hey JymmmEMC , got those drives all fired up yet?
a-l-p-h-a is now known as a-l-p-h-a
a-l-p-h-a is now known as a-l-p-h-a
a-l-p-h-a2 is now known as a-l-p-h-a
a-l-p-h-a is now known as a-l-p-h-a
03jepler 07TRUNK * 10emc2/src/hal/drivers/probe_parport.c:
change "no PNP ports detected" into a warning, so that probe_parport
can always be loaded whether or not the parport being used requires it
a-l-p-h-a is now known as a-l-p-h-a
dot matrix bicycle printer... row of paint cans on rear of bike print messages on sidewalk http://www.bikesagainstbush.com/blog/
today i thought that alex's tri-kins could eliminate the wobble in hektor.
I've already googled "Trig for dumbasses" and no luck, any suggestions?
Jymmm: do you have a specific question?
cradek: Thanks but no. I'd actually like to learn trig so I can understand things much better.
this looks nice: http://www.clarku.edu/~djoyce/trig/
I've dl'ed a few things, abd seen some on the web.. but just didn't hit me too well.
Hmmm, maybe I need to step back to geometry
I'm bad at learning from books/websites unless I have a problem I want to solve - only that keeps me interested long enough
cradek: Besides the CNC stuff, I'm also doing some GIS things too, both got trig, so I figure I better learn it.
sin^2+cos^2 = 1
COS 0 = 1
that and some algebra gives you almost everything you need to know :)
understanding the unit circle is nice
I think you need sin/cos = tan too
COS 0 = 1 <---- Not that I actually know what that means
Jymmm, you'll have to be diligent in reading and understanding some tutorial like the one cradek linked
as cradek said, it isn't easy to get through, especially if you don't have a target in mind
SWPadnos: It's already over my head... goin down to geomtry now
personally, I think I find trig easier - geometry in school was all about proofs, and I can't stand doing those
I've wanted/needed to learn trig for a long time, so as long as I can get thru the medial stuff, I'll be ok.
SWPadnos: eh just pick out the similar triangles and you're done
yeah - prove it! ;)
"corresponding parts of similar triangles are similar"
circular proof! circular proof!
whats a triangle =)
I wonder how much trouble it would be to enroll in just one college or community college course
those funny hats they used to wear
that's a sure way to stick to it
heh - pay $1000 to get interested
cradek: But I have much more free time available when I'm at work
Jymmm, do you have access to a Linux machine?
* Jymmm reaches for knoppix usb stick in pocket... yep!
good luck Jymmm
G'ngiht cradek, thanks =)
ok - there's a program - I'm looking for the name - that lets you make geometric constructions
see you cradek
like click to make a point, then make another and you can construct a line between them
it has stuff like arcs, vectors, lines, half-lines ...
and you can drag a point around to see what happens - all the stiff dependent on the point will update as you drag
cool... remember the name yet? I know it isn't a food name, that's for sure =)
[03:30:22] <Jymmm> http://en.wikipedia.org/wiki/Interactive_geometry_software
one of those?
[03:31:21] <Jymmm> http://edu.kde.org/kig/screenshots.php
Hmmm, lets see if it's already on the kubuntu livecd...
doh, let my backpack in the car.
I don't think kig will teach you much - it's not a tutorial program or anything
but if you read through some other tutorial, you can do the constructions with kig and see how they change when you move stuff
I used that to make diagrams for cutter comp
(when trying to figure out how it should work :) )
time for bed. good night (and good luck)
or read the machinery handbook
fine. CRC hand book then :P
* Skullworks-PGAB might start a war, but...
WAR WAR WAR WAR
which is better under RT - Nvidia TNT2 M64 or an ATi-9200
I could win it easily, but I won't be a jerk.... or will i???
jerk all you want...
* Skullworks-PGAB feels safely out of range.
just building up a sim box to start porting Fanuc macros for EMC use
pew pew pew
have a small pile of graphic cards to choose from
but then the sys is built on an Athlon (mobile) XP2200+
again more old stuff
Skullworks-PGAB: Wanna Bet?! HAHAHAHAHA
alex__ is now known as TheBall
ok -recently mfg old designed stuff... :)
a-l-p-h-a is now known as a-l-p-h-a
a-l-p-h-a2 is now known as a-l-p-h-a
a-l-p-h-a is now known as a-l-p-h-a
a-l-p-h-a2 is now known as a-l-p-h-a
a-l-p-h-a is now known as a-l-p-h-a
a-l-p-h-a2 is now known as a-l-p-h-a
g'day mr Joni
got one manuscript submitted yesterday - so the boss is a bit relaxed for a while again :)
worked a bit on zooming/rotating/panning the OpenGL/C# thing
these guys have a 30-day trial available http://www.devdept.com/eyeshot/
so I figured they probably thought about the design and the API a lot
so it makes sense to copy their functionality to some extent
a-l-p-h-a2 is now known as a-l-p-h-a
a-l-p-h-a2 is now known as a-l-p-h-a
jepler are you there?
alex: do you know what is the "TO_PROG_LEN" macro (if it is a macro...)
a-l-p-h-a2 is now known as a-l-p-h-a
it is a conversion macro from internal units to program units
if you have G21 programmed, it converts to mm
the _LEN part of it tells you that it refers to linear units
there are some macros referring to angular units aswell
I ask because I've a problem with my nurbs patch...it works if the home is equal to the workpiece origin...if not, I've problem with it...but I think jepler knows more about it because he guied me when I wrote that patch
can you put the patch online?
jepler will probably be online in 1-2 hours
xemet: just a guess, but the function "find_ends" applies the relevant offets such as tool offset, g5x offset, and g92 offset to the coordinates found on a particular line of gcode.
