SWPadnos: Full T1 $399/mo
lerman______ is now known as lerman
That's just weird... I powered off the DSL modem for 3m and still maintained my irc and shell sessions
Upgraded DSL (effective today)... I aint seeing it yet (okey, they said 4h to tune). Figure I'd DL a 4GB ISO to help them out =)
that should only work if you have a fixed IP
(the maintained connection thing)
SWPadnos: I know... weird huh?
Heck, I unplug a ethernet cable for 1/2s and I loose connection
SWPadnos: Even my shell was still connected.
that makes some sense on the client side, but not so much on the host (internet) side
I can imagine IRC not timing out (or ssh for that matter), but you'd have to get the same IP when the modem reconnected
they often do that so it's not very surprising, but it is a little surprising :)
I think I got a different subnet actually, can't be sure though.
ok, it's avg about 485KB/s, that sure aint 6mbps, and not much improvement over 3Mbps
SWPadnos: You know how you always hate my bandwidth?
no, I hate *my* bandwidth
SWPadnos: You'll hate *ME* now... 18:27:35 (10.86 MB/s) - `slamd64-12.0-mini.iso' saved [17967104/17967104]
that seems pretty fast :)
it's about full bandwidth for a 100mbit network
intertubes baby, intertubes =)
I wonder how slow this RAID will be with the "green" drives in it
that's fast all right
Not my DSL, colo
of course :)
I know DSL doesn't go that fast
Actually, it does now... 15Mbps
(top speed of 7.1 Mbps if I remember correctly)
that's bits, not bytes
yeah, I know
I'm under a mile from the CO, but there are no current plans to improve the speeds here
limited upload to 1M too
I can get an OC48 if I want to spend a years pay every month
SWPadnos: Mmmmmm dark fiber =)
if only I knew someone at the other end ;)
that'd be a looooooong fiber
SWPadnos: Yeah, but we could peer and just sell bandwidth and nothign else
SWPadnos: we'd have to get a BAR though
BAR = Big Ass Router =)
I'd raise the BAR
or think outside it
that's BOX, not BAR
well, it's hard to think inside a bar - they're too loud
you shouldn't even try thinking in a bar, that's kinda the opposite reaosn they exist
lerman______ is now known as lerman
*gasp* jmkasunich quit?
cool, i just found an apartment on craigs list , in taichung :) maybe i can get out of the crappy bed-as-hard-as-concrete hotel this trip.
lerman______ is now known as lerman
[08:11:41] <alex_joni> http://www.techamok.com/pics/08/oct/google.jpg
which hal module can i use if i want to change the spindle speed value to 6 bit output on parport
from max->63 to min->0
micges just wrote a hal comp to do that, i guess it's on the mailing list
i'd suggest using PDM or PWM instead though
if you can
what is the pwm frequency range and pwm resolution?
when filtered, a voltage
resolution depends on your BASE_PERIOD and what frequency you choose
for a BASE_PERIOD*2*8 frequency would give 8 bit resolution
at 1/(BASE*2*8) Hz
so if BASE = 20000ns, f = 3.125kHz
uh, but i think it does dithering too.. lemme look
yes, set pwmgen.0.dither-pwm TRUE, for higher apparent resolution
i have no idea how to calculate dithered resolution
i wont use dithered output. I will convert pwm to binary output with microcontroller
so dont want the pwm frequency duty cycle to change for a selected spindle speed
just thought, someone did the gearbox speed chang perhaps thats a better staring point
i will not actualy controll the spindle, but laser power
ah then see what micges does
fogl: wait moment
fogl: I've sended to emc-users mail with that module
I'm using it also to control laser power
What is the procedure to load a custom hal module. Where can i read more anout this?
