03jepler 07HEAD * 10emc2/src/hal/utils/scope_disp.c: draw the triggerline only for the right trace, not all of them
beer's good m,key
agreed. beer's good. :)
so is rum and coke.
yup, I think GT is next though
GT? Gin and Tonic?
03cradek 07lathe_offsets * 10emc2/src/emc/canterp/canterp.cc: pass lathe tool info through all the layers
03cradek 07lathe_offsets * 10emc2/src/emc/iotask/ioControl.cc: pass lathe tool info through all the layers
03cradek 07lathe_offsets * 10emc2/src/emc/nml_intf/ (canon.hh emc.cc emcops.cc): pass lathe tool info through all the layers
03cradek 07lathe_offsets * 10emc2/src/emc/rs274ngc/ (5 files): pass lathe tool info through all the layers
03cradek 07lathe_offsets * 10emc2/src/emc/task/emccanon.cc: pass lathe tool info through all the layers
03cradek 07lathe_offsets * 10emc2/src/emc/usr_intf/ (iosh.cc keystick.cc): pass lathe tool info through all the layers
wow, it even works, we have lathe tool offsets
we're very, very close to being a useful lathe controller
[02:45:40] <cradek> http://timeguy.com/cradek-files/emc/lathe-offset.png
this is radius compensation with x=.5 z=.5 offset in the tool table
(axis doesn't know about tool offsets)
what are the two lines? to passes?
the top is the actual programmed path, the bottom one is with radius comp turned on
well, top and bottom of each pair
you can see the crazy entry moves on the bottom line
ok, so what are the two pairs (white and red)
the white is ignoring the tool location, the red is the .5/.5 offset
axis doesn't read those offsets from the real tool table, so the preview doesn't line up
oh, ok... I was confusing tool offset and tool radius comp
yeah it's several different things, all working :-)
you've been busy
me too: http://home.att.net/~jmkasunich/Pics/KE01052109.pdf
turning the OD of the gear blank now, then gotta do the throat
really wish I had CNC for that, one arc and done (well, a few passes actually)
cool, I saw your photo of this setup from last time
the teeth are this weekend's project
cutting gears is NOT fun without cnc - it's so easy to lose attention and screw it up
I just want to get the blank finished today
you made the cutter yourself too?
you're way more of a machinist than I am
did that back in 2001 when I did the first jig, don't need to do it again
two cutters actually, the drawing shows the gashing cutter, the hob is just a piece of the threaded rod with flutes cut in it and hardened
Been looking for my wirewrap tool for three hours... if it was a snake it would have biten my nose off *sigh*
Jymmm: my grandma used to say that
doncha hate it when that happens
jmkasunich I'm telling ya...
And the REALLY sad part... the last time I used it, I put it in a place so I wouldn't misplace/forget where I stored it!
that will get you every time
that's a sure way to lose stuff
put it somewhere "special"
and I looked RIGHT at it 45 minutes ago too... but did I see it. Nooooooooooooooooo!
been there done that
spent an hour looking for something that was in my pocket once
I can beat that, I've looked for stuff I was holding in my hand
just think what it'll be like when we're old
I've done that at least once.
that's good of you to admit.
usually it's when your working on some POS crap, so pissed casue it's not doing what you KNOW it should be and are ready to start destroying something
I wish my lathe was done to test all this stuff
cradek: get to work on your lathe!
I keep telling myself I won't unpack the coffin (nist lathe) so it encourages me to finish mine instead
jepler: did you see my picture?
[02:59:42] <cradek> http://timeguy.com/cradek-files/emc/lathe-offset.png
cradek: yes I just looked at it
cradek: does this include tool shapes, or just two dimensions worth of offsets?
looks like we'll have to get axis to handle the tool offsets one of these days
just two dims worth of offsets
but I think the rest will be "simple" math in one place
Take for example when you have headlight issues that's ohming out correctly but when you go to turn them on no light. Test the lightbulb and socket seperatly all good, then say fuckit and slice open the harness and you find (this is no joke)... a corroded compression joint of copper wire in the harness where they used 100% pressure to "bind" the wires together.... no crimp ring, no solder.
cradek: the shape information will just treat some kinds of compensated curves as causing gouging, right?
yes I think so
so, you think you can make axis show the tool shape instead of just a cone?
also there's the tool origin offsets that I was talking about today
jmkasunich: SMOP :-)
cradek: while you're changing the tool table you should add enough extra columns so axis can show the mill tools right too
for mills I kept compatibility with the old tool table format
diameter and length, right?
I'm sqrt(2)/3 serious
what more does axis need? (I guess you might want to say "ball mill" or "end mill" or even "tapered 22.5 degrees"
jmkasunich: yes, and FMS, whatever that is
jmkasunich: I think that's exactly what he means
I should just provide a way to load a 3d model of the tool
cradek: very cool work .. when do I take you out for your beeer?
cradek: or maybe not until you finish that lathe
not sure... soon
what is left to do on the lathe?
all the mechanical stuff
motor mounts and such?
you're going servo, right?
got the pulleys and belts?
whacha waiting for then?
