er, i mean, 1048576
now why would somebody make a 17.4mm diameter thread?
because it's close enough to 18mm when you chop off the thread peaks
its probably supposed to be 18, but I wish I could be sure of that
18mm x 1mm pitch
broken part for a camera tripod
did you measure the pitch diameter? (three wire method)
don't have wires/too lazy/won't work well on plastic/etc - pick your excuse
regular threads have flat tops so they'll fit female threads with rounded bottoms
* fenn wonders how to rephrase that
the plastic part screws into an aluminum tube, and the fit is sloppy, so duplicating the pitch diameter of the plastic part won't be helpfull
I know what you are saying
but the nominal measurement for threads takes into account the flat tops
this will be like many repair jobs - get the pitch right, then cut till it fits
you can measure the pitch diameter then the actual diameter and find out exactly what thread engagement you have
which will tell you exactly how to cut the mating thread.
the equations are in machinery's handbook.
not that you have to do it that way obv, it's just faster than guess and check
the part I need to engage to is the internal thread - I have no way to measure that pitch diameter
you should stil be able to measure the id and ballpark it
my ballpark estimate is that its a 18mm dia x 1mm pitch ;-)
which doesn't tell you anything about the shaft you need to start with, but whatever.
I'm gonna start at about 17.8 or 17.9mm
the ID of the female thread is 17.1mm
so stay with me for a second here
if you know the ID
you can go look up the theoretical thread engagement of that ID for that pitch
and then look at the mating shaft information
and just... just...
cut the thread once and be done with it
we must be talking past each other
I fail to see how thread engagement (theoretical or otherwise) has anything to do with this
opps....news channel please
it tells you exactly how to cut the mating part.
not if I don't know the size of the thread I'm trying to fit into
if it's not a standard thread, then no, and my aplogies.
18mm is almost 3/4", and 1mm is finer than 24 tpi, so its not a common thread
it may or may not actually be 18mm, might be something really stupid like 17.5mm
if it is indeed 18mm x 1mm, then I agree, Machinery's can tell me what the pitch diameter needs to be on my mating thread
well, MH says that standard bolt threads don't come in 18mm - the nearest bolt threads are 16x2mm and 20x2.5
American National Standard Fine Metric come in 18mm x 1.5mm
nothing around 18x1mm
how sure are you that it's metric?
the pitch is metric - its finer than my 24 tpi gage (by a lot), coarser than the 26 tpi (by a little) and a perfect match to my M6x1 tap
diameter, not sure at all, although an english diameter and a metric pitch would be pretty perverse
made in Tiawan 20 years ago, so I wouldn't rule anything out
you could measure the lateral slop in your existing too-small PD, then measure that PD and add
easier said than done
then cut the PD you want, file off the peaks until it fits
the existing male part is plastic, sloppy, and broken
surely so, good thing I'm on the saying end :-)
ah I figured you had a complete part, but just badly made
no - it served well for a long time, but I was at the park a couple weeks ago, it was cold, the plastic was brittle, and I dropped the tripod
fortunately the camera was not attached
this part screws into the center upright tube of the tripod and had a 1/4-20 boltish thing molded into it that the tripod head attached to
the force on the steel part broke it out of the plastic molding, leaving about half the length in the tube
I drilled a couple holes in that and did a spanner wrench thing to get it out
it's surely not M17x1? I see that exists
no, minor diameter of the female thread is 17.1
major diameter of the plastic is 17.4ish
and although I can't measure it to 0.1mm precision, I wouldn't be surprised if the slop is the better part of a mm
I'm guessing PD of M18x1 should be about 17.29
I'd just cut that, take a little off the peaks, and see if it futs
or fits even
can you make it between centers? you could even test fit then.
I can test fit while its in the chuck
well then cut away!
you could be done! :-)
well, I said a while ago that I was gonna just cut the OD to 17.9 or so and then sneak up on a fit
do you have your 127 tooth change gear handy? haha
yo yo yo
jmkasunich, any toying around with the mesa lately?
you mean sully my American machine with a G21? Nevar!
you mean cut a metric part in the most natural way? Nevar!
actually I guess G21 makes sense for this
I worked very hard on letting you do that :-)
I guess I should try it then
the hard part is figuring out all the darn metric feeds and stuff
guinea pig is kinda gross... the show.
that reminds me, I need to update to 2.2.4
G21.1 - metric lengths and ipm feeds!
