Ok, if I want Z to clear material, then move XY to a "home" location, I have to put those two sets of gcode on seperate lines?
uh oh.... I just heard my steppers stall, but emc didn't catch it =(
ah - why would emc catch it?
emc knows all
or does it!
the emc knows that you are a doubt
OR DOES IT
also 20 hp lathes
make a terrible sound
when you manage to overload the spindle
terrible is relative... that's the sound of money for machine repair people ;)
no it didn't break
it just made an awful din
guess it wasn't overloaded enough!
like a rocket tearing itself apart
the whole shop heard it
didn't they say, feed harder if it makes noises }:-)
that was the problem in the first place
5" slug, using the facing roughing cycle
copper doesn't like that so much, apparently
the stuff was just building up in front of the tool
and eventually welded itself back to the slug
at which point we heard "BAM, WWWTTHHHCCCCCCCKKKSSSHHHH"
a live version of the terminator?
it was so loud that we weren't sure it even came from the lathe
and that is the end of my story.
we grabbed a on inch end mill and thought it was carbide. It wasn't and promptly turned it into a glowing mess.
it when we estopped it welded itself to the part
is that where you use estop rather then feed hold?
if you are talking to me, the e-stop on this machine doesn't do anything
I was told to use feed hold for most situations so when is it appropriate to use estop instead?
you have to hit feed hold and kill the spindle
at the same time
yes, this is definately an e-stop moment
anytime you are actually crashing the machine
e-stop is appropriate
Well, maybe I need to find an encoder and tune each axis giving emc feedback
making it closed loop iirc
I wonder if there's anythng within emc that could do that.... run it thru it's paces, pushin the limits a lil more, than making adjustments as needed.
I know velocity can be changed dynaimcally, not sure about acceleration though.
lerman_ is now known as lerman
i need a 1/8 shank flycutter... hmmm
i jsut had a 200 dollar order on enco ready to pay.. then the site goes down or some crap?
is everyone busy cnc'ing?
we are all ducking from the bits of the 1/8" shank flycutter flying around ;)
lerman_ is now known as lerman
lol.. i made one out of the head of a woodscrew
it even cuts aluminum in like .010 passes
but.. it's not the best cutter i ever made
i'd liek to have a real mini flycutter, with a carbide adjustable lathe bit :D
good luck with a 1/8 shank
i thought about an 1/4 shank one at one point and decided against
try the woodscrew thing
a drywall screw would even be hard
and hold up a lil better
what machine are you putting it on?
i settled on a 3/8 shank one
[04:37:05] <chr0n1c> http://ohiopctech.com/cnc/
my mini mill...
it's cutting everything i throw at it right now like a champ
minimill is R8, why not a standard 1/2 shank?
i re-did the z though
oh.. it's uh dremel
it's a self built mini mill
Ohhhh not the seig minimill
lol, 3 precision slides i mounted on a 8 inch mini drill press fram
why not machine something that screws in place of the collet closer?
eh? (can't tell if it is sarcastic or not)
an entension to hold /4" endmills would be better than facing stuff off with 1/8
just thought it was obvious (taig/sherline folks been doing it for ages)
it took me 2 hours to face off 11x4" with the flycutter i made out of a woodscrew at 4 ipm
i haven't had my machine together long enough to come up with cool trix like that
i only used real cnc mills before this...
real as in big boy controls like Haas/Fanuc/etc?
i have all kinds of tricks for knee mill stuff
my fav machine i ever ran was a milltronics control on a mori sekei mill
i think that how you spell it
never had the opportunity to see anyting other than haas and fanuc (limited to what the JC had)
it had a 32"x12"x12" cutting envelope
12" with the knee moving also
maybe a bit more
what kinda machine you got emc running?
how fast can the knee move?
as fast as you can crank the handle.....
it didn't have power.. so only 6" of automated, lol
I have a taig and it is in the process of being emc'ified; still using the crude stock control
oh... it is 2.5-d machine
no full 3
but if you want more than 6 on z you'd have to move it by hand
not if you want to drill a 10inch hole ;)
i had to do that a few times
or build a stair case > 6 inches
I made a quick attempt using a dremel on their drill press as a micro milling machine and the results was disappointing
now i gotta figure out what thread my dremel has on the end of the spindle.. thanks for the tip ;)
it is metric, IIRC
the dremel i have is amazingly sturdy
it's not a brand name dremel
the only drawback is my moutn for it is not sturdy enough to cut CRS yet
I was trying to square a peice of 1/4" 6061
i have it up to cutting about.020 a pass with a carbide 4 flute 1/8 endmill in aluminum
the result was I can file it square better then the dremel
the secret is not to push it past a certain point
0.020? heh... that's a stalled spindle... think I was doing 0.005
like .010 or .005 passes
you mean 0.020 DoC?
i swear i did it...
i milled down a p2 slot 1 heatsink to use on the new driver card i am building
don't doubt that... the question is how square is it =)
jsut a heatsink.. who needs square
filing is quicker!
but.. it's more fun watching chips fly on the mill you just built all by your self...
i put it all together here in my basement withlimited tools... except the x/y slides.. i indicated them and doweled them together on the mill at work
and it does beautiful work... i couldn't be more impressed with emc2
I still got a lathe to finish up
that's my next toy.. i'm going to buy a 7x10 i think
i got it down to harbor freight or enco
be prepare to do a lot of work
on the machine?
why not a 7x12 from Cummins or Homier?
i have read they need tweaked once you get one
well.. i won't use the lathe as much as a mill...
yes, I got a 7x12 when they showed up at a truck sale and I decided I need to turn stuff before I can finish my own lathe :/
i'd rather invest more money in a bigger mill
i am finding the opposite, despite the 1234324234234 tweaks it needs, the lathe gets more use
and go with a wimpy lathe
the 7x's isn't that wimpy
about going all out and building my own 7x10
once you get it aligned, tweaked, it does okay work... I was doing 0.050 DoC and it didn't stall
make it run on linear bearings...
lerman_ is now known as lerman
a few brackets out of 6061 or crs and a chuck.. you got a lathe
you want to CNC it?
the idea has been.. to cnc the brackets on the mini mill once i get it solidified
liquid mill?! heh
oh, it's out of ice, I get it now =)
chr0n1c: where you at?
if you are going to do that, make sure the head stock passes enough for a 5C
San Jose, calif
you mean just make the spindle a 5c spindle?
