sigh. the junkyard only pays $120/ton for my minivan
better then trying to sell it..
even in "poor" shape, it's theoretically worth >$2k
uh - really? what year model?
the only ptroblems with it are rust related, but the inspection rules here state that you can't have any "sharp edges"
1996, ~150k miles
Plymouth Voyager SE - same thing as a Caravan (the Dodge)
yes - 3 liter mitsubishi engine
I wanted to give it to my sister-in-law or father-in-law, but we never got it out there
you don't have inspections there, right?
nope - well not that I know of..
I don't know if the big cities do - but I really don;t think so
if you'll pay my expenses, I'll drive it out to you :)
heh - we stay away from minivans like the plage
you can't beat minivans for cargo capacity to fuel ecnomy ratio
yes - but I just don't like them and we can leave it at that..
I can carry a refrigerator in it (as long as it can be laid down :) ), but I get 25 MPG or better on the highway
of course you can't tow shit with them
funny - I just got an email from the local Jeep dealer - they're really cheap these days :)
'08 Grand Cherokee, $20k
they are all hurting..
especially those selling gas guzzlers
big 8's and propane are great
I wonder if mine could run on propane
or a little nukular plant ;)
the small scale thermonulear devive is a semi-realistic option at best ; ) peace OUT
tabletop nuclear reactors existed in the early '60s
safe ones even
and handheld nuclear bombs, for that matter
yeah. still classified as to how they did it
from close to 50 years ago
very low power stuff yes.. My sister was born in Elliot lake where the surface rads equal that old stuff
I think the handheld bombs were in the 5-20 kT range
meant for the Orion spacecraft
no one would launch that hand held
when you need to carry 20000-50000 bombs, size is important
no by then all thats important... is KILL EM ALL Let God Sort E'm Out
I liked the nuclear powered propultion..
well, if you use the bombs as instruments of war, rather than specific impulse for a spacecraft
seems to be viable on heavy lift lighter than air deigables
there's a great book on the Orion spacecraft, written by Freeman Dyson's sonGeorge
[00:46:59] <SWPadnos> http://books.google.com/books?id=r_Gu4f0QxrkC&dq=orion+dyson&pg=PP1&ots=F37V2wP12Z&sig=tcTYjbI4VjfvWnXmpUBlmzCA84s&hl=en&sa=X&oi=book_result&resnum=1&ct=result
arn't they calling the new craft orion?
fascinating subject matter, boring as hell to read
it's quite adichotomy
yes, there's a new Orion craft
yay - a friend can use the tires
you should be able to piece out most of it if you have TIME
I don't have the time, but if I can give a friend a couple of sets of good tires instead of getting 6c/pound for them, I'm happy
likewise... wont it run??
I'm half temptedto just fill the rusty parts with foam and tape over them. that may make it pass inspection
it's mechanically sound
the battey is dead because it's been sitting for a year
I think the E-brake cable needs replacement. otherwise, it's in fine shape - probably has another 150k miles in it
the stuff you nee is called STUFFIT'
that's the stiff foamy stuff, the expanding insulation?
its a fiberglass - microbubble stuff
NOT the insulation stuff
this is workable the other is NOT
you can only lay the insulation on and hope it doesnt fall off when you cut it to shap
you can spread and lay the other stuff up
the expanding foam would probably push the door apart - it's pretty rusty
that's the annoying part. it would take maybe $3k to fix it up right - new quarterpanel and new sliding doors, plus a paint job
you wouldnt want to fill anything weak with it
environmentally, that's the thing to do, no question about it
but financially, I'd be putting in as much as I could get out of it, and "nobody in their right mind" would put that kind of money into a vehicle this old
have you got NO scrapyards with 1 punched in the front
unless it's a classic, which a 1996 Voyager will never be :)
there aren't many parts for it on the jynkyard web search engines
really few near here - the closest was 90 miles I think
im still running a 1996 gmc ASTRA minivan
well, I'd be running this if it didn't have some rusty parts :)
the inspection criteria are pretty tough here
panels are cheap if you can weld??
err - well, that may be true
I don't know how to weld at the moment
though I do have a regulator sitting in my garage
i used to blaze things in on the cheap
passes safety... but ain't pretty...
I think the quarterpanel is somehow part of the frame though, which is why I'd want to do it right, not with foam
QP.. will NEVER be integral to the frame... even on uni-body construction
it may wrap some of the overhead stuff ..but the panels are for looks
well, with my mad torch skillz, I'd probably cut through the frame getting the panel off
or make the gas tank explode
what's this "off" stuff?
just weld some sheet steel over the top
the panels (both door and quarter) are curvy
so bend the sheet steel around the nearest tree-trunk first
but the trees are the wrong size! :)
pound to fit, paint to match
HE SEEMS TO WANT TO RESTORE HIS 1996 CRYCO MINI VAN
yeah, repaint as primer ;)
hey - I think I may have discovered a way to reliably crash Windows Explorer
just open a folder from the start menu, then close it
* skunkworks has left #emc
heh - just kidding
SWPadnos - we do air craft sheet metal - bring it to Wichita - we will pop rivet some panels on - :)
all I need to do is get out of Vermont without being noticed
and avoid accidents ;)
that's hard when you are driving one
it's too bad I forgot to bring the title to Wisconsin when I was there in August. they could have registered it and sent me the plated
send me the title - we don't have inspections here
ok. email me your address
Swpadnos: body dolly, body hammer, welder, grinder, primer, sandpaper, paint
an hour with each and you'll be good as new
plus a day to learn how to not hurt myself with each ;)
other than the paint, sandpaper, and primer
or spend 5 months building a CNC sheet metal rolling machine...
or 5 years "retrofitting" a Bridgeport
you could try riveting a panel in place?