xemet: if this function is not used, and block->x_number is used instead, you would not get the right behavior when there are offsets in effect
I got up early today
seems like it
at the office before 7AM -- not too common for me
looking at STRAIGHT_TRAVERSE() I see:
x = FROM_PROG_LEN(x);
x += programOrigin.x;
x += currentXToolOffset;
linearMoveMsg.end.tran.x = TO_EXT_LEN(x);
oh -- maybe what I said was mistaken, then
I see now that Interp::find_ends is in charge of *removing* all those offsets if the move is in G53
this is all happening in emccanon.cc
interesting -- I didn't know g53 "canceled" tool length offset as well. I wonder what the documentation says.
it's unclear from the docs whether a tool length offset should affect the destination of a move under G53
jepler: 7am .. hmm, I got here at around 10:15 today :P
alex_joni: yeah but it's well known that you are a lazy good-for-nothing bastard.
you're probably right :)
might be a reason why I'm still at work at 6-7pm
xemet: so now I think alex_joni is probably right -- you need to add the offsets to the coordinates yourself, inside the canon call
jepler, alex: thanks! I'm going to read the STRAIGHT_TRAVERSE, I hope I will undersand
ah, the problem is with Z
if the Home Z is the same as th workpiece Z origin, everything woeks
if not the Z start to go at sideral distances...
where is it defined the find_ends function? How much parameter can I put inside?
sorry, I found it
jepler: I see that I've used the find_ends function...the problem is with the Z. The fact is, the G.2 or G.3 doesn't modify Z, because the nurbs are on plane. It should take the current_z position and pass that to the ARC_FEED right?
Also, the axis preview is in the right place, and the quotes for X, Y and Z dimensions are ok and within the limits
So I think I've a problem with the ARC_FEED function
xemet: I really think you need to update/convert the position from your function
yes, but the problem is that in my function Z is not nominated
hm. ARC_FEED is going to add on the offets *again*
you either have to subtract the offsets then, or split ARC_FEED into 2 functions
Guest618 is now known as skunkworks_
jepler: I used your function arc() that is the final passage that calls ARC_FEED after biarc calculations
I guess my function wasn't perfect, then
here is the patch...probably it sucks because I'm NOT a programmer and maybe should not try to do those things...:) http://www.pastebin.ca/640476
it includes the spline patch from jepler
at the 591 there is the arc() function that lcalls ARC_FEED, it is the only part where a Z is given to make a move
what is "canonEndPoint.z"?
I think that's the target Z position
canonEndPoint is the last programmed end point, stored in case it's
needed for subsequent calculations. It's in absolute frame, mm units.
xemet: emccanon.cc has some descriptions of these at the top of the file
also the macros for various conversions are described and implemented there
that doesn't mean you shouldn't ask in here when there's something unclear ;)
wow, I'm understandind something...
somthing about the error
X Y moves are correct
I wish I had time to try to help, but my real job is calling me today..
jepler, don't worry
it is like every arc the Z is increased
my best guess is that you need to subtract the offsets from z that ARC_FEED will add again
I'm getting something like the elical moce!
look at ARC_FEED, find what it does to modify Z, and then do the opposite (subtract instead of add) in the function arc()
I will try!
I had to subtract programOrigin.z from Z when calling ARC_FEED
you were right
alex do you have access to the linuxcnc.org website ?
wpuld you like to add a link to the italian live CD in the italian version
jepler added it in the wiki
but in 15 minutes.. a bit busy right now
ok, I tell you the page: http://www.linuxcnc.org/content/view/21/4/lang,it/
at the first point where it says: "Basta scaricare l'immagine ISO (EU mirror) e masterizzarla su un CD. " You could add: "E' anche possibile scaricare una versione in italiano del LiveCD" and link it to: http://learning.dtpm.unipa.it/emc/emc.html
argh, I think my smbd has a memory leak
wow. I think Windows Vista has really had a good effect on memory prices
even laptop memory (SO-DIMMs) is around $40/gig
you mean vista uses less so there is an over supply
err - no :)
hehe I thought not
it is about time for an epoxy plant to blow up..
shhhh - don't tell anyone
trouble is I need new boxes to fit all this cheap ram in
yeah - "legacy" SDRAM isn't as cheap
* alex_joni runs away for a while
run alex_joni run. ;)
could be a run lola run reference - or a forest gump
03cradek 07TRUNK * 10emc2/src/emc/task/emccanon.cc: oops!
the log message shouldn't describe the change; it should describe why you made the change... This does an OK job of that actually
lerman: "lerman", mailbox "Mail/emc-commit": Can't write to mailbox
deliver: can't create lockfile Mail/emc-commi.lock: No such file or directory
lerman: ^^ you've been unsubscribed from -commit because of these fatal bounces
you know, it's really cool to be able to vnc into your oscilloscope
SWPadnos: you have a scope that lets you do this?
yep - the MSO6104A
hm that's a little out of my price range
"Manufacturer list price: $16,500"
ah well just have to wait for the secondhand price to lower a bit
I got my first oscope recently, I'm still learning to use it.
* skunkworks_ is still learning. Don't use it enough
I just recently ebay'd an old digital storage scope, I hope it doesn't suck too badly to use.
ooh which one?
HP/Agilent 54502 (2 channel + 1 trigger, 400mSa/s, 2K memory per channel)
jep, I got the 54001D
it seems ok so far, but I dont really have anything to compare it to.
HP is generally nice stuff
* skunkworks_ still uses a 2 channel B&K
there's a pdf of the manual online somewhere.
yeah I found some PDF manuals for this scope on agilent's site
I was reading the manual last night, seems like it's worth the time to read.
jepler: looks cool, hope you like it
jepler, did it come with input pods?
the 5000 series is mostly the same as the 6000 series, a little lower bandwidth, smaller size (which is nice), but no option for digital channels (the MSO option)
I didn't realize they were seperate, then I had to find pods elsewhere.
bill2or3: this is an "A", so it doesn't have the capability. unfortunately.
there's only the two analog channels
that's a feature, if you dont have to spend another $100 on pods. :-)
darn, I want a logic analyzer once every year or two
or another $2000 on the scope ;)
cradek, I've got an HP 54622D I can sell you
a logic analyser is very usefull
but I'm not at the point where I want to let it go for a song yet :)
bill2or3: if you didn't look at the information on agilent's site for your scope yet, you should do so.
