logger_aj: you suck
I'm logging. I don't understand 'you suck', alex_joni. Try /msg logger_aj help
poor logger aj :/
oh, wrt to the "include an editor in axis" debate, would something like a button that launches <program> and loads <current nc file> be simple? and when you save & exit (or possibly just save) axis reloads the file?
i think the issue is that you can click on the line, it highlights, but you cant edit it
its kinda frustrating
anyone rememebr the url for that place that has the cnc video?
you dont need to open an editing program to edit one line of text
yeah those are pretty sexy
i think =)
I'd like to have a automated cycle button in axis(there might be one already)
for doing a run of 10 parts or so
anonimasu: you can program that using I-words
[08:46:25] <alex_joni> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Oword
oh, now that jepler wants last-minute add-ons to axis, how about an up/down arrow history thing for MDI mode? :D
I'd still like to be able to tell axis to run it 5 times..
anonimasu: do while #1 LT 5 ?
don't think that's too hard to do :)
not it's not but still it requires editing the file :)
wtf is meant by "thrust" force..
on the table..
the motor's maximum force in translating the table?
im calculating with a table weight of 100kg in motioneering..
with direct drive it ends up at 1,031 N-m
that seems really low..
maybe thats 1031 Nm ;)
somehow I doubt it
Gotta love it....
dslreports.com speed test 2006-10-26 05:38:39 EST:
50967 / 15978 (Kbps)
(6221.6 / 1950.4 KB/sec)
I want that upload speed! (and download speed too for that matter)
Note that is a CAPITAL B, not a lowercase b
oh, wrong test results, nm
dslreports.com speed test 2006-10-26 05:48:25 EST:
52859 / 10168 (Kbps)
(6452.6 / 1241.2 KB/sec)
I'd still want that upload :D
dslreports.com speed test 2006-10-26 05:49:12 EST:
33194 / 44011 (Kbps)
(4052 / 5372.5 KB/sec)
100KB/s is kind of sucky
still no call from ostergrens
* Lerneaen_Hydra pokes ostergrens
That's 5MByte/s if you noticed
seems like farnell is the best bet for a sutiable stepper
[09:50:37] <jymmmmm> http://www.dslreports.com/im/17685577/3534.png
* anonimasu sighs
farnell is way expensive :)
they are by far cheaper then any of the reatilers I have..
actually I only got some chips from them
my normal supplier sells steppers at 360eur..
LS7266 iirc, at about 10-16 pounds
180 for 3.2
and they are < 5$ in the US
farnell's 5.2 is 200eur
seems like quite a bit :)
what current / voltage?
the ones from my retailer?
I have no idea..
[10:04:14] <anonimasu> http://se.farnell.com/jsp/endecaSearch/partDetail.jsp?SKU=9948147
4A current rating
that puts it within my PSU range too
hmm.. probably that picture is the wrong one
that doesn't look like a 5Nm stepper
although the measurements look a bit better :)
measuring from the size of my stepper it has 1,1nm
well, you can always easily measure it
why's stuff so expensive in sweden?
shipping cant be that bad
are there import tariffs or something?
fenn: "handling" fees
and.. it'd probably cost about the same in the US
brand new from a place like farnell
fenn: I wanted to get a camera backpack
170$ in the US
330 EUR locally
even overnight shipping is cheaper than that :/
fenn: money is more expensive in sweden too (100kr for a swede is worth more than 100kr for an american). which is something that I noticed when I moved here.