if you can't weld it
sure, I can do that
I can probably learn to weld, or get my wife to do it
(though it's been a while for her)
rivet the corner which seems sturdiest, then work it down the seam with a body hammer to match the panel contour
you will probably beat it all to hell but thats what bondo and primer is for
the trouble with this vehicle is that I don't really need it. I have a replacement, I had just planned to give this one away
which is proving harder than I thought it would be
alternatively tack weld it in a corner and beat it with a body hammer until it roughly matches the shape, then finish the seam and beat in the weld with the back side of the body hammer
craigslist will usually get rid of it if you are in the states
heh. I was hoping to help someone I know, but you're right about that
you can probably part it out for more than you could sell it for anyway if you'd rather get cash
or try registering it as a farm van?
yep. I may even be able to make a generator out of the engine (or something)
if only I had a farm
not strictly required
at least around here anyway
ehh maybe not a good example
they don't even do inspections here though
spd-si shipped me some pulleys and were nice enough to include a FREE pocket knife
here we have california emissions and yearly inspections, which include "no sharp edges" or similar wording
I just got a new bore gage - amusing (and sad) contrast
I have a Federal brand gage that covers 2.25 to 4", but I needed one for the 0.7-1.4 range
bought a cheap one off of ebay, I knew it would be an import
you probably got ones as crappy as I have :)
the federal has diamonds in the fixed and movable anvils that contact the bore
the import has _soft_ steel, plated with copper, plated with cheap-ass (not hard) chrome
ONE measurement wore thru the chrome and revealed copper
well, you wouldn't want to scratch your work now, would you?
a couple more revealed (and started wearing) the steel
that is the fixed anvil
the moving anvil is a wee bit better
you didn't measure anything hard with it did you? it's for measuring holes in marshmallows
I think its moderately hard steel, no plating
it's not the manufacturer's fault if you abuse it
that said, I get more repeatable and easier measurements with this gage than with the indical - its much easier to keep it centered
must not have read the instructions
yeah. some instructions say "do not use in harsh environments"
others should just say "do not use"
I'm thinking about drilling out the little half-dome contact point, dropping a steel ball (bearing ball) in there, and crimping it over
I'm also gonna keep my eyes open for a good price on a real bore gage in that size range
ebay is disappointing - a Mitu went for well over $300
this one was $35
was it from the 800watt guy?
no, but similar
I think I've either attepmted to or actually ordererd from him, and I don't recommend it
I can't remember the details, but I have a memory of "bad things"
[02:42:06] <jmkasunich> http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&rd=1&item=300256486938
the 800watt one is only $25
but I didn't pay shipping on this one, the seller is 10miles away
i've bought two items, one good, one bad
ah - JTS. they have a reasonable selection of taper adapters, if I remember correctly
place winning bid at lunchtime, pick up at 4:45
I think I got some R8 to MT2 and MT3 adapters from them
they have quite a bit of stuff in their catalog (printed)
including the elusive dial metric/inch calipers for $19
hmm the timing pulleys fit right on. in theory i have a cnc lathe now
those should work by having one needle, and having one of the dials spin with the other fixed
it would be cool :)
they work fine the way they are
I think the spinning dial would be hard to read
bah - 2 needles. it's too much like an analog clock!
yes - the estimation would have to go the other way
hmmm - no, it wouldn't
since the mm dial is 2mm per rev, it really helps to know that even mm are at the top and odd at the bottom
if the dial spun, that wouldn't work anymore
so spin the inch dial instead
don't mess with mah inches
man i wish i knew g codes
quick reference (with links to more details) http://www.linuxcnc.org/docview/html//gcode.html
still a ton to do on my build, but it is actually running 2 axis right now. no limits, e-stops, spindle speed control though
yeah i've been reading through the excellent documentation there
its the best on the web that i've found
but it sort of glazes over when you don't have a machine to try it out on
makes sense but doesn't really settle in
sort of like trying to learn to program assembly without writing a few trial programs
so i guess the procedure is to chuck something up, take a trial cut in MDI, then measure it and set the offsets?
or use a simulator build and watch a fake machine with vismach
dareposte: for non-critical work, you can set offsets in easier ways
like a home switch?
for example, if your stock is 1" in diameter, bring the tool point just under 1/4" from it, hold a 1/4" dowel pin between tool and work, and jog away 0.001 or 0.0001 at a time until the pin drops thru the space between tooltip and work
then set X to work radius + dowel pin diameter
maybe not so accurate though...
my 3-jaw is only repeatable to a few thou
note that setting the offset while jogging away is easier (jog till pin drops), but if you have any lash, cuts made moving toward the part will be a bit off
if the work is concentric it is extremely accurate
yeah in that case, best to do it after a cut
it's also likely to be fine for just about any roughing (for multi-setup parts)
so if i need to turn an accurate bearing journal on the end of a piece of bar stock, what is the "best" procedure to get it within .001"?
i'm a bit confused about some of the offsets
turn quite a few other parts first ;-)
experience rules then?
yes set offsets the best you can, write a program, test run, measure, tweak tool table
i would think there's a good method that should be fairly reliable
seriously - you should make a few "don't care" parts, to get a grasp of what the machine can do
use g92 offsets?
I don't like G92 personally
no, you have to use the tool table for a lathe
use G54 (touch off) to locate Z=0 (end of work)
I use the touch off button myself, but thats because I don't have spindle speed control
if you want CSS to work right, you can't use G54 for X
dareposte: the key is measuring the machine's expected vs. actual as you creep up on your final dim
what kind of tool holder do you have?
i have a very rigid one
repeatable quick change?
made from 3/4" plate steel with a qctp on it
you will end up with a tool table entry for each tool
phase 2 qctp, so not the best but better than the rest of my lathe
once it is set, you can load a tool, apply that offset, and cut a part to size
so set the Z with touch-off, and then use the tool table for x comps?
tool table will have Z offsets too, so all tools will be located on the part
so when i power the machine down and the steppers lose position, do i have to re-do the tool offsets each time?
that's why you really want homing
i got some high accuracy limit switches
you should only have to find the first tool
on X at least
cool, you can home to those then.
i haven't tested the repeatability yet on them
so i don't use indexable tooling usually, since i can't find any with enough positive rake
so if i regrind a tool bit then i will have to edit the tool table offset?
they make some inserts for aluminum...