SWPadnos: here's one on ebay for $2200, so I'll preemptively say "no thanks"
bill2or3: not sure where you found the manual you looked at..
* bill2or3 goes to look
I think it was at hp's site, actually.
or aligent, but I think it was an 'official' pdf, not just some scans.
I'd love to replace the 6104A with an 8104A, but those are $23k+
I do ok still with my analog storage scope, I like it and it's rare that I want more than it can do
bill2or3: the stuff on agilent's site is "just scans", but at least it's not some jerk on ebay who wants to charge you for them
in fact for the 54502A, one of the pages had somebody's notes written on it
for some older stuff I have manuals
I *think* that if I can scare up a GPIB interface for my PC, it's possible to read the waveform data out of this scope, which would sure be neat
I like that you can plug in a plotter.
anybody know anything about GPIB interfaces and linux? can a USB or PCI interface card that works with linux be had for less than an arm and a leg?
go back in time 10 years and say something, and I have one you can have
not that I have one, but it amuses me.
where are you going to find a GPIB printer, though?
(cradek, do we have one in the office basement somewhere?)
jepler: I don't think so anymore
in the antique section?
I have my old pet gpib printer
jepler: and my plotter has stopped working...
PET as in Commodore?
cradek: drat and drat
* bill2or3 wants a plotter.
to label cd's & stuff.
I want a usb to gpib
huh, I didn't know Commodore did anything with GPIB.
looks like there's a pretty long list of interfaces that are said to work: http://linux-gpib.sourceforge.net/doc_html/x259.html#HARDWARE-MATRIX
it was standard even the disk drives were gpib
[16:33:11] <bill2or3> http://sine.ni.com/nips/cds/view/p/lang/en/nid/201586
there you go.
archivist: strange, I only remember the clocked serial of Commodore 64 / Vic 20 floppy drives.
good luck with the drivers. :-)
jepler too recent it was the PET range
eek no way I'm spending $529 for that
jepler: if you're interested we could try again to fix my plotter, it must be a pretty basic problem but I didn't find it in 15 minutes of trying
the "desktop" 11x17 one
here's plans for a DIY usb-gpib converter: http://lpvo.fe.uni-lj.si/gpib
it says ~US $50 in parts
cradek, is it an X-Y plotter, where the paper is stationary?
archivist: 7750 I think
why have I not heard of gpib? before my time?
bill2or3: ooh, that's much more interesting to me. looking.
cradek I scrapped an HP one a few bits still lying around
jepler, there's some on ebay too, for $50ish
skunk, it's certainly before *my* time.
archivist: do you know if it worked?
[16:38:53] <cradek> http://www.mjs-electronics.se/images/Data_tele/hp7550b.jpg
archivist: I checked all the obvious sensors/switches, so it might be something on the main board (there's just one big board I think)
yes similar to that dunno if the pcb is still at home
was just one pcb
before you bother doing anything, let me try to find what is wrong for sure
ok as its 3000+ miles from you
I was in a "I need the space" mood and to see what motors were in it
"All of Commodore's post-PET/CBM 8-bit machines, from the VIC-20 to the C128, utilized a proprietary 'serial IEEE-488' for peripherals" -- wikipedia
I have a few intruments I would love to connect up to a PC
they are apparently not rare
PCI are expensive
yeah, no way around the ISA thing I bet
then you need a USB/ISA converter
I still have an ISA machine together
yeah ISA slots are still everywhere
wonder if there's linux support for that exact card
I keep an ISA box just for PCAD a pcb design program
archivist: if it doesn't need special dedicated hardware, have you tried qemu or vmware to run it?
qemu plays "ISA machine" very nicely
(I use freedos with the ms lanman client)
it writes /reads a dongle
cradek: "hp_82341.ko" is one of the modules listed at http://linux-gpib.sourceforge.net/doc_html/x259.html#HARDWARE-MATRIX
and direct to vga
it's fairly easy to write pretend hardware in qemu, you can hook into the port IO
jepler: I'd sure go that way then, it seems like a fair price
I'm also tempted to find enough information about HPIB to write an interface for 5i20 or pluto-p .. then it's just a matter of wiring the connector
I have the driver info for PCAD but never get the time to really get into it
no, it's a matter of finding the connector
thats the easy bit
for me anyway
I have enough cables to rob a socket off
true I forgot the cables are both genders
and I have a socket on an old Epson printer adapter card
TX80 and RX80 had IEE488 cards available
looks like you can still track down connectors: http://www.l-com.com/item.aspx?id=4326
I'm pretty sure I have a few of the cables at home ... somewhere
jepler nice find , RS in the uk were describing type 57 as IEE488 for a while
i love how they dont show the actual connector part in the photo
its like a portrait of the back of someone's head
fenn: did you have any sketches of the hexapod your making?
jepler thats not real IEE488 the screws are metric
thats a probably a screw lock type 57
its an amrican thread in the standard
pew pew pew
at work, i am in the diamond turning room
it is very similar in appearance/layout to the assembly rooms
and sometimes assembly guys work in there because there's no space
such was the case yesterday
except the assembly guys are party animals and everyone in our room wants it dead silent at all times
this dude brought in a boom box
i am suprised nobody killed him
not his day to make new friends.
it is so hard to work with music going
i just turned it off every time he left the room
he eventually got the message
a bit passive agressive?
i just didn't want to talk to him
what do you think - sell or keep? http://www.e-engraving.com/machines/new_hermes/New_Hermes_Vanguard_7200.htm
i guess it's just the room, after you are in there for awhile you just start avoiding conversation
like that day we went six hours before someone said something
sounds like heaven.
it freaks people out
s/diamond turning room/Need to get a life room/
it is a very nice place to work, i love it in there
it's cold, quiet
you can read
lots of spare time
sounds like a cemetary to me
is it at 68F?
they try to make it 68f
they pretend it is temperature controlled
are cardboard allowed in there?
skunkworks does it work?
cardboard is allowed, why?