Lerneaen_Hydra: that's true here too
yay.. really love listening to War of the Worlds
the radio version?
not the original though
live-cd creating is easy with uck
flo-h: just remember to start from the existing live cd with emc2 on it
yes, I used the emc2 live cd as basis, until now I changed the language and played a little with the package management, but it works
flo-h: when it's done I can mirror it
ok, thanks, but before i have to do more testing, how things work
flo-h: probably uck sets default LANG and other env variables properly
so emc2 should start with german interfaces when available :)
and if it doesn't it needs fixing
[12:52:51] <skunkworks> http://www.cnczone.com/forums/showpost.php?p=210520&postcount=51
boy his attitude sure has changed eh?
hmm.. the pminmo guy could probably pull it off to produce a decent stepper driver
alex_joni: - I have 2.6
all 8 disks
we used to run sbt accounting software which was based on foxpro.
torque was rpm/9.6*power, right (SI units)?
oh, wait. 9.6*rpm/power must be correct
where does that 9.6 come from?
breaking it down to base units, 1N is 1 kg * m / s^2
1W is 1 kg * m^2 / s^3
so if you multiply RPS (unit of 1/s) * Nm (1 kg *m^2 / s^2), you get watts
1W = 1N*m/s
power = torque * speed
i highly recommend the "units" program
heh - I have an HP calculator that does that for me ;)
i tried out qalculate today and wasnt impressed
i wish units could solve for an unknown
it looks like it could look nice (without the cursive an pink theme)
qalculate isnt strict enough with its units, and it applies prefixes so you always get a number like 50.1234
prefixes? do you mean it always says something like 50.1234 kN, vs. 50123.4 N
so for example "($2.29/gal)/(35mi/gal) to $/mi" gives me "approx. 65.428571($ / kmi)"
which is obviously not what i asked for
ah - ok
of course, it could say something like 6.54285714286E-2 instead ;)
[14:04:38] <jepler> http://blog.modernmechanix.com/2006/10/25/key-caller-speeds-phone-service/
I used to do things the other way - repeatedly clicking the hook switch will also dial
(much slower than the dial was ;) )
I couldn't reliably dial a number that way, though I seem to recall trying
it was definitely harder
the real pain was dialing my mother's number - it had a lot of high digits
9998754, in decreasing order
did she work for the phone company or something?
thats a strange telephone number
that was in decreasing magnitude order, not the correct phone number order
she's paranoid about getting phone calls from numbers she doesn't recognize, so I figured I shouldn't post her number on teh intarweb
and you didn't say the area code either
there was a photo of my credit card on my webserver for like 10 months and was never downloaded
yeah - Vermont is 802 ;) (and 0 = 10 clicks)
but there are only 840 orders for those digits, some of which must be illegal
then one day some guy randomly looking through my files, the first person to see this photo, says to a crowded irc channel "hey look his credit card number" and posts the url
I don't think there are any illegal numbers any more (some are reserved though)
heh - I'm sure that helped
I sure don't know the phone number rules
cant start with 1 or 9
cradek_ is now known as cradek
it used to be that any 3-digit sequence with 0 or 1 in the middle was an area code (at least, the first 3 digits dialed)
that's why you could dial long-distance within an area code without dialing the code, since no exchanges could have 0 or 1 in the middle
we should just switch to ipv6
call ben.lipkowitz.com for my work number
yes operator, I'd like to place a call to 0001:FFED:1254:4563:9935:ABF6
hmmm - maybe the 10-digit codes are better ;)
why should i have to call long-distance to my roommate's cellphone?
err - because the cell phone is from a long distance away?
so it might just work as-is?
03jepler 07HEAD * 10emc2/tests/save.0/ (README expected test.hal): test of 'halcmd save'
03jepler 07HEAD * 10emc2/src/hal/utils/halcmd.c: allow simulator to pass test save.0
03jepler 07HEAD * 10emc2/src/hal/utils/comp.g: as a step towards making 'comp --compile' work on sim and non-kbuild kernels, don't assume the extension for modules is always .ko
03jepler 07HEAD * 10emc2/src/hal/ (hal_lib.c hal_priv.h): allow things created by 'newinst' to have entries in 'show comp' and create lines in 'save'
03jepler 07HEAD * 10emc2/src/hal/utils/halcmd.c: allow things created by 'newinst' to have entries in 'show comp' and create lines in 'save'
03jepler 07HEAD * 10emc2/tests/save.1/ (README test.hal): test of 'newinst'
03jepler 07HEAD * 10emc2/tests/save.1/expected: whoops, expected output is reqired for this test
how come i get this when i come on,ERROR]You need to be an operator in #emc to do that.