I don't have experience with them
okay so why does G92 even exist then? i had imagined using it for all these offsets
g92 is sometimes useful
is it for mill work?
g92 reprograms the current offset to where you are
g92 moves all the coordinate systems some amount
yes, based on where you are
it's really not what you want, especially for X
from the documentation it seems that G54 is just a coordinate system but someone said it was touch-off?
touch off is the procedure/button for setting g5x coordinate systems
g54 is the default
oh i see
so you would say g54 z0
if you want Z to be zero where you are, poke touch off, enter 0
is the convention to have the chuck jaws be zero?
or the end of the part?
often the end of the part, but do whatever you think seems good
man there's so many details
FWIW, I use Z=0 at the end of the part
so if i were to get some part code from someone, it wouldn't necessarily be in any given coordinate system, just whoever chose to program it would make it however they want?
bore .5" deep: g85 z-.5
dareposte: you bet
dareposte: if you're lucky it'll have some nice setup instructions saying how to use it
so then the zeroing method is really as important as the code to make the part
if it's someone you work with a lot, you'll learn their style
yes it sure is
sounds a bit dangerous, but i guess i will find out soon enough
and certain tools, and setting tool offsets, how the part is mounted, etc.
maybe make some chips tomorrow with it
is there a "g" code that does whatever the touch-off button does?
yes but it's esoteric to use: G10 L2
i mean if you were using a non-axis interface there would have to be a hotkey or something instead
the other interfaces all have a different way of doing kind of what touch off does
because I designed it, I obviously think AXIS does it the best :-)
nice work, it's a sweet interface
a big step up for lathe in my opinion, especially for less experienced types that don't necessarily have the background to understand the curses interfaces right away
usually jepler does the dirty work to make all my crazy ideas work
AXIS is the only interface that has lathe-specific features currently
i was also comparing to other available controllers
like turbo or some of the dos based ones
oh, I don't really know any of them
although now its pretty clear there really is no comparison
I am generally not interested in non-opensource software
i am interested in using it sometimes when there isn't an alternative
hard to beat some of the cam packages that are commercial
yes, in that case I just go without
or write your own?
yes sometimes I write stuff to solve a problem I'm having (and then I usually give it away for better or for worse)
i have used linux for years, and also windows, and usually wound up having to use windows for several things that just didn't work on linux, but ubuntu 8 has pretty much resolved most of those now
REALIZE came out of a situation like that. I think it's very handy. unfortunately it works in autocad.
the more i use it the better I like it
my silly 2.5d cam software
hmm i may need to check into that..
works great for most of what I want to do on the mill
only bother if you have/use autocad
yeah i do
i'd like to find a good autocad replacement though, i've been using qcad and its not bad but its hard to do many things that autocad makes easy
yeah I tried very briefly to switch, but it wasn't for me
no 3d, no polylines
the only open source i've been able to contribute so far is a good ballistics engine, since there was none before and i needed one
and no 10 years of experience using it
yea i've been using since R13 came out
can't seem to kick it
heh, I learned on R11, I own R12
you've got more than 10 years in then...
it was probably about 94 I was using it (for work)
I bought my r12 in 98 I think
i was thinking r13 came out around 98
my sister is an architect and got all excited when it came out, and introduced me to it
could be. it was an old version when I bought it cheap
they would sell the previous versions cheap to students
no support, no manuals!
i'm not one to use the support much anyway
i find google groups usually knows the answers
the amazing thing is after all this time using it, I don't know of any bugs it has
and the question has almost always been asked before, just gotta find the answer
you still use R12?
my sister gave me her copy of R13 when they upgraded, but work uses some 2006 version now so i don't use it much at home
the small differences are annoying once you get used to the newer version, trying to go back and forth
but to their credit all the "old" commands work just the same as they always have, just new clutter piled on top
well at least all the "old" commands I knew
which is about 10
good luck with the lathe.
thanks for the food for thought and guidance
* fragalot returns with 3 L298's, a 24VDC switched PSU, and 3 1F capacitors because he liked how they looked
oh this sucks. it's got a 110V/220V input select switch that is currently set to 110V, but to be able to switch it to 220, I need to break open the warranty label >.>
fragalot: your reseller has screwed up then
mr_boo: stuck a screwdriver in, flicked it that way, :p
fragalot: hes supposed to adjust it to the customers voltage
mr_boo: yeah,.. sht happens..
well it's not really a warranty label, it's a QC Inspected label, but meh
btw, how much current can a small 95 ohm/winding 7.5 deg stepper be expected to handle?
well what voltage was it rated for?
all i know is that its small and have 7.5 degree steps size and 95 ohm windings
i know its not in the amp range
but do i dare 250mA
sounds probable, my 75ohm 1.8 pulls .39A
(if that helps :p)
makes somewhat sense
this is a pancake shaped stepper with a diameter of around 40mm but a length of only 14mm
its scavenged from an old fax machine
mines ~ 42.10 x 47.74 x 42.12
ah, those found in old 5.25" floppy drives
o.0 were floppy drives that big? because last time I checked, a 5.25" bay is a bit smaller
maybe yours is a smidgen bigger then
that would precisely fit inside a 5.25" bay
or just barely
That means i need to find some floppy drives of that size then.