JymmmEMC: don't know yet - picking it up today.
cardboard fibres are suppose to be bad if there is any precise inspection going on
any fiber is bad
but like, this isn't a real precision area
it's like, a faux precision area
it's way more precise than most places have, but it's still not really "right"
so there are tombstone blanks instead of granite tables?
skunkworks it's max 25", keep it, you could always make PCB's
hahah actually some of our surface plates are weird shapes
we have two really big plates from boeing, one is 12' x 3' x 3'
JymmmEMC: that was one plan for it.. :)
that got turned into a machine
the other one is like 6' x 4' x 3'
how do you move 12' x 3' x 3' of granite?
sounds hard to sweep under.
there is no "under," it sits right on the floor
ds2: there's a weird rigging forklift
for very heavy machines/etc
isn't 12' too much overhang for a fork?
it's not a real forklift, it's like, a mining truck with two very large vertical hydraulic rams
and on top of those rams are "forks"
and they run i-beams on top of those forks, 12' long
skunkworks: optional vinyl sign cutter
and cradle it that way
skunkworks just be sure to get the original software that it came with.
108 cubic feet of granite just sounds dangerous
I don't think the one I am getting is 'open archetecture
ds2: it's pretty big, yeah
that's 4 cubic yards!
skunkworks That's why I said get the original software (on CD)
probally more then what a concrete truck can carry
and it has a dinky little cutting head that rides on it
it came on a lowboy, as i am told
50 ton lowboys can hold a lot of stuff
JymmmEMC: for the price I paid - I really didn't know what I was getting - I was buying the x/y/z/ table. I really don't know what comes with it. (univeristy surplus) I will find out today. :)
you can probally survive a nuclear attack by staying behind it!
skunkworks be sure to ask if it came with a security dongle too (just in case)
I will definalty try to get the software if it isn't there
dad used to work there - so he has some pull :)
skunkworks what you pay for it?
skunkworks_: wouldn't it be easier to emc retrofit it?
lerneaen_hydra: depends on if I sell it. I am thinking it is worth quite a bit more than I payed... That is the problem
skunkworks Oh, it accepts HPGL - cool beans
skunkworks_: ah, I see
JymmmEMC: I don't know if the software from the site is what comes with mine.. I think it is older.
I will totally give you $140.
that's $10 pure profit!
skunkworks If they have two, I'll take the 2nd one for $150
:0 they only had one. sorry
skunkworks Ok, I'll take the 1st one for $150 =)
* archivist sets up an auction
I bet I would get more than one bidder ;)
too rich for me, it's all yours.
looking at it quick yesterday - it seems to have rolled ball screws
worth > $140 just for parts.
but then labor disassembly!
JymmmEMC: I have no clue if it works.. yet
JymmmEMC: no - later though
hmm, why would it have ballscrews? the reported accuracy is only 0.15mm, easily acheiveable with acme screws
no clue - as I say - it was just from the short time I looked at it yesterday
I could be wrong - I have found little or no info on the web
* skunkworks_ is hoping there is also manuals with it..
lerneaen_hydra: where did you see that - the specs say repeatablity of .0006"
skunkworks Take LOTS of pics!
skunkworks_: hmm, must have misread the small text ;)
repeatability and accuracy are so different
JymmmEMC: I plan on it...
skunkworks: If you could, toss a steel rule up againest the cutting tool
may have repeatablility of .0006 but accuracy of .01 ;)
JymmmEMC: say again?
accuracy depends on material and speed too
skunkworks put a steel rule next to the cutting tool when you take pics.
[18:40:54] <skunkworks_> http://minneapolis.craigslist.org:80/bfs/356902482.html
i think it's safe to assume someone's talking about free tool accuracy when discussing the limits of a machine
unless they specifify otherwise, of course.
hmm 120ipm - cool
I don't think I have the vaccum table option though come to think of it.
vacuum tables a+
now there's repeatability.
hrm.. drop-cutter is sort of working(-ish) http://imagebin.org/9632
still some strange bug left...
might be the left vs right handedness of the ordering of triangle vertices
skunkworks: i might have made some sketches but nothing worth looking at really
awallin: what's a "drop cutter" ?
JymmmEMC: a way of calculating finish or semifinish toolpaths for STL models. you imagine dropping down the cutter along the z-axis until it touches the model
did you do it in the vector realm or did you scan convert the model?
(I wrote the latter and it works but is pretty memory intensive)
it tests against the three vertices, the three edges, and the facet. the cutter representation is 'exact' (forumulas for cylinder/sphere/torus)
for people cutting with donuts
a filleted endmill
awallin: Is this to overcome something specific to STL files?
mmmmmmm - dnonuts
something between a flat endmill and a ball-endmill
JymmmEMC: ? I don't understand the question
the reason the engraver has ballscrews is cause backlash is way more important at those scales than raw precision
I thought of that, but didn't want to say anything ;)
if your backlash is twice the size of the engraving bit *snap*
awallin: I don't under it's purpose (drop cutter that is), compared to just taking an offset for a finishing pass (as example).
JymmmEMC: think about the shape of the tool
ball, end, or either?
any shape of tool needs a path that's not just an offset
you have to offset in 3d
JymmmEMC: ah. the reason is that it becoms very very hard (some would say just 'non-trivial' :) to calculate an offset surface from a model consisting of many different surfaces (think planes, spheres, nurbs, whatever). So internally most CAM programs convert everything to an STL representation before calculating paths
it's sort of a 'stupid' brute force approach. but that is what most commercial programs use I understand
geez - I hope they're not all that stupid
it's not even the brute-force-ness of it, it results in a loss of precision sonce the STL is a sampled version of the mathematical surfaces of a typical CAD model
awallin: Ok, gotcha. And you are trying to accomplish what exactly? Or maybe I should ask "in what"
yeah that's the scary part
(most parts don't actually have a bunch of triangular facets)
not that I disagree about that probably being what most programs do
so you make some faceted bumps and then emc rounds them over and you're back where you started :)
was reading a mag about an aircraft part maker and they hand finish because of the errors
[19:40:56] <JymmmEMC> http://rpdrc.ic.polyu.edu.hk/old_files/stl_introduction.htm
well, if the triangulation is good to 20 or 5 microns, what does it matter in the end.