[12:50:28] <SWPadnos> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?ContributedComponents#How_to_compile_and_install_a_component
that's just about using comp, it doesn't mention that you need emc2-dev and build-essential installed
fenn: I think that to figure the resolution of pdm or dithered pwm you have to say how long you want to integrate over. If you have a 25uS BASE_PERIOD and an integration time of 10ms, then you have a resolution of 10ms/25uS = 400 levels
i think i would be useful to write those pwm dity cycle/frequency equations to man page of pwm hal module
* jymm almost set off the fire alarm for the whole bldg =)
MY motion activated fog machine works WAY TOO WELL =)
I wonder how much precsiion you need to make a cycloidal gearbox
the fun part is internal gears
for a cycloidal?
I were under then impression they were gearless
cycloidal is a gear form to me
oh, then you dont get what I mean
err cycloidal drive.
[13:26:49] <anonimasu> http://www.darali.com/page17.html
pins and a shaped tooth
I wonder how close you need to machine one to make it work fairly well
err the profiles :)
same with any gear mechanism I assume, any error can easily become a step error in output
that's not really a answer to the question,
_any_ error is how much :)
0.01mm or 0.001?
we measured a few gears here, nearly a degree
and most error is over a short radial distance
cycloidal drives seem immune to backslash
cause turned out to be from the worm dividing
I wont believe till I see and measure one
worst source would be the eccentric pin in the middle
you'd use a preloaded ball bearing for the center pin
i've been thinking about this a bit, and if you had two gears 180 deg out of phase, they could preload against each other (and also balance the whole assembly)
pin placement must also have tolerance
pins can be adjusted (eccentric pin)
Im thinking of all the other pins
and they dont seem to mention backlash amount as a feature
other sites say "ZERO backlash" in many places
hm.. I saw 20 arcsec of backslash for some drive
Im a realist how could they maintain 0 in a running one
internal flexure.. high internal preload forces combined with rolling bearings
are there any other rolling element gears?
yes ish the one that flexes the ring
harmonic? that's the same principle basically
yup thats it
cycloid is just easier to make
harmonic springs the gear teeth in/out
hm, cycloidal must have less backslash as it's easier to make round holöes then gear profiles..
archivist: it uses the gear-has-one-less-tooth-than-ring-gear principle, like a vernier
you could do the same thing with two solid spur gears 180 out of phase in a ring gear
(for zero backlash)
but rolling elements has better efficiency and wear
[13:51:42] <archivist> http://en.wikipedia.org/wiki/Harmonic_drive
note the bending
yeah and that's exactly the same as the pins bending in a cycloid drive
some CNC mill heads are driven that way
I need 0 in my trunnion
yeah, i'm just.. curious on how thye manage 0 backlash in such a simple concept
im curious how they ever manage to make gears with near zero backslash..
zero in a worm drive is easily possible
that's more fantastic, where only a few teeth ever meshes
at the same time
archivist: how about a "daisy" made of ball bearing cartridges, being driven by a really coarse worm with a ballscrew profile
not to mention one is flexible
um nevermind this is too hard to draw
fenn getting close to the traditional steering recirculating ball
i dont get it.. that's just rack and pinion with a ballscrew actuator
how do you take the slop out?
hmm i could use 3d cad right about now
i never realized how bloody hard it is to find a parallel cablke
fragalot: welcome to the post-industrial world
if i do find one, it's 25 euro shipping
I dont get how you get backslash out of it
or adjustment or spring
neither will work really well for a table
archivist: yea that's the idea, but the rollers are arranged radially on pins about a disk
agjustment springs are neat..
err sorry spring loaded.
until you load them.
* archivist contemplates recirculating ball worm and wheel and has that 1969 patent run out
i dont think that's called recirculating ball
all the other images have a ballscrew with rack stuck on the side
[14:11:47] <archivist> http://www.google.co.uk/patents?id=l-J0AAAAEBAJ&dq=recirculating+ball+worm+drive+3468179&printsec=frontcover&source=bl&ots=DRkiOYAGf2&sig=DzE6mTBdm22XWc3lF3Eo_LOzncw&hl=en&sa=X&oi=book_result&resnum=5&ct=result
the real thing
good luck.. :)
Kearny and Trecker own the patent
patents.. what a bunch of crap
i have that exact same drawing in my notebook
or is it owned past tense 20 years?
dosent patents last 50 years?