'cos i still need 2 more
owell, gotta make the circuit first i guess
oh ffs you have got to be kidding me
multisim just refused to open one of it's save files
well... There go all my schematics,.. I spent DAYS ON THIS BLOODY THING
sounds frustrating to me
time to eat
fragalot: good thing you do frequent checkins to version control
Vq^: apparently multisim refuses to open by double-clicking the file. i had to open multisim on it's own, go to file, open, select the file, and click the "open" button for it to work
oh, i thought it was some multisim bug
ms9 sucks, lol, i have to edit every single part seperately to change the footprints,.. wish I had ms10, where they have a nice list where you can select all simular componements and do it in one go >.<
is multisim free software provided with their hardware?
mr_boo: unfortionately, no.
this is just an old student license i had laying about
i've heard plenty of complaints on multisim
but the UI is more usable than anything else i've tried sofar..
eagle just annoys me, and so does kicad
PCAD V5 on dos is the bestest :))
googling that just locked firefox up
under a different name these days iirc Altium maybe
mr_boo: in multisim it takes about.. 2 days to draw out a circuit, and PCB for a 12VDC->5VDC supply that outputs 150mA
(eg. just 3 capacitors, 7805, and 2 headers....)
good hand layout is about 2hrs per 14 pins ish
give or take
autorouting usually being a bit the crap side
i never autoroute
always fails anyhow
kicad went outside the border
Guest457 is now known as skunkworks_
hi ray, sam
1:37 /me eating dinner, some people get up late :)
wow you're gmt..
wait - I don't think I did the math right..
no - thats right.. we are -6 and it is 7:40 here
well the meridian is in england
is the whole of the uk in the same time zone?
one hour off from almost all of europe, crazy brits. http://commons.wikimedia.org/wiki/Image:Timezones_optimized.png
hehe we set THE meridian everyone else is wrong
poland: it's 14:47 here (+2 gmt)
huh, I didn't realize there was a +13 timezone (tonga)
or +12 3/4 (Chatham Islands)
I thought the main reason that england was 0 was because they had printed all the maritime maps
jepler: what step range are you running on the pluto?
(and how does that work for slow speeds? I assume it cycles between the lowest hz and off?)
+1GMT == ftw.
well, +2 atm,.. CEST..
huh - I thought the pluto step had 1 pwm generator.
must have been something that had been discussed
does EMC have a charge pump output?
no, pluto_step doesn't have any pwm generators
nm I found it
[13:49:59] <skunkworks_> http://www.linuxcnc.org/docs/2.2/html/man/man9/charge_pump.9.html
now for the real question is it in the stepconf wizard?
I use the highest step range (312.5kHz) even though the actual top speed is 32kHz. I don't see any problem with slow step rates, the pulses remain approximately equally spaced
(so I might drop the speed range feature if I ever rebuild the firmware for some other reason)
is the charge_pump at least 10kHz?
I would assume it is at the speed at which thread it is in.
sounds like it is 1/2 the thread speed.. (I might be talking out of my ass)
yes, I think skunkworks_ is right
[13:53:01] <BigJohnT> http://www.cnczone.com/forums/showthread.php?p=502060&posted=1#post502060
stepconf has a charge-pump option. it connects user-enable-out to charge-pump.enable.
the Gecko 540 requires a charge pump to run
it places the charge-pump function in the base thread
if your max step rate configured in stepconf is less than 20kHz, then the BASE_PERIOD will be set too long to give a 10kHz charge-pump output
Marcus says that most will work down to 5kHz
the longest BASE_PERIOD that stepconf will write is 100000us, which gives 5kHz for charge pump and 10kHz for step rate (stepconf doublestep mode for stepping, but not for charge-pump)
We test it here with a function generator running at 10kHz, and it can go as low as 5kHz on some G540s.
cradek_ is now known as cradek
"on some" I wouldn't want to bet on
thanks for clearing that up for me :)
so you might need to hand-edit the inifile written by stepconf to set a BASE_PERIOD of 50000 or lower to be sure to get a 10kHz or higher charge pump frequency
I'm a little surprised by the decision to put a charge pump in a low-cost step driver
and I'm surprised that it is mandatory to use
I love mach people.. 'I noticed this the other day in the cheap drives thread. It would seem that the G540 will only work with Mach3, due to the charge pump being mandatory.'
you gotta love ignorance
many people guess and generalize when they aren't sure, instead of just stating the facts they ARE sure about
<3 I'm so effecient. Wasting 4Watts into pure heat. :p
it's common; don't assume there is malice
wait, no, i'm wasting 12
how do you like the "ads by google" stuck in between the threads...
I guess I don't see them
I just started seeing them
a locked thread by sponsor
locked thread by sponsor?
yep with links to ads
in the EMC forum its the second thread
well, sf puts advertising on all the e-mails so it's hard to get all offended about ads in cnczone.com
yah - I guess I am used to that.
(why the heck is is "EMC+closed loop" a sticky thread in the cnczone emc forum?)
because someone who knows how to glue threads thought it was the most important thread
no - one of the admins put it there...
let's make emc look bad by asking the wrong questions
I think it is because it always comes up - closed loop with stepper.s
It is a pretty good thread - fenn had a really good reply
hmm sounds like someone from the EMC group needs to be a moderator not a Mach guy
* fragalot nods, even though he has no idea about anything
how about somebody impartial. like me. xD
i've made my first wire winding test now
ok, go for it :)
only 38 turns fitted onto bobbin rather than the calculated 49
mr_boo: how goes it?
that messes things up a bit
i'd like to post a photo but my cellphone is dead for some reason
its supposed to recharge through usb interface cable
sounds like you need to verify the scale of the axes if you wanted 49 turns but got 38
or is it a wire diameter problem?
could the chargepump module be re-written like the doublefreq?
* skunkworks_ thinking out loud
its certainly a wire diameter problem
that's a good thought skunkworks_
i've adjusted the scales now
if i tell the traverse axis to travel 40mm it does now
skunkworks_: well, there's the question of what you can do with stepconf, and what you can do in general.
gauge meter says 0.15mm diameter of test wire
* skunkworks_ is more of a big picture person..
I think the 'constant' block + doublefreq would make a charge pump
skunkworks_: if you hook emc's user-enable-out to a parport pin, and turn on the corresponding 'reset' parameter, you have something like a charge pump.
oh, or that
maybe i should buy some 0.1mm copper wire
or a constant block, like cradek says
yeah, any true signal
I don't like the bus system of Eagle... htf can you determine which line connects to what later?
but it wouldn't be near 50% duty cycle
nvm, found the label thing
that is the issue I was thinking..
argh cradek is typing faster than me
hard to say what the exact requirements are
yeah you'd need to look at whether the duty cycle matches what is required
.. asssuming it's even documented.
