzis: it depends on the USB driver you need to write for EMC. I don't think anyone uses anything with USB right now
or you could look at how Ubuntu responds to USB/HID devices
alex_joni:i am trying to make a led bit with pyvcp and hook it to the lube level,and read the input on pin15 on parport
i have the led in axis already
jlmjvm: so what's the problem?
i dont know how to link it to hal
can you see the correct behavior with halmeter (look at the parport pin)
lemmee put a breakout board on here so i can have a switch hooked up to simulate it
wont i have to link iocontrol to a parport pin before i can see anything in halshow?
if you have emc running with a parport config I would assume the parport driver is running and you can see all the parport pins as HAL pins
k,i can see the pins,and have a switch hooked up to pin 15,and can see it change states
now you want to hook the parport input to the LED so that the LED shows the status of the pin?
anonimasu: you still there?
jlmjvm: the line of HAL code you want is probably something like this: ' net lube-signal parport.input.pin pyvcp.led.indicator'
with the correct pin names
k,working on that now
and you want to put that in a textfile myfile.hal and have that file when AXIS has generated your panel. do that by refering to the file on a HAL_POSTGUI line in your ini
hope that made sense :)
yes,read about that earlier,just couldnt find any example of the lube level being used
you can try it just by firing up emc and typing it on the commandline: 'halcmd net lube-sig parport.pin pyvcp.led'
the textfile and modifying the ini is just if you always want to have it do that automatically
jlmjvm: any luck?
still plugging away
should it be lube-sig,or lube_level? iocontrol shows lube_level
the first word after "net" is the name of the signal, not the name of a pin
you can name it frodo if you want
well.. frodo failed http://www.iist.unu.edu/~antonio/Home/fun/frodohasfailed.jpg
awallin: im here
anonimasu: just wanted to remind you that you should put anything you work on in svn - so I don't spend time working on the same issues as you
jmkasunich__ is now known as jmkasunich
i can get docs on mux2 by ' man mux2'. there's a thing created by motmod named axis.N.blah. how do i get man for axis? or what alternative docs?
tomp2: 'man 9 motion' in emc 2.2
not sure why it's 'motion' and not 'motmod'...
making gEDA schma, and didnt have that component
alex_joni:i am having trouble trying to link a switch to a led in pyvcp,cant get them connected
my 2.2 is run-in-place so the man pages 'are not there', but mc (midnight commander) displays the man format nicely :)
tomp2: you can also use . scripts/emc-environment
then you can use man
oooh, thx, i wasnt running emc now, good idea
jlmjvm: if you can be more specific, I might be of help
i have a switch on pin 15,can see it work with hal show config,and have a led in axis,want to link them together
just want the led to light up when i press the switch
halcmd net mysignal parport.pin pyvcp.led
did you try that?
[19:14:16] <jlmjvm> http://pastebin.ca/769897
net signal <= parport.0.pin-15-in-not pyvcp.led
find the right name for the led
added that to the hal file
what is it called when you look with halmeter or halshow
joe.hal:43: pin 'pyvcp.led.0' does not exist
thats what it shows\
and this is called from the HAL_POSTGUI=joe.hal line in your ini file?
been having no luck
no,its just in joe.hal,not post gui
comment out the line in the hal file that causes problems. start emc, look with halmeter or halshow what the pins are called
is what it shows
jlmjvm: how did you call the hal file from the ini?
good, now that pin is not created before AXIS runs and creates the pyvcp panel. that is why you can't hook up signals to it from the normal hal file
hooking up things to the pyvcp panel needs to happen from the HAL_POSTGUI= line in your ini, not from another hal file
alex_joni:HALFILE = joe.hal
HALFILE = custom.hal
POSTGUI_HALFILE = custom_postgui.hal
,is what i have
ok, the HALFILE = joe.hal is wrong
you want the stuff to happen inside the POSTGUI_HALFILE
so you will need to take the "net ... " and move it to custom_postgui.hal
hey its working now
thanks for the help guys
this will work perfect as a lube sensor
jlmjvm: this will work perfect as a lube sensor - My squirt can is wireless...