Does enco have a retail outlet?
i would maybe use a grinding spin fixture for the headstock anyways...
if you want... MT4 is also doable
they might somewhere
i know msc is right down the road from me
and j&l used to be
J&L got bought by MSC
don't think enco has any more retail...they used to have one near the brokaw frys many years ago or so I was told
oh yeah i read that
Enco got bought by MSC too
harbor freight, msc, about 15 steel plants... all within 15 minutes from me
chr0n1c: oh ok, as long as oyu dont have to mail order it
msc is getting fat
easier to return/exchange in perosn
MSC is nice to deal with
yeah.. i usually obsess over details for weeks before i buy tooling
so i don't take stuff back often
mcmaster carr is somewhere close also i think
Well, when you buy china iron, you just might have to
McMaster is less nice
yeah the mini lathe form harbor freight just might have to go back from some stories i have read on the net
*if i go that route
[05:03:07] <Jymmmmmmm> http://www.practicalmachinist.com/cgi-bin/ubbcgi/ultimatebb.cgi?ubb=get_topic;f=1;t=022185;p=0
Jymmmmmm: you could try KBC
KBC tools.... short drive for you
where they at?
i foudn a cummins indutrial tools lathe earlier, on mysimon.com
they are sold on amazon
anyone heard of them??
hayward I think
chr0n1c: that's a 7x12 right? I have that one
it was 399... 465 shipped to my front door
or buy it off the truck
is it the same as the.. grixxly 7x12 i think it was?
ds3: A lil far for me,
think it has more accessories
but if I HAD to...
Jymmmmm: but you were considering newark close
ds3: are they like Western, everything behind the counter?
ds3: Because I work near fremont
Jymmmmm: yep. 1" thick catalog....there is a tiny tiny showroom with a few machines
ds3: do they carry thing in stock though?
not everything...but a lot of stuff
I know they don't have O2/W2 drill rod
sounds like HF retail
yes, kind of but much more machinist centric
they also have a close out pile there
we ahve a place.. dayton tool crib...
and dayton tool supply
and we can't afford that stuff, goin for the china ABBA rails and such =)
i'm in love with SOME of the stuf at harbor freight
99% is junk tho
HF sells 'raw' materials and kits
ok, i'm gonna try one more time tonight to give enco my money.. then i am giving up...
test indicator... and some mini carbide 2 fluters...
the tes indicator will fit right into the collet of my dremel ;)
i will be able to.. tram parts/the table/a vise.. the slides... everything on the machine that way
+add a surface plate to that just for fun
free shipping sale?
no i just need stuff...
all my good tools got left at my fathers shop when he fired me
i vowed to never work at another tool n die shop again
but i worked at a few food places and missed it
now i work for a reproduction car parts palce
and have my own mini mill
sooo.. i need tools again.. :(
i left abotu 10k worth of shit there
i was really pissed off at that man and never wanted to go back
ehh.. that was a long story...
but... i need all those tools again...
tmi tmi... i know
i'm rambling and not even drunk
there is a spider.. building a web in front of the infrared light of my front door camera..
it's a little crazy watching it on the monitor
hey so my dad's shop is out of biddness since i left
so does that mean when you ARE drunk, dyou dont ramble? lol
Besides, it's all good =)
i'm working on registering a general cnc chat room on this server
ds3: thanks for letting me know about KBC, too bad they don't have crap.
i started it # cnc
i jsut gotta register it i think
* Jymmmmmmm really misses a GOOD hardware store... I can't even find metric set screws locally.
mr metric in san jose
for small quantities
not what it used to be, but still not bad
petev: delacruz area?
they may have moved, but they used to be off of 101 around 13th
ah 10th st, thanks
oh wow... looks like freenode needs real legal entities to register a channel?
is this right?
or maybe i should read the website more when i'm not half asleep
#CNC MUST LIVE ON!
... anyways i registered ##cnc
if anyone wants to help keep that room alive
i haven't figure out how to make it permanant yet...
the channel #cnc was registerd two years ago
oh this machine has been running all day, I should get the temp of the north/south bridges real quick
befoore I forget again!!!
got a laser temp probe?
No, I bought one of these, works great http://www.harborfreight.com/cpi/ctaf/displayitem.taf?Itemnumber=91778
about a couple months ago, for $40
lol, then the answer would have been yes to my q ;)
Well, it's non contact IR, with a laser pointer
yes, we had one for work...
to check spindle temps and stuff
Yeah, I bought it to check the temp of my steppers - works great, especially for $40
$60, I probably wouldn't have bought it. Not THAT much need for it, but it does come in handy, and fairly acurate too.
how hot do they run?
mine stay at about 70 when it's been on for about an hour
Oh, dont recall. havne't run them hard in a while. Stil setting up emc
you can touch em but not too hot
Well, I have been having so many issues for months, that I wanted to eliminated the possiblity of thermal issues
i would like to use one to check the temp of my motor wires
it's only cat5
and 1 is an old playstationcontroller cable with the ends cut off
any idea of what's safe for a stepper motor?