"our board expects what a certain version of mach does when you set it up as shown in these screenshots of a half-dozen dialog boxes"
(note: screenshots may not actually be included)
what diodes would I use for the freewheel diodes on a small 75ohm .4A stepper running at 24V ?
anything that says "fast" in the name and looks like it has suitable ratings
"fast" is the important part (hint: not 1N4001)
MUR440RL ? "ultra-fast recovery rectifier"
rated for 4A, incase i decide to use bigger steppers later, so i can use the same board..
fragalot: did you get l293 or l293d? or am I thinking of someone else?
jepler: traded in for L298N -- I /HAD/ L293's which worked, but couldn't run 'm as choppers
fragalot: ok -- just wondering because the "d" version of l293 has integrated diodes
my l298 datasheet says "D1 to D8: 2A fast recovery diode (Trr <= 200ns, Vf <= 1.2V @ I=2A)" for the stepper motor circuit (page 8)
* fragalot pulls it up
no, for the external diodes
meh, aslong as the footprints match
in the MUR440 datasheet I find trr 75ns, Vf 1.28V
so yeah I'd say those diodes match the requirements
fragalot: oh, in eagle you can use the "label" command to create a label with the same name as a particular signal, then place it on the schematic in a place that makes it unambiguous to you.
fragalot: I don't do a lot of stuff where buses seem like the appropriate tool. I often just just break the net and label both sides
using 'ports' when it helps clarify things
oh can you just break it up?
.. owell too late
I'm almost finished, just making the interface (buffers..) on the board itself, and then a breakout board to distribute the EMC2 signal
using the commandline you can specify the net's name (if you don't, and you don't get buses, the nets get names like N$33)
found that out ye :p
[14:50:59] <jepler> http://emergent.unpy.net/files/sandbox/labels-and-ports.png
<-- here's a schematic I have been working on recently that shows use of labels and ports. "port" is the pointed box thing around the net names at the bottom
does any1 here have a board that takes the printer port signal and turns it into something usefull? (basically just a print from port->headers would do, lol)
for the other spot where SCK/PDI/PDO are, there's not room for a port symbol so I just put the net's label right by the net itself
i'll finish this and put it up for you to quickly review, if you want to, that is
I'll glance at it but I shouldn't spend too much time, being at work..
k then i'll let you go on with your work. :p
jepler: how do you add a "power" thingy for a multi-part device, like a hex schmitt trigger inverter ?
a power thingy?
well, an IC needs to have Vcc connected before it can do anything
multipart devices usually have power pins on the first part, but not on others.
fragalot: "invoke" (type it in the commandline or choose the icon, which shows 4 similar tiny gates), click the gate that is already shown, and then choose from the list
i can only find how to connect thmm, i can't find those pins
jepler, does "invoke" create a par/subpartt
SWPadnos: in eagle terminology it will "call a specific symbol from a device".
so a hex inverter is (probably) a device with 6 inverter symbols and one power symbol
oh. that makes perfect(ly no) sense
each symbol has an origin that can be moved independently, so you can put the inverter signals where it makes sense
er, inverter symbols
now to figure out htf to set the PCB size :p
fragalot: you do it by dragging the boundary in the board layout
the board boundary is just lines in a particular layer
so you can even remove them and add ones where you want
but how do you figure out the actual size?
or do you just count the grid lines :p
fragalot: properties gives you line lenghts and such
but now you know how to do it :)
eagles ratsnest needs improvement, imho
ratsnest is supposed to look like shit
so you have motivation to get rid of it
I mean that it should jump to the closest object, and not remain fixed onto one object far far away, thats not really related to it directly
oh, when you're routing signals?
is there an "optimize nets" function in Eagle?
SWPadnos: no, there's really nothing "between" autorouting and hand routing
optimize nets isn't a routing tool, it jus minimizes the connection lengths of the ratsnest
no, there's only "ratsnest"
heh, ok :)
I have seen variable quality of ratsnests
some can switch off power and gnd rats
I'd like that feature
usually optimizing lengths gets close to that - many connections between pads on the same device are effectively invisible (just vertical/horizontal lines, often coincident with the body outline)
stupid eagle and it's inches :p
most chips you'll work with have 0.1 inch lead spacing
unless you're doing surface mount
get used to it with PCB design most is inches
fragalot: you can switch the grid to mm if you want
archivist_ub: i'm used to working with mil for PCB design, ... crud i just realized 1mil is 0.001"
just typing "grid mm" will maintain the current spacing but show mm in the user interfce
you can also "grid mil"
I think we should all count ourselves lucky that it's 'mil', and not 1/96 inches or something
mm would 've been nice too tho,.. except since all parts are dimensioned with inches in mind.. :p
I like 1/96 better. It is a fraction of both twelfths of inches and sixteenths of inches
is it me, or once you've drawn a line, you can't delete it?
surely if it's good for the inch to be 1/12 of a foot, it's good to have some other unit that's 1/12 of an inch
so why did they decide on 1/16 of an inch?
fragalot: if you routed a signal, use "ripup" to get rid of it. if it's any other kind of line, "delete" should work.
or /32 /64.. was that binary
jepler: just found it, thanks :D
skunkworks_: well, I like binary fractions too; I think that's why 1/96 is a great subdivision of the inch
it's 1/3 of 1/32
so you can get third lines for nice photos
of 1/32 inch things
see, 1/12 is the fraction that just keeps on giving
indeed - factors 1,2,3,4, and 6
1/(a bakers dozen) screws things up a bit
ooooh - scary
60 is a nice number too.