Dallur: fyi, i began gEDA version of your THC odg schema... http://imagebin.ca/view/c2oCnG1.html
because I'll use servo, not stepper, and wanted to drag stuff around, & get 'net lists'
tomp: can you get some useful output from this also, or is it just visualization?
more recent http://imagebin.org/11668
no useful output yet, hope to get draggable, which lets me 'see better' now, and netlists for auto-.hal generation in the sci-fi future :)
tomp2: great, i'm on the road atm, I will take a better look when I get home
I'll wait for sci-fi then...
del.icio.us plugin seems to have broken firefox a little
fixed in new version
i did a new config with stepconf and my external estop isnt working,is there something that has to be changed like in 2.1.7
do you have to comment out anything in the hal file
ext-estop is working in hal show,but the gui doesnt respond
I'm still stuck trying to figure out what my estop should do
power your drives off
and stop the spindle
rather disengage the latch that powers your drives ^_^
guess I have to figure out how to dump the dc bus on the drives
big honking relay
relay the power supply
or just stop the signal going to the drivers
drives are self-contained
they keep going for a while after you stop the ac line
we usually have big brakes on the servos
when someone hits estop it first engages mechanical brakes
then dumps excess voltage on the bus into breaking resistors
that's out of my price range
well.. a motor is usually 3-5k$
mine were originally that much
bud due to the magic of ebay....
yeah, that's always nice
unless you need to replace one with the same model
my impression is that almost all servo brakes are for holding, not stopping
well.. I can assure you on our robots they are for stopping
robots are different
I have some of those
but I don't want to use dc servos
these are AC nowadays
mine are dc
alex_joni: is there something you have to change in the standard hal file to make ext estop button work?
jlmjvm: yes, but it is described there
brakes are for holding the axis or spindle in place. they're usually rated as having the same torwue as the motor itself
so if you hit the brake, and you power the motor in reverse, you'll decelerate twice as fast or thereabouts
I think with an AC servo, dumping the dc bus will have a pretty good stopping effect
eric_1: I don't really think so
no, I'd imagine it's the drive that will be the main determinant of stopping speed
they used to do that for DC motors
but for AC (revolving field) you need to keep the field in synch in order to obtain a breaking effect
with a DC servo, you stick a resistor across the leads and that's a brake - not quite as simple with an AC servo
what drivesdo you have?
ok, permanent magnet synchronous motor
do I have to type that every time?
or can I call them AC servo like everyone else does :)
I have DDM-019
no, steppers (synchro motors) are the same as AC servos in that you have to move the drive field in synch ( or a little slower) with the motor to get good braking
ok, I don't know those
I have some Yaskawa AC motors/drives that have settings for braking current
very similar drives
the setpoint defaults to 8x current for braking, and it can go up to 10x (1000%)
but if I throw a relay, the drive doesn't get to play
that's the problem
if you want the motors to be actively stopped, you either need to leave the drive powered, or you need a brake
at least, that's the limit of my meager knowledge on the subect ;)
so what happens on a run-away motor?
synchronous motor probably doesn't have that problem
meager... (good joke)
darned good question. that's left as an exercise for the integrator ;)
eric_1: I didn't mean the motor went haywire.. maybe the servo did
one reason why I don't want to use cd motors
some internal thingie craps out and gets stuck to full power forward
maybe emc's output..
emc could be a problem
PID's have a tendency to turn full power if feedback is broken
ferror should cover that.. but still
probably want to dump the dc bus and take my chances
it will put a big load on the motor and stop things, eventually
the runaway danger is more of a problem at standstill, when PID may still be correcting for minor errors/integrator windup/whatever, but the motion controller isn't expecting any motion
if you never use the full power of the motor you could calc the max reversing load and put a crowbar circuit on each amp...
well, if you short some 3-pole resistor to the motor leads, that should still brake
that's nine poles of big contactor
or N 3-pole contactors
for each axis
I like the heidenhain in that respect..
yeah, and still you need some big care
it has position monitoring..
but I think it's not recommended to put contactors between the drive and the motor
lots of drives tend to blow up if you remove the motor from them
that's why I am going to put it on the bus
alex,where is that described?
i cant find it
I like the idea of using a 500W halogen floodlamp bulb as a loaddump/breaking resistor.