I take it the magnets don't like too high temperatures
lerneaen_hydra: I think many have temp specs
under Absolute Maximum Ratings:
hmm so pretty fscking hot
I had to turn off JS to get that link, dman frames!
i.e. if you can hold your hand on it for over 3 seconds it's safe
but those are the AMXIUMUM ratings
hell even one second should be enough
usdigital sells the cables seperately - bastards
lerman_ is now known as lerman
lerman_ is now known as lerman
lerman_ is now known as lerman
lerman_ is now known as lerman
cold and drizzly
sunny here :)
great day .. 26 degrees, clear blue sky, spring is sprung!
we had that last week
today its about 8C, gray, and wet
bought an great new welding torch this week, ran 300 amps into a 300mm long weld on 15mm plate ...
and the torch tip is cool enough to hold in your hand
water cooled :)
i just got back from mendelsons
i bought 2 steppers...
one is bigger than i can mount on my mini mill... but i can make some kinda adapter
then it's not too big to mount on your mini-mill!
i got a slo-syn dual shaft.. nema-something er other
dude i have no idea what any of that means
mount the mill on the motor instead
the plot thickens!
i don't know the frame size.. 450 oz. in holding tourque
it weighs like 10#'s
slo-syn stuff tends to be a bit ancient
yeah.. they said there was more of them in boxes at the top of one of the shelfs...
Ive got nema 34, dual stack motors that develop 720 oz ... and do it at speed too
(i got this one for 15 bux) but he said the others wold be 150 bux cuz they had the box and the paperwork with them
150 woudl be too expensive for a 450 oz motor, esp a slow one
15 is a good deal though
i kinda gave them them i'm not gonna buy this unless you give me a great deal look when he said 140
then he was like "i guess i can let that go for 15.. then winked..
i was like uh
how many amps?
2.21 volts 7.0 amps
i figure the volts wont matter so much when i hook up my chopper drive
serously .. a modern motor will be smaller, faster, produce twice that torque and use half that current
yeah... i went down there
go for about 60 volts of drive on that motor
they have a whole-store 20% of sale
i got another tiny dual shaft.. 17 or 23 frame maybe? for a spare for the ones i have now
i think we figured i had 23 frame
is there a nema stepper frame chart somewhere?
60 volts should be ideal on that motor, assuming you have a Geckodrive
i'm building a hobbycnc 4aupc
chr0n1c: I paid about $110 for a 990 oz-in NEMA34, brand new, modern high speed design
i can go to 42 volts
don't want to pay even 100 for a slo-syn
15 bux ;)
15 was OK
but ti will be slow
thanks for the heads up
just don't go back and buy the other ones
i seen a few of these blue-slo syns on ebay for more than 15 so i was ok with it
I love the Sanyo-Denki ones
$15 was a great deal for messing around with, and if you don't need high speed, that will be a fine motor
i am prolly gonna have to mount this to the side of my slide.. then gear it to the ballscrew
[17:11:22] <jmk-solo> http://www.kelinginc.net/SMotorstock.html
the others he has on a shelf for $150 in boxes have probably been there for 30 years, and I suspect at $150 should be left there for another 30 years
there is no way this will hang off the end of that slide without tipping my whole mill over lol
I wouldnt gear it much
keling has modern 400 and 465 oz-in motors for $69-79
i saw some there...
i was gonna buy some from keling. but mendelsons is so much fun!
and i wanted to see if they had a good stash
keling also has NEMA23 (much better fit for your machine) with 425 oz-in for $54
i couldn't leave without buying something
i coulda built a whole new machine standing there in the aisles with all the junk they have
number 3 on that HTML page, single shaft is the baby
they had a boatload of tiny steppers...
would be nice for a probe machine or something of the likes
for like 3 n 4 bux a pop
avoid the old stuff though
wow 906 oz in
do motors go bad
or just weak?
like the 30 yr old steppers they haveon the shelf..
designs have imoproved a LOT
new magent technology etc
lower inductance etc
hey! check this out
i got a job for the mini mill engraving some flasks for a wedding comin up soon
it's awesome being able to use this beast to make money
i stole the job from a real engraving place
cuz.. megan loves me!
robin_sz: I got the dual shaft version of #3
they are nice
ahh... i think i was looking at the #9's maybe it was
I've had them up to 3000 rpm on the bench with a gecko and 50ish volts
someone reccomended them to me
i had a dream that i built a machine with geko's
and... it was fcuking awesome...
but.. only a dream so far
dremaing of building machines.. how sad
usually don't people dream about gettin laid?
[17:25:01] <chr0n1c> http://cgi.ebay.com/Superior-Electric-SLO-SYN-Stepping-Motor-M093-FD-8204E_W0QQitemZ160111708781QQihZ006QQcategoryZ58145QQrdZ1QQssPageNameZWD1VQQcmdZViewItem
that's what i got for 15 bux
that is even mendlesons selling them
ii bet this one was only out of the box for them to take a pic for that listing
..and i just found the schematic for it
schemtaics for motors are usually a bit simple :)
uhh... i'm fine with two wires maybe three
anymore than that i get cornfuzed
hmm.. can i hook up a motor with more amps than my driver card can output?
card 3a motor 7 a
and would i need resistors to change the 5 volts to 2.27 volts.. or can i just do 5
it's not a chopper driver...
motor drawing more amps than drive is a bad combo usually
NOT a chopper driver?
i been trying to find some info on google
then i got sidetracked into pics of other machines again
eric_u, no, you are confused .. the drive is a 3 amp drive, it will only supply 3A to a drive, the drive is CAPABLE of 7A, it will only get fed 3 from the drive, so it will be weak
what sort of drive is it?
