ken is now known as ken_
which emc2 is best to send info to the machine ? mini..stepper...stepper-1
i've been useing mini (haven't got mill yet) to learn the codes but not sure if it's good to use with the mill
anybody have experience machining sand cast aluminum? I think it might be 356 alloy
we do these bizarre ring things at work
I just ran 5 of 16 parts for a friend, and I'm seeing significant wear on my uncoated HSS endmill
is that stuff more abrasive than the usual 6061?
very, very much so
there's a lot of sand inclusions in it.
we use exclusively carbide insert tooling for it
the harder grades, at that
you can see it kick off a spark once in awhile.
carbide endmill is $56, HSS with TiCN coating is $15, I'm having trouble deciding which way to go
the carbide will last you the job and probably longer
but it will also take a beating
since this is a friend job, I can't really just pass the cost on
jmkasunich: darn, wish I could loan you mine
you can always send an uncoated endmill to get a resharp
that helps the cost of carbide tooling
also a good carbide endmill in regular aluminum will haul ass forever
I doubt resharp makes much sense at this size (5/16" dia, 3/8" shank, one size longer than standard (because I need the reach)
if you run it fast
eh, we do have some resharps in that size
I'm running 2400 RPM, which is as fast as my machine goes - thats about 180 SFPM
in 6061 uncoated HSS would cut all day
not so in anything sand cast.
sand is unavoidable in the castings.
I suppose I should check somebody other than mcmaster - if I can find a slightly better price for carbide it sounds like the way to go
has any one done any speed runs using emc2??
can you get by with 1/4? They are not expensive at all
you mean 1/4 dia with 1/4" shanks?
what kind of carbide??
I don't (currently) have a 1/4" holder
hmm, you'd need even more spindle speed too I guess
dmess: the kind that cuts sand cast aluminum without getting dull
there is very little carbide that does high silicon content alu very well.... it is harder than glass
it lasts longer than hss tho
cradek: 5/16 with 3/8 shank vs. 5/16 with 5/16 shank is a big difference - I imagine that 1/4 even more so
jmkasunich: the pump is unstuck and I didn't overtly break anything!
true... is this a production job?? or 1 off
I wouldn't be at all surprised if it runs tomorrow
semi-production - 16 pieces for a friend
after 5 pieces, my uncoated HSS mill is starting to bog down, and visibly worn
thats production for me
the "for a friend" part means I can't just go by a $50 end mill and pass on the cost
hey a carbide should finish it up..
can you wait a few days... i'll send ya 1
[03:02:15] <cradek> http://www.use-enco.com/CGI/INSRIT?PMAKA=322-5024&PMPXNO=7776407&PARTPG=INLMK32
lots of choices at $18-$20
have you tried using varsol on it??
(I've been using WD-40, which is mostly kerosene)
varsol is thinner
leaves a nice finish... but dont smoke around the machine..
i'm not sure how much coolant will help aside from finish, he's only able to spin it at 180 sfm
dmess: I'm using a shoptask - a rather lame mill
2400 RPM on a 5/16" cutter
do you need the 5/16 or can you use straight 3/8?
5 ipm when cutting part-width, about 2-3 ipm when cutting full width
are you profiling or facing this thing
the machine is NOT happy about cutting full width
I could use a larger tool - I'd have to revise the program a bit is all
if you are facing use a flycutter
4 flute or 2??
and just resharpen it
do you have any carbide? I bet you could use a flycutter
flycutter isn't really an option
flycutter fould face it off best... single point 1 pass
this machine doesn't like low speeds, unbalance, or interrupted cuts - its very flexible as mills go
the job isn't entirely facing - I use the same tool to outline a port
a toolchange is in order
its an intake manifold for a small engine
facing a port seal area with anything other than a f/c is unacceptable if its MY engine..
toolchanges are manual - so I'm trying to avoid them
toolchange is in order, man
but to make a good part you should
you need to balance a flycutter and face it that way, you'll way save on endmills
because like, you know, a 20 dollar grinder from HF
and you have a completely abusable and resharpenable facing tool
how big is the part??? aprox machined area l x w x d
dmess: I was just measuring it ;-)
1.5W x 2.5L x about 0.100 deep
processing parts is my day job too
with a 0.812 port in the middle that is cast a bit off center and needs rounded out
is this port CUSTOM
or would a form cutter do
3-d contouring is kellering
part fixtured: http://jmkasunich.com/pics/IMGP1995.JPG
I face it, clean up the port, and drill two holes
face all the parts
the next step is flipping it over, mounting it by the newly drilled holes, and do the other end, haven't worked out the details of that end yet
toolchange, clean the part
fun ehh.. some of my part hav 60-70 machining operations and 30-50 processing ops..
flycutter is sounding more and more like a good idea
dmess: you get paid a lot more per part too
this is definitely a low end operation
different business lad..
you make landing gear, right?
the only 'trick,' if you even want to call it that, from doing 1 part to 10+ is to change the part, not the tool
then when you finish all whatever you change the tool and start over
toastydeath: that assumes that the part goes into the fixture consistently every time
i perforn the same duties for many low end operations that need it right the 1st time or they loose the farm..
jmkasunich: true, but if it doesn't there's usually other issues as well.
and a little creative process engineering can go a long way.
yes... landing gear now... but ive made some of everything... steel mills , cars , cruise missles , knobbly washers is what i call them all
this first step is supposed to give us a flat face and two good holes - then it won't be so hard to fixture for subsequent operations
then have it clamp upward against stops
rather than clamping down.
or to the side, or what have you.
or just keep going how you are, whatever is working
so face the bottom end 1st... off the flater larger facer
well, I have two evenings into fixturing and programming so far
how much longer in machine time and tooling?
evenings are cheaper arent they??