I've seen it
hm, well, they can break..
couldn't tell you where
I wonder if I'm going to need external braking resistors for general use
table weighs a lot
eric_1: inertia's a bitch
jlmjvm: you need to decide what you want to do with estop
it's usually what kills you after the chuck has stopped..
that was the chuck key
like workpieces comming lose..
jlmjvm: usually you have an external estop circuitry
which you connect to emc
jlmjvm: the standard_pinout.hal has this:
# create a signal for the estop loopback
net estop-loop iocontrol.0.user-enable-out iocontrol.0.emc-enable-in
you need to comment that line and figure out how to connect iocontro.0.user-enable-out and iocontrol.0.emc-enable-in
user-enable-out is an output (going from emc to the outside of the PC, e.g. a LED or bulb)
hmmm. somebody needs to come up with truly smart tab-completion
emc-enable-in is an input for emc2, comes from the parport, and tells emc that it's safe to come out of estop
SWPadnos: what for?
smart enough to know what I'm thinking about, and only complete terms that are related to that ;)
to reduce typos :)
On synchronous AC drives, DC bus dump will brake the motors, the IGBT bridge being a three phase bridge rectifier when off...
eric_1: listen to PeterW, he knows what he's taling about :D
SWPadnos: my zsh completed configure parameters ;) that's samrt
anonimasu: bash does it too on ubuntu
hmmm. I wonder what needs to be done to get halcmd completion to work at the bash command line
alex_joni: this was on slackware 4 years ago
or maybe 5
I know some programs (scripts anyway) work
Windows Vista should have DASH ( Died Again shell )
Damaged Anti-functional System Hell
what's this vista you are talking about?
that's what you see when you look past the Windows :)
in a world without walls and fences ..
Who needs Gate?
I have that T-shirt
not too sure - but rumor has it that it all started somewhere in New Mexico...
from the first Linux expo I think
"In a world without fenced, Who needs gates?"
we know what you mean
what did he mean?
I only smoke meat
* alex_joni goes back to his book
trolls & gnomes are mentally more stable than this :D
it's bad when the channel is so boring that the loggers leave the room
I smoke tires (when I feel the local Law Enfarcement is not around).
got red ones?
CNC building has been put on hold - I have to rebuild my roof - structural damage to the tune of about $15K.
I just stop in to try and stay current.
that's no fun
net estop-ext => iocontrol.0.emc-enable-in
yeah - I have a bunch of geckos and some servos - need encoders and ballscrews. But I don't really want to use the geckos
i added that and it worked
commented out the original 2 lines
jlmjvm, that's half the puzzle (the easy half :) )
jlmjvm: your stompin all your bugs today.
what do you mean
if you want EMC to be able to stop the machine, you also need to connect the stop output of EMC to the estop hardware chain
you mean that wont stop the software?
jlmjvm: you miss the part connecting that to parport
net estop-ext <= parport.7.pin-31-in-not-always
i have a button hooked to the parport
SWPadnos: you ever use SNMP?
JymmmEMC: I only used snmpwalk a couple of times
and snmp-based graphs
i have 2 lines in my hal file ,net estop-ext <= parport.0.pin-10-in and net estop-ext => iocontrol.0.emc-enable-in
alex_joni: any recommended reading? I can only snmpwalk from console, not remotely
JymmmEMC: sorry.. gotta pass there
and it appears to be working properly
alex_joni: I understand =)
JymmmEMC: I mean.. lots of recommended reading
is that correct
but none is snmp-related
jlmjvm: if it does what you want it to do.. then yeah :D
alex_joni: No, I don't need to read your collection of Playboy's for the articles ;)
JymmmEMC: heh, got your own collection?
i press the ext button and it presses the gui estop and stops the software
alex_joni: Nah, sold em all =)
oh, that's why prices dropped on ebay lately
i figured thats what it was supposed to do
JymmmEMC: actually I'm enjoying some terry pratchett again :)
ah comic books
he released a couple as comic books aswell, but I don't have any
[21:24:03] <JymmmEMC> http://en.wikipedia.org/wiki/Discworld
four elephants on a giant tutrle sounds like a comic too me =)
"satirical parallels with current cultural, technological and scientific issues."
sounds like a horror book to me from that desciption
it's fun to read
you could get one cheap from ebay
alex_joni: it's all good =)
alex_joni: The last fiction I read was koonz
and that was a few years ago
this isn't so much about the fiction part
it's just fun to read
alex_joni: I consider myself a slow reader, don't enjoy it too much.