I know how that goes
I guess I felt like programming instead
thats what happened to me - now its three years later and my conversion project is still stalled
I hope mine doesn't stall that long!
probalby won't, your lathe is smaller
I have a feeling I'll really want a tool changer next
for the lathe?
the rotating changer on that overpriced lathe at workshop was pretty inspiring
what about a QC toolpost and "pause" in the g-code?
that would definitely work too
don't know if they can be bought that small, maybe I would have to make it somehow
one problem I have with toolposts and tool changers is that they take up space
a QC might not be that much bigger than what's on there
the smallest standard QC is about 2" square without a toolholder on it
the holder adds another 3/4 or so
[03:10:30] <cradek> http://www.mikestools.com/2250-Sherline-Quick-Change-Tool-Post-And-Holders.aspx
looks nice but expensive
can't tell what size it actually is
they say it's the sherline parting tool
so the holder is largeish
I bet it is 2"
[03:13:27] <cradek> http://www.sherline.com/images/2250fig1.gif
this gives a really good idea of the size
[03:13:40] <jmkasunich> http://www.littlemachineshop.com/products/product_view.php?ProductID=2477
[03:15:09] <jmkasunich> http://www.littlemachineshop.com/products/product_view.php?ProductID=1147
I bet that one could be re-designed to be much narrower
it uses a t-nut kind of thing to pull the tool onto the block
you could probalby make the block half as wide, and fasten it to the table with two screws that straddle the tee-nut screw
[03:16:57] <jmkasunich> http://www.littlemachineshop.com/products/product_view.php?ProductID=2401
this is pretty much what I have
that $75 sherline setup looks very nice
I don't currently have any way to mount a round boring bar
round boring bars are great
yeah I bet they fit in round holes better
I've made small ones from endmills
cool, that's a good idea
chip one flute of a 2 or 4 flute endmill and its no good for milling
(I keep all my broken tools too)
but grind off all but one flute and you have a boring bar
a nice one too
I'll remember that
I also have some real 3/4" shank bars with reduced tips that can bore down to about 1/2"
and a homemade one that is 3/4" diameter x 6" long and accepts a short piece of 3/16" square HSS
that one is for holes over about 1"
I doubt I've ever bored anything 6" deep
actually its 8" long, but with 3" in the toolholder, max depth is 5"
not sure what the deepest I've bored is
the bore for the worm in the gearbox has got to be up there
maybe 3.5 or 4" deep
diameter is only a hair over 5/8 tho, that bore is a fight with chatter
yeah that's really deep then
[03:25:28] <jmkasunich> http://www.metalworking.com/DropBox/_2001_retired_files/Worm02.jpg
you can see the bore in that pic
I bore it from the left side (large dia for the first bearing, then straight thru the rest of the way
what's the tolerance on the small part?
the OD of the gear is 1.99", so I'm guessing the block is about 3.5 wide (drawing is on the doze PC)
slip fit for the bearing
I aim for plus 0.001, minus nothing
the big (left) bearing takes the thrust
your lathe is big enough to hold a 3.5" block in a 4-jaw?
shoptasks (like most 3-in-1's) have a large swing
I always forget people's machines are all bigger than mine
something like 16" over the ways, 8" over the table
actually I could do that on the 6" if the jaws don't stick out much
wow that is big
I honestly don't remember if I turn that gearbox housing in the 4-jaw or on the faceplate
its been a couple years since the last time I made one
pain in the ass to machine
first bolt all the pieces together, drill and ream for down pins
then drill for the bolts that hold it together
then put it in the lathe, drill and bore the hold for the leadscrew/gear
take it apart, put it back in the lathe, and bore the larger diameter steps for the gear and thrust bearings
then put it back together, turn it 90 degrees, and bore the hole for the worm
damn I can't spell for shite
"down pins" = "dowel pins"
heh, close enough
"bore the hold" ;-)
boar the whole
[03:34:21] <jmkasunich> http://www.metalworking.com/DropBox/_2001_retired_files/Worm04.jpg
is the best view of all the pieces
very nice finish on everything
you make the crank too?