let me guess
not worth 10c
i was thinking the drive should only supply 3 amps
you havent got a drive
lol, it's been working fine on the mini mill
sorry, but you are wasting your time with that
i got the 4aupc chopper almost done
then i'll replace the sp3 drive with it
and use the sp3 for a fancy door lock or something else
someone should sue the stepperworld guy for trade descriptions and misleading adverstising
who is laughing?
it does work flawlessly
but... not for actual work
just playing around mayube
you should try a stepper drive
even a L297 one (which are generally considered to be ancient) will be light years ahead of that
yeah i got a hobby cnc baord
but.. i have to finish soldering it
buy a gecko :)
bah.. i don't need pretty
i jsut need it to work
the hobby cnc one will do 7A at 70V?
you need that for the motor you just bought
it will do 42v and 3A
well, there you are then
more amps would be fun...
it will be safe to use with that motor
just slow and weak
after i make some money with this machi8ne i might buy some geko's
you are stundent or something?
* chr0n1c vows to not give up till the machine makes him money
no.. but i got bills and live on my own
i'm a student of life
how many hobby cnc drives did you buy? 3?
it's one.. 4 axis drive
$69 diy solder
cheap enough I guess
i think it was 69...
unless of course it smokes on first volts :)
at which point a G201 for $114 begins to look good value :)
Where's everyone hiding?
It's Saturday, and past noon... So, who has a hangover???? LOL
<- doesn't drink... except a bottle of tequila on my birthday every year
and that bubbly stuff at ney years mabye
what ya up to today jymm?
Just waiting for the guys, wanted to find out if there's anything I can do within emc about my steppers stalling
I know it's a tuning thing, but also wanted to run an idea by them I had.
my motors were stalling.. i was trying to push them too hard
the max_velocity bits helped... tweaking them way down from default
Well, here's the thing... under TCNC I was able to run tem at 75, but I can't even do 60 under emc.
could it be because of the linux overhead
even though it's realtime?
But even beyond that, I am more concerned about accuracy than speed
they explained to me that the motors are only rated at so much tourque
and trying to push them too fast at the beginning was holding mine back
ramping up the speed
I know there's a "tuning" aspect for my machine, but I thought that if I add encoders that maybe there could be a way to automate the tuning process.
you mean acceleration?
that's over my head for now...
yeah velocity and acceleration
Yeah, the accelration might be high on mine. Reading thru the user manual now to see if can get some insight.
fellas-- is there any reason for ismod.modutils in the top dir to be marked "broken". actually a link is broken i believe.
lerman_ is now known as lerman
sooo.. i'm engraving on the radius of a loaded bic lighter right now...
hope it's trammed in right
i gues i coulda used an empty one to test with
where's the fun in that?
lol.. my buddies were here.. they decided to leave before i cut it
pu$$ies i guess...
my wife is a bench jeweler - uses a small torch all the time
It's always good to know people that are smarter than you are
her co-worker managed to hit a bic with the torch flame
she wasn't seriously hurt, but she did need to grow new eyebrows and eyelashes
i tossed one in a campfire once... BAD IDEA THERE!
ok, the lighter is finished... i'll clean it up and post pics...
the nascar on the radius....
Jymmmmmmm: no, there's no way to automate tuning
anonimasu, thats not exactly true, but in emc there is no way (currently) to automate tuning
robin_sz: stop being a smartass.
ok, why the insult?
smartass is a compliment in some countries
several commercial controllers DO automate the tuning, eg my Baldor nextmove stuff
and, anything they can do, we can do better :)
robin_sz: I guess im just tired.
well, just chill dude :)
trying to.. my head's stuck on a stupid algorithm
I'll grant you without encoders theres no way, but as Jymmmmmmm said "I thought that if I add encoders that maybe there could be a way to automate the tuning process." .. hes right, there are ways to do it.
pid tuning is not a simple thing to implement..
the baldor doe the tiniest of jogs and seems to be able to get it pretty damn close to perfect .. it really is a microscopic jog
takes about 10 seconds usually
were we have had it fail is when using a drive train with some backlash, and linear encoders on the bed
we cured that though ... by eliminating the backlash ...
which was easy, we just got rid of the gearbox and the rack and pinion ;)
actually, we got rid of the servo motors too ...
* robin_sz nods
normal people would buy linear motors ... thats way too easy :)
* robin_sz points to the big box of neo. magnets
what price are they at?
oh, you have to do a sum to find out
magnets.price + copperWire.price + pottingCompound.price + tube.price
mmm industrial automation linear motors
I have one at work that I never managed to get running
we buil an x/y pick n place arm with some shiny silver things... it had an air gripper on the end of the y
they were long and flat linear motors
if you build the motors yourself, they work out cheaper than rack and pinion
sounds liek afun project
are they stepper like?
you just light up different coils in sequence?