Though I have read an 800pg tech book in week
surprised the shit out of me too =)
I usually read about 3-400pg/day if I'm in the mood
better you than me
I can't get past ch1 of my CCNA book,
read it 3 times so far
VERY dry reading, and you have to know the cisco way to pass the exam
lerman_ is now known as lerman
alex_joni: Murhpy's law says it'll be one simple paragraph that I miss and fail.
will there ever be a version emc for ubuntu 7.10
we'll probably skip straight to Hardy Heron, version 8.04
got a new laptop,6.06 wont install,but 7.10 will
there are experimental packages for 7.10, but they're just that - experimental
JymmmEMC, no, not really except for messing with it a little when it first got popular
really just wanted to be able to run the simulator
jlmjvm: try 7.04?
oh, for simulator, it's a lot easier
but I don't know the specifics of how to install simulator, since I've never done it
may try 7.04
<SWPadnos> the runaway danger is more of a problem at standstill, when PID may still be correcting for minor errors/integrator windup/whatever, but the motion controller isn't expecting any motion.
jlmjvm: Did you MD5 the iso before you burned it?
The broken encoder runaway problem is fairly easy to fix by monitoring the PID output, if its full scale for more than N samples, (say N= 25)
you are running open loop, engineer him dead, time to shut her down!
I don't think that's any better in terms of having pre-made packages
PeterW, sure, there are techniques to get around it
or if you have output at some multiple of "requested torque" for too long, etc
the iso is good,it trys to install,but hangs up on x because of the onboard video
does it boot but not install, or not boot?
ie, do you get a desktop?
jlmjvm: did you burn the iso at 4x ?
JymmmEMC, it's probably not the ISO, new hardware can be problematic
ive loaded 2 more machines with the iso since i tried the laptop
but the 7.10 iso loads all the way up
newer drivers im sure
SWPadnos: I understand, just seems the more simplistic things are the most annotying typically =)
ok, so it's not the install that fails, it's actually booting the CD that fails
there are two things to try in that case. one is to explicitly set the screen resolution (sometimes the system works but you can't see it because the monitor wasn't detected), the second is to boot into "safe graphics mode"
which I believe uses the VESA driver, which "should" work with just about any video hardware
hmmm,may try it again with the vesa driver
however, you're probably better off installing a later Ubuntu, since you'll also want everything else to work, like BlueTooth, WiFi, etc.
and figuring out how to install the simulator ;)
wish i had never bought it,should have got a kurt vise instead
the wrost thing would be to install 7.10, and then get VMWare Player and do a 6.06 install into a virtual lmachine
or a power drawbar
ok, that's getting into the realm of a nice laptop now :)
or you were talking about an 8" vise or something
musta been a cheap laptop then
btw,my killer latency came back
err - bummer!
hmmm. do you suppose it happened after some system update?
would a pentium system be better for this than amd
ive change the vid card and ram,no change
well, you said the latency was down when you first reinstalled. I wonder if some update made it bad again
what about a pci parport card not being used
i had the updates before i ever checked it
and have another box here with all updates that doesnt do it
can a motherboard do it?
sounds like it's the hardware then
the chipset, specific configurations of the chipset, CPU, filesystem type, etc
then there's video, USB, ...
this is a nforce chipset,wanna change to via
that's not the direction I'd go
u like nforce better
via tends to suck for many things
yep, I do at the moment
i ahve an identical motherboard and cpu that i can try
well, let's hope you get identical results
or not, depending ;)
well, having the same hardware act differently isn't a great thing ...
the msi motherboard with the via chipset has no issues
should i even be able to run the axis program if my latency is 135000
axis doesn't have anything to do with latency
AXIS doesn't care about latency. it's the realtime part that cares
i figured it would jam up
the RT part may, which will lock up your machine
but AXIS is just another userspace program, it has nothing to do with RT latency
im gonna plug this hard drive into an identical box and see what the latency is,didnt think of that
that's agood plan
in fact, you should be able to plug it into any box that has a similar video card. the chipset drivers are all included
that will rule out the remaining hardware with the least work
yep,ive done that
if you move the video card and hard drive to the machine that works OK, you may get some good info (like "this video card sucks")
whats a good card?