I wised up since I made the first one - $25 to McMaster gets me a pre-made crank
(the crank in the pic I made)
the knob is a nice touch
I also buy the shaft coupling (again, the pic has a home-made one)
the mcmaster crank is even nicer
it has a knob, and the OD of the wheel is knurled
I've never tried knurling...
it usually works fine in only a few tries
(I've had mixed success)
Actually, kinda cool... http://led.linear1.org/led.wiz
I bet such a big diameter would be a challenge
I'm quite happy letting McMaster do it for me ;-)
you'd be pretty committed by the time you could tell if your knurl was going to line up
shame I can't buy the gear (or the whold gearbox)
that couldn't be any fun
I bet it takes me 30 hours to make that gear and box, maybe more
I'm not surprised
if I could buy it for $200 I'd do it in a flash
it's easy to use an hour to bore a hole
I probably have 8-10 hours in the two gear/screw assemblies already
cut gear blank, face one side, drill, bore, tap (tappings a bitch), cut screw to length, clean up ends, turn down first 3/4" to 3/8 dia, thread gear onto screw and locktite, then pin, cut 5/8 sleeve to length, face both ends, drill/ream, and press onto screw
then face both sides of gear blank, put steps on it to center the thrust washers, turn the OD (where I am now)
turn the throat
then gash teeth, hob, and spend a hour deburring
how many hours (if any) would cnc save you on this job?
hard to say
the only thing that really screams out for CNC so far is the throat on the gear (an arc cut)
there are a lot of places where CNC would let me sit and watch instead of turning the cranks
but the actual cutting time wouldn't be any lower
have you every used HNC (human)? calculate the steps and dial them in?
might be longer actually, doing it by hand I can push the depth and/or feed more
I considered that for the throat
still haven't decided what to do
HNC is in my bag of tricks but I haven't used it for a real project
I rarely find myself cutting the kind of profiles where CNC (or HNC) really shines
usually its a straight diameter, a step, and another diameter
maybe several steps, etc
yeah CNC doesn't get you much then
unless it's an hour of slow boring
diameter and depth are critical, but there aren't any curves or even tapers
the limiting factor on the bulk boring is clearing chips
damn things wrap around the boring bar and make a mess
if I could get the chips to break it would be much quicker
thats one nice thing about the worm bore (as opposed to the gear bore)... since the housing is split and I'm boring along the split line, there is a natural break in the chip every 180 degrees
at least it's not chattering if you get those long chips
oh, I can get chatter and long chips at the same time :-(
[03:53:40] <cradek> http://www.metalworking.com/DropBox/_2001_retired_files/untwisted.jpg
chinese drill bits
costs you a drill bit AND your work
jmkasunich No, they are HSS... Hunk of Shit Steel
chinese cheddar cheese steel
[03:55:47] <jmkasunich> http://www.metalworking.com/DropBox/_2001_retired_files/Facemil3.jpg
this is what I'll be doing in a few weeks
that's quite the tool
actually I'm a little bummed - I have 4 pieces of 3/4" and 8 pieces of 1/2" to do, so I'll need two stacks
twice as much time
that tool cost me all of $10
at the $2/lb sale
5-1/2" diameter, 12 teeth, it moves some metal
I've gone as deep as 0.250 while cutting nearly full width (that pic is probably between 0.050 and 0.100 deep)
I've also cut 5/8" deep x 1" wide in one pass with it
see pic 4
the chips go "clink" when they hit the floor ;-)
what sucks is that it takes me longer to break the sharp 90 degree corners on the plates than it did to mill them
12 plates, each has 8 long edges, 4 short edges, and 8 corners......
I wonder if a router table, and a chamfer bit with a ball bearing pilot
still probably leave a burr
heh, they really stretch the meaning of the word "portable" http://www.metalworking.com/DropBox/_2001_retired_files/facemill96.jpg
I agree it would be possible to move it from one place to another
I remember when all those pics got posted.. we were having "facemill wars" - one-upmanship
started with a 2" one http://www.metalworking.com/DropBox/_2001_retired_files/facemill1.jpg
then this one
[04:06:26] <jmkasunich> http://www.metalworking.com/DropBox/_2001_retired_files/Face_mill1.jpg
then finally the 96" one... everybody pretty much shut up after that
wth is that last pic?
about a 12" face mill
no thanks, you're not my type
what do you use that on?
things that need to be flat
heh, no I mean surely not a BP
jmk what type of gear are you trying to cut
no, not a BP
big old cincinati horizontal or something like that
roltek: worm gear
[04:09:39] <jmkasunich> http://www.metalworking.com/DropBox/_2001_retired_files/Worm04.jpg
I'm using 5/8 - 5tpi acme threaded rod as the worm
so pressure angle is whatever it is
[04:10:59] <jmkasunich> http://home.att.net/~jmkasunich/Pics/KE01052109.pdf
is the drawing, it has some of the specs
i have hobs but think i can help you with that
I only have to make 2 of them
I've done it before, have all the tooling already
if you look at higher numbered pics (worm05.jpg thru worm09.jpg) you'll see my tooling and setup
5 & 6 are tooling, 7 & 8 are gashing, 9 & 10 are hobbing (free hobbing)
pretty much amateur hour compared to the kind of work you do ;-)
not really just another way to skin a cat
the slow way
you do with what you have
when your gashing worm wheels , i take it your indexing your gear so you will have to be concerned with indexing error
I'm using a dividing head, if thats what you mean by indexing
IOW, don't fsck up the cranking...