I don't know, but I get the same thing (also using chatzilla)
I get that too (chatzilla)
I suppose someone could run a sniffer / IP traffic monitor and see what it's trying to do
it may be a default user mode setting that isn't allowed
i tried turning on an output on the motenc board using the reference given last night. newsig TRUE:setp TRUE 1:linksp TRUE motenc.0.out-02
sets not setp
could be why. get back in a few.
what command does the set?
sets TRUE 1
so to summarize:
newsig TRUE 1
newsig TRUE bit
sets TRUE 1
linksp TRUE motenc.0.out-02
got the signal , got the link , just couldn't get it set. i'll try the sets.
bbl.. running home
thank you alex out-2 is on-----finding these commands is like pulling hens teeth. anysource for the complete command structure.
halcmd help <command name>
emc2 documentation PDF
(need more? ;) )
i was using the manual setp was the only thing i found that set anything it set param i needed set a sig. i'll go back and see if i can find a sets.do you mean the manual on pdf.
wiki wiki woo
logic i should have realized. sep was for parameter . to set a signal would have been sets
halcmd -? should print a list of commands with a short synopsis of their functions
or "halcmd -h" - I don't recall
* anonimasu sighs
I've decided, I'm going to wind my own steppers.
and I'm going to do it using stolen HV cabling off the power company.
thats what I did.
except I rolled my own wire from solid copper sheets.
nope - I made the machine to do it :)
im going insane.
mtedad: man halcmd should also work
and the HAL documentation describes this in detail
is the HAL info in the doc that's in the menu?
some of it
or is it harder to find?
ok. it seems that several people are having a hard time finding docs
sure, but many EMC machines aren't net connected
right.. we probably will include most of the docs for 2.1
maybe more docs should be installed by the packages, with links in the menus
ok - good
[17:48:45] <alex_joni> http://www.linuxcnc.org/docs/HAL_Documentation.pdf
i'm working in hal show . command line does't recognize ?..-h -help HALCMD ?... HALCMD -h HALCMD help
tells me i'm in the wrong place.
halshow uses a different mode for halcmd. try help in the command line
I'm not sure what it will do with the output though
the commands we're talking about are meant to be used in a shell
should i be in terminal to do this?
it's not strictly necessary, but I think more developers can help you if you're working in an environment that's familiar to us :)
mtedad: usually we open a terminal to do this stuff
I personally haven't used halshow, because I'm comfortable with halcmd
SWPadnos: I think I opened it once or twice for watch
I looked at it, to see that the menu link from Axis worked, but that's the extent of it (recently)
SWPadnos: the user guide, which is (I think) all the docs put together, is on the menu
cradek: not quite all the docs, but probably all a user would want
ok. I waasn't sure if all the halcmd stuff is in there (or the hal tutorial, etc)
maybe some separate docs would be good - it
it's hard to search a 175-page document if you don't know what you're looking for
(ie, if you don't know that "halcmd" is the command to manipulate HAL ...)
it has a decent table of contents that is like hyperlinks in evince at least
mtedad, did you look at the EMC2 documentation from the menu, and if so, can you offer suggestions on how it might have been easier for you to find the information you're asking about here?
maybe the GUIs should have menu links for the documentation as well
it's all confusing. i have the docs printed and i agree it 's hard to find anything when yuo don't know what your looking for.
ah, good question
in fact, maybe that question should be asked on the user list
if you print them, you lose the usefulness of the hyperlinks.
ie, ask for user suggestions on the "system" rather than the "functions"
is the pdf that alex linked to the one that's included in the distribution?
no, that's only HAL
ah - duh ;)
the user manual is also in docs/
that one is included in a similar form (the version from the 2.0 branch)
this one? http://www.linuxcnc.org/EMC2_User_Manual.pdf
yeah, that one is the one included I think
and the linuxcnc.org/docs/EMC2_User_Manual.pdf is the one from HEAD
err.. this one: http://www.linuxcnc.org/docs/devel/EMC2_User_Manual.pdf
the manual in the devel dir, just downloaded thanx.