I can do about 5-6 parts per hour
so two more hours would finish this end
I really don't want to build a different fixture for this end - that would feel like starting over
also is it just the tips of the endmill
that are worn
if rpm 's are limited you may be better to stick with hss
yes, just the tips
grind the EM back
like a drill.
except you know, don't point it.
slow it down then rpm wise they will last longer
when i'm beating on an endmill i just let it get a little dull and grind it till the dul part of the flute is gone, which isn't far.
hmm, I should try that
take a little off one flute at time, and spin to the next one
so they're all somewhat even by the time you're done
usuall its at the depth of cut line
if you don't care if you end up cutting mostly on one flute, I bet you can freehand it
how much mat'l is your deepest up the e/m
DOC is 0.050 or less
cradek: i am talking about freehand
I've been thinking about trying to go deeper, so it would take fewer passes = less wear on the corners
are the chips sticking??
jmkasunich: that's another good tip, the less shear area a cutter has to go through the longer it lasts
slow the feed and go... keep it to about .002-.003 per tooth...
dmess: right now I'm at well under that
that's uh, bad
but drop the rpm to 2000
If I had scrap pieces to experiment with I would
why is that experimenting
use the firs .05 to experiment.. DRIVE it... break the chip
but I'm afraid to go heavier with his parts - not sure of my fixture rigidity, and the machine itself isn't very good
its a 5/16" cutter... 1 good kick checks the fixture...
did you see the fixture pic?
and stand by the machine dont just plug and go play with youself
the vise is doing all the clamping - it bears on about 1/4" of the rounded ends
dmess: I'm not stupid
i know... i just work in a union shop and you gotta say it..
I stand there, and spritz lube on it, but if something shifts and digs in, I'll have a ruined part or broken tool before I can even start to reach for the estop
or the say he didnt say i couldn't...
what's that one saying
I realise I'm kinda chicken-shit - I'm still learning the machines limits
the customer has the explicit right to keep both halves of the broken part?
i like that toast...
the friend paid an arm and a leg (tooling cost especially) for the parts that I really don't want to ruin
then why is he trying to get them don on the so cheap
jmkasunich is a much better friend than i am
cause he's alot like me
i'd be charging anyone who wanted sand castings done, at least for tooling
and i'd want a spare for setup, especially if i'm making 16 instead of 1
tools and time is by best deal
and 5% scrap allowace ofer 10 pcs.
I think I'll try sharpening that end mill and run another part - see if it runs nicer
keep in mind that 5 ipm and .002 chip load is an ultra light cut, especially at .050
for the future, anyway
the machine may very well work better once loaded a littlee
i bet 80% of your wear is that you have you chip load under .002
and that's giving the sand a chance to really grind the tool down
right now its just rattling and clanging under a lose spindle
there's small tools ( <1" tool shank) that use inserts. you can buy different inserts ( hss, tin, carbide ) and handle different materials and handle chipped tools ( longer term solution)
FYI first profile cut should be conventional milled not climb milled
i'm pretty sure his machine has backlash and thus precludes climb milling without shenanigans
lets your tooth dig into base metal insted of a coating of sand encrusted crap
the sharpened tool doesn't look too bad under a 10x lens
I'm gonna drop speed, up feed, and see what happens
or just drop speed!
if you are concerned about forces.
feed rate determines the stress on the fixture, not rpm
not feed per tooth?
force DECREASES with increasing feed per tooth
it's called the feed coefficient with a .010" per tooth/rev/whatever as the standard
.002" has a 1.20" coeffcient, which is bad
.060" per tooth is 0.79
weird - I sure didn't know that
shear is the major force
how can ripping off a 0.060 chip generate less force than shaving off a 0.002 one
the fewer shear planes you generate in a given distance, the less force it takes to cut that material
obviously you do reach a diminishing return
where the bending force takes over
but it's very large, at like .100"-.120"
that's why they use .060"
that's the most economical feed.
.060 on big lathes and whatnot, that is
you need a pretty big facemill to get .060 on a milling machine
but a lot of mills will do .020" easy
damn I'm good
flute lengths match within 0.001 - freehand sharpened
also this may interest you, in small machines especially
(DTI on a height gage - my tool length pre-setter)
the chip load determines the chatter damping factor
the depth of cut is the gain, and the chip load is the damper
so especially on a small machine like a bport, it's important to maximize feed to get the most out of the machine's limited rigidity.
bport and smaller, obv
a compared to a shoptask, a bport is a big machine
and so loading the flutes down becomes even more important.
the less rigidity you have, the more pedantic about chip aspect ratio you need to be.
feed controls the power required. RPM and chip load are just things you can screw around with to change the surface finish and efficiency.
so increasing DOC is probably NOT a good plan....
(that is what I was considering doing, so I'd have fewer passes and less wear)
i'd make changes in steps, so i didn't get confused when something didn't work
first drop the spindle speed until you get .002-.003 per flute
then once you know that works, change the depth of cut to .100 or something.
I actually have two kinds of cuts - some are full width, some are part width
on a 5/16ths endmill .100 should be fine
the full width ones are the nasty ones
or use a squirt bottle and a light stream of coolant
its not a chip clearing thing - only 0.050 deep remember?
chips clear fine - but I swear I can see that cutter bending
full width slotting shouldn't be an issue then, and i'd wager it's only an issue because of a too low chip load.
I've been running the slots at a lower feed
sounds like thats the wrong thing to do
that's a big problem then.
dull cutters need progressively larger feed rates to cut.
because the edge is rounded and you have to get enough of the tool into the metal to get around that edge radius.
or else it just rubs
otherwise the wear rate just gets worse and worse
and rubbing in this case really means grinding, because you have abrasive media in contact with the endmill.
and that is my sermon on feed rates.