I shouldn't say that in public
otherwise the G21 won't last when I switch from Auto/MDI to manual
yeah let me know how that works...
I hope you will have fewer problems with TLO too
I should air-cut another fusee
see if it sounds different with the bobble gone
at fest this year our first (probably only) beer together will be to celebrate that tough bugfix
cradek: *ONLY* beer ???
hi cradek ... whats the closest airport to fly into for fest??? Chicago?
don't know about closest. closest cheap one is probably chicago
hm, 3 hrs from chicago though
peoria is the closest "meduim" city
what city/town is fest in?
Chicago is several hours drive away, but if you're flying you might as well drive from there since Peoria is close to an hour away
then probably Iowa City or Cedar Rapids, but they're almost as far as Chicago, with a far smaller number of flights
JymmmEMC: galesburg, IL
KGBG Galesburg Municipal Airport
Galesburg, Illinois, USA
yeah, Chicago is the closese hub, and with the time it takes to change planes then fly to a smaller city, then drive from that smaller city, you might as well just drive from Chicago
actually, the shop is within sight of the "airport" there
JymmmEMC: if you have a piper cub, you can fly into galesburg
think you'll find a direct flight from LA to galesburg? haha
direct flights from Peoria maybe
maybe, if you know someone
[01:34:17] <JymmmEMC> http://www.rentplanes.com/results/index.asp?Loc=62682&c=1
or you have a Piper cub ;)
I bet someone from OshKosh could fly there
hmmm. I wonder when Roland will actually put content on the sire
that's all the necessary content for repeat offenders, err, visitors
heh - but no web payment form
oh right, we have to pay
yeah -that is the bummer
I think we should be treated as an attrraction
hmmm - not so far
* JymmmEMC counts SWPadnos' fingers... 7, 8, 9, hmmm 9.5
SWPadnos: "I told you the plasma cutter is NOT a Star Trek ionic shower"
I thought it was the transporter
SWPadnos: Well, it transported half your finger to the floor didn't it
it cuts while it cauterizes
$377 round trip SJC to PIA
chicago I assume?
no, wiat... wrong flight
Actually, depends on which flight times.... either DFW or ORD
[01:49:07] <JymmmEMC> http://www.cheaptickets.com/App/ViewFlightSearchResults?gcid=C16036x268&WT.mc_ev=click&WT.mc_id=ec2127&retrieveParams=true&z=f50d&r=3j&z=f50f&r=3l&lastPage=interstitial
[01:50:32] <JymmmEMC> http://www.google.com/search?num=100&hl=en&safe=off&client=firefox&rls=org.mozilla:en-US:unofficial&sa=X&oi=spell&resnum=0&ct=result&cd=1&q=sjc+to+peoria+il&spell=1
What do I know, but on one of Mariss' whitepapers he mentions using a cap when using a switching PS. My PS is email@example.comA I have 3900uF@400V caps that I was going to wire two in series w/ the PS to the drives. Does this sound about right?
series reduces capacitance
damn, I did it again <rolls eyes>
so does 7200uF sound about right?
<rolls eyes, again>
what's the motor current?
(I know the PS is 4.5A)
looking up now
~1.1 ohm, 3.1V
[02:35:47] <JymmmEMC> http://www.xylotex.com/StepperMotor.htm
ok. IRC :)
SWPadnos: OMG you were .03V off... WE'RE GONNA CRASH !!!!!!! lol
you'd have to look at the parker maual for guidance. I don't know if the topology they use hasthe same guarantees as the Geckos
SWPadnos: Explain yourself Plutonian!