3 phase coils, yes
on a set of coils around a tube packed with neodymium magnets
you do need linear encoders to go with it though
i have a linear bolted down inside my x slide on the mini mill... i'd have to take it all apart to get to it, so i don't know what type it is yet
about 12" long, and has a gold alum case.. like a plain old digital readout encoder
but i can't see much of it
usually cpacative pickups
I forgot about the potential for backlash, and encoders wouldn't pick that up.
on servos with a direct attached encoder, it doesnt matter,
I got steppers
mount encoders directly on the back of the stepper
That's what I was thinkg, but that wouldn't account for backlash in the auto-tuning aspect.
backlash is seperate from tuning the motors
one is mechanical and the other is electrical
Sorry, when I say "auto tuning" I mena even measure ay backlash and accoutn for that as well.
since it is an entry in the ini file
got a dial indicator?
or even a pair of calipers properly clamped
to the table and the z
Right, but I'm talking auto tuning, where you run a script, it determines max velocity, acceleration, backlash, everything related to your machine.
emc is hardcore we don't need that fluff.. lol
but i see where you are going...
but my machine aint
Jymmmmmmm: you are trying to put an engineer in a program
the real world doesn't work that way
jmk-solo: Not really... just have a script "push the envelope" back off, and see what variables are "best"
maybe it takes 30 minutes to run
and what does that script use to observe the machine?
how does the script know when its settings are overheating the motors, or shaking the machine to pieces, or....
If you eliminat the backlash aspect for a moment, toss encoders on the back of the steppers
there is no substitute for an intelligent human!
waay to much wiring there
jmk-solo, whilst I agree that you cant replace an engineer with softeware, there are controls that do a damn good job of getting basic PID tunings automagically
heat sensors current sensors... encoders...
jmk-solo, I agree on the no substitute bit though
Jymmmmmmm: anybody willing to put so many sensors on a stepper machine to avoid reading some datasheets and doing some math and test, deserves the results they get from their wasted money
* robin_sz nods
so many sensors? no, just encoders
for steppers, its not worht it
for servos, well you have the encoders anyway
Jymmmmmmm: why not tune it by hand/ear?
if you are using steppers, its because you can't afford encoders, or don't have the hardware to read them, or...
my machine humms liek a bassline from a roland tb303.. it sounds awesome
anonimasu, it can be hard for newbies, you cant impulse test it like you can a servo drive
Jymmmmmmm: for accel I set it to a value that sounds good and that I know gives a sane start/stop distance..
anonimasu: ok, so when axis.ngc stalled my steppers, why?
then I increase my velocity, to where they start stalling
Jymmmmmmm: one of three things:
Jymmmmmmm.. it was moving them to fast..
decrease it a good bit..
you tried to go to fast
to many steps/second
you tried to accel too fast
and then increase the accel.. until it accelerates too fast..
you had a disturbance in your step waveforms
and slow it down untl I have some margins..
or your voltage dropped..
i know my machine will stall out when i push it past 10ipm
jmk-solo: Right, so that's where the auto-tine comes into place, so instead of guessing what the problem moght be by trial and error, use things within emc (halscope?) to look
so i have max set at 9 ipm and i have a little bit of % over that in the over-ride
Jymmmmmmm: it's a 30 minutes thing
Jymmmmmmm: it's not rocket science..
anonimasu: No, it's not. For me it's been 18 month thing.
I think if you have experience of stepper systems, these things are so obvious, you can look at a macine make one move and go "ooh, thats accelling way too [fast|slow]" ... and its hard to put in words that feeling ... its very much a seat of the pants thing I guess
there is only so many things that could be wrong
check them off the list one bby one...
robin_sz: well, my stepper goes "whirr" when I go too fast..
now I'm running twice as fast and no stalling
before you got real motors?
I didnt yet..
robin_sz: the problem was the rectifier remember?
I've got 6 damn big caps on there now..
there aint no replacement fo displacement
what do you mean?
and im running a battery charger in series.. to get 65V
"there aint no replacement for displacement" .. its a motorsport thing
yeah.. though I wondered where that came from ;)
same as using big caps
Jymmmmmmm: when there are three things that can be causing a problem, you divide and conquer
women, money and work?
money & women then work..
Jymmmmmmm: set accel low, and gradually increase speed, you'll eventually find a speed that causes problems
then back off (maybe 20%)
and gradually increase accel
you'll eventually find an accel that causes problems
then back off (maybe 20%)
jmk-solo: I can do that, the problem is that the stalling isn't always repeatable.
what are you using as a test?
right now, just axis.ngc
not g-code I hope
running a program is not "divide and conquer"
its "run the whole bleeding mess at one time and try to guess what is wrong"
you want a test program? here:
jog it round the block a few times... at different speed
jmkasunich: I dont have problems jogging the machine aroun, just when I run a program.
repeat 50 times
Jymmmmmmm: bolt your machine base to a steel frame?
well then the feedrate in your program is set higher than you machine should run...?
Jymmmmmmm: could it be deflection or something like that?
chr0n1c: no, his machine is misconfigured
anonimasu: sindle is off, and not touching anything at all.
if the limits in the config are set right, then he could command F100000 in the program and the machine would run at the machine limtis
and if the machine limits are right, it won't lose steps at those limits
personally don't even use EMC at that stage
put a square wave from siggen into stepgen
start with a 0.1Hz 1.0 amplitude
in hal scope?
you don't need halscope
with steppers, nothing inside the computer (including halscope) can see what is really happening
you gotta use external sensors, also known as eyes and ears (and eventually a dial indicator)
set stepgen maxvel to say 0.5 (30 ipm) and stepgen maxaccel to 0.5 as well (1 second to reach full speed and one second to stop
i ordered a test indicator from enco last night! and a surface plate and some 1/8 carbide 2 fluters for aluminum
the 0.1 Hz square wave will move 1 inch every 5 seconds, first one way, then back
i'm most excited abotu the indicator though
One of the reasons I thought about encoders is because I ran a job, then ran a second pass. On the 2nd pass it skipped about 0.010" and ruined the piece. I never caught it till the job was completed. I'd like some way to have known that it had missed steps.
I'll admit, sure I could have my config mucked up, but some kind of feedback would be nice.