ALL RISE FOR THE MACHINIST'S CREED
niagra' advise for sand cast aluminum 1) machinability charts http://www.niagaracutter.com/techinfo/millhandbook/speedfeed/mach.html
2) chipload charts http://www.niagaracutter.com/techinfo/millhandbook/speedfeed/tooth.html
i'm still reading 'em, but toasty's advise sounds good
just raised the feeds to 5 ipm, and will be lowering speed to 1200 rpm, that will give me 100 SFPM and 0.002 per tooth
you can basically throw sfm out the window
that low it should't really change anything give or take a couple hundred
obv. -200 sfm is a real number
pleses add #include <algorithm> #include <stdlib.h> in /emc2/src/rtapi
gnix_oag: stdlib cannot be used in realtime code (RTAPI is realtime stuff)
toastydeath: dunno if its the newly sharpened cutter or the feeds/speeds, but its working nicely
if the cutter doesn't dull in 5 parts, it's the feed
if it does, oh well you tried
how do i make the .mo file from the .po? put it into .../emc2-trunk/src/po and run make?
gnix_oag: thanks, i do have msgfmt
tom1: rs274_err.po tcl.po not work
how not work?
I send you email for new axis.po
#2 done, no wear visible yet
I eliminated the first roughing pass, so the DOC was higher this time, and it didn't mind
took the feed up to 6ipm (from 5) on #3, and it was happy
don't think I want to go higher though
no need to push it
like you said earlier
#3 done ;-)
part time is down, thanks to the higher feeds and one less facing pass
better tool life _and_ better productivity - gotta love it
I think the rest can wait till tomorrow (after midnight here)
use msgfmt to make the .mo from the .po, then create the local directory & put the .mo there, then 'export LANGUAGE=zh; LANG=zh; scripts/emc', then pick sim-axis... and 'hao hao hao hao' you got Chinese EMC2 and AXIS (for 'hao hao hao hao' see http://www.youtube.com/watch?v=rOyj4ciJk34)
AH! i used big5 ! doh!
xie xie nin!
i was using the wrong fonts!
weird when it uses ascii to try and convert
hmmm, cant get AXIS to use the fonts all other apps seem to be using. eg: irc uses a pretty Chinese font, axis uses ugly stick figure Chinese font this is from terminal (gedit:28393): Gtk-WARNING **: Locale not supported by C library. Using the fallback 'C' locale.
loaded gnome Chinese support package, restarted X, same ugly stick font in Axis, great fonts in gedit, irc, firefox, desktop... save this for another day. goodnite all :)
tomp: axis uses tkinter, not gtk
well.. tk is usually ugly :(
how to tell tkinter not to be so ugly?
"stop being ugly" ?
or louder :)
but thx, i understand now
i think you can use the x resources database
do somethiing like *Axis*Font: -*-courier-bold-r-*--100-*-*-*-*-*-*-1
stick that in a file and do xrdb -merge <file>
hey guys - anyone here to talk to me about cad and files and stuff?
what did you find out
not much - he wants to convert to STEP and IGES - not too worried about GCODE...
If his version of ACAD won't do it then the only thing I know about is high dollar software
there i snothing written to do it without high dollar s/w?
not that I know of
you might try on the #cam
lol ur on cam already! hehehe
yea, it's another favorite spot to hang out
dmess: i dont understand what you mean.... lol
[Global Notice] Hi all! It appears we are having some problems making our EU and US hubs speak with eachother, I'm going to do some noisy re-hubbing and hopefully we'll be back to normal shortly! Thank you for using freenode and have a great day!
[Global Notice] Hi all! It appears we are having some problems making our EU and US hubs speak with eachother, I'm going to do some noisy re-hubbing and hopefully we'll be back to normal shortly! Thank you for using freenode and have a great day!
ptc pro/e for linux
pro/e for linux?
[Global Notice] Hi all, Sorry about the disturbance there -- all should be back to normal-ish now. Have a good day.
[13:29:23] <gnix_oag> http://forum.ubuntu.org.cn/viewtopic.php?t=3425&highlight=
cat did a tour on the key board last nite bt the looks of it..
dmess: you mean that wasn't you?
finally started to make sense..
yeah... go figure....
keyboarding is about as good as mine
jmk... you up and about??
not before noon
33 mins before noon
a new record!
jmkasunich: you sicj? should I call an ambulance for ya?
that big yellow ball in the sky (that never appears during a cleveland winter) shined into my window and woke me up
how'd you last few parts go .. or have you made them yet??
I did three last night - worked nice
I haven't done any today yet, I have some errands to run
good to hear.. learn anything in the process??
yeah, keep a minimum of 0.002 chip per tooth ;-)
oh hey baby speed is your friend in aluminum
or rather FEED
thats what I thought, but in this case, dropping the speed is what it took to get the chip load up
I'm tempted to bump the speed up again, and raise the feed to match
but that will be later this afternoon
at the sfm you were running it really couldnt hurt... and would deffine=itly help tool life
if its working and you only have a few to go.. wait till the last one then bump it to the limit
I went from 2400 RPM and 2 to 4 ipm to 1200 RPM and 6 ipm
dmess: thats silly
so max is 2400 and 12 ipm
the reason to bump it is to make the parts get done faster
I'll probably sneak up on it - do one at 1800 and 9 ipm, see how that goes
well safely correctly is my moto/creed
you have something you KNOW works.... work it
ruining the last one is just as bad as ruining one in the middle
if I bump up and there are problems, I know I can go back to 1200/6ipm
but it may also wipe out your set/u
i hate setting up again for the last 2 parts
true - but I'm developing more confidence in the setup
dont get over confident
I have 1450, 1800, and 2400 available (this machine uses v-belts for speed)
my past experience has been that increasing speed and keeping feed per tooth constant actually helps - maybe at higher speed the weight of the machine helps smooth things out
well its actually the feed thats loading up the 'weak' machine
yes and no
and keeping it from rattling itself to bits
feed determines the average force, which "loads up" and deflects the machine
if feed per tooth is the same, force per tooth is the same - more speed means more teeth per minute and more _average_ force
ti tightens it up for lack of a better analogy
but the individual forces are what makes things vibrate
you ever use and variable helix mills?