Mariss says the geckos will never draw more than 2/3 the rated motor current
I don't know what parker says
(I don't know why he says that, but I'm pretty sure that's the number)
if that's also true of the Parkers, then you could need as much as ~6A for all 3 motors
is that becasue of the BEMF and all three motors will (probability wise) not be deaccelerating at the exact same time?
no, I think that was for each drive
but I'm not sure
maybe that is the deal though, that the recommended current is 2/3 the total for all motors - I don't remember the specifics
Mariss' formula is 80000 * I / V I believe
480000/48 = 10000 uF
so 7800 is pretty close
there are also caps in the supply, and the total needed is different because his formulas are for bulk supplies, not regulated
[02:42:52] <JymmmEMC> http://www.compumotor.com/manuals/oem/OEM_manuals.htm#OEM750
I never knew you could tune a stepper motor before
SWPadnos: Could the 2/3rd's amperage be related to peak vs RMS ?
I don't think it was
no matter though, because that's the gecko recommendation and these are Parker drives
incidentally, the "new" gecko formula for optimim voltage says you should have a 60V supply (V=1000/sqrt(l) )
you will proabbly want an inrush limiter on the caps - they look like a dead short when you first power on the supply
which would be?
what would I use as an "inrush limiter" ?
an inrush limiter - an NTC (negative temperature coefficient) resistor
SWPadnos: pg32 http://www.compumotor.com/manuals/OEM750/OEM750_Entire_Rev_B.pdf
I read lots of it, they say nothing about power supply ripple or capacitors
SWPadnos: no way man, only PTC for me!!!
then you'r escrewed
SWPadnos: Sounds like a thermister to me
sort of, but one that's meant to take several amps current
SWPadnos: MOV ?
search digikey for "inrush"
what's the initial resistance? (cold)
of the NTC res?
I assumed you were looking at some :)
well, something like 10-20 ohms if it's a 5A one
they range from .04 ohms and up
that's probably hte "hot" resistance
[03:04:02] <JymmmEMC> http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=570-1018-ND
0.18 ohm * 4.5A (max supply current) = 0.81V drop
that's not so bad
SWPadnos: What if I just turn on the power switch REALLY s l o w l y instead ;)
h m m m m m m m
Now, what if I toss on bleed resistors on the caps instead?
(it was something I was considering anyway)
the inrush limiter goes directly on the output of the supply, in series to the rest of the circuit (including the caps)
SWPadnos: Hey, on those tiny relays I got from you... do you have any that are DPDT?
PS + -> inrush -> cap + (= positive supply for motors)
and they're contactors ;)
frequently called motor starters on ebay if searching
not these :)
SkinnypuppY these are fancy ones
they're general-purpose contactors - pilot duty
not motor starters, there's no easy provision for overloads (and they're 2-pole)
[03:11:37] <SWPadnos> http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=570-1003-ND
less $$, 8A, 1/2 the resistance (0.09 @ temp)
Is the inrush current from the xfmr being powered up or back emf from the load ?
it's not a transformer, it's a switching (?) supply feeding some bulk filter caps
I figured an apparent dead short at startup might not be ideal all the time
well, the tripod thread worked
no idea what size it is, but the new piece is screwed in tight and bottomed out - ain't comin' out
on that note - sleep is good
any okuma programmers hanging out
lerman_ is now known as lerman
okay so guys
im doing pcb with .004 Y and .0002 X backlash
and its milling pcb like it tested with the indicator
taig X = almost toolroom shit
taig Y = typical sadness
hmm, a sort script to force the gcode to be unidirectional
to avoid backlash errors
upgraded to 2.2.4 finally ;). No problems.
morning alex. How is the day?
skunkworks: the day is fine, the night was short though
we crashed at 8:00 last night.. We needed it.
it was a bit after 9 I think.. but using US timezones
we did get up a bit early - had to get trash togather and the wife needed to get some homework done.
renesis: can you rig up a spring? cutting forces for PCBs are nil. You can't fix the backlash in software, but there are simple hardware solutions that should work fine
renesis: for a while I was using two nuts (rigged up not to turn) with a spring between them... it's easy
that sounded odd - she is getting a BS degree online though one of the local colleges
hmm project!! http://cgi.ebay.co.uk/CNC-lathe-Bosch-CNC-Alpha-2-not-Colchester_W0QQitemZ120232705947QQihZ002QQcategoryZ112399QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
heh, I have a BS degree too
archivist: how close is that to you?
about 25 miles
THere you go - you need another project :)
I need space and a lorry for that!
that is close enought to go physically look at.
although - I would think anythink in the uk is close enough to look at ;)
price of fuel is making it harder
£1.05 per litre
you lot are still on cheap fuel
lerman_ is now known as lerman
Yes - but everyone still complains.