the arms really help there, but you still gotta remember to do everything the right way
turn the arms, pull the pin, turn the handle N turns, then continue till it gets to the arm
i have a ? on your encoder ratio hal mod
can i hook more than 1 axis up at a time to follow master
just use more than one encoder ratio component
you can route the master encoder signals to as many components as you want
was there any correction factors built in for counting of steps and slowwing down or speeding up
the main limitation will either be having enough inputs for the encoder signals, or running out of speed (this is a software encoder counter)
not sure I follow you
the intent was that the master encoder signal goes to one input, the slave encoder goes to the other, and the output is position (tracking) error
send the error to the PID to close the loop
if the master speeds up or slows down the slave does the same
some of electronic gear box's i have seen or egb will have counting correction's
guess I don't know what they are correcting for
i am not sure but will find out more
if its something that is needed, and you can explain it in a way I understand, I can probably add it
i have to do more studying on egb
i do know that they use phase lock looping
see you guy's later
man, solding just kills my back!
need a better bench layout
yeah, and a 3rd hand wouldn't hurt much either
but when your soldering bench is also your buffing/snading/saw bench, it makes it difficult at times =)
3rd hand like this: http://www.rjsintl.com/productListing.asp?category=Soldering&cat1=Tools&cat2=Third+Hand
I was thinking more like this, but I guess that'll do =) http://elfwood.lysator.liu.se/farp/hand/Untitled-28.JPG
What I'll end up doing is just making a fixture instead
Hey, have you used a wire wrap GUN before?
a long time ago, and not very much
do yo remember if it automatically strips the wire too?
I don't think it does
do you think ANY of them might?
I haven't done any wire-wrapping in maybe 10 years
heh, me neither. I have a manual tool.
wake up gents :D
Can anyone think of a "reasonable" way to hold a 9V battery in a pocket that doens't require any tools to replace the battery? The only thing I could think of was using a brass disc that would rotate over the battery itself (bent slightly to hold the battery snag).
Jymmm: is it correct "Thanks to Giacus for share this video." ?
or "Thanks to Giacus for sharing this video." ?
the 2nd one
I should probably get going to work :)
only if you like to eat =)
well, I already ate
and sleep under a roof
well... I just woke up, and it seems I'm under a roof :P
not if you dont pay the rent
over here it's not really custom to pay rent
most people kinda own their places
as it should be
nobody seems to sleep much in these days
WE can sleep all we want when we're dead
I had a battle with mosquitos latest night
wake up at 6:00
there's a new sort of tropical mosquitos here since some year
called mosquito tiger
they looks like condors
eat garlic for snack
[08:16:23] <giacus> http://www.comune.torino.it/ucstampa/2004/article_562.shtml
No, make a garlic concentrate an d spray around windows, doors, house, etc.
yes, I need something like that
giacus: how bad is it over there?
alex_joni: for ?
with the mosquitos I mean
* alex_joni is thinking about sicily :D
should I pack tons of spray?
no, you can find all there around
the better I know is called Vape magic
giacus Make some garlic oil, but dont use a vegetable oil that will go rancid on you
mosquitos become very aggressive especially with peoples that comes from places far away
it seems they are able to understand the difference of skin
giacus: To make garlic oil, use LOW heat
Jymmm: I'll try
ok.. later guys
regarding the "dapper" problems of installing EMC... do they extend to compiling the source too?
how do you mean?
just compiling the source won't help
AFAIK it's the RT kernel that's causing the issues, but you'd better ask someone that knows about that better (cradek, alex_joni, and several others. check #emc-devel)
emc needs realtime extensions in the linux kernel
Ah ... thnx :-)
because it needs certain parts of it to work very fast (every few microseconds)
Lerneaen_Hydra_ is now known as Lerneaen_Hydra
do you know anything about emc & dapper?
ok, I'll wait untill alex or someone else who knows comes
* Lerneaen_Hydra2 pokes alex_joni
[10:49:21] <Lerneaen_Hydra2> http://emc.pastebin.com/728667
<- this is the second time I run the script, the first time I didn't run it in a terminal and it silently died
[10:54:03] <Lerneaen_Hydra2> http://emc.pastebin.com/728671
<- the modified script I ran
unfortunately it seems to take forever to load
yes, it did for me too
just use regular pastebin then
for some reason, even though AFAIK it's on the same server
hmm, pastebin seems to be very slow to upload
ok if I flood?
gksudo -m "Enter your password to install EMC2" true
gpg --keyserver pgpkeys.mit.edu --recv-key BC92B87F
gpg -a --export BC92B87F | sudo apt-key add -
grep -q emc2 /etc/apt/sources.list || sudo sh -c 'echo "deb http://dsplabs.cs.upt.ro/emc2/
dapper emc2" >>/etc/apt/sources.list; echo "deb-src http://dsplabs.cs.upt.ro/emc2/
dapper emc2" >>/etc/apt/sources.list'
sudo apt-get update
sudo apt-get install emc2-axis
is my script
and when I run it it proceeds with the apt-get update fine, and seems to add the reps correctly
does your /etc/apt/sources.list already have stuff in it that says breezy?