* fragalot really doesn't like eagle for PCB editing :(
fragalot: there are alternatives, but I've never spent enough time with them
* fragalot ponders
tried to import in KiCad, and in ultiboard,.. neither work
1400m of 0.1mm of laquered copper wire for around $15, is that cheap?
mostly post cost possibly
that's not included
* archivist_ub contemplates doing a wire stock list
end bill would probably land on $45
I'd love to answer that question, but crazy americans measure wire diameter by "wire gauge" which is a sort of exponential thing, and instead of length you buy it in pounds
be careful of wire sizes, wire plus laquer
jepler: isn't it measured in circular-nautical-mile-feet or something like that?
so I can tell you that mouser.com has 1/2lb of 36awg "magnet wire" for $61.47
cradek: no, you're thinking of "circular mils"
and small sizes do cost
(but you probably knew that)
seems that in 36ga wire there's 13210 feet per pound, so that would be about 2000 meters. sounds like your deal is not a bad one compared to this wire.
(36ga wire is 25 circular mils, which is 5 mils, which is .127mm)
* skunkworks_ 's head is spinning
(that's according to http://www.interfacebus.com/Copper_Wire_AWG_SIze.html)
(yes, SIze not Size)
also remember the tap drill for 1/4-20 is "7" and clearance hole is "F"
* cradek snickers
(and no I didn't have to look it up)
is circular mils a mil^2?
no, but you did have to check when it came to #4-40 screws
jepler: true enough
jepler: I did know they are the next size smaller than .120 though
because I remembered that #5, the screw that nobody uses, is 1/8"
(this table always gives "circular mils" = diameter squared, so it's not actually the cross-sectional area of the wire in square mils ... it's just the square of the diameter in mils)
bah. eagle2kicad_sch.ulp isn't really doing a great job ...
oh - that is just odd
fragalot: could be eagle's fault just as well as kicad's, who knows
I use #5 screws :)
quick what's the tap drill?
jepler: it's a "user" program
oh yeah it could be the fault of the .ulp, I forgot that option
I don't have to remember it is on my drill tap stand
I would never leave them visible - you have to hide the good drills
jepler: apparently, it doesn't like the hex inverter
makes it easy
I'm designing a machine at the moment and the slide uses 5-40 screws :)
what kind of machine?
cradek: you could also just trust your coworkers or employees.
fragalot: home shop - you'd be surprised who wanders around in there looking for drills :-)
ah yes,.. I know that problem
the neighborhood cats and llamas for instance
* jepler tries to remember whether there actually are llamas in cradek's neighborhood
I know there are midget horses and runt donkeys
yes there are llamas
are they normal size?
the local accountant has llamas in his yard...
we just have 3 cats - one of which is sometimes mistaken for a midget horse.
cradek: it puts the breather cover on lawnmower engines and runs in 4 screws automagic
[16:27:51] <skunkworks_> http://www.electronicsam.com/images/house/cat.JPG
BigJohnT: tack welding is faster :D
I'll put up a pic of the 3-d cad when i get closer
hope it'll run HAL/emc2 :P
but alex_joni tack welding is tough when you have steel part and aluminum cylinder and a gasket in between
tricky, though not impossible :P
panasonic plc and servo drive
the gasket makes it hard
would be convenient with a home axes button
mr_boo, section 18.104.22.168 of the manual (or near that) tells about HOME_SEQUENCE
it also makes it hard to replace the seal..
i was more thinking of a button in Axis that zeros all three axes on the fly
most applications wont need such a feature but a wire winder is likely to
I think the homing button does that... Doesn't it?
(if you don't have any switches setup)
yes it would
yes, and you can use HOME_SEQUENCE equally on machines with and without home switches
the "home axis" only does it for each axis one at a time
if you want a "return to home" button you can do that with pyVCP
mr_boo: you didn't follow SWPadnos's url or advice
Home all is a button on the mini interface for non switch users.
"home all" is also a button in axis. on machines with a HOME_SEQUENCE it replaces the (single axis) "home" button.
cradek: this is the part transfer http://i47.photobucket.com/albums/f163/johnplctech/transferassb.jpg
looks like solidworks
I see that mini does not use the home sequence. It should be updated to do that in the "home all" case.
btw, can one declare variables in g code?
#5330 = 20
and that too
actually I don't see a home all button in mini, only ALL ZERO which does other things too
would be nice to have named vars on subroutines
I thought we had that now, but I don't see it in the docs
is that a hint :)
I think we have named vars (2.2), named subroutines (TRUNK) but I don't recall subroutines with named parameters
oh, I misread the request, sorry
BigJohnT: no, just showing my confusion
I may have foud a bug in subroutine vars as well I referred to a non existent #6 and it did not complain
That zero all is only intended for non-switch machines.
I don't think that's necessarily a bug - all variables "exist"
#6 still exists, it's just not a passed in parameter
rayh: I see - in that case using the home sequence would not matter.
missing named ones cause an error missing #6 not, thats an error
just about locked the box up
Perhaps we should look for a home sequence there and switch to that if it exists.
I'll put that on a todo list.
Thanks for pointing that out.
eagle light (which is what i have) appears to be limited to half a eurocard for PCB's
.. that sucks
4x3 inches or some such
they're pretty up frontabout the limitation
fragalot: yep, that's true
I only need a LITTLE MORE and it's done, lol.
(I bought "eagle light", which is the same as eagle freeware but without the noncommercial requirement, and just try to live within the limitations it imposes..)
I just gave up, tossed them all together, and hit "autoroute"
archivist_ub: it can't do it :'(
it finished at 87% complete.. while before while it was busy it had 96 .... bloody.....
need to get best package before the router is ever allowed to make a mess
.. I can only use one layer anyways,.. darnit
I meant package layout
BigJohnT: in trunk we now have O-repeat and it could use a doc too
sorry I never write docs for stuff I add.