So basically, everything in a STL file is (basically) all triangles?
i bet its much less of an issue now with 1GB of ram and GHz's
JymmmEMC: basically, yes
technically, the STL fomat can use any polygon, but triangles are the de-facto standard
i thought it had to use triangles
(ie, it seems nobody does quadragons or higher)
nope, the standard allows any number of vertices (>=3, I think)
SWPadnos: Sounds like too much of a PITA to
yes but the number of passes gets silly with small triangles
the number of passes and sampling rate depends on your cutter size and what scallop height you can tolerate.
You could make a ball using triangles?
it's more complex to use anything other than triangles - the polygons are supposed to be planar, and you can only guarnatee that with triangles (since rounding can make a 4-gon or higher non-planar)
that's identical wether you calculate the path from a triangulation or from an analytical offset surface
Could you make a ball using triangles?
is Dallur around?
JymmmEMC: even a penguin! http://www.anderswallin.net/wp-content/2007_07stl/tux.png
JymmmEMC, see: geodesic domes
nice answer :p
alex_joni: I just get really mad when people start spewing bile like that
[19:47:07] <alex_joni> http://eneas.juve.ro/~juve/weekend%20764.jpg
SWPadnos: Not very effect though it seems.
computers are good at doing lots of ineffective things to simulate something effective
JymmmEMC: it was only twice as big as me :P
alex_joni: If it wasn't for that shirt, looks like you're ready to go work on the ranch =)
it was my second time on a horse
(and the watch too)
the first time was a day before :P
alex_joni: Ah, gotcha. saddle sore?
alex_joni: keep you heals down =)
way less than I expected
alex_joni: I like horseback riding... haven't done it in years though.
well, I like it too.. decided I'll do it again
alex_joni: Did you get a chance to "break away" from everyone else?
only for a short while
we were all beginners
[19:54:31] <alex_joni> http://eneas.juve.ro/~juve/weekend%20485.jpg
alex_joni: it's all good. There's a place in SoCal that rents, and you can go anywhere in their 30-50 acres of marshland, trails, streams, etc.
these guys had some nice offers too
you could explore some mountains nearby with a guide
about 5-6 hours of riding.. but the guy said it's a bit too much for first timers
so maybe next year
there ya go
until then I'll practise around here
this was actually about 400 miles away from home..
hmmm. is that epsilon-lambda-lambda-alpha-sigma? I'm not so good with lower case greek letters :)
or that stylized font anyway
ok, I guess I got it right the first time :)
that's the greek name for greece
I wish I could find my book on greek ancient history and stuff
but I think its ancient history :)
there's such a difference
between ancient greek history and the present day greece
this was soem book I had when I was maybe 5 or 6, and it had the alphabet and a bunch of interesting information about the ancients
nice pictures too?
I think so - it's been 30+ years since I've seen it :)
might have been a reason to read it :P
I could be confusing it with the book on ancient egypt though
bet you didn't learn that alphabet
alex_joni: Mmmmmmmmm LUNCH!
they had about 300 horses
about 10 very nice stallions
(which I didn't get to ride :P)
alex_joni: next year =)
think I'll wait a couple more years for that
mommas, don't let your babies grow up to be cowboys.
alex_joni: you don't pick guitar or drive an old truck do you?
alex_joni: Got Boots?
JymmmEMC: nope, not yet either
Case mod complete... I REALLY need to find more things to use my angle grinder on
make a cast iron surface plate
make a toothpick
make a spare angleginder
I try to go for distance
so far, 20 ft
I miss having Les around.
how long does it take to make a factory anyway
the workers take the longest imo
"Sim Grinder - Create a living world using only an angle grinder"
if you had a 18v cordless one you could use it all over town.
bill2or3: I think we have a failure to communicate!
bill2or3: did you ever see the movie?
uhm, I have no idea what you're talking about.
but I just got back from lunch, and probablly lack context.
"...like a child that wanders into the middle of a movie."
LOL, I'll take that as a no, and that you've never heard that refernce either.
where am I?
bill2or3: Cool Hand Luke
use the force
at the VERY beginning actually
I'll netflix it.
Where he takes a pipe cutter and cuts off all the parking meters
just s/pipe cutter/cordless grinder/
oh, I think I vaguely remember that.
bill2or3: it's a good movie
* alex_joni is gone to bed
good night all
alex_joni: See ya in a few!
Gawd I want a laser!
No sharks, just a laser
200W CO2 would work
then you can make your own robo-sharks.
The 200W one has manual Z, WTH?!
The china import is only 120W =(
how're you supposed to make a passable roboshark with only 120W?
bill2or3: Not out of wood that's for sure.
plastic, no problem
good, more waterproof.
I was going to buy a used laser from a place that makes custom wooden urns becasue he needed Ug Ug Ug MORE POWER!
After seeing the work he does, that made me want more power becasue it was simular I wanted to do.
Man did he do some AWESOME work too.
This is the one I've been considering (120W) http://imagebin.org/9634
lots, I bet.
~ $7000 with spare parts
That includes the freight costs, and I found out that we have a a free-trade agreement, so I *think* there isn't any duties involved.
I'm just chicken/leary of wiring money to China and having to wait 30 days for it to arrive.