Building dependency tree... Done
Some packages could not be installed. This may mean that you have
requested an impossible situation or if you are using the unstable
distribution that some required packages have not yet been created
or been moved out of Incoming.
Since you only requested a single operation it is extremely likely that
the package is simply not installable and a bug report against
that package should be filed.
The following information may help to resolve the situation:
The following packages have unmet dependencies:
emc2-axis: Depends: emc2 but it is not going to be installed
Depends: bwidget (>= 1.7) but it is not installable
Depends: bwidget (< 1.8) but it is not installable
E: Broken packages
is the end of what happens when I run
I'm not sure
Lerneaen_Hydra2: you need Universe
because bwidget is there
I forgot to add them
universe is enough
I'll test that soon
ooh. 2.0.1 works nicely
or at least, sim does
anyway, I'll test cvs later today or tomorrow
cool, so except the bwidget there was no other problem?
could someone suggest a good way of adding trim-pots to this design so i can trim the lowest and highest voltage from this voltage regulator application? http://www.carmi.se/misterstarshine/img/4.gif
right now naturally it swings between -0.1 volts and 18.7 volts
i'd like to have between 0 and 15 volts
is there a clever way to do this?
add trim pots above and below Rvar, so you have 3 in series instead of only one
oops - one needs to be in parallel, one in series
turn Rvar all the way down, and adjustthe series trim pot so you get the lowest voltage you want
turn Rvar all the way up, and adjust the parallel trim pot for the highest voltage you want
03jepler 07HEAD * 10emc2/src/hal/utils/scope_disp.c: draw the baseline for the samples even when there are no samples
the trimmed minimum output voltage seems to be incredibly sensitive to which mode on the multimeter i choose
on the 1000 VDC it says the minimum val is 11 volts and on the 200 volts mode it says 2 volts all the way down to the 200mV mode where it shows zero
i thought the output impedance of a multimeter in voltage mode was high
the input impedance should be around 1 or 10 megohm
alex_jon1 is now known as alex_joni
SWP_Away is now known as SWPadnos
it is 1 megaohm on all modes
and the voltage value shows different for each mode. this is totally whacked up!
is it a crappy meter?
I suppose I should rephrase that - it
i don't know. but it should'nt be this bat
it's probably a crappy meter
the output voltage doesn't seem to be very stable either
it's flickering around all over the place
remember - the meter has the same input voltage range and resolution, regardless of the range setting. there are just resistor dividers put in front of the A/D
gotta look at the output on the scope i think
so small amounts of noise will look like large voltages in the 200V range
set the meter for AC, and see what happens
i've looked at the output voltage on the osc.scope and there is hardly any ripple or noise at all
hook up the meter and the scope, and see if there are any major changes when you switch meter ranges
i'm now testing with two multimeters, one is newer than the other and possibly more reliable
the other multimeter shows 20 volts on the 600V mode and 10 volts when in 200V mode
and what does mr. o-scope say?
it didn't do that when i was trying it with a 9V battery
o-scope doesn't show a lot of ripple but these strange behaviour must be due to ripple
lets zoom in the ripple real good!
as a matter of fact, it looks like the ripple exceeds 0.2 volts in magnitude!
but how can the multimeter read 20 volts on 10 volts
the signal looks awful
there is a low frequency sawtooth overloaded and a 100MHz signal with 0.2V magnitude p-p
my 1000µF smoothing caps appears highly inadequate for the task :(
one can hardly obtain ripple-free 1 amps current from a 1000µF smoothing cap
i seriously need some monster caps
this explained the saw-tooth shaped ripple on my output signal but it does not explain the aggressive 100MHz signal with 0.2V p-p
could someone refresh my memory regarding smoothing caps for power supplies as a function of the load current drawn
Bo^Dick: a 1000uF cap probably won't take out a 100mhz signal, you need to add some smaller bypass caps
but.. where the heck is a 100mhz signal coming from?
it's probably from a nearby radio station
they have radio stations in sweden? :)
oh right - maybe not ;)
the 100MHz signal puzzles me
first of all i need to recall to that formula that calculated the needed smoothing capacitance as a function of the desired current to be drawn
dv/dt = I/C
if i remember correctly the formula also incorporated the ripple level which in my case shouldn't be larger than 0.5 volts max
at 1 amp that is
are you on 50 Hz or 60 Hz?
ok, so assuming that the ripple is from line frequency (possibly a bad assumption):
dt = 0.01 second
dv = 0.5V (specified by you)
i'm only talking about that ripple now, not about the 100MHz noise
I = 1A (specified by you)
.5/0.01 = 1/C
yep (left as an exercise for the reader)
that's a monster cap!
you havea n extra zero in there
could be worse
time to re-assume some assumptions
alex_joni: you there ?