03jepler 07HEAD * 10emc2/src/emc/usr_intf/axis/scripts/axis.py:
if the file ~/.axisrc exists, it is treated as python source code and executed.
in this way, users can customize axis to some degree, for instance by making
calls to root_window.bind(). however, the details about what can be written in
~/.axisrc are subject to change without notice.
03jepler 07HEAD * 10emc2/docs/src/gui/axis.lyx: document axisrc
good night all
see you ale
is connecting the spindle load meter to emc and auto adjusting the feed override when roughing a bad idea?
anonimasu, no - I think that's a good idea
I'm going to connect a pressure sensor so I can see the load of my spindle.. anyway..
and since the plc will talk nml in maybe a week :)
a 4kw motor is way too heavy to fit on the mill :)
ok. just make sure you get the sign right (higher pressure = lower adaptive feed override ;) )
I'm going to connect my jogwheel.
and see if I can hook up the hal component for it
the question is now how i link the encoder of the ppmc to the jogwheel.. thingie..
*starts reading about hal
halcmd show pin | grep -i jog will tell you a lot
I dont have a jog thingie in my hal yet..
I think ;)
you probably need to know the pins to connect though, and that should help
I found the encoder :)
I dont have the "jog" hal stuff loaded
motion has some jogging inputs - those are the ones I don't remember the names of :)
if you run EMC2, you do have some jog stuff, I think
though it may only be in HEAD
I guess it's upgrade time..
I'm not sure - it was added around Fest time
stuff works so well I have to break it.
that's the spirit!
it's not in the motion in the "hal" configuration dialog
if it was added during fest, it's probably only in HEAD
2.0 was getting only bugfixes at that time
ok. I thikn it was in response to erratic jogging on the NIST-lathe, so it's probably in HEAD
hm, what packages do I need to compile head?
actually linking it to feed override would be fun for now
[20:49:59] <SWPadnos> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl/emcinfo.pl?Installing_EMC2
* anonimasu heads inside again
It's a good project for tomorrow afternoon
logger _ aj ...
one more time
log<tab> search ;)
Try this address for searching the logs: http://126.96.36.199/cgi-bin/search.cgi
commas, not colons
Try this address for searching the logs: http://188.8.131.52/cgi-bin/search.cgi
I need to find a bigger/better motor
for Z first..
that's my main priority
oh - just slow today ;)
maybe we'll get another one ;)
it has been a long day so far
it ain't over 'till it's over
I went and got new glasses today
there goes the servo for Z $
something like 400 eur..
if you can gear down the stepper more, as RObin suggested, you may end up not needing a servo
I tried spinning it today fast..
it went up to something like 2000rpm..
holy crap its nice having a search now - thanks alex
ok. that's plenty fast enough (and about as fast as the gecko can go)
well, half the speed the gecko can attain
or thereabouts ;)
it would go faster if you had a higher voltage supply
that moves the corner frequency towards higher speeds
I were looking at farnell's catalog
[21:05:54] <anonimasu> http://se.farnell.com/jsp/endecaSearch/partDetail.jsp?SKU=9948147
and im pondering buying that to replace the one I use now
SWPadnos: what do you think about htat?
it's pretty high torque compared to the one you have :)
also the lower voltage rating is good, you should get better speeds from it
I'm pondering if I could get 2m/min with it..
or perhaps a bit more
actually I need accel more then I need speed
what's your Z travel range?
something like 0.5m
ok - that's fairly long
SWPadnos: having the z almost as fast as x/y makes it nice when doing 3d work
that part I posted yesterday should have taken something like 6 minutes but due to the Z slowing down every other axis it took 40..
today ... I discovered my spot welder was faulty
was disicussing weld problems with a tech sales guy
2m/min at 2mm/rev is 1000rpm
"do you keep the tips well cleaned and dressed?
oh, when they need it ...
how often is that?
oh, every 50 welds or so
arrg ,,,, its broken ... it should last 5000 welds on a tip easily
turns out the water cooling system isnt getting the water right down to the tips
SWPadnos: is that a insane idea?