14 years i think
anonimasu, thats copyrights iirc
I wonder if milled form would be good enough
in the us
19 years past sell by date then :)
eh /me doesnt bother scanning notebook after all
would be interesting to make
haha here's a good entry: sail blimp.. tethered to a keel
heh, got to this page from xkcd random-> http://xkcd.com/404/
alex_joni, did you go there april the first this year
they swapped the dns for about 4 funny sites
was funny getting the wrong sites
another worm doohickey: http://www.cdxetextbook.com/images/350px-Wormgearbox.jpg
formed worms become much more of a possibility with cnc
is that what the hourglass shape is called?
i think it looks like a tazmanian devil myself
some of the weird forms used in bevel gears come to mind as well
you can "hob" them with single point cutters
that hourglass shape makes the backlash adjustable
they can move the worm forward or backwards to adjust the backlash, while the machine is running
Im intending somes mall sprung movement in that direction
you... you are designing a spring... into the axis of a worm shaft?
to keep it in contact
so you are only expecting limited force then
such that the spring pressure won't be overcome
that depends on tooth angle and yes
will be on a rocking plate with the drive motor, which is in loaded bearings in the case
add up to lots of bearings
Hello All.... Question regarding pwm modes ... Am I right in assuming that pwm / dir mode works by setting dir pin high / low for direction and send 0% - 100% for 0 speed to 100% speed where as PWM up / down mode is as follows 50% = stationary, 50% > 0% = turn in one direction increasing in speed aprouching 0% PWM and 50% > 100% turn in oposite direction increasing in speed aprouching 100%?
although step/dir and pwm are used for different things
step/dir is for driving position (each step is a certain positional change)
Paragon: I think you are describing pluto's pwm/dir mode correctly
pwm is used to drive a certain velocity (like spindle speed)
Paragon: up/down also uses two pins. At 0%, "up" and "down" are both off 100% of the time. Otherwise, for positive values, "up" is on for a fraction of the time; for negative values, "down" is on for a fraction of the time
Sorry Chaps ... Yes I am talking about pluto currently going through the pluto_servo.comp file ...
pluto doesn't directly have a "50% = stationary" mode, but you could get it by using offset to add 50% and just using the "pwm" or "up" pin, ignoring the "dir" or "down" pin.
Thanks for the info ... So what is pdm (interleaved pwm) ?
(iirc, the pluto firmware controls which of two pins gets the pwm value and which gets a constant 0 value, then each pin can optionally be xor'd with another value)
Paragon: pdm makes the output change more often than pdm. it's the mode to use if you're generating an analog waveform using an RC filter.
it works by reversing some of the high-order bits of the pwm counter before doing the comparison
you can show yourself how this works by imagining a 3-bit counter being compared to a 3-bit value. If the comparison value is 4, then you get successive pwm outputs of: 00001111
if you transpose the top two bits of the counter before comparison, you get something else .. I think you get 00110011
it's a technique I learned about from peter wallace; it's used in the mesa line of boards and is recommended when using their analog servo board to produce +-10V signals
(technically I think pdm and interleaved pwm are different techniques but the result is similar in terms of giving smaller on- and off-times than basic pwm)
Paragon: this is the pluto outputting pwm up/down closed loop with an encoder and emc doing the pid tune. The pwm up is 0-100% for cw on one pin and 0-100% for ccw on the other pin. http://www.youtube.com/watch?v=fGF07oVkl6M
Nice explanation jepler, Thank you
Thanks for the vid and wiki link alex_joni / skunkworks ... :-)
skunkworks_: is that the new bridge?
Things I have learned: Copper + Lathes = bad idea.
That said, the art lab is thinking about selling the lathe anyway.
I think cradek learned that recently too
They want $1,000 for a beat-up old South Bend with a bench grinder motor held on with plywood and door hinges!
it's ok, but you have to know what you're doing (which is what my problem was!)