O100 repeat 
peat and repeat :)
very useful to use G91 in the loop, you might do that if you do an example
and Ive been using counters......
if i could just export this to ultiboard, i'd be a happy camper
I just sent myself an e mail :)
here's the board I've been working on compared to the working area of eagle light -- I don't often get near the limits. http://emergent.unpy.net/files/sandbox/at90usbboard.png
i /COULD/ probably split this into multiple boards, but i don't want that hassle
I guess i'll just do that..
the board that I showed earilier is the max eagle lite will do
time I did another large board
is somebody nave the electrical drawing of the pluto board?
the at90usb82/162 may be cheap (cheaper than ft232 + comparable atmega) but unlike the latest ft232-family chips it still requires an external crystal or clock to meet USB timing requirements, and there's no ADC
it's interesting to note that the D-/D+ pins on it can also be configured in a PS/2-mouse/keyboard compatible mode
(not that it's complicated -- it's just an open collector with integrated pull-up)
hm, i miss being able to fiddle with EMC
argh, when's the last time I went a day without doing that (typing an eagle command into IRC)
I have the #mysql bot respond to unix commands for fun
rm -rf /
how do i create another desktop in xfce
hehe I just got someone in #mysql-ndb with an ls
yeah -- rats
"bash: yeah: command not found"
my print is FINALLY starting to look like it could somehow ... work? :|
[18:58:08] <fragalot> http://omploader.org/vcjVv
granted, it doesn't look good at all, and thers lots of wasted space, but i wanna see if its doable or not first, lol..
single-sided layout can get pretty frustrating
are you basing your work on the pminmo l297/8 schematic, or some other source?
yeah,.. made more complex ones, with the wrong footprints, and some things that ended up not beeing connected in the schematic.
jepler: the L297/297 i found on the emc site somewhere
it may not help layout any, but if it does I'd be tempted to lose the JP4 for switching the direction of rotation -- you will be easily able to set that in emc
hardly makes a difference in the price, operation and layout of the print, really
besides, i could re-use it for other things :p
I'm currently in the process of trying to fit everything under the 298 and 7805
hmm the SGS book has a sample layout I could take a pic and put it up
[19:07:41] <fragalot> http://i38.tinypic.com/icuste.jpg
archivist_ub: sample layout.. for?
l298 etc single sided
archivist_ub: could 've mentioned that sooner :D
I have the circa 1984 data book in mah hands (just went to car to fetch)
I think i still have some books predating that from my dad,.. But i don't wanna go near them, as they made a "voltage regulator" by just hooking diodes up in series (forward)
current datasheet still has the same layout
[19:16:31] <archivist_ub> http://www.st.com/stonline/products/literature/ds/1773.htm
better that you get current PDF direct
yeah that board layout looks like it could be as old as somebody's dad :-P
actually looking closely they have spread the gnd plane around
good old tape up design
yeah i saw that circuit, but it looked like it would give a lot of interference if they weren't closed together
so i decided to go for the one i found on the emc site, and adapt it a bit to my liking,.. just because. :p
granted, his board looks better, prolly works better too.. but who wants that! >.>
crud. wrong footprint for the pot... I guess i can force it in...
PCB design means reworking bits till its right
i really need to read up on coil winding
i think i'm starting to get the hang of the basics in Axis and g-code
archivist_ub: soldering sometimes means "push harder" :p
i guess most of you have cnc-mills
no it means improve your technique
maybe some lathes too
am i right?
I have cnc mill (home brew)
mr_boo1: it's safe to say that's the shared interest of people in this channel
for several years didn't have a mill of my own, though I do now
although emc can be used to control any kind of machine of course
some have retrofits
[19:35:51] <archivist_ub> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Case_Studies
i'm probably gonna fail with the coil winder but in the process i've at least got familiar to one of the greatest software on earth
i mean, ok it needs a particular kernel to get the performance needed but on the other hand, that performance is staggering
never seen a pc computer output flawless pulses as those rates before
early teething troubles are a good learning exercise
emc is truly amazing
all creds to developers
mr_boo1: it's nice to hear some praise, thanks for the kind words
see stuarts 5 axis cinci
what i look forward to do is to retrofit my grandfathers old mill
need to get that bulky thing to my place
no1 big irons http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Videos
I don't have one of my own ... yet.
[19:40:12] <fragalot> http://nl.youtube.com/watch?v=4XYakTeQahA
this machine scares me.
heh nice and fast that one
I wonder what happens if one of those tools flies loose
note the window has bars
archivist_ub: for a reason, i'd assume, lol.
even after working with CNC machines for nearly 5 years now,... It's still impressive how fast they can lodge those heavy parts arround, lol
those mustve big servos
[19:45:38] <fragalot> http://nl.youtube.com/watch?v=sG22Z2_-Yhk
on this thing it looks like the coolant is just taking a leak isntead of cooling o.O
cooling is probably the reason it breaks at the end
[19:51:45] <toastatwork> http://nl.youtube.com/watch?v=QgsAyIVA75s&feature=related
the cooling in the machines i work with at my summer job is insane
4 jets squirting at the tool, and 2 big waterfalls surrounding that
toastatwork: pretty! :p
hm, I was pretty disappointed by that first video
jepler: the toolchanger?
er, the "crash" one
i was just making fun of his coolant
[19:56:43] <fragalot> http://nl.youtube.com/watch?v=_4W2wSEG0jE&NR=1
lol.. I wouldn't even dare to do it that deep in one go
at least whatever happened here apparently knocked the cameraman over :-P http://nl.youtube.com/watch?v=QwZMSt6GdfY
worst crash i've ever experienced was in highschool with a manual lathe
bigass 4-claw chuck came loose at 2000rpm, clipped my left ear
it had no issues going trough ~1cm of plexy, going over my shoulder, to continue knocking somebody over ~20m away
reversable lathes with screw on chucks is silly
got a small one like that at my summer job,.. it's one of 2 machines that run 24/7.. aslong as you slow the RPM down graduately it's fine
our swiss lathes are screw on and reversable
most of the machines there are taiwaneese.. they do the job just fine, and they are great in terms of maintenance... everythings nicely accessible, and can be fixed "DIY" style.
the daewoo's cost ~4x more, and are a NIGHTMARE to fix if something breaks
are there machines that have protection against crashing?
mr_boo1: can you be more specific?
like telling the machine accidently to smash tool right down to board
something as simple as soft limits can be a defense against crashing
thats a solution of course
soft limits, and some machines also know how much force is applied on the tool
if there is too much force, they stop and give an alarm
but the machine cant allways know you put a long tool in
i can imagine a scenario when one presses the start button and leave machine
archivist_ub: IF you tell it that it's a long tool, that shouldn't be a problem.. if you forget to do that, or if the machine goes side ways and picks the wrong tool like I once had,....