I lost the url of your videoclip (my toy) :(
alex how could be named that ?
was running with emc2 I suppose
just to be sure :)
Is Anders Wallin here ? Or does someone know if he comes by here ?
I don't recall ever seeing him here (at least not by a name that I recognize)
jmkasunich: Thanks, I got an email from him and thought he might use IRC to, I will just drop him a mail .. thanks
he does use irc, but I forget the name he uses
great, thanks :D
Does anyone know of a way to dynamically adjust classicladder timers/monostables ? I was hoping to make configurable timers through VCP when float support is available :D
what dos emulator can i throw at ubuntu??
I don't think that's doable (without major hacking on classicladder)
isn't there a wine package for ubuntu ?
Dallur: if you need timers who's delay is adjusted by a HAL signal, I could probably write such a hal component
that seems like a hack tho
the classicladder timers are already there
jmkasunich: Yup, I think it would make sense to do it with classicladder
jmkasunich: I think I might have found some ref. to a T[...] variable but I still have no way to pass float/int to classicladder, I'm going to poke around a bit more
the problem is that we don' t have anyone who really understands classicladder
(with the possible exception of petev)
what kind of granularity are you looking for?
today, CL limits you to 100mS, even tho the code is running every 1mS
(an artifact of CL's original, non-RT design)
yup, which is fine for what I am doing
I have a pierce delay timer which really should be adjustable for sections of different thickness
and I have a timer which triggers an estop if an arc is not established within a given period from the torch being turned on
100ms is just right for the pierce delay and more than good enough for the estop timer
dmessier: I use qemu with freedos, there's also a free (free as in money) vmware now
dmessier: the vmware is a little more challenging to install, but it runs faster
dmessier: I highly recommend -> http://www.winehq.org/site/download-deb,
I use it at work when I have to run IE on my laptop to test IE only web stuff
I think wine does not run dos
dmessier: although I run Centos on my laptop :D
if you have a real copy of dos (or freedos) you want to run, you need a machine emulator, like qemu or vmware
cradek: It is true that it will not run dos but it emulates dos and as such you can run almost any win32 application on it
Dallur: how well does wine work these days?
ahh i see... does freedos let you set dos version??
dmessier: not sure
I tried it once, long ago, and gave up in frustration
i need an old dos ver..
I want to run a doze CAD program, easycad, on my linux box
tryin' to resurect a postprocessor..
I've tried wine on and off for many years, never seemed quite good enough to be useful
dmessier: why does it depend on dos version?
I bet wine is good now, it's a very old app
locked in the code SOMEWHERE..
dmessier: that's nuts
fine wine improves with age?
comes up tith incorrect dos version error
cradek: it is pretty decent by now, runs the entire office suite, IE and almost anything I have thrown at it, hmm I think I need to try Mach3 just for fun one of these days :D
Dallur: that's nice to know
Dallur: it doesn't require you own a windows license, which is a big advantage I think
windows license? whats that?
(I run w95 on the one remaining doze box I have)
jmkasunich: it is like a tax on every computer sold that goes to a company in seattle so they can spy on you
this post is a fill kinetic carrying post that eats apt cl
dmessier: do you remember the name of the program that tells dos to fake versions for certain apps? I can see if mine has it, but I can't remember the name of it
ive tried to fake it from win2000 ,XP , 98... doesnt work... also it seemed to like a 486 best..
dont ask... its old stuff... i did say "Resurrect"
I don't see a dosver
want to borrow my AT?
want to try some software??
actually I recently unhooked it because I needed the desk space
why not just put dos on an old 486?
figures... EVERY ones trashin them... and NOT giving them to me... i have 1 in the basement but its a pita
or do you want to actually use this software as-is
my AT is running a recent freedos too, so it probably wouldn't work
i would like to have it web based... send in apt get back g-code
maybe you could disassemble it enough to find the dos version check, and disable it
you build YOUR machines post.. on this server.. and post to it... for g-code
I doubt it actually depends on a certain dos version.
i've looked and couldnt make it out..
me too.. but its something
linux just seems to be the way to get the networking i need IF i can get the s/ware to run somehow
dmessier: is it just an exe? I can try to run it for you on freedos
looks like you can set VERSION in the environment
I think it's setver, not dosver
cradek: thanks for suggesting HNC last night
just tried it on the first gear blank, worked great
0.005 steps, all I needed was a tiny bit of 500 grit to smooth it up
(I was cutting a 7/16 dia curve, so I wrapped the sandpaper around a 27/64 drill bit shank, the 1/64 undersize is to make up for the paper thickness)
I think the result is better than I got in the past (where I set up a crude "ball turning attachment" by mounting a tool and swiveling the compound to cut the arc
about 80 steps?
that's another creative way of doing it
0.250 to 0.500 by 0.005
feeling math challenged right now...