SWPadnos: the speeds/stuff with a stepper?
robin_z: or maybe you know
dismantled the weld heads and discovered the thin brass tubes stop like 50mm short of the tip .. they shoud go to within 5mm of the tip
steppers, yeah .. shoot
1000 rpm may be a bit high, from what Robin was saying yesterday, bnut it'll depend on your supply voltage
robin_z: my little stepper moves it today..
5 to 600rpm shoudl be a good aiming point
and it's very little..
oh it moves it very well, just slow
is it spinning real fast?
~3000 RPM unloaded, ~150 loaded, geckos + USC for steps
150 rpm is a bit slow
you mentioned yesterday that higher gearing would help, and (after clearing the brain fart I had) I agree :)
I presume it stalls at like 200rpm?
something like yes
that sounds very right
you gear up .. so now it needs to do say, 300rpm for the same machine speed it did at 150rpm before
you should be able to spin it to 600rpm and get twice the machine speed you did before
I dont have the torque to do that
yes you do
the gearing lowers the torque requirement from the motor
gear up, and yo only nbeed half the torwu right>
and getting it up past the corner frequency means you're getting full power all the time
so .. say the motor has its "corner" at 600rpm right?
you kow what I mean by corner?
where torque starts to drop?
where the power is constant
look, at zero rpm power output is?
let me find a graph
power at zero is always zero
but for the corner frequency
and at 600rpm you get full power
whatever that is
hm.. no in the datasheet..
so at 150rpm, you are getting 1/4 of the motors maximum power
just before it stalls
you cant use the full power of the motor
its like your car is locked in 4th gear
yeah, you dont have the torque to accelerate up to where you have the power band of the motor
or something like that
I kow it sounds backwards, but .. its right enough
[21:21:12] <SWPadnos> http://www.geckodrive.com/photos/Step_motor_basics.pdf
yeah, mariss explains it well
with graphs ;)
of course, thats for a reasonably modern moter
some of those old crap things had really low corner freq
how I hate this wireless router
swp = !wireless. I only use it when traveling, because most airports and coffee shops don't have ethernet jacks
it sounds so reverse
Do you guys think I should buy the farnell motor?
no - I think you should try gearing this one down first (assuming that gearing is much less expensive than the motor)
I'm gearing it down now
make it 6:1 or 10:1, and then see what happens
the motor is 210eur
what would a different gear cost?
and double the size of theese ones..
(or a pair)
I think I have stuff to gear it to 6:1
aren't the servos 6:1?
they were on the old machine with other screws
make it 6:1 and see what happens. you'll probably get at least 1.5x - 2x the top speed
remember, the motor can go 20x the speed you're starting to lose steps at
you mean right now?
yeah - when you remove it from the machine, you said you can get close to 3000 RPM
and you're starting to lose steps at 100 mm/min, which is 50 RPM at the screw, or 150 RPM at the motor
I'll try gearing it down tomorrow and see
ok - sounds like a plan ;)
if it dosent work I'll order the larger stepper and gear it suitably
that makes sense
as a servo costs 500eur minimum
I bet the bigger stepper would be better with the 6:1 gearing anyway
it's quite a margin..
heh - is that with or wothout the encoder?
that's the cheapest servo I can buy in the US ;)
the small ones are 400eur each..