Maybe I do need servos or whatever, but I can keep tossing money to fix the problem only so many times.
call a blador rep
they will sell you some fancy stuff
with all the tuning built in
and shiny covers...
we used it for machines... industrial automation
i' wasn't tryin to be funny
in all seriousness. that stuff just gets plugged in and does the job
but costs thousands of dollars
* robin_sz nods
well, if it's $2k it's still a cheap thing
compared to a cnc control..
but it is lovely if you can't figure out how to use emc
depends on whatyou have more of, money or (time + clue)
last machine i ran for a shop was a 6,000 dollar knee mill with 25,000 dollars worth of controls
& motors i 'spose
actually, that might be (time * clue)
chr0n1c: but it made money right?
I think I need a better understanding of acceleration
oh yeah.. paid for 4 people's homes/familys food
Jymmmmmmm: what do you need to understand about acceleration?
inches per second per second
jmk-solo: I can't put my finger on it, but I think my steppers/machine has a issue with it.
if your max velocity is 2 inches per second, and your max accel is 4 inches per second per second, it will take you 0.5 seconds to reach top speed
if your drive/motor/machine combo can't do 4 inches per second per second, you will lose steps
the ONLY way to figure out what your machine can do is to set up a SIMPLE test, without a lot of variables, and try it
simple means one axis at a time
that's how i tweaked my ini file...
simple means don't waste time with G1 moves, you only care about moves at the limit
x first.. then y.. then z.. until each was working 100%
if you want you can use a dial indicator..
to make sure its coming back to the same place every time
to see when you start missing steps..
make a rapid move away from it
and g1 slowly back..
at a sane speed..
if you got balls you can rapid back.. lol
Guys, when it DOES skip/stall, it's not always repeatable. sometimes it's 90 minutes into a program.
Jymmmmmmm: do can you isolate the moves that lead up to it?
that part of the program tells it to move faster than possible without losing steps then?
does it happen at the same place in the same program, or totally random?
anonimasu: No, it's not evne in the same location.
does your motors/drives heat up?
No, I measured them
if you lower all of the limits in the ini file by 20%, does it still happen?
did you get the problem with x crashing fixed?
jmk-solo: That I don't know. I can't repeat the skip/stall each and every time. I fI could then it be musch easier ti resolve.
Jymmmmmmm: does anything heat up?
does it happen more if you RAISE all the limits in the ini file by 20%?
makybe the cut is too deep.. something that simple
take smaller passes
see if it still does the same
anonimasu: : No, I have an non-contact IR thermometer, and have been checking for that all along.
well, I'd start by cutting air, so you can run the program 50 times if you need to without wasting material
the motors, the driver board, the computer
jmk-solo: I have been cutting air axis.ngc and it stlled yesterday
how many times did you cut it and how many times did it stall?
I can HEAR when it stalls with spindle off.
Jymmmmmmm: can you put a dial indicator on/below a rail beam
5 times and stlled 3 times in different spots
ok, you are doing some smart things: turning off the spindle lets you rule out electrical noise from the spindle, and it lets you better hear the problem
anonimasu: if its happening at random places in a large program, an indicator won't help
fwiw, my stepper cables are shielded
especially if he can hear it when it happens anyway
jmk-solo: I'd still want to see if there are any flex or anything
At SWP suggestion, I turned down the amperage too, that helped.
anonimasu: he's having enough trouble staying focused on one thing, don't introduce more ;-)
Jymmmmmmm: that was days (or weeks) ago IIRC?
jmk-solo: I'll save it for later is this dosent work
how fast are you running it?
(the current change)
Jymmmmmmm: sorry I really hope you get your machine running
I seem to recall you having fits in the last few days trying to get a computer that runs RTAI with good latency
EVERYTHING before that is total crap and can't be trusted
step by step
first step is a computer with decent (and more important KNOWN latency)
if you KNOW the latency is 30uS, even though that is not great, its better than THINKING its 15uS when actually its sometimes 20uS
next step is base period and stepgen timing parameters, to make a step waveform that will NEVER violate your drive's timing requirements
jmk-solo: Yeah, and I switched to a completely different computer all together.
ok, forget everything you did before the switch
you have a computer
you've beat it up with latency testing
you have a number, that is the worst latency you've EVER seen on that computer
you have stepper drives, xylotex I think?
even did your spreadsheet thing
where are the driver timing specs?
* jmk-solo searches cxylotex.com
A = step setup time = 0.2uS
A = dir setup time = 0.2uS
B = dir hold time = 0.2uS
C = step length = 2uS
Required Step Length2
Required Step Space1
Required Direction Setup0.2
Required Direction Hold0.2
Worst Case Latency15.12
D = step space = 1 uS
ok, what did you set your base period to?
I'd probably try 20000, some margin never hurts
you are running 2.1.4, right? no old stuff?
I have no sluggishness, even when rotating the axis display
is it easy to disconnect your motors from the leadscrews? or a pain in the rear?
I found the timing issues that resulted in me making that spreadsheet by running just a motor and hearing some funny noises
if the motor is running at a constant speed and the driver misses a step, you can hear it
If I run the motors w/o anything connected to them, they stall instantly
I have to hold the motor in my hand to prevent them from stalling with nothing connected.
at what speed to they stall? at what accel rate do they stall?
if the motor is sitting on the bench and I say go, they stall.
"go" at what speed and with what accel?
the equiv of 40ipm
and with what accel?
accel dont recall.
when did you do this test?
days ago, weeks ago, months ago?