I guess that wouldn't take long at all
much less than the setup time the old way
I don't think my compound slide is at all rigid when it's not tightened down
mine isn't that rigid either, but I take light cuts
I ordered that QC
the $75 one?
it'll be interesting to see what size it actually is
once you get it, let me know, I'd like to compare the dovetail measurements with mine
I got a couple extra turning tool holders, so I can mount all three of my insert tools (left, center, right) and a threading tool
we recently had a request from design engineering to make the I/D of a set of bushings spherical...
but simple with cnc and a boring bar
after the bushes have been installed into the main fitting... ;(
just get out your spherical reamer ;-)
d'andrea head attatchment on our new okk SHOULD handle it... but what a cluster fack of a BAD idea
dumbass design engineers
i just DONT see the engineering theory in this one... the pin is still a cylinder..
sounds very odd
spherical as in bigger in the center than the ends?
next they'll ask you how to get the sphere in there
amd we are talking about hte main PINTLE pin that holds the L/g the the aircraft
how spherical? (I mean, how much bigger in the center than the ends - a few thou, or a lot?)
its been mentioned.... i went ballistic as did the SR. methods guy... we went of on need ing gravity intensifiers to do it... and shite.. was a hoot..
about .024 over a 3.5" bush from center both ways...
and dont get to the nitro way... ive been there..
and the ID is bigger in the center?
Jymmm is now known as Jymm
smaller in the center...
oh a torus shape
thats not spherical then
* cradek points at jmk
and I can even (maybe) understand why
under heavy load, the pin is gonna bend a little
I can see a torus, it allows error in alignment
in a straight bore, that will concentrate all force on the edges
doghnut yeah.. torus connonical form torus/x,y,z,r1,r2
1 edge point in the center..
but i think the plane will wiggle EVERY time it lands
I bet there are other points of support
words fail me...
is the gear hanging from only one pin?
there shouldnt be a engineer that stupid alive.. not to have additional support..
1 pin... an actuator... and a side stay brace..
the side stay brace will prevent the wiggle then
the main pintle pin is ALL that holds MOST commercial and buisness jets l/g on
hm, might the donut have to do with preventing long term fatigue
but the short ter..high stress profile in the center defeats it...
ther will be CLEARANCE Clerance.. by the 2 landing
the bushings are 17-4 ph at 42-26 Rc
clearance when unloaded
what about the pin?
and what diameter is it?
it WILL mash the highspot off the bushing i believe
300m.. pre- machined....h/t...finish turned...nital etch inspect... mag particle inspect .. out for chrome and cad and another mag.. final grind on chrome...lpi of ground chrome... mag on ground chrome... final dimensional and visual insection
lpi = liquid penetrant inspection
they really really don't want cracks
Can anyone think of a reasonable way to hold a 9V battery into a pocket w/o using tool to replace the battery? I was thinking of using a small disc that rotated over the battery, but I'm not sure I'll have even room to do that, I'm workign with limited space.
I DO have side room, just not depth to play with
cover and a zip tie glued OUtside??
edges of the battery are rounded, can you make something that goes over only the edge, not the flat part?
maybe one flathead screw on each side of the battery
jmkasunich like a spring tab of sorts?
the screwheads are flush with the top of the battery, but the edge of each head holds down the battery
super glue.. no screws... ; 0
could you mould it??
They have to be able to change the battery, preferably without the use of tools.
Jymm: can you visualize the screw based design I described?
jmkasunich The screw head overlaps the rounded part of the longest edge of the battery
2 slots to catch the cover on the sides...
if you don't want screws, then do something similar with springy tabs
either way, it can be effectively zero depth (other than the depth of the battery itself)
jmkasunich I have no room on the depth of width of the battery, I DO have shitloads on the length though.
a few minutes ago you said you had side room
jmkasunich My bad... I was looking at it laying on it's edge
on one end you have the terminals, right?
I was thinking of some way to take advantage of the clip as part of the holding
so that end ain't gonna pop out
do you have _any_ depth at all?
0.001", 0.005", 0.010"?
likewise on the sides?
If I go more than .065" in deoth, I'll break thru the wall.
what about on the sides
a little can of exaning foam... some fiberglass cloth... and a little epoxy resin is all it takes to make a cover
line the longest edges with weather stripping, and just compression fit it in there
vacuum bagging is cool too... but it must be right
jmkasunich .125" on the width and I'll break thru the walls
what material are these walls?
you got OOOdles a room.. .125??? LOL
i thought you were dealing with mi usual .03 wall thikness
dmessier in wood =)
0.030 in wood?
after epoxy.... yes...
Jymm: I'm having a real problem visualizing this
there is wood under the battery, right? (assume battery laying down flat)
jmkasunich Take a 1x4" and make a 2"x1" pocket to hold the battery.