they are nice
yes the can pitch the chips
seems like they'd only help for profiling
they cut down on vibration
slotting is magic too
great for pockets too
I use lakeshore carbide
hanita and international mini-cut are good ones
I had some 1/2" high helix three-flutes that I picked up surplus - those were _nice_
unfortunately I broke them
this was on the van norman, not the shoptask
lakeshore puts a radius on the edge and most others only chamfer the edge
the collet wasn't tight enough, and the high helix made it pull out and get deeper
the radius lasts longer
3 flute IS the dream cutter for aluminum
the VN is so rigid I didn't have a clue what was happening until snap - it was the better part of an inch deep by then
a little SUCK out eh'
so limit your cuts to 3/4" deep now
the collets on that machine suck
on the vn
what kind are they
[15:44:21] <jmkasunich> http://jmkasunich.com/vannorman/VN_Spindle.htm
someday I want to make a new spindle for that machine, with an R8 in it
as spindles go its not that complicated - no splines or anything
put a RapidSwitch 200 in it
once you set your Z on each tool you can take them out and put them back as the job progresses
sometimes you can find the master with your shank and just insert that
here's one for the r8
[15:48:07] <BigJohnT> http://www.tools-n-gizmos.com/store/agora.cgi?user2=yes&cart_id=1905425.4183*582LM7&p_id=20995&xm=on&ppinc=1a
since the VN isn't CNC, Z repeatibility isn't much of an issue
I rarely use end mills in it anyway. mostly use it for facemilling
ok, I forgot it was a manual
[15:51:45] <jmkasunich> http://www.metalworking.com/dropbox/_2001_retired_files/Facemil3.jpg
I did this drawing several years ago when I first started thinking about an R8 conversion
[15:52:41] <jmkasunich> http://www.metalworking.com/dropbox/_2002_retired_files/SPINDLE.gif
I guess the toughest part is grinding the taper in the inside
drilling the drawbar hole wouldn't be a picnic either
only 12" long short and ridgid
yeah, they are great machines for their size
but that does not have to be very precise
only 1800 lbs, but significantly more rigid that a bport
what kind of spindle speeds do you have on the VN?
70 to 1400 in 9 steps
1.5HP spindle motor, I put a VFD on it
the machine has taken cuts that made the VFD trip on overload, and the machine didn't care
well, TTUL John nap time for me
nap time? I just woke up!
* jmkasunich <== night owl
I've been up since 5am
Saturday and Sunday I get a nap
I better head off to the bank before they close
see you later
looks like a nice machine for hogging Jmk
looks like it might be a flying day..
WSW @ 6 km/h.... not our ideal launch direction but we can work with it...
now to locate all my gear...wing check... bucket check...x-tra biners check... bridle check...vario check... gps n/a - staying local... looks like we can lock and load...
dmess: glider, hang*, para* ?
hi! I am having trouble using the parallel port. it doesn't seem to be talking. I've read the FAQ, but it didn't help. Can someone point me to a good place to troubleshoot my parallel port?
and I've tried to load par_port in the config file too.
para.. but my buddy has a few hangies i fly reglarily too.... my vario seems to be acting flacky on reset..
Dallur .. you Fly?? real aircraft...need no engines...
sorry! I got kicked off again.
looks like i'm doing the annual inspection this afternoon so i can fly tomorrow...wing inly had 6 flights in 2007... NOT good for her or me...although i did fly ray's MAX 3-4 times as well as to solo 2 x... for a total of 6.5 hrs...so not really a BAD year in southern ontario..
4.8 hrs on varous hang-gliders too
you have a picture of the Solo?
dmess: I flew hang gliders for a couple of years, some ultralights and then my dad and brother are both commercial pilots so I get to play with big toys some times :)
hmmm chk out www.uflyontaio.com
there should be some pics of it in the gallery.. i believe its still for sale... test driving.. you know...
i like the smaller powered things... i want a float capable super ultralight..
we have NO hills to hop off... so we TOW...
I like even smaller things... like model aircraft. (I am afraid of heights)
ah. finally, the server kills my ghost
the best winches on the planet are in southern ontario... ive worked on the design of 2 of the 4
tow by a truck?
height isnt an issue.. so am i... after 80 feet the perspective is gone.. speed is your frien... ; )
no hydrostatic winch
is there specific place you can fly?
as in?? controlled .. NO.. we are in VFR teritory
dmess, did anyone reply my question (as Alphonse_) before I (as _Alphonse_) show up?
I was looking for a place to fly large model aircraft.
and I don't know the area in Ontario.
and 1 site is designated a training area with gliders on the map... BUT every one comes to train in our airspace...
I know a few of my internet freidn there, and I like to go there.
what does VFR stand for?
not good when we can have up to 5000 feet of string attatched to a glider on downwind..
visual flight rules
where are you??
aka fly by the seat of your pants (no instuments =)
vario for alt... otherwise its eyes WIDE OPEN ALL THE TIME...
gps for positional airspace if your really close to an a/p
I like to keep my feets on the ground. :p
GPS gives about 15m of accuracy.
i like to have em in the air
ah. my second ghost got killed as well
_Alphonse_: you can always kill them yourself.
hell if im 15m into bad air space and turn around i dont think they'll be scrambling the cf-18's
you are too slow to be shot down by cf-18's!
_Alphonse_: /msg nickserv help
I'll kill myself so no one answer my qyestions while I am kicked off.
yes but 1 pass and his turbulance would WIPE me out of the skY
my router is not liking IRC.
can you fly up to 5000ft?
a friend made it to 10000' last summer only stoped cuz thats our cieling....he said it was a BOOMING thermal kicked his asss and 20 minutes after he landed it turned into an anvil
to about 75000' he is military pilot so ONLY flys by the rules or he can lose his real job
what does he fly?