any one that works ;)
there have been successes with nvidia cards, using the nv driver (that's what you have, I think)
old matrox cards are good
most people have problems with onboard video, but there are some that work OK (don't remember which at the moment)
would be nice if any video would work
some do, some don't. it depends on a lot of factors
well im outta here,gonna check this in another machine
thanks for the info
is o-word variable label gonna change anytime soon?
the # thing is confusing
there are no plans to change it that I know of
cuz you nest and # start becoming other # and it all starts getting random =\
heh, like im writing the code out c style on paper so i can translate to funky # labels
like G1 X[#3 + sin[#4+#5]/#6]
cuz i cant just be like o001 if [ y GE ymax - dia_offset - edge_finish]
with 2.2, you can use the alternate confusing form: #<named_var>
oh wait i can do that?
but you need the # and the <>
wait i really need the <>
nyah nyah! :P
i was totally about to like punch my crt
oh, go ahead, as long as you aren't wearing brass knuckles
so params are prob still passed to the sub as #n
so i need to be like #<real_variable_name> = #n
okay thats not so bad
i think you need to do [#<foo>]
shouldn't need to unless you want #<foo>+#<bar>
Any plans to map the G33 rigid tap function to the normal G84 Fixed cycle?
what does that mean?
oh - now I get it
not that I know of
Oh well - there were so many variants of tapping cycles a few more won't hurt
Yasnaq used IPR (Incrament per revolution, Fanuc/Haas uses FPM (feed per minute) Others used an unsigned Z value since all tapping was assumed to be "down".
good thing there's a standard
well strangely enough all used G84 - for R-hand tapping - its just the meaning of the word addresses and how they are applied that was different
so in other words, you can't port code between the machines, even though it looks like you can ...
well G84 Z-.5 R.1 F.05 on a Yaznaq = G84 Z-.5 R.1 F50. on a Fanuc with a 1/4-20 tap @ S1000
so yeah mistakes can be exciting
F.05 != F50
one is Inch per revolution - other is Inch per minute.
SkullWorks-PGAB: none of those are actually rigid tap are they?
worse yet - more recent versions of these controls allows you to swap modes - so you could use either on the same machine, if you remembered to set the proper Gcodes in advance
someone (?) told me that G84 is not rigid tap, it's for floating holders
[23:14:02] <renesis> http://pastebin.ca/770169
this should work?
a lot of people call that rigid tapping
SWPadnos: oh let's not go there again
I forgot about that trainwreck
no, just pointing out the terminology is "fungible"
S1000 M29 - sets rigid tap mode for a max spped of 1000 rpm
then a normal G84 line follows
without the machine maker rigid tap M code ( M29 is fanuc default ) G84 is used for floating holders
on Haas a machine parameter sets all tapping to rigid or not
ha our instructor would kill us if we did tapping cycles on the haas minimill
the reverse spindle thing scares him or something
I use 3 VF0's
one is an old "Tupperware" machine - all rigid tap
is that minimill a 40 taper?
heh, it's funny how things are relative
in the hobby world, a mini-mill is the next step down from a mill-drill - maybe 150-200 lbs
and a micromill is a sherline or tiag, pick it up with one hand
never heard of a minimill with a 40 taper (or even a 30)
I know Hass has a small mill on casters that is sized so you can fit it through a doorway.
I think it uses some mini 20 taper toolholders
but it does have a 10 or 12 tool ATC on it
so it goes something like this from large to small
VMC, haas style "mini-mill", bport sized knee mill (2000-2500 lbs), mill-drill (600-1000 lbs), mini-mill (150-200 lbs) and micro-mill (30-50 lbs)
I can't even guess at weights for the VMC and haas style mini-mill
I think the Haas mini mill (the one that fits thru doorways) was intended for the graphite EMD electrode makers
OK - they call it the Haas Office mill - http://www.haascnc.com/prod/CNC_large/72992_lg.jpg
Office Mill; 12" x 10" x 12" (xyz), 5 HP, 30,000 rpm ISO 20 taper spindle with 20 tool carousel, includes Visual Quick Code, 1MB, Beacon Light, coolant pump, high-speed machining, 1-year Warranty. 208-250VAC 50-60Hz, 20A, single-phase power only, with NEMA L6-20P plug.
1350 lbs - lighter than a bridgeport
and costs as much as a regular Cat40 VMC
on casters... put a big shell mill in it, keep orienting it, see if the cabinet spins :)
a "BIG" shellmill would be 1.5"