[20:41:56] <a-l-p-h-a> http://www.use-enco.com/CGI/INPAGE?DCMP=EMC-1115705&PMPAGE=/specials/890-9841
This isn't too shabby... http://www.logicsupply.com/products/d201gly
you should get that router working
fenn: I just took apart the X leadscrew... trying to figure out what kind of jig to make so I can shave off and square up the al block
hrm isnt a DTI supposed to read .0001's?
or am i not understanding their usage of 'graduations'
cheap stuff .001
I have one at .000001 iirc
wow is it electronic or something?
no a mechanical
on a very solid stand
i read about a dti that was based on sensing the capacitance between the probe and the work
it was like.. .001 to .000000001
I like measuring toys
hmmm. a capacitive probw should have a hard time being that accurate, since it would be very affected by things like humidity
well, that was the resolution
as a comparator maybe
it was used for measuring runout anyway
a local company does that sort of stuff for gas turbines
Dynamic resolution 0.0000001 inch p-p
[21:10:55] <fenn> http://www.globalspec.com/FeaturedProducts/Detail/mti/Capacitance_Displacement_Sensor/10
that wasn't it btw
Capacitance systems can make measurements in 100 µsec with resolutions to 10-7 in. (0.001 micron).
as quick as the freq its running at and resolution depends on the counter
Do most 1-2-3 blocks come with bolts?
(okey, the cheap imports that is)
my cheapies didn't.
mine have no holes
I just thought a 123 block would be a cheap way of me gauging when I have to file down this al block I need todo
JymmmEMC, are you trying for flat surfaces?
hmmm. squareness is a different matter
flatness tou can do with any 3 pieces of stock
I'm not sure how to do right angles
SWPadnos: all my stock sucks =)
But for $15 I thought 123 would be the way to go
machine a cylinder to make a square
any 3 pieces can be gotten flat to within your ability to (a) grind and (b) measure
SWPadnos: I'm more interested in square, than flat
I think you can't reliably measure squareness without also having flatness, but I could be wrong on that
SWPadnos: I took my ballscrew/nut to the mfg plant. From what they saw, he thought the wear was caused by side pressure on the nut
SWPadnos: After looking at the ballnut mounting block, I can see why - totally fscked up.
how big is it?
SWPadnos: No two sides are even square with each other. 2" x 2" x 0.5"
a cylinder accurately made is by definition square end to side (concave machine the endfaces)
archivist: no mill.
angle grinder!!!! ;)
JymmmEMC: what have you got? sandpaper and your hands?
I hear they're under utilized.
sandpaper and files
bill2or3: you got that right!
The REALLY suck thing, is if my machine was running, I could fix it on it.
Even if I shaved off too much (but was square), I could always shim it up.
or just anglegrind down the other end.
bill2or3: aluminum == no 20' flying sparks =(
[21:37:01] <JymmmEMC> http://www.chicagobrand.com/50301.htm
well - it worked for a bit.. then one of the transisters on the drive board smoked :)
get fixing then
it is a 486dx board with a hermes drive and powersupply
boots off of the hermes board
they are l298 drivers
[21:56:18] <skunkworks> http://www.electronicsam.com/images/engraver/rip.JPG
[21:56:27] <skunkworks> http://www.electronicsam.com/images/engraver/engraver.JPG
[21:56:54] <skunkworks> http://www.electronicsam.com/images/engraver/drive.JPG
this was made in 92 - I wonder if the caps are shorted out.
it power cycled a few times before it smoked - jogged it around and everything
check the caps
they look bulged
to me anyways
i bet that's what caused it
this would be one hell of a emc machine.. It has limits on each end of travel.. spring loaded nuts - the only thing is the steppers are not bipolar.
steppers suck anyway
why worry about how much they suck
be a good use for these :) http://www.electronicsam.com/images/KandT/servostart/ampmess.JPG
jymmm do your calipers not work or something?
fenn: yeah, they work =)
do you have a file and a vise?
tiny vise, but yeah
so, bolt the thing to a flat surface, measure the distance between the ballscrew and the surface
measure in two places along the length
file off a bit, shim back up to height
fenn: cant measure between ballnut and flat
zen and the art of cnc maintenance
hell you could use a ruler
fenn: it rotates
but the length isn't the issue, it's squareness
keep it still while you measure
archivist: heh, easier said than done
JymmmEMC: squareness of the bearing mount determines the angle of the ballscrew relative to the mounting surface right?
I hope that pic explains things a bit better
if you cant measure to the bottom of the channel, then measure to the top of the channel
which one's out of square? the unbolted one?
fenn: the channel is just a dust guard
its straight enough i bet
or you could use a separate straightedge
fenn: "striaght enough" has been the problem
and piece of wire/string from a side to screw at each end for frame to screw squareness
the ballnut mount attaches to the underside gantry plate (seen left)
what exactly is the problem again?
misalignment of the ballnut to the gantry plate
yea but.. does it not run or something?
yeah, it cause wear in the ballnut
That's what the guys at the mfg plant said.
what are the symptoms?
fenn: read the last two years of logs
fenn: or scroll up.... (14:30:05) JymmmEMC: SWPadnos: I took my ballscrew/nut to the mfg plant. From what they saw, he thought the wear was caused by side pressure on the nut
what i want to know is why you are worrying about this in the first place
did you see what I just pasted?
i mean i can understand trying to improve a crappy tool, but you keep saying it doesn't work
boy - it did a number to the circut board - is it bad I like the smell of burned electronics?
fenn: stalling for the last two years
skunkworks: I like it too, as well as the smell from using an angle grinder =)
the cheap chinese discs smell the worst
jymm what if you were to bed the nut in epoxy?
fen instead of al ?
well, shim it to the correct angle, then take it apart and bolt it back together with epoxy between the surfaces
fenn: and no, I'm not using epoxy tyvm.
fenn: no polymers
just shim it and get it over with then
fenn: DUH, but I have to shave and square it first.
no you don't
just stick the shim under one side
take 1 beer can
carefuly slice into strips
insert strips under nut until square
taking a look at that should help explain
bolt nut down, check squareness
heh in watches just add a burr to a corner or edge
it will take me a while to read that
they have 2x2" ones as well, for $7
fenn: It'll take me longer to explain it.
the 1-2-3 blocks were a better deal i think
SWPadnos: you think that's better than a 123 block?
then all you need to do is put the motor mount holes in it, which I'll bet will be easier than grinding something yourself
no, use that as the motor mount
SWPadnos: Oh, the NEMA34 stuff.