OMG $3.25 a gallon...
Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2008-03-18.txt
wouldn't want to support features like that :)
I think it's a clever hack
with our paradigm of picking the line in the gui, we can't do a similar thing
I had wondered in the back of my mind.. but there are always ways around that..
If I wanted to do it right, I would have the interpreter keep a history of the initial state and the state after executing each line. Then, picking a line would give you the state and you could get there without stepping through everything.
[14:21:22] <alex_joni> http://www.pugetsystems.com/submerged.php
It's a little more complicated when you have loops and subroutines, though. The gui would need a way to move through the history, not just the source.
lerman: I bet it is
lerman: To get the Okuma behavior, I don't think you need that. You could run the program from the beginning and start executing when you hit line L the Nth time
Not that bad though. Keep a history file that includes the state and a pointer to the original line.
(for "start at L.N")
not that I think this is a great solution, mind you
Yes, but the Okuma behavior is a kludge that assumes the user knows which invocation of the line is the right one.
In the case of a crash, you can look at the history file (if you have one) and see where it last executed.
Ideally, though, you should be able to go there and step backwards to find the proper restart position.
[15:09:50] <cradek> http://www.youtube.com/watch?v=EKL6elkbFy0
lerman: and in the case of recursive o-word procedure calls
the user is probably lost counting more than 10 invocations
That was great :)
gasp.. mach. but still cool
[15:16:34] <skunkworks> http://www.youtube.com/watch?v=A0kf1D49ces&feature=related
yikes, gloves around a running lathe
anyone use geckodrives here? I use G320 servo drives, and i've been having a few fault issues, which is a seperate issue, but how do I get EMC to detect the problem and stop the running program when a servo drive faults out? there is the err/res terminal, which increments 0.04 volts for every step it's out of sync, and if it loses sync gets to about 4volts, obviously i realize parallel port pins are not analog, but it might work as an i
even if it does work as an input, I can't seem to find the doc page to setup the HAL file to take that parallel port as some kind of incoming fault switch..
sounds like you could use an analog comparator to generate a digital signal you could feed into the parport.
that could cause an estop, or could reduce the feed rate by half, or whatever else you care to configure in your HAL
in an ideal world, you'd replace the geckos with real servo amps that take velocity or torque analog command, and send the encoder feedback directly to EMC
yea, i need it to cause an estop
hmm, you can send the encoder feedback straight to emc, i didn't know that
it can read quadrature encoders?
yes in many different ways
to read encoders quickly you need some specialized hardware. there are several options.
(of couse you need specialized hardware to generate the analog +-10V signal too)
but when you're done, you have a closed loop servo system like real cnc machines have
oh, so you need some hardware, yea
are you working on a large machine?
the gecko drives don't have a fault out pin?
skunkworks: it has a ERR/RESET pin, which is an analog output of the error
but other errors can occur which would causea fault that would not show up with that
like a limit error
i mean, i could open it up and just connect something ot the damn LED that shows fault
well, that's what I really want, and it's more or less digital, on or off, and detects all their different tyupes of faults, i just don't know what voltage it is
i might open one up and see how much is involved with adding a wire to it, i'm sure it's not to much, i'd rather not modify it... but oh well
if you want to spend a little money - you could look at jonE stuff. It is made to take the gecko encoders and give you actual position.