75K?!?!?!??!?! Got Gravity??
but to see where he came out of turn to that was awe inspining....
its calle cloud suck
is it a jet engine plane?
paraglider.... he didnt go to 75000' the cloud did...
no engine aircraft
his vario showed some spots of 1800fpm UP
spiking to 2800 for up to 5-7 seconds
hum. the emc2 update puts in a new version of config file...
its an elevator ride to heaven when and if you find it
what did I missed?
ah it worked!
what did I miss after "elevator ride to heaven"?
how much does it cost to maintain the plane?
maintain... my para has cost me 2 new carabiners when i bought it 9 yrs ago...
Ray's falcon 170 may have cost a few hundred dolars in down tubes but we,ve taught 50-75 ppl to fly on that wing
why am I keep getting disconnected. :(
so it is pretty low maintance cost...
many of my things will break just sitting there. :)
the cheapest thing that the FAA still regulates
you have to have certain licences to fly abofe 750' agl in canada
how much is the training?
our place for $1500 and the purchase of a new glider.. you get about 2 seasons of flying for the bang
gliders run from 1800-4500
I want to learn to fly!
but understand NO-ONE in ther right mind will sell a newbe a 4500 dollar glider... he'd be the next statistic
I like that odd! hahaha
so donate time and work it off that how i did it... i knew nothing about these towing winches when i showed up... but by the end of my 1st visit i knew how it worked and WHY the needed it - step towing -
I have to find out how to stay connected.
I'll come back later.
I still need to fix my parallel port.
it just won't talk.
swap connections on the m/b
parallel port or the ethernet?
are you sure you are using the right address?
the disconnect is probably the router.
I am pretty sure.
I check the BIOS,
and also the config file.
and I updated the emc2 just now
pretty sure.... is a no
too many changes all at once.
somewhere you have a line that says "loadrt hal_parport cfg="<some_port_address>"
the address might be 0x03f8 or whatever
it MUST match the address of your parallel port
I checked it is 0x3BC and changed it in the con fig file.
checked the pin out.
do a loadrt probe_parport
put the pin on a real scope.
please let me stay connect!
please let me stay connect!
I can't help it!
well.. it was fine, until you closed it..
interesting - this time you have a different ID
last time: email@example.com
anyway, I checked the BIOS... it is 0x3BC
this time: firstname.lastname@example.org
yup, probably a DSL reconnect
new ip address
yeah, I have to run to another computer... but this means I am away from my machien that has EMC2 on it.
are both machines in the same building?
thank for your understanding!
yep. this machine is actually a wireless.
one address is at a university, the other is DSL (probably residential)
the wire machine keeps cutting me off!
so I have to run back and forth now. :(
why is one machine connecting thru ucdavis.edu and the other thru what looks like residential DSL?
I hope this stays connect... I am almost out of range.
yep. someone seems to find a way to do a remote sharing on their wireless... I am sharing the connection over remote desktop
ok, I am using this IRC program over remote desktop.
... it doesn't make sense does it?
should I try again?
it doesn't have to :)
are you at ucdavis? or at home?
I am still at UC Davis.
_Alphonse_: I'm sure you'll figure it out
so the sbcglobal you are using is NOT your own ISP?
and it is the UC Davis computer keeps killing me
we can be of help regarding emc
SBC is my home computer ISP.
why does comcast stink so bad?
I wish I have comcast. they don't do it in this area.
my home area
anyway, back to EMC
I need it to work... but parallel port just doesn't like to talk.
about half the internet is not available right no
on the Dell PC, I checked the BIOS to be 0x3BC.
it is the default.
another one of my ghost.
anyway, I am using the sample-config of the stepper_mm
_Alphonse_: some parallel ports need another module before
so I changed the core_stepper.hal to match the address
I tried putting the loadrt probe_parport also.
ssh -D ftw!!! =)
oh? alex_joni, which one? so the default config still need more work?
JymmmmEMC, are you hacking my computer?
I was meaning try to add probe_parport, and see if you get any difference..
nope. it didn't work.
I put it... let me check...
_Alphonse_: Nope, from what I hear, you're doing a fine job of that all on your own ;)
I put it in core_stepper.hal, didn't work, and also tried in standard_pinout.hal, and it didn't work.
JymmmEMC, yeah, it seems I am staying connected!
you should set BIOS to EPP, right?
just got my pico systems universal stepper control plus gecko board... now i only have like 4 thousand wires to hookup, fun :) Gotta get a box to put this all in first
klickrr, I had to use something given to me.. so I can't buy a pico system. :p
but, it still takes a stepper signal like pico.
but right now, I have it hooked up to an oscilloscope.
_Alphonse_: you need to place it before the loadrt hal_parport line
and I have the right pin! pin 3
_Alphonse_, try SPP or bi-directional if you are just using step and dir from the port
alex, yes, I put it before loadrt hal_parport cfg="0x03BC"
_Alphonse_, if you have a smart board like the pico attached you may need EPP
I am on EPP now. I tried ECP also. but didn't try bi-directional.
should I try it? I'll run and try it.
_Alphonse_: you're trying to load a ppmc board?
Jon Elson had a small program on his site that tests communication, you should try getting that one to work first
its unusual for the port to be at 0x3BC... that was the typical address when it was on a mono graphics card
I think it even does something to the EPP port
alex, nope. it is just the standard parallel port on the motherboard
then why do you want EPP?
standard address for parport is 0x378 usually
and the second one at 0x278
I read one of the archeived files on EMC site, and it was said to use EPP
yep, this computer defaults to 0x3BC
i can change that too.
I am on Jon Elson's site... let me see if I can learn somethign there.
so I should be on bi-directional?
epp/ecp "should" work - it acts the same as a bidirectional port unless you use fancy multi-byte writes
I've tried bo EPP and ECP. but haven't tired bi-directional
what's the difference between FERROR and MIN_FERROR? what units are they in?
awallin_emc: ini units
awallin_emc: ferror is the max error allowed at max velocity
one is for halted, the other for max vel.
the allowed error increases proportional to velocity
should be mm/s or in/s
alex, which on Jon's website is a test program? it seems the listing are all hardware.