NEMA23, NEMA34 - who casres. it's a square piece of stock :)
SWPadnos: the motor is on a belt
so 2x2" isn't the dimension I asked for :)
SWPadnos: I thought you were asking of the ballnut mount dims
nope - the angled piece
SWPadnos: That is/was the angled piece that isn't square
so the angled piece is 2x2" x 0.5" thick?
i really doubt a belt-drive stepper needs to be mounted on a cast iron angle block
SWPadnos: Yes, you DO need glasses, no the pic isn't fuzzy!
os, is that the piece you want to grind or replace? - the one with the two screw holes facing the camera?
ah i figured out the problem
all that grease is slowing it down!
is that really grease? looks more like it came out of a bird
the bird was ill at the time
SWPadnos: Yes. The two outter holes are what the underside gantry plate bolt to. The center hole is a set screw to prevent the ballnut from unscrewing from the ballnut mount.
is the ballnut tightened in the aluminum before you tighten the setscrew?
fenn: the ballnut is threaded into the al block.
~ 7/8" threaded hole
what holds the ballnut parallel relative to the bottom of the mounting block?
the thread surface?
the thread ...
that is common
JymmmEMC: how would squaring the block help? it looks like the threads themselves are not prependicular to each other
ds2: you can't see the threads in that pic
skunkworks, it's not that common - on a Bridgeport, the ballnut has a flat flange held on by allen bolts, and you can shim it
JymmmEMC: I am assuming the guy threading it made it prependicular to the surfaces
well, it amy be common, but it's not everywhere
right - I am talking hobby
oh - nm :)
ds2: You're assuming the surface was flat too =)
i dont see why "hobby" has to be "stupid bad design"
I don't think it matters if the surface is flat. what you need is for the ballnut mounting thread to be parallel to the direction of travel
2 bolts hardly sounds expensive
it's the flanged nut that may be more expensive
the flange adds money I suppose
with ground surfaces ...
SWPadnos: the mount is too high for one, putting upward pressure on the ballnut.
if the nut is thread bound it could push it off square
jymmm i recommend grinding that mount down with your angle grinder and then shimming it back up to the correct height/angle
and then bedding it with jb-weld, but you wont listen to me
SWPadnos: so instead of the ballnut being like --- it's like this \
(so to speak)
Would would also explain the wear marks in the ballnut only appearing on one side, but not on the other.
it may not just be squareness
could be basic parallelism beqtween the screw and side rails
JymmmEMC: were all the parts machined or cast?
archivist: No, I understand that, but 1) I HAVE to reduce the height (that's a gimme) , and 2) I need to make sure that I dont make it worse when I file/sand the mount
it was made on a VMC i believe
ALL al parts were machined (fucked up mind you), excpet the 80/20
A haas actually
damn you gene haas!!
so it is not an attempt to compensate for draft then
But wehnt eh fscking operator doesn't even bother to clear the swarf to place stock on a flat surface, you get shit like that.
heh you dont put draft on bearing surfaces
if it was cast peice...
You shold see the box of replacement parts I have from them.... there is at least 4 copies of some pieces that they couldn't get right the 1st, 2nd, or 34rd time.
How do you bolt 123 blocks together?
I've never used one before
the ones I have are threaded (some holes anyway)
probally takes 3/8-16 bolts
oh, ok. I wasn't sure of that or not
yep - I think that's it
counter sunk holes ?
but some of them have the unthreaded holes drilled out to 5/16
I got mine from that eBay seller I linked to
SWPadnos: I want to head to weatern tool today and pick one up
normal SHCS should work
I'd get an angle block and a flat (maybe a 1-2-3 block)
SWPadnos: $15 for a 123 blk
then you can see the angle
each, or for a pair?
I suppose you can figure out if it is square with a ball anvil mic and lots of trig
[22:50:16] <JymmmEMC> http://www.chicagobrand.com/50301.htm
[22:50:18] <SWPadnos> http://cgi.ebay.com/PAIR-123-BLOCKS-1-2-3-ULTRA-PRECISION-0002-HARDENED_W0QQitemZ330151842782QQihZ014QQcategoryZ41934QQcmdZViewItem
you can see the pattern of threaded holes in the photo for that auction
a good level
is shipping to you ~$10?
a good long straight edge as well
shit, they close at 5pm
you can make it
archivist: my luck, I'd bump it and not realize it =)
they are easy to check
straight edge or the level?
levels be reversal
SWPadnos: ah, yeah. I see the threads thee.
easier to find a straightedge than a good level
you could do something silly with mirrors and a laser i guess
fenn: fuccccckk HD most expensive leveel $130 is cheap al
and not even level!!!
a mirror and a collimator
for $130, get a machinist level
well it's called "home depo" not "machinist depot"
archivist: piece of dental floss
$93 for 6" starrett level
* archivist would go to a second hand tool shop
jymmm you know people make cnc machines out of cardboard
* ds2 would like to know where to find second hand tool shops
ds2: last one I knew of closed 3 yrs ago
JymmmEMC: that's why I ask
need one that is .0005 per foot
on Bascom ave. I've even looked in the pawn shops too
triangle seems to be the closest to it around here and well...
ds2: We need to make an LA run
I have seen them in antique shops
ds2: would make a killing reselling them up here
at least in LA, I know of one place
JymmmEMC: esp. if we do a little restoring before reselling
and army surplus dealers
archivist: only one in the area
Mnt Surplus on Bascom ave
Mtn View Surplus in Campbell, and in Mtn View plus stevens creek
ds2: name another that sells military surplus?
ds2: Nope, Stevens creek is all new product.
it's CALLEd surplus, but it don't actually sell surplus
JymmmEMC: last I was there, they had a good supply of used mil stuff
ds2: Where? the lil room to the left?
it is as you go in on the left side, there are 2 rows of ammo boxes, shovels, belts, etc
yeah, the little room
better selection of ammo boxes then the Mtn view one
ds2: when was the last time you were there?
ds2: Hmmm, maybe they started carrying then. When all I saw was new, I never went back.