I am NOT pushing the accel limit. when I first got them, and when K2 replaced my machine.
the machien mfg
days ago, weeks ago, months ago?
you changed your computer last week
step by step... how many times do I have to repeat.... you build bigger things on foundations that you know are solid
motors that stall on the bench are not solid
computers that don't pass RTAI are not solid
so far, the only thing you know is solid is the latency numbers that you figured out eariler this week, for this one PC
I spoke with Mariss months ago on this, and he said it wasn't uncommon for that t occure
you are microstepping the xylotex, right?
ok, thats the same as I do on my stepper testbed (which is unloaded steppers)
I've never had low speed stalls on them
high speed sure, when I run out of torque/speed curve
but thats what the limts are for, so you never run out when making a part
Let me pause you just a moment....
Would the stalling/skipping steps be resolved by using geckos instead of xylotex?
impossible to say
a more accurate statement would be "a given set of motors is likely to have a higher usable top speed with geckos than with xylotex"
but if the stalling is being caused by bad step waveforms (for example), gecko is MORE sensitive to that, and might be worse
iirc geckos have circuitry in them to prevent midband
right, but you don't KNOW that your problem is midband
it could be anything
* chr0n1c jsut took apart an old cash resister that a restuarant threw away and found $1.05 inside it
chr0n1c: keep looking and you'll find $100 bill
that was a good dumpster dive i don't care what anyone says .. lol
i find cash in dumpsters all the time...
(i get bored at 3 am sometimes)
and good parts for my cnc machines!
JymmmEMC: are you starting to get why I keep harping on step by step? if you buy new drives because geckos have midband comp, and midband comp isn't your problem, you are wasting money. If you buy new drives because you KNOW from doing a specific test that midband comp is your problem, then you win
I really do understand. The biggest issue I have right now is that I can't repeat the problem successfully, so I can't determine what/wher the problem(s) are.
a chip on the ballscrew?
you said you ran axis.ngc 5 times and had 3 failures....
that is a semi repeatable test - it may take hours, but its a test
the goal is to find a better test
better = one that shows the problem sooner
if its accel or speed, then you want a program that does a bunch of G0... G1 at some slow rate will never fail, and just wastes time
But the thing is that I have nothing to base thise failures againest.... if the motors over heat on the last three test, sure. (as example)
you don't KNOW if its accel or speed, but you can write a test program that does lots of G0 and run it for a few hours
well, do several thousand g0 moves..
if no failures, then maybe its not accel or speed
like a square?
a big 12" square
which axis loses steps?
mostly X, sometimes Y
more like a single axis forward stop forward stop forward stop..
if its mosty X< then run mostly X
and jog back to your dial indicator..
rapid to it and jog slowly the last 5mm..
anonimasu: I'm not sure repeatability is a factor at this point.
use basic or python or shell or whatever to make a file with 3000 lines of
JymmmEMC: it's related.
anonimasu: Right, but down the road a bit
JymmmEMC: we're not measuring repeatability, we're trying to detect a lost step
if you can hear it, then you don't need the indicator.
jmk-solo: then I need to grab a digital indicator or switch and feed tha tback into emc and marka a flag somehow
the problem I have with even the G0 X0 G0 X3 approach is that you already have many many things in the system
do it simple
if you can hear it, then you don't need the indicator.
you said you can hear it
also the idea is to make your drives heat up.. if they do..
jmk- but not for hours on end.
lots of accel/decel should make them heat up faster..
dinner is ready.... bbiab
and rapid should make them heat up faster than G1 moves in axis.ngc
in his shoes I would start with siggen driving stepgen
and not move beyond that until it can run back and forth (and later in circles with sin/cosine) for hours at a time with no lost steps
jmk-solo: you probably should try to guide him how to do that
I'm not sure I have the patience :-(
and uncouple the motor even if it is difficult
if you have time and patience
2 things, the noise & he can stall it w/o load ( weird symptoms )
we don't even know if he has the motor current set right
I really find it odd that xylotex 8x microstepping stalls unloaded, unless he's simply running out of torque/speed
lerman_ is now known as lerman
does the amp get hot ( we asked about motor )
the xylotex drives have a pot for motor current - set it too high, and you surely will get non-linearity which leads to roughness, resonance, and stalls
I don't know the current he is running at
xylotex recommends fans over a certain current, don't recall what that is
duh, I still have the xylotex manual open in a browser...
* jmk-solo checks
pot varies available current, maxV 35=, minV 8V= , seems 'typical' is 24V, (how did he adjust the Vref? ), how well did it pass "setup & operation" ?, "Use of a cooling fan is recommended for systems operating at or near the maximum current rating "
use the "setup & operation" with just hal supplying step (siggen -> stepgen -> parport), maybe just a switch supplying 'dir'
does he mean what we think when he says 'stalls', becuz the docs talk about 'locking up' in "setup & operation"
locking up just means that there is current in the motors, and you can no longer spin them easily
stalls means it was moving and then it stops when its not supposed to
yes i understand, my question was what is he seeing (is what he says a good description)
you'll have to ask him
I don't think he's confusing "lock-up" with "stall"
bbl thanks all
I just found a post on the Xylotex list where he (jeff @ xylotex) recommends a fan for anything above Vref = 2.0V, which is less than 1.5 amps
I wonder what current JymmmEMC is running at
I need food, back later
I have 6 fans blowing on the xylotex board
The current is at 2.5a
@ ? voltage?
30, which I had a long discussion with Jeff @ xylotex about and he suggested it.
so your up near the upper limits on both V A
should run fast
no, I turned down the amperage from 3 to 2.5
was at three till SWP suggestion
he explained it to me, but I forget his explanation
it had todo with peaking iirc
thought 3A was over max for a Xylotec
not over, just the max
I can't prove this, but I think these steppers have an issue with acceleration.
can't say you are not getting cooling
yeah, and loud as hell too
* Skullworks_PGAB giggles
Well before I bought the xylotex, I kept reading about the over heating, so I made sure that wasn't an issue.
what steppers you using?