* archivist sells his southbend for 999.99c
Is it just me, or is this rather a lot for an ancient beat-up lathe?
hard to say. if it comes with a lot of tooling and accessories, and lathes are not plentiful nearby, it might be ok
define beatup, they can do nice work when worn
That's the same as boxford here in the UK from memory...
taper threading ftw
mine has the taper turning attachment :)
cradek: It's beat up, rickety, and hasn't got much of anything beyond a chuck and a tailstock.
[16:40:16] <spasticteapot> http://cgi.ebay.com/South-Bend-9-Precision-Metal-Lathe-Model-A_W0QQitemZ300268606537QQcmdZViewItem?hash=item300268606537&_trkparms=72%3A1205|39%3A1|66%3A2|65%3A12|240%3A1318&_trksid=p3286.c0.m14
The one in this auction is very similar, but in MUCH MUCH MUCH better shape.
although the saddle has dropped 20-40 thou it can still do good work
The one in that auction is very clean indeed.
breaking news: e-bay seller optimistic about prices
Pickup only.... I'm on a plane .... ;-)
mine cost me about £150
20 years ago
[16:43:26] <Paragon> http://www.lathes.co.uk/southbend/index.html
similar to mine http://www.lathes.co.uk/southbend/page4.html
I love the south-bend's
bed well designed
no way, double-V design is overconstrained
Like the apron...
To be honest the lathe I have is a flatbed ....http://www.lathes.co.uk/granville/index.html http://www.lathes.co.uk/granville/page2.html
more than one turn is a ball screw overconstrained?
not if right!
uh.. if it has more than 5 balls
archivist: do you have the collet changer also?
I saw a great tutorial video 'which I can't remember' with a chap and his southbend.
that's a nice bed paragon
fenn: Would'nt want to sleep on though ;-)
Would'nt want to sleep on IT though.... talk about screwing up a joke .... lol
the rectagular beds get front to bad slop you cannot adjust out
I used a worn Myford like that once
One thing I like about the Granville Senior is that all Myford ML7 part fit it. Basicaly it is a Myford ML7 but much more heavier. I paid £250 for it five years ago. There is some wear in the leadscrew (rounding) and a little wear at the front of the bed (which could be ground) but other than that it runs very true. Heck it's old enough to smoke right...!
On the Southbend I have to adjust the depth of cut to get parallel, but its old enough for a pension
best ways for a lathe is inverted V in back and flat/half-dovetail in front
Best of both worlds ...
I prefer inverted v at front as the force then forces alignment
causes wear that affects front/back alignment though
I took my first steel metal cut on my denford starturn last night .... Very impressed is I...
It cut like a hot knife through butter ...
Can anyone here recommend an accurate way to cut copper pipe?
Disperses heat too fast.
spasticteapot: try negative rake?
My dad used a pipe cutter. If need be fill the pipe with sand to stop it crimpng...
Copper is very sticky / graby right?
methinks negative is the wrong answer for copper and its work hardening
LotLots of soapy water may help...
yeah nevermind, just a wild guess
supposedly milk is the coolant to use for copper
sounds like a waste of milk to me though
Yeah I think Iv'e heard that somewhere else also fen...
OK if you have a dairy farm ;-)
spasticteapot: what sort of problems are you having?
It takes eons to cut through copper on a lathe.
And I don't really have many other tools.
Machinig Copper :- http://www.pupman.com/listarchives/2004/July/msg00343.html
spasticteapot: Sharp tools, Fast Speed and Milk
sharp tools insane speed and coolant, drink the mick and user usual oil
Worth a shot.
Of course, it would help if I could lathe worth a damn.
spasticteapot: Have you been cutting at a slow speed thus far?
is high speed to keep heat from going back into the work?
Anyone wanna sell me some aluminum scrap I can practice on? McMaster Carr is too !@#$ expensive for something I'm going to chuck out anyway.