jonE stuff? you have a link to what you're talking about
i'm actually building a much bigger much more expensive machine which I have no problem with spending some money on
this is my old machine, but I keep getting faults for whatever reason, and I was just hoping to easily detect it in EMC. Like my X axis kept stopping, and then of course once it started the Y axis move, it was in the wrong position, and it's very scary when it tries to cut through way to much aluminum
so i was just hoping to shut it down, and hopefully fix the problem in the next machine with new servo drives
gecko is supposed to come out with new servo drives in a few months, I had figured I might use though, but i'm open to something that integrates with EMC better
I think you need this http://jelinux.pico-systems.com/univstep.html
and this http://jelinux.pico-systems.com/gecko.html
But I would email him.
you will hear many opinions, but people who have software that can do better, or who are working with larger machinery, generally agree that step/dir servo drives are a poor solution
* skunkworks doesn't like step/dir
they are a good solution for an incremental improvement of a stepper machine run by stepper-only software
If your starting from scratch.. Mesa card with servo amps/encoders.
why is step/dir bad? because ocne yer moving it's technically very un-efficient? since yer sending thousands of pulses?
klickrr: limited pulse rates, and no real feedback into the software (like you are discovering)
with a closed loop servo system you would be able to graph and tune your exact following error throughout the move
and - of cource - emc can do this also.. :)
following error trips when handled in emc are proportional to the velocity - in other words if you are moving fast, more FE is allowed without a trip
gecko does have their G-Rex 100, or whatever, it's a system to watch all the encoders like you are saying, kinda like what you are showing, although i doubt anyone's integrated it with emc, so maybe the other choice would be easier, probably cheaper too
the g-rex still does not involve the PC in the servo loop. it is not "the emc way"
oh and they have sample configs
sample configs for what?
i was just saying it was nice that the one drive he linked to above has sample configs for it
i see the second link is really almost what i'm asking for
No - not integrated with emc - G-grex is a motion controller.. Anything that takes motion out of emc - is not a good match for it.
it's really exactly what i'm asking for
I think you need both - one is a hardware step generator.. the other is the gecko encoder interface.
the ERR/RES terminal on the geckos is not a proportional output
it's a little weird to use because it's both an input and an output
there were some schematics how to hook it up to the PC
I think in the files section around the geckodrive list at yahho
you're supposed to be able to just hook all the err/res terminals together, then run that back to the PC, but I think it doesn't quite work if you do
also, the geckos drop encoder power when you hit that terminal, so you'll lose position
so is pico-systems.com run from someone in this channel?
he's not on now, but he's an early adopter/contributer to EMC
contributor, that is
He is very friendly - give him an email.
cradek: interesting, neat (sprung nut)
so what drive systems do you guys usually use? doesn't seem like gecko is high on anyone's list eheh
* alex_joni uses geckos
and I quite like them
yea, i like mine
but as with any system, you gotta know what the intended target is
can't compare them ,steppers to AC servo motors
i used their stepper drives years ago, then switched over to the servo drives, worked for years, just recently i've been having some fault issues, but i'm sure that's probably really a tuning issue I need to take care of
there isn't a lot of tunin gyou can do with the geckos - that's one reason they only get a lukewarm reception here :)
EMC actually has the capability of closing the servo loop itself, so using step/dir servo drives isn't necessary
it's only necessary if you have an open-loop controller that can't do the PID itself
so geckos are a solution in search of a problem around here
ok, so do you suggest using the pico-systems.com thing with gecko? or something completly different
that's the system I bought - pico systems USC + their gecko servo interface
if you have the geckos, it probably doesn't make sense to replace them
yea, that's what i'm planning to get, looks good to me
looks like a lot more wires to hookup eheh, but oh well, i guess if i want it to work right
you'd still need a servo drive and an interface to it (like a velocity-mode drive and an analog / PWM hardware output)
well the USC from pico-systems looks liek it hooks straight upto EMC right
yes, there are sample configs included
has all the configs right there
the only problems people have are with the parallel port - it could take a little fiddling in the BIOS to make sure it's in the right mode
when I bought mine, he sent the USC and GSI already wired together, so that made it a little easier :)
well i use the parallel port right now directly to the gecko's, or does that not mean it always works
ahh, yea, i may ask for that
that would be nice
parallel port for step generation and parallel port for EPP data transfers to the USC are different problems
oh, it needs to be epp, ok
I think so
Did I mention how cool emc is? I think it has been a few weeks.