[22:55:25] <Jymmmmmmm> http://www.xylotex.com/StepperMotor.htm
part of the package deal?
3axis board, 3 steppers, and ps
I wish I could send one of these motors to Mariss to run on his machine,
but i dont think he'd do it for free
SWP is right you don't want to over current the motors - just makes more heat and in some cases reduce the magnets strenght
No, he had another explanation that made a lot of sense.
something along the lines of the magnetic field taking too long to collapse
sounds likme they jusyt dont have enough "oomph"
* Skullworks_PGAB yells Henry
they go, just dont like to stop
those are a good sized Nema 23
Jymmmmmmm: why does it with for other people with the same setup?
anonimasu: say what? lol
anonimasu, you think anyone lese bought one?
I have 280 oz-in and 425 oz-in
but I run them uni-polar which drops them down to 200 and 300 respectively.
anonimasu: dont mind robin_sz, he's just being an ass
personally, if I bought a supposedly working machine froma manufacturer, and it didnt work, well, I;d want money back
i know i want my money back!
sounds like the motor/drive combo simply doesnt have enough torque to cope with the friction losses inthe machine
what kind of leadscrew?
Jymmmmmmm: are you running geared/ungeared?
5tpi ballscrew on 2: gear
up or down reduction?
2 in 1 out right?
equte to 10tpi
Z is 10tpi acme direct driver, never have problems with it.
chips in the threads?
no metal cutting at all
wood or plastic
wood can still clog a ballnut
nothing in there, already took apart the ballnut
mine are smooth... i can spin it by hand with the motor off.. effortlessly.. even with a table and stuff on it
to give us an idea, if you have the motor running at mid speed, and grab the shaft with two fingers and squeeze real hard, can you stop it?
so it has at least some torque in reserve then
same at slow speed? - its strong?
do you talk mean to the machine?
i mean do you cuss at it?
they have feeling.. tel it you are sorry for the mean things you said
I keep my grinder nearby..
it keeps the machine in working order.
i really thought i could eat all 20 nuggets that i just got from mcd's.. but i'm stick at like 13
* Skullworks_PGAB sets a can of clover within sight of the machine...
if you have thermite that works too
chronic - you finish the HCNC board?
i my current setup goes over 10 ipm... the tourque drops instantly
no.. i am going to build a bigger machine and put that on it.. so i stalled in finishing it
i started collecting the parts for a bigger 24"x48" machine
i can't take mini mill much longer
setups are a pita.. and not much cutting room
i'm leaving the mini mill like it is
damn - there was a link I wanted to show you.
it will be soon
when i get the new machine built...
Nema 23 280oz-in driving a router at like 500 IMP
* chr0n1c closes eyes and prays about the new bigger machine
i'd like to check it out
OK - I will search for it.
"search" sounds so labourious...
i feel bad for asking now
well - I had it marked... then forgot where I put it...
what would ..."you" want to see engraved on a bic lighter.. i mean since i got the setup for it already...
The surgen general...
they already have that.. ona sticker
like the mudflap girl!
i lvoe her
"Not Approved for Airline use."
"do not dispose in fire"
"i didn't burn that building down, i swear"
"tear on dotted line"
* Skullworks_PGAB gets to run the Airport security gaunlet next week.
i'm only going .010" deep... i wonder how much plastic is left uinder the engraving
* skunkworks hopes your not milling into the plastic case..
sounds fun Skullworks_PGAB
yeah.. right onto the plastic lol
i know.. not too safe
atleast my local airport provides free 802.11b service...
but it seems like they made them "extra thick" for safety
but it should be OK, right?
so there is room for me to chop out .010
if i don't sell them to idiots...
im sure that wont create stress points. you'll be fine
The ultimate rubbing of two sticks + compresed butane.... gorra love it
i did my buddies lighter with nascar while he was here a few mins ago
* Skullworks_PGAB once made the mistake of Laser engraving a FULL can of Budweisher.
i got the program down to 2.5 mins
"i had it in my posket when ..."
+1 min setup
i put the nascar logo on it
had a mini beer fountain... in the office...
what could possibly go wrong ...
ssetting up the pic for the radius only takes a few mins now i got that down pat too
* skunkworks is starting to think chr0n1c isn't long for this world
Skullworks_PGAB: so, how long did it take to clean out all the beer?
i know i know.. but butane by it's self isn't gona explode..
* Skullworks_PGAB hands chronic a spare Halon cylinder
you;d have to explode the butane out.. and have a spark in your pocket
chr0n1c: your running a brushed dremel motor 7" above the lighter, right?
not long - the pin holes were only .002" and there were not too many
maybe it could explode in like a motorcycle crash a tad bit easier.. when the bike is skidding on top your leg... creating sparks after the lighter just smashed
soda/beer cans are just toooooo thin
so let me get this right.... you have a plastic cylinder containing compressed butane, that people hold in their hand, probaly close to their face, and in proximity to a naked flame. you plan on cutting away "some" of the sidewalls, and then selling it, to people in the only country with more than one persoanl injury litigation lawyer per 100 citizens ...
lerman_ is now known as lerman
i'm not selling them lol
and it's only .010... which is actually only .005 when i set it on zero with a pc of .005 paper in between the cutter and the lighter
i could sell them.. and feel ok about it
i mean there is a pretty good chance the lighter will blow up anyways.. without engraving
im sure the class action suit will agree with you
a special LLC company founded just only to sell engraved bics
chr0n1c: better to make inserts
ok fine i wont sell them
right bedtime for this girl
just give them to people i dont like
need my beauty sleep :)
gnight n stuff