SWPadnos: I have no idea really
ebay is your (or at least my) friend there. I paid $350 or so for 3 servos, shipped
27 in-lb continuous
my main issue is shipping
yeah - that's true
I'd need to ship them to a friend in the us that mails me them
well, ask away if there's anything you're interested in
it's the only reason I havent bought any servos off ebay
[21:38:00] <SWPadnos> http://cgi.ebay.com/New-API-Controls-Servo-Motor-for-Automation-CNC_W0QQitemZ170043424538QQihZ007QQcategoryZ78195QQrdZ1QQcmdZViewItem
that the 5.7Nm one?
that one would be perfect..
actually, it says nothing about AC vs DC, and it looks like a lot of pins for a DC servo
might be encoder and tach..
the right side connector has at least 4 pins, so I think it's an AC servo
[21:43:26] <SWPadnos> http://cgi.ebay.com/3-AXIS-DC-Servo-Motor-System-ametek-CNC-drive-mill-LQQK_W0QQitemZ200040465181QQihZ010QQcategoryZ100184QQrdZ1QQcmdZViewItem
but, the servos didnt have too great specs
950 oz-in is reasonable, even if it's peak
hm, larger then my servos..
remember that torque doesn't drop off for servos
I may have one of those motors around here actually
I tried counterbalancing my z axis
with some shocks off a car.. with 2 of them it got a bit easier to go upwards.. but not much..
not enough to cause it to go whirr..
well, I really hope you are right about the gearing :)
you probably are though
yep - I do have a motor that's nearly identical to those
no encoder though
do you want it?
got my stuff up and running.
if it'd be a good match for my Z :)
just need to measure the backlash and see what it does.
SWPadnos: it'd probably be perfect
though it's odd that they are so low rpm
but it might be the size :)
mine also say 20V, but there isn't much else on them ;)
they can probably be run at higher voltage though, as long as the current limit is correct
the back shaft is 5/8, so you'd need a relatively expensive encoder
or a lathe tool and a DC power supply ;)
that's kind of like machining ballscrews isnt it?
hmmm - not quite, I think :)
hehe the "if I screw up there's no way to repair" feeling
well, sure :)
those are BIG mofo motors. http://cgi.ebay.com/3-AXIS-DC-Servo-Motor-System-ametek-CNC-drive-mill-LQQK_W0QQitemZ200040465181QQihZ010QQcategoryZ100184QQrdZ1QQcmdZViewItem
nah - only about 6-8 pounds each :)
SWPadnos: if you would like a new home for it I'd be happy to take care of it
oh, thank you :)
ofcourse I'd buy it off you :)
suitable encoders, I think: http://www.usdigital.com/products/e6/
but I'd have to see if it'd fit for the z axis
im so crappy when it comes down to calcing load/force/stuff
* anonimasu nods
want some good advice?
never whistle with a mouth full of muesli
its OK to sneeze on Piza
but only your own
SWPadnos: reasonable price also
well, if you'd like to pay for the motor, I'll take $25 off your hands (plus shipping) :)
is spinach back on shelves yet?
want a pizza tonight.
craving cheese, feta, grilled chicken, and spinach.
damn GOOD pizza.
SWPadnos: that sounds like a great deal
especially from the place behind me.
I have another one that fits nicely in the shipping box, a 2HP treadmill motor
SWPadnos... how much for the motor?
for you, $50 :)
2HP... DC? or AC?
probably good for a spindle, come to think of it
SWPadnos, that's what I'm thinking.
for like my mill.
SWPadnos: it's a deal then unless you change your mind :) and _if_ it'll work for my z
use PWM to it.
SWPadnos... fifty, shipped? :D
well, I'd hate to pay for shipping both ways
it's not too expensive..
the stuff that ends up expensive is the customs..
A-L-P-H-A, no, but I'll deliver it if you're buying at Rodney's
SWPadnos, uh... delivered... to me free, and I'll exchange you for Rodney's... that's like $50-100CDN... unless you're a pig, and have a bottomless stomach.
and a wife
but she doesn't like Oysters, so ...
only if your wife is hot, and puts out for me. ;)
I am a pig with a bottomless stomach, by the way :)
then there's only crab and lobster.
those damn things are expensive there too.
oysters don't really fill me up... just makes me more hungry
heh - everything there is expensive
I could probably easily have like 24 of them.
hmm...24 x $4/each... urk.