* archivist warms emc a bit, hey emc it hot!
so cool it's hot
* BigJohnT uses Geckos
Gecko stands behind their product 110%
some guy hacked up a couple of drives with side cutters and twiddled with the pots till it wouldn't run and Gecko still gave him is money back
Then Marris checked out the drives to make sure they were ok and had a contest on CNC Zone and gave them away...
Save 10% on enco weborders for a week code WB389 if anyone wants
it's hard to choose between 10% off and free shipping
depends on the budget :)
lots of neat stuff in this video http://www.youtube.com/watch?v=qCmWZY_00_I&feature=related
cnc metal spinning and forming
Spinning is pretty cool
I like when they make the serpintine style belt pull with the gear
Seeing metal flow like that still impresses me. Looks like a trick
i always wondered how toroid coils were made: http://www.youtube.com/watch?v=BdvoUAta92w&feature=related
well... that's like magic
looks a lot better than the straight shuttle manual method
when the local transformer company shut we missed out on the toroid machine
I like - smoosh the fiber tape roll so it fits thru the center of the toroid
doesn't work so well with fat solid wire
wire is a different matter entirely..
here's an explanation of how it works (about 1/3 down) http://www.physicsforums.com/archive/index.php/t-130135.html
i dont really see why you would want a smaller diameter hoop
So - you wind it off onto the big loop - then on to the toriod?
yes, but since the toroid is not on the hoop center the distance from the toroid to the point the wire exits the hoop will change
so you need some mechanism in the hoop to reel the wire in and out a little bit each time it goes around
i think just a spring arm coming out of the inside of the loop should work, but i dont see anything like that on the videos
The tape winding head though - pulls the tape from the outside of the ring.. Seems to be 'realtime'. (you don't have to put it on the ring first - then onto the toroid)
you do have to put it on the ring first
notice the scrap he pulls out after cutting off the end
So.. He is putting the tape on for the next run?
You can see the tape being put onto something :)
oop i just watched it again, you're right
Heh - yes - you actaully see him feed the tape thru the rig right onto the toroid
but the tape gets cut because the distance traveled around the hoop is much larger than the distance around the toroid
So I was right - it is magic ;)
i think the wire gets loaded a different way
there are some other videos
iirc,mount toroid, put the correct length of wire/tape on the ring, wind, remove
ah ha. Here is one showing it being put on the ring - then off onto the toroid
[19:42:24] <skunkworks> http://www.youtube.com/watch?v=_EJ0nzvgN-E&feature=related
nohi: what is up?
How is poland these days?
spring soon, snow and rain :)
heh - same here
for the record same in southern sweden :)
OoBIGeye_ is now known as OoBIGeye
when i logged back in, there was a message about registering my nick, how do i do that?
there is an faq on the freenode website
or /msg nickserv help
but when i do as the web pages says, i'm already registered is the response, so i guess i must be ok
I finally got the box wrapped around the PMDX-106 and the VSR from the mill, so now I have to finish building the swarf shielding I guess
and as near as I can tell, its all worrking.
However its pouring it out of a boot here, so I think I quit for the day. One could say its beer-thirty :)
anyone worked a lot with multi-threading ? (programming, not the machining kind of threading :)
a lot not, what's the problem ?
micges: I was just wondering about multi-threading recursive functions. is that normally done? if so, how?
I imagine not normally but could be done with extreme care
It's suicide I think
well, there's other stuff I can practice multi-threading on first. Things like function calls inside a for-loop which are fairly independent of eachother (not much need for communication btw. threads)
I can think of a few applications, just the launcher would need to wait for all threads to finish before it returns/finishes
did jmk advertize this video he made? http://www.youtube.com/watch?v=ACvRilmIKDQ
yup saw it the other day
anyone know why g83 doesen't work adn g84 does after i upgraded to emc2
oops got that backwards...g84 doesn't
does two sided milling work with EMC ?