My lathe has two speeds: Very slow, and very, very fast.
find a local scrapyard
archivist: Hmm. Worth a shot.
where I get mine from
ally can be sticky as well, wax and speed again
find any company that has some lathes... they always has some crap lying around
we have crap lathes yes
Or buy some carving wax in stock form. Cut it, then collect all the wax swarf, melt it down and form back into stock. Almost an endless supply.
no some material for the lathes, it seems to acumulate around machines..
wax isnt good for learning how to cut metal
interesting to watch the two of us here, I use top speed a lot more than boss
buy some scrap aluminum, melt it down and form back into stock. Almost an endless supply. :)
I use wax as lube on ally
just enough to stop built up edge
fenn: True but it will teach co-ordination and give the student confidence to progress onto metal. It will also give the student an idea of how the lathe fuctions as a whole.
it will also give you silly ideas about plunging, roughing, etc
lol .... got no answer for that ....
[17:15:43] <jymm> http://www.youtube.com/watch?v=oUH3m6ZyW-g
machining wood and plastic will teach chuck use, watch for flying lumps
jymm: Cool ...
this has got to be the most useless feature ever.. http://www.youtube.com/watch?v=DLZcGDyacHo&feature=related
they want to waste your cpy cycles, so you need a faster better....
Ive stopped at XP and win2k
M$ has lost the plot
but what about the emotional connection!
I iz disconnected!
pew pew pew lasers
hah. yeah i read that
pew pew pew
omg, we can't fly anymore.. http://xkcd.com/107/
[17:48:04] <skunkworks_> http://www.xkcd.com/163/
hmm xkcd lolcat and youtube friday :)
isn't it great?
I think we need a little dilbert though
less worksafe also nickscipio.com picture of the day
i am debating getting tested for a learning disability
I am not sure if it would be worth it, though
good night all
toastatwork: it has it's advantages for some courses
well it's for math, but i have a system worked out that does well
but it could be derailed depending on the teacher's test/class rules
dmess: i have'nt had a chance to go at the lathe again.... but it still looks fachked to work from the key pad...
(16:53:08) archivist: him not here, added the link he mentioned?
(16:53:10) dmess: gonna have to tie in emc JUST to get spindle control...
no such thing as a learning disability.... its an IN-ability to TEACH
we have a good example in #mysql at the moment
somewhat below standard
learning disabilities do exist im sure oif it
it is me, for sure
not the teachers
I've been contemplating if there's a test for that kindof thing over here, too
I always find math formula manipulation hard
there is, you have to go to a psychologist or neurospychiatrist who specializes in learning
"discalculia" is the umbrella term
I study study study and study and some shit dont stick.
anonimasu: less to wipe then
all you really need to learn, is how to find a reference/book/person to get your answer
archivist: slow down.. it isn't a race .... it took me till i was 30 to figure out the #'s worked
I have found that it's easier to just learn the basics from which most complex things are derived
I collect books to save learning
and if you forget those complex things, it's easy to find them back by just thinking it over
for people with math disorders, it's the simple things that are difficult, and the complex things are easy.
I keep forgetting complex shit I need.
calculus is easy, it's adding 3 and 4 that i have problems with
numbers look the same, etc
toastatwork: i've got ADD, it's near impossible to sit down and study :p
no its not enuf #
on my calc exams, all the math is correct
and the difference between + and - or * and / kill me
it sometimes happens that i make numbers up and add them in the formula as i go along
"so.. thomas.. where did this 3547 come from?" 'uh.. I dunno?'
but what about in an equation??
dmess: doesn't matter where it is
even on price tags in shops the numbers shift about randomly
I got 100% in the exam and 0% in the coursework for a tech maths course
I can't read properly either because the letters walk arround randomly, but for some reason, I don't have that issue if it's on a computer screen
archivist: ouch :p
one confused lecturer
have I mentioned I see dead people?
had school maths teache sout at me fo just writing the answer
* anonimasu hides
same for the engineering class for me
if i'm interrested in the subject, i can skip the whole work and just give the answer in most cases
math itself.. I don't care about,.. math beeing applied to something i care about, no problem
so the fact that ive watched about 3 hrs ov savant's with similar simptoms tells your special... and highly functioning
saddest part for the teacher is that he can't say the answer is wrong, since the outcome is always correct, so he can't hold back points for not giving the method i used :p
dmess: yes, for the things we care about, we're "highly functional", for all others, we're useless
now that i have successfully killed the conversation, gnite!