SWPadnos: how much does one motor like that weigh?
give me some home fries with that to fill me up.
a 2hp motor? or the popbottle sized ones?
the popbottle sized one
damn they are large.
oh. geeze... umm.
probably a good 50lbs, so yeah... roughly 20kg.
one sec - phone
if not more.
on my motor, the body is about 100mm in diameter, the face is ~135mm square
length ~175mm, not including shafts
hm really large :)
and it weighs 5.5 kilos
yes, it's in the medium size range
[22:20:24] <anonimasu> http://ircalc.usps.gov/intl_speed.asp?CID=10411&MailType=package&Pounds=12&Ounces=1
for airmail parcel post
fair enough :)
I'll be back in a bit im going to have a hot bath
oh good :)
if you do want to get it, we can order geckos and encoders, and ship it all in one package
I've got a spare gecko :)
though no encoder
ok, just the encoder then
do you have a paypal account?
ok, then we should be set
I'll remind you about it later someday(within a week or so) need to check how much I really need for the Z axis
I'm having major issues finding anything about thoose calculations
I could swear we've had this discussion before
about weight.. but I cant remeber it
did you re-read the gecko white paper?
yeah just quickly
regarding the torque / voltage / speed stuff
but I get the idea about it
it's like robin said about a car gearbox
im not near the torque peak :)
yeah -stuck in 2nd gear with a 1mm fuel hose ;)
I can actually call my stepper retailer up and ask them about the torque curves for the motor
oh - that's good
one of the advantages of buying from a retailer ;)
that's probably a good thing to do
I buy my pulleys and belts off the same place
they sell galil things also
hm, I wonder if I can find this formula collection for calculating screws/forces
it's relatively easy, actually
if you know the transmission efficiency of the drive train (usually 90% for ballscrews and 50% for acme)
gearing and stuff I understand.. that's easy
oh - are you thinking of cutting forces, or thrust?
ok. cutting forces I have no idea about
the plunge force will never be that great..
thrust is relatively easy - consider a 1 Nm motor moving your table
at 1m radius, you have to move a lever the diameter of a 1m radius circle, or 6.28 meters per revolution
brb in a sec.. grabbing a pen
one rev of the motor gives you 2/3 mm of table travel (2mm/rev, but only 1/3 rev of the screw per motor rev)
so the mechanical advantage is 6.28 meters / (2/3mm)
hmmm - I probably have some factor wrong there, but that's the thought process anyway
also, you have to factor in the losses in the transmission system
SWPadnos: why isnt one rev on the motor 1 rev no the screw?
because you have 3:1 gearing
oh yeah but disregarding the gearing..
well then it is ;)
you may have to take the screw diameter into account as well - I don't remember well
that's where the fun comes in
the basic idea is to figure out the ratio of motion on the motor shaft vs. motion on the table, and that gives you a ratio that you multiply the motor torque by to get thrust
maybe I should have used the screw diameter instead of a hypothetical 1m :)
that's what I dont get..
since if 1nm at the motor equals 1nm at the load I wouldnt move my table with the stepper right now
Neglecting screw efficiency: Lbs = pi * TPI * in-oz / 8 - Mariss
the 8 is 16 oz/lb / 2 (from 2 * pi)
and what is the mariss variable? *grin*
just being silly :)
hm that should translate to nm/mm
yeah, so I guess I was right: newtons = pi * 2 * TPm * Nm
tracks per meter, that is
you can solve that reverse..
100 * pi * motor torque for you
newtons would be 1200 for 120 kg..