Jymm: you can push the battery from behind with a spring..
noting that a 1x4 is actually .75 x 3.5"
ah, got it - a pocket in a solid piece
0 is the connector...
you said you have a lot of length
I'm sure you've seen it before
anonimasu any suggestions on the spring and how to keep the connector end from falling out?
how much? can you install the battery from the end?
the spring pushes the battery aginst the connector
jmkasunich I have about 2" on each end on the battery
you can get model airplain plywood in 1/32 and even 1/64 thick
pocket the hole for the battery, then pocket out 1/4" all around and 1/32 deep, glue in a cover of 1/32 plywood
stuff the battery in the end
jmkasunich so sitting on the table, would you be able to see the battery hole?
I have no idea how it sits on the table
you have any pics or drawings? words aren't enough
think of a tbale lamp, where the bottom of the lamp is where the battery goes, hidden underneith
ok, when the lamp is sitting on the table, is the battery flat or vertical?
so its a big but thin base
but only 1/8" wider than the battery
but 4" longer than the battery
and you want the top surface to appear as untouched wood
but the bottom can be open, or ugly, or whatever
as long as it doesn't fall out
what kind of wood? fine grained and strong, or coarse and crappy?
could be all the above, but if I have to limit what can be used, so be it.
how about this: use a cutter that can undercut just a tiny bit (think a 3/8" diameter sphere on a 1/4" shank)
mill a pocket that is twice as long as the battery
the end near the connector is undercut, so it extends just a tiny bit over the rounded corners of the battery
the far end isn't undercut, it has straight walls
ok, like a weird keyhole slot on half
drop the battery in the straight side, then slide it toward the connector
drop in the battery, slide it over
still gotta figure out how to hold it once its in there
I could get some battery clips that mount to the wall of the wood
it'll make it a reall bitch to wire, but might be doable
why can't you make the wood a little thicker or wider?
btw, how thick is a 9V battery?
1" is common, cost goes up dramatically past that. .680"
1" nominal, 0.75 actual, I assume?
how wide is the battery?
and any sanding drps that .75" down too.
1" x 2"
1" x 2" x .680"
how wide is the board?
so basically, you want to stick 1" x 2" x 0.680" inside 1.25" x 6" x 0.75"
* Jymm nods
which way is the grain running? you said 6" wide and 1.25 long, is that true?
grain runs the short way?
ok, so 6" long and 1.25" wide
It's being cut out of 1x4's
1x4 is 3.5 wide, not 1.25
or are you trying to get two side by side?
No, that's the MAX that that battery can have, past that is other stuff
going beyond 1.25" and you break thru walls
either to the outside of the item, or into another "compartment", right?
well, hitting 1.25" and there is no wall =)
I assume that the thing powered by the battery is in the other compartment?
yes, and I have to drill a 1.25"+ hole to get from that compartment into the battery compartment to run the wires.
hmm, I was starting to get a mental image, and now its gone
1/8" diameter, 1.25"+ long
duh, I was assuming 1.25 dia
tis ok, it's the weekend =)
brb,, need ice tea refill
if the wall is so thin, why is the hole so deep?
Just incase I need to use a walwort
this is really tough
because you obviously have a mental image of what we are talking about
and I/we don't
Let me see if I can draw this up...
I gotta figure out how to draw the "lips" =)
here's the 'experiment' http://www.giacus.org/en/videoclips.html
text scrolling on video might be changed
* robin_sz wonders what a "walwort" is?
codec should be fine, for latest mediaplayer vers. too
and file size small enough
how are you jmk?
robin_sz: did you shou a video of you machinaries running ?
why not ?
would be nice
jmkasunich, we got busy again ... were quiet last week, but run off our feet this week
coolant/swarf systems for Mazak
I'm making gears
actually, I'm procrastinating, but I'm supposed to be making gears
giacus, one comment, lose the cartoons. gotta be 25% of the download.. nice as it is .. put it up seperately
its one thing to have a 30 second intro to a 10 minute video
but when the video is less than a minute, a long intro is a big much
* robin_sz is enjoying a nice pint of beer
we could also remove transitions and video delay :D
weird, we still do beer in pints huh?
just back to the original format :/
you have to impress linucnc in the mind of peoples ..
that would be the target
less is more
depends if you are targetting engineers or housewives
have to get out of the shell a bit ..
engineers aitn fussed about cute cartooons.
users are the power
you can develop how much you want
your power are users
doesnt matter is engineers, hobbist or other
well, unlike a cash-based business, the developers can ignore the users if they want,
open source projects are so
how can ignore they ?
well, imagine there are two tasks to be done in the code
one the users are really wanting, but is dul, tedious and boring to code
the other is just an idea that the developers have had, but is sexy and exciting to code ...
which one will get done?
no .. its a different approach
may you're missing what I meant
must be ...
you need user for testing, debug
users cant submit bug reports usually
This nickname is owned by someone else
If this is your nickname, type /msg NickServ IDENTIFY <password>
especially for debug
shrug .. they are of some use
you should