(in the scenario that mr_boo1 described)
or a large fixture or ......
machine makes 50 pieces flawlessly,.. then decides to take a measurement thingy tool (dno what it's called in english) to drill that 6mm hole...
when one has the coffee in the next room, *bang*
mr_boo1: depending on the machine, it wouldn't care and just continue without the tool that just broke off :p
"probe contact closed unexpectedly" might be a perfectly good reason to abort the running program ..
a good system would be some kind of tool monitoring either optically or some sensor thing
jepler: jepler with a dial
jepler: theres a lil' pin sticking out, and a dial telling you the offset comparing to where you zero'd it
fragalot: ah, not what I was thinking of then
I think that's called a "dial indicator"
haha why didn't I think of that
mr_boo1: well, some machines know what force is applied to the tool, and when that force should be there, and when it shouldn't
if it suddenly finds an abnormal spike in that, it stops and starts whining
(ofcourse, thats when it's too late)
you should never run a machine without a dry run, and somebody babysitting it for atleast the first piece anyways
i found out that winding coils is much more difficult than one could imagine
first of all, since there are no feedback the machine must know *exactly* how many turns that fit onto bobbin
if wire diameter would differ the slightest there would be a notch that would mess up following layer for example
often its done by hand
only "coils" i've ever made on a cnc machine was with 8mm steel rods, lol
which was also by hand.. tedious task
we have coil winders here but a lot of hand work is built into them (these are old ones)
maybe it would be advantageous to wind them with space between each turn
rather than having them lined up next to each other
mr_boo1: how would that help? means you're creating a gap to let the next wind pull that open and mess it up
what these do is sense wire reaching a side then reversing
however, my winder doesn't know when a side is reached
well my print is almost done.. just a few more tricky lines, and i'm tired
and the operator would fiddle as it wound
add switches to sense
or optos to reduce load
having turns close to each other would be the most efficient method wouldn't it?
what is the nastiest material you've ever had to work with?
any thing that work hardens
mr_boo, yes except for very thin wire its not practical
depends also on quality of back tension
gonna try this looser spacing concept
if i had been smart i would've placed the traverse guide as close as possible
good night all
a major rev of scilab out today 5.0 ( skylab has fallen! scilab is up :)
will a huge heatsink raise current rating on a stepper motor?
a bit not a lot
cuz its bundles of wire
like, very small wires with enamel insulation
while the heatsinks will help cool the outer wires
which will kinda help cool the inner wires
they can stand quite a high temperature anyway
the inner wires will still prob get too hot and the insulation will prob fry before the heatsinks can dissipate the heat
if i can touch the motor w/o burning my fingers, do i know the motor is allright then?
thats why the heatsink wont help a ton, inside them theyre already pushing insulation limits =(
just because its below 100C outside doesnt mean the inside isnt burning up
mine get warm but theyre def nowhere near close to burning me
motor can be safe if too hot to touch (varnish curing temp is 150 deg iirc)
yeh thats the other thing
some of them just meant to run way hot
so no way to generalize anyway
I try to keep to 50 deg C ish
mine get cooled by the machine mass
like it might be fine giving you 3rd degree burns, no way to know without a datasheet with temp over ambient @ dissipation specs
but in general, i doubt heatsinks will help the situation very much
my stepper looks almost as this, http://www.smcelectronics.com/MOT35.JPG
i've equipped it with a large heatsink
heh, yeh thats prob got wires like tinyfilaments
they rarely get hot
the heatsink will help, im just not convinced it will help significantly
too high a resistance
likely, thats not very high current
i think i'm overcurrent this little sucker
the current limiting is prob based on the individual fillament current handling capability
if i crank down pwm current limit it starts to emit audible noise
again heatsink might help but it wont help a ton
an audible (but high pitched) noise is not uncommon for pwm-type stepper controllers
on its own, it doesn't indicate a problem
anything switched involving magnetics will usually be audible, heh
if i crank up this one to over 300mA it gets quiet
just normal, mine make like a released pressurized gas sound
very low level
mr_boo: heh, weird
mr_boo: might just be an inaudible vibration, might just pushing things together enough they cant vibrate as much
magnetics is weird yo
could be the driver that behaves this way too
wonder if this little sucker is supposed to handle 300mA
microstepped drivers use pwm, high speed switching is almost always noisy
switching supplies in consumer electronics and wall warts are examples
also why you should overspec motors
just because it works doesnt mean it will work in 6 months
part of the ratings might involve insulation reliability over time @ temps
like at higher temps the tiny vibrations might rub away the insulation quicker
higher current ratings = more $$$ for the manufacturer, theyre not gonna be overly conservative
*shouldnt go overspec on motors
hm -- my motor's datasheet has an item for "operating temp. range", but it's not at all clear whether that's the max ambient temperature, case temperature, or core temperature ..
its prob an absolute temp
and then they prob have temp rise @ wattages
that would make sense, but I haven't found that figure yet
so you can add that to ambient and make the absolute temp specs make sense
hehe, its weird when you can actually do the thermal maths on discrete parts
and you realize that they mostly suck
like a 50W parts isnt actually 50W when youre putting 50W into it
unless you have some sort of god cooling setup
are higher ohm stepper motor coils likely to have thinner wire and thus lower current rating?
would be smart to have the coils dipped in epoxy with thick laquer on them
interesting -- this stepper driver chip (which I've never heard of before) promises "sensorless stall protection". trying to interpret the shortspecs pdf, it works by detecting a rapidly oscillating load (which is either the motor current or a value derived from it) http://www.linengineeringstore.com/trinamic/tmc246.aspx
(its current and voltage specs make it otherwise a pretty uninteresting chip for most cnc milling applications)
faq has some good pointers
thanks for all support