_Alphonse_: his website is only about his hardware
which you don't have, so don't waste time on it
oh! found it!
er... mm and in, not /s
should Stepconf work right the way when you test the x-axis?
am I still connected?
_Alphonse_: if you're debugging the parport by seeing whether the motors move, maybe that's the wrong way to do it
i would get a multimeter and look at the voltage
actually, I have it on an oscillocope.
how are you testing?
that's why I feel sad!
I can't see anything sqare wave coming out!
what do you do in emc/hal to test the port?
fenn, I first use the Stepconf to try on the x-axis.
how many encoder counts do people usually set DEADBAND to? I now get humming and hissing, perhaps it's too low?
awallin_emc: at least 2
then, I tried using the stepper_mm and change the config files.
_Alphonse_: do you know how to use halcmd?
fenn, I've chaned/checked the BIOS, the address, do the loadrt probe_parport, etc.
fenn, actually no. I tried that to unload the pin and toggle it, but it doesn't change. so I still don't know if I am just not doing it right, or something is worng.
well, you can see what hal thinks the value of the pin is with 'halcmd show parport*'
er, halcmd show pin parport*
should I do it on the command prompt?
bad. I got lots of messages
RTAPI: ERROR: could not open shared memory (errno=2)
so the rtapi kernel module is not loaded.
what does 'uname -r' say?
using /etc/init.d/realtime status...
shouldn't hitting the ESC key result in an E-stop by default? (I'm using AXIS)
everything is not loaded.
_Alphonse_: this is with emc running?
nope. emc not running.
I reloaded the moudle just now
by using /etc/init.d/realtime restart
well, start emc or stepconf or whatever you're testing with
and call halcmd show pin parport*
awallin_emc: esc aborts motion, but leave amps enabled, etc
and it returns, well, empty. here;
Owner Type Dir Value Name
try just halcmd show pin parport (no star)
and that's it.
if you onle restarted realtime, then parport is not loaded
the levels of "on-ness" are estopped (everything off) , machine on, and moving - esc takes you from moving to machine on
you can load it by: halcmd loadrt hal_parport cfg="0x378"
jmkasunich: ok... another question: what's the time-constant fir the I-term in PID? I've tuned my pressure controller at work and it seems the time constant is important
but you still need to set up threads, add functions to the threads, etc
should I unload, then start stepconf?
IOW, an I gain of 1 means that an error of 1.0 that lats 1.0 seconds will result in 1.0 of output
jmkasunich: ok. trying to tune a motor without load here. maybe not worth it as it has to be re-tuned when in the machine anyway...
OK, I started stepconf wizard, then called halcmd show pin parport.
but it gives me the same RTAPI: Error: could not open shared memory (errno=2)
awallin_emc: can be a good learning experience though, if you have the time and don't mind doing it over again later
_Alphonse_: i havent used stepconf, but i think you have to go through all the screens to get to the 'test axis' screen
fenn, oh. I can do that.
jmkasunich: yeah, trying to get some reasonable values so I can dry-run the whole shebang (servos, VFD, spindle encoder, coolant, jogwheel etc)
_Alphonse_ fenn why is stepconf suddenly in the picture
I thought he was trying to test a parport
jmkasunich: because it's easier than setting up threads and functions
if you say so
* fenn shrugs
ah! a good sign!
(I have no idea what stepconf actually does, so I'd be working blind)
OK, after getting to testing x-axis, actually getting to "Test this axis", and do the halcmd show pin parport, I got a long list!
if you don't know what it does, why did you tell him to use it?
perhaps its a case of the blind leading the clueless.. or is it the other way around?
_Alphonse_: so now look for the pin number in that list and see if TRUE/FALSE matches up with reality
hahah! thanks fenn, JymmmmEMC. I wish I can do better with Stepper_mm... but for now.. here is what I see.
uh, please dont paste all the lines in the channel
on the step pin... pin 3 it said:
6 bit IN FALSE parport.0.pin-03-out <== step
how should I decode that?
is there any rule-of-thumb to go by when you need to mill a round/rectangular hole for something to fit in; gap wise? pilot lamp in metal, plastic in plastic, etc
FALSE means the pin should be at 0 volts
so if a pilot lamp is .750" what size the hole should be
fenn, multimeter said it is high at 4.81 volt!
on pin 3!
dmess: total, or per side?
now and 4.82 volt. :)
this is bad
_Alphonse_: I bet you have the wrong port address
I don't care what the BIOS says
total is best... and they SHOULD all fit inter-chagably
jmkasunich, should I restart to change the address to the EMC default 0x387?
lamp is punch... size the die...
dmess: Well, I have a rectangular circuit breaker that's .750" wide, so the total width should be .760" or .780" ?
whats the tolerance on the .750"
JymmmmEMC: if you don't know the size of the breaker (including tolerance) how do you expect to make a hole to fit it?
if you have only one breaker, measure the damn thing
.005 true position of the walls to the center of the bore.... $1000 invoice on the way..
jmkasunich: breaker size I know, it's tolerance I don't
if you have only one, tolerance doesn't matter
make the hole 0.01 bigger than the part that is gonna go in it and call it done
Now, it that for metal or plastic?
wait a minnit
why do you think that matters?