we make bad teachers
im with YOU
NEgative... on we make bad techers
I wouldnt dare teach any maths
i can learn any learn any of MY son's to run a boring milll ANY time
and i do and have taught math to my collegues... but somethine CLICKED... and MATH is me very good friend these days
sometimes showing someone else your problem will teach you
used in software a lot for quality control, called a walkthrough
not in the s/ware in our place ;)
I personally do it with all new plans and procedures prior to sighning...
but the correct ppl need to be there
lerman______ is now known as lerman
hi every one, ...
hi one from everyone
hello archivist, ...
i've got a problem with rtai, ...
archivist are you present, ...
i have some questions about linux ... and for emc2 rtai...
as kaway. if we can answer we will
in few words when i start on rtai kernel i've got a blank screen. after login but when i choose the opion ...safe gnome... or some think like that before login, rtai kernel -< gnome and emc2 too,
with option safe gnome on parameter login, all run very well, emc2 in simulation, hmi interface gnome too.
sorry, I don't understand the question
what can i do, in order to do a normal boot and don't have a blank screen
so you're talking about the GRUB menu, where it shows you the possible kernels to boot (or memtest), or are you talking about the login screen, and after login the screen goes blank?
methinks error messages would help and trying Ubuntu forums for odd screen problems
it could be a driver + RT kernel issue, in which case it goes here and there
year, when i boot rtai in order to use emc2, i have a blank screen, ... sory swpadnos i don't understand your last post...
the login screen is whereyou type in your user name and then password
so you don't get a login screen when you boot the RT kernel?
does it go blank after the grub menu right away or do u see anything at all before going blank
i have a login for booting in the kernel rt ... when i log with the option, safe gnome, it doesn't boot in blank screen mode .... :-D and when i do a uname -a its tell me that i run rtai.... but when i boot whitout ... safe gnome.. i have a blank blank blank screen... ------< i have to loggggg
simsim, u could try a ctrl-alt-f6 after display goes blank, should kill all X and give u a command promt
sounds like maybe its booting with wrong resolution after grub maybe?
thanks, but for using emc2 it's better to have a hmi, ...
yes that it but... i have look in xorg.conf but one question are for all kernel the same xorg.conf file'
here we go, http://ubuntuforums.org/showthread.php?t=258484&highlight=kernel+booting+parameters
you need to be sure you have the correct drivers compiled for any kernel you want to boot
the conf file is the same (as far as input devices and resolutions), but the drivers need to exist or you'll have problems
:-[ 1... the correct drivers. are there different drivers between normal core 2.6.24-21-generic and rtai... 2 ...
for the os the hardware is the same, is'nt it...
i will try ctrl-alt-f6... but when i will in command line what can i do... start x or ,,,
there may be an error on screen when you drop out
yes, the drivers have to be compiled against the running kernel. there are drivers provided for the RT kernel, you just have to install them
archivist, do you mean when i logout out... swpadnos, ok drivers are compiled ...dinamycally... for the os, but what have i to do to install the driver for rt ...
what video card do you have, and what driver do you use?
(hint, for RT performance, you probably want to use the vesa driver)
incidentally, do you have a wide screen monitor?
the video card is ati x1400, the ...........ati specific property driver................ is installed, 1024x768, thinkpad t60, not wide
ok, you'll want to test well before using that driver in RT
I think the package name will be linux-restricted-modules-<your kernel version>
I don't remember what the kernel version is exactly, but it's possible that the following will install it:
sudo apt-get install linux-restricted-modules-`uname -r`
what is it...
a test software...
that may install the proprietary ATI driver
I'm not positive it's included though
sorry for my english, i have the proprietary ati driver.
so you have video problems with the stock kernel and not the RT kernel?