(neglecting the 3:1 gearing)
1200N ona 120kg table would be 10G acceleration, which is highly doubtful
err - nevermind - closer to 1G, once I divide by 9.8 or so :)
hm,so 100N would be 1G of accel
F=newtons, M=kg, a=m/s
1 newton is the force it takes to move 1kg 1m/s^2
1 G is 9.8m/s^2, so 1200N will accelerate 120kg ar slightly over 1G
that sounds right
that's a absolute maximum
if I'd get the axis going that fast I'd have a hole in the roof :D
but that's only enough to accelerate upwards at slightly over zero, since you have to counteract gravity (unless a counterweight / clockspring is in there)
yep - pretty close to that anyway
a high performance machine might be able to accelerate at 0.5-1G, and I think 0.1-0.2 is closer for many
you actually would want ~1400N in Z, I think
it's a overstatement the actual weight is more like 70kg..
after factoring in drive inefficiency
ok, then ~850 is closer
hm *ponders how to apply that formula to it*
you think you have the 1.1nm motor? or the 1.6-ish
you're right - you can back calculate, if there's only one unknown (out of drive ratio, motor torque, and required thrust)
ok, so motor torque / 2 * pi * 850 = drive ratio
err - 1/x, maybe :)
and motor torque would be 1nm
*writes this down*
for you, it's ratio = thrust / (torque * 2 * pi)
or 850 / (2 * pi * 1.1)
oops - got to factor in efficiency loss
what kind of screws>
I have no idea
but they dont have acme thread either..
heh - acme you think? (or just threaded rod)
not threaded rod for sure
*messed up what he was writing*
hmmm. ok. let's assume acme-like efficiency then - roughly 50%
so divide by 0.5 and get 245 ratio
the screw is already 50 of that, so you need a drive ratio of 5:1 or better to move that axis
so try the 6:1 and see what happens ;)
lets try this
3.14 that is.
I think that's inverted
thrust = pi * 2 * ratio * torque * efficiency
so ratio = thrust / (2 * pi * torque * efficiency)
thrust is the 850 number
is it my threads per rev?
2 pi is the circumference of a circle ...
the ratio is final drive ratio, including gearing/belts, and thread pitch (in threads per meter)
ie, how many turns does it take to go one meter
I dont get where you derive the torque from..
from the motor?
you either calculate the motor torque you need, or you use the rating of the motor to calculate thrust (or ratio) ...
that's the formula i'm interested in
it's 4 variables (torque, thrust, ratio, and efficiency)
the one to calculate the torque you need..
algebra, baby :)
I've tried to forget as much as possible of it..
I have no problem with it it's just keeping track of where you get variables..
this is the whole thing, you just have to isolate the variable you want to solve for
thrust = pi * 2 * ratio * torque * efficiency
850 = 3.14 * 2 * 1 * T * 0.5
ratio of 1?
you've got a 50:1 ratio from the screw, since it's 2mm pitch
that's without gearing / belts
in that case T would be 80
isolating T, and using only the screw reduction, T=850/(2 * pi * 50 * 0.5)
hm, I get 4,9
but you are probably right
you used 1.1 Nm, not 1
1.1 is correct though
I left out the actual torque, I guess
on the bottom23:21 < SWPadnos> isolating T, and using only the screw reduction, T=850/(2 * pi * 50 * 0.5)
hm.. I did the same..
though im cheating..
using the solver on my calc ;)
I could do the same, I guess
I forgot the other term - but oh well, it's close to 1 anyway ;)
but that tells you that the motor torque * belt reduction needs to be 5 or higher
the 4.9 is the n/m required
so you need a 1.8Nm motor with the 3:1 reduction, or change to 5:1 with the 1.1
assuming 1:1 motor:screw ratio
that's where you start at :
if you have a 5:1 motor:screw ratio, then you only need a 1Nm motor, which you have ;)
or with 6:1, you'll get ~1000 N of thrust instead of 850
bbl - dinner time
bath time for me
thanks for all the help!
ok. see you tomorrow
sure - I hope I'm right :)
have a nice dinner