I'm trying to accoutn for thermal expansion
you are a trip sometimes
but some STUPID assembly dwg's call up unmeasurable stuff.. and on an FAI you say this was this hence that is that alot...
is there an easy way of doing a screen-cast in ubuntu? i.e. record an AVI or something of what's happening in AXIS ?
awallin_emc: not really
snapshot yes, video not so much
awallin_emc: most screen cast methods dont play well with opengl
thermal is a crop a crap
awallin_emc: jepler had something on his blog how he did it
especially when you have 0.010 clearance on a 0.750 hole
unless its thin titanium or SS
fenn, jmk: oh, ok... have some load-meters in AXIS which are kind of nice.
jmkasunich: Seriously. I don't know metal characteristics, more familure with thermosets, and they do expand quite a bit. I was just inquiring if there was a rule of thumb when a comonents tolerance isn't known
if you were working with tolerances 10x tighter then it might matter
awallin_emc: you can install libgl1-mesa-swx11 (software opengl) and use xvidcap
hell my van fit thru .010" if i hit it fast enuf
JymmmmEMC: define "quite a bit"
then do the math
dmess: elongated missile effect?
but they are bismuth based (mosty) specificly for that reason
google says aluminum expands 25 ppm per degree C
jmkasunich: by quite a bit where the component will bind to what it's being fastened to w/o fasteners =)
and googl know EVERYTHING right...
if your breaker heats up by 50C (damn hot IMO) thats 1250ppm, which is less than a thou on a 0.75 hole
alex_joni: thanks for the link. looks complicated...
jmkasunich: I'm not saying just a breaker - that was just one example, but in general.
dmess: when google finds a couple of tables of material properties and they agree within an acceptable margin, yes ;-)
interesting: all the pin signal is the opposite of what halcmd reports to be!
eh? maybe i've got it backwards then
we have to let parts coming in off a truck sit for 3-4 hours before we can do inspection control sheets properly
_Alphonse_: try monitoring the dir pin (direction) and change it back and forth with stepconf
not sure... let me tell stepconf that It is opposite.
a couple tables... i have 60 yrs of material tables to draw on... steel or alu...
i have a table - it's made out of maple
with 1 guy still around from the START...
i draw on it sometimes
dmess: if I was doing something that mattered (besides telling jymm to stop worrying about stuff that just doesn't matter for his project) I'd look it up in Machinery's or a reference book
alex_joni: I'm missing the build tools on this machine, what's the package for that? emc2-dev? emc2.2-dev?
_Alphonse_: try this: from a command line, do "cat /proc/ioports"
and see if there is a line in there that says parport
it has some of our stuff but no where near ALL of it
mine says: "0378-037a : parport0"
fenn, it doesn't change
jmkasunich, let me try
"0378-037b : hal_parport"
I still have Stepconf running
am I still connected?
jmkasunich: I'm surprised it says that
_Alphonse_: yes you are
am i still awake?
I hope so!
(this is my desktop machine, not running EMC)
jmkasunich: by default once you installed the emc2 package parport & co can't be loaded
ahh.. ok, then I'm not surprised anymore :P
now I installed the emc2-dev package but I still don't have 'make' ??
I just run stepper_mm in EMC2, the direction pin just won't go low. :(
anyway, _Alphonse_ has "0378-037b : hal_parport" which explains his problem
he's loading the parport driver with 0x0378 as the address, even though he has told us 12 times that his BIOS says the address is something else
I restarted and change the BIOS!
because I may have all people/software confused. I changed the bios to the EMC default 0x378!
well, you sure have the people confused
personally I would change the software to match the hardware, not the other way around
over an hour ago I wrote: <jmkasunich> somewhere you have a line that says "loadrt hal_parport cfg="<some_port_address>"
<-- SWPLinux has quit ("ChatZilla 0.9.81 [Firefox 126.96.36.199/2008032514]")
<jmkasunich> the address might be 0x03f8 or whatever
<jmkasunich> it MUST match the address of your parallel port
yeah, I did that but I didn't help. :( so I tried the other way.
yep. both times I matched the software and hardware address.
I've tried matching 0x3BC in both software and hardware... and now 0x378.
ok. me shuts up
I did that very carefully.
did you read item 4 here: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?TroubleShooting
that's OK. I modified the stepper_mm hal file, I changed the Stepconf during the startup, etc, etc.
yep I did that too.
so, in the stepper_mm, I changed the...
too manychanges at the sam time
I changed standard_pinout.hal, and put loadrt probe_parport before loadrt hal_parport cfg="0x378"
dmess, sorry. I do keep track of things I do. I just didn't report it on IRC.
but then you tested using stepconf
jmkasunich, I also followed item 8 on that URL.
stepconf has no clue about the changes you made to standardpinout.hal
yes, because the HAL communication screen in Stepper_mm didn't help me. so I had to try another apporach.
I don't understand your last statement
yes, so I have tried to run Stepper_mm in EMC.
"hal communication screen" what is that?
ok, on item 8 on http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?TroubleShooting
I launched stepper_mm in EMC
then, I go under, Machine, show HAL configuration
then, I found the pin3 in the parport, and unlinked it
I tried toggling it up and down, but it won't show on the scope
or the multimeter
so I tried to do my own config with Stepconf. but I don't know if it will do loadrt probe_parport
so I tried a few things, and also on different matching address.
I tried EPP/ECP, and now it is on bi-direction
but I didn't ry EPP/ECP with 0x378 yet.
EPP/ECP should not be needed to run regular parport stuff
it won't hurt, but you don't need it
for sure it is not working.
the parallel port is really very simple - if the address is right and the stupid operating system and/or BIOS is not getting in the way, it will work
some of the stupid operating system stuff that can get in the way is plug-n-play, etc
I agree. that's why I am sad! I can't even get the simplest hardware known to man to work!
does the parallel port work for printing?
not sure. I never configure to print.
I just load the entire OS straight from the CD after burning the ISO file.
is there an item in the BIOS that says something like "Plug-N-Play OS - yes/no" ?
awallin_emc: apt-get install build-essential emc2-dev; apt-get build-dep emc2
I think there is...
I can restart and check... can you wait?
it should set to no,right?
that would be my guess, but if NO isn't working, you could try YES
PnP is the invention of the devil
is this a motherboard parport?
that would be MS
fenn: Build-Depends dependency for emc2 cannot be satisfied because the package lyx cannot be found
awallin_emc: it's in universe I think
ah, yet another reason to put docs in a separate package
i think probe_parport only needs to be done once per boot
uncommented 'universe' in sources.list and now build-dep emc2 works.