SWPadnos: thanks for the links, I'll look more when I get to work tonight.
they even had old MREs (was there looking for camping gear)
ds2: there is the one on El Camino IN MV, but I haven't been there in over 5 yrs
there's some article on the web, written by Jon Elson I believe, that tells you what you're in for if you want to scrape a block flat/square
you should plan on spending something like 20-100 hours on it if you want to do it right
SWPadnos: close enough to shim will (have to) be fine.
JymmmEMC: they are the same as the Bascom one...just 2 branches of the same store
unless someone has a working mill they want to play on
if you can wait a few months, I can probally run it on a mill for you
ds2: I've waited 24 months too long as it is.
*nod* that's why I didn't bring it up sooner
too busy to even do my own machine work and I got a long list of projects for them
I might try to find a local shop bring the block and a case of beer with me =)
let him drink the the beer AFTEr it's been machined =)
JymmmEMC, isolated from the screw would be a waste of time its the pair that matters, hence a shim job may be quickest
or if you know the instructor at deanza...he has a night class 4 days a week
archivist: the block as it stands now is about .040" too tall, putting upward pressure on the nut
add distance to the gantry to compensate
JymmmEMC: couldn't you get someone to sneak you into the TechShop for a quick milling?
or lift the rails 40 thou
lots of ways to correct
nah it would be so easy to put shims under where that big plate bolts to the "trucks"
right next to the white nylon cable holder
5 mins work done
but check error both ends first
shim screw ends till identical
then shim gantry
wow thats a nice heatsink: http://jmkasunich.dyndns.org/shoptask/geckos-heatsink.jpg
then shim nut to correct for any squareness error
fenn: isn't that a bit... excessive?
I know enclooks like it may have been symetrical at some point :)
looks like it may have been symetrical at some point
hey - those can pump 1600W each - that's more than an Intel Core 2 Quadro Mega-HT
or something like that
SWPadnos: processors are all dissipation though, hopefully gecko gets better efficiency :)
70-80% or better
6 wire steppers can be used with bipolar drives - correct?
they would be concidered - series?
unless its five phase, it's a delta right?
I think (but I have not meassured - 2 pairs center tapped
oh lookie here http://www.piclist.com/images/member/RB-ezy-Q33/6wire.GIF
exactly.. let me measure
yes - that is what the steppers are.
so I would just hook A- and A+ to the drive and B- and B+ to the drive
archivist: I already checked the end mounts. they are actually true. So I don't want to mess with them
You call that a heatsink?!
they may be too high/low at the ends check the 40 thou at each end of travel FIRST
archivist: I already checked them againest each other... they are true
against the gantry not each other
looking for slope error
archivist: the fram eyou mean?
archivist: I took the two end mounts, place a bearing between them to align them, they are even
rail v screw
ignoring frame, they are true to each other
nothing is exact in this world
archivist: I'm nothing going to zero-gravity to confirm
any way off home now
or can I only use one of the coils from each pair?
skunkworks: i think the idea is that it's "sorta bipolar" because the coils that share a common wire are wound opposite directions
so in effect they would be wired as a series stepper?
might have a use for my stockpile of oem650 drives finally ;)
just to play... :)
fenn: you like my heatsink, eh?
I know its overkill, but it was free
i'm a fan of oversized heatsinks in general
same here :)
it wasn't symmetrical in the first place
there were power devices mounted on the flat where the geckos are now
jmkasunich: I had a good score this week.. http://www.electronicsam.com/images/engraver/engraver.JPG
the drive smoked thoughhttp://www.electronicsam.com/images/engraver/drive.JPG
[23:46:49] <skunkworks> http://www.electronicsam.com/images/engraver/drive.JPG
simple l298 driver though - but really emc would turn circles around it
it is a 486 dedicated controller.
[23:48:43] <mess> http://cgi.ebay.ca/Hardinge-HXL-CNC-GE-550-cabinet-for-electrical-parts_W0QQitemZ110154174861QQihZ001QQcategoryZ25290QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
[23:48:46] <skunkworks> http://www.electronicsam.com/images/engraver/rip.JPG
that heatsink came from a dead one of these: http://www.ab.com/motion/servodrives/images/mainimage_1394digital.jpg
some nice old hardware
* fenn squints. looks like a computer case
the big box contains brains and DC bus supply
the littler ones each have one axis of motor drive
the widest of the littler ones is 5kW, the others are 3, 2 and maybe 1kw
three phase AC servo stuff
ge 550 is classic old school g-code box
another pic shows how they go together: http://www.ab.com/motion/servodrives/images/mainimage_1394overview.jpg
the DC bus and control sigs pass down that black stripe in the middle
skunkworks: got pics of the cutter?
or the bits?
skunkworks: the cutting tip and the full view of the cutter
skunkworks: I'm interested in the tooling specifically
skunkworks: I got the RPM's and the HP, just might consider doing soft metals if I can find the tooling
soft isnt always easier..
JymmmEMC: looks like this (camera is dead atm
[23:57:04] <skunkworks> http://cgi.ebay.com/Hermes-diamond-tip-engraving-cutter_W0QQitemZ140142583279QQihZ004QQcategoryZ26245QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
mess: But I won't do steel/iron period here. Want too easy to get cuaght on shoe and tracked in house
and I have some of these also http://cgi.ebay.com/2-Cutting-Tools-For-New-Hermes-Power-Engraving-Machine_W0QQitemZ250147497675QQihZ015QQcategoryZ46744QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
the head is adjustable - I might remove the head and make something that holds normal 1/8 inch shank tooling..
skunkworks: I'm interested in THIS tip... http://www.e-engraving.com/machines/new_hermes/img/New_Hermes_Vanguard_7200_Spec.jpg
mess: we have cockatiels, and dont want them to eat metals
pretty sure that is this http://cgi.ebay.com/2-Cutting-Tools-For-New-Hermes-Power-Engraving-Machine_W0QQitemZ250147497675QQihZ015QQcategoryZ46744QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
skunkworks: can't tell, no close up of tip