_Alphonse_: this link describes in generaic terms what I think the problem is - only I don't know the proper solution
[19:05:24] <jmkasunich> http://tldp.org/HOWTO/Plug-and-Play-HOWTO-2.html#ss2.4
for a PCMCIA parport board, I had to manually load the regular modules, then unload them, then loading the hal modules worked
jmkasunich, no plug-and-play OS option in BIOS. this machine may be too old. :)
old is probably good
the older it is the less weird stuff the BIOS is doing
do you know that the parport works? (for example, if you were running windows on the PC before you put linux on it, could it print to that parport?)
also true. :(
yes, it worked as parallel printer port. it worked with an Epson Stylus Color II printer.
ok, so the hardware is good - that's one step
I just restarted it back up... Linux is now on
darn. pin 3 the entire time never goes low! :(
1 and 17 are still low
humm. I just did a cat /proc/ioports
0x0378 is not in the list.
was emc2 running when you did it?
is hal_parport in the list?
you just contradicted yourself
there it is. it is back after launching EMC
when I first told you to do cat /proc/ioports, I misunderstood what it does
it is now on "0378-037b : hal_parport"
I thought it would tell you where the hardware is, but it actually tells you what addresses the drivers are using
so hal_parport is using 0x0378
jmkasunich: it only tells you the address if you convince it to load the linux driver
thats why it doesn't show up when emc isn/t running
(which atm is disabled..)
so is that a good sign?
hard to say
_Alphonse_: try this
stop emc (and everything related)
stepconf, or any other attempts
it tells us only one of the two things we need to know: 1) what address the driver is using, and 2) what address the port is actually at
all programs are off
ok, now open a terminal
enter 'sudo -s'
and supply your password
jmkasunich, when parallel port is not loaded with the OS, that's a good sign, right? it is not suppose to.
you'll notice the sign before the cursor should have changed to '#'
that means you have root access now
_Alphonse_: just follow alex's instructions right now
now do 'mv /etc/modprobe.d/emc2 ~/.'
that will move the script called emc2 from /etc/modprobe.d/ to your home folder
this was preventing the loading of the regular parport driver
hum... one sec
(if you get an error it might be because you already have emc2 configs in your home dir)
yep. should I delete it?
in that case do 'mv /etc/modprobe.d/emc2 /tmp/'
no, don't delete.. just move it to /tmp/
ok, now you can try to 'modprobe parport'
I did 'modprobe parport' OK
do the others aswell
when I do 'modprobe parport_pc', it gives me a Segmentation fault
that's not good
check 'cat /proc/ioports'
what am I looking for?
nothing like that
in the list
try parport_pc again
parport_pc is the one that detects the parport address
hummm... it is waiting
_Alphonse_: try rebooting
I bet something might have been borked up
yup.. don't be gentle :P
booting up now
_Alphonse_: after it booted do a lsmod
and see if anything related to parport is in that list
ok... still booting
there is lp
ther eis parport_pc
ok, lp it is not used by anything..
same as parport_pc
but parport, it is used by 2 lp, parport_pc
now check /proc/ioports
0x0378 is not in the list
no no no
0x378-037a : parport0 is there!
ok, so the 0x378 is probably the right address
now do the sudo -s again
goooooood! I am not drunk!
then rmmod lp, parport_pc, ppdev
ok, so rmmod lp, rmmod parport_pc... one at a time?
rmmod ppdev I got an error: ERROR: Module ppdev does not exist in /proc/modules
ok, now try the emc test again
running stepper_mm or Stepconf?
the scope gives me some that that may be a square wave on the step pin, but less that 5V... the direction pin doesn't go low.
that isn't very promising
oh, I still have the loadrt probe_parport in the stepper config file. did that cause a problem?
should I restart?
_Alphonse_: I would try to move the file back from /tmp to /etc/modules.d/
then see if it still works (run stepper_mm, and check the voltages on the pin)
if it works less well then now, then I'd say you have to do this procedure on each startup
do you know what files should be in /etc/modules.d/emc2
I just peeked inside /tmp/ and there are things there including an ssk- file.
_Alphonse_: it's only one file
if it's gone from /tmp
you can do this
echo "install parport_pc /bin/true" > /etc/modules.d/emc2
'touch' himself.... hmmm
ok... let me try
sorry alex i couldnt resist
touch: cannot touch '/etc/modules.d/emc2': No such file or directory
I have a few emc.* file in /tmp, but no emc2!
_Alphonse_: are you sure you did the sudo -s ?
then try the next without the touch
how long have you had this insallation??
or if it's easier type 'gedit'
about a month... I haven't be able to get it to work since I installed it. I use the ISO file to install.
it will open an editor which you can use to create a file called 'emc2' in /etc/modprobe.d/
ok.... let me try
sounds like a sour iso... or install SOMETHING has gone sour
not sure. emc2 shows up straight the way after the installatoin...
alex, is it: install parport_pc /bin/true
I should put inside the text file.
in a file called emc2. I put "install parport_pc /bin/true" in there... and it is saved
restart and try again?
humm... all the emc.* files in /tmp are now gone!!
at reboot everything in /tmp gets deleted
I haven't restarted yet.
back in Linux now
nope. the pins are still high
let's take a break. :) and do lunch then come back.
thanks for all your help.
everyone working ?
on a saturday evening?
i sure hope not
* alex_joni is
can someone tell me which configuration to use to run the mill in emc2 ?
kennethj: depends on the mill
but it's most likely that stepper_mm or stepper_inch is close to what you might need
sherline mini mill
click the button
no, the other button!
i mean mini...tkemc...axis ?
that's up to your personal taste
it'll be stepper_inch
didn't know what the differents were
where do i set the tool list file ?
it would be stepper.var for stepper_inch.ini
in the emc folder ?
in the file where the config is
usually when you start you start with a sample configuration
and it asks you if you want to make a copy in your home folder
(which you should do..)
ok, then it's in /home/<username>/emc2/configs/..
got it thanks
why doesn't g92 work with y and z ?
oops z...y works :)
good night all
g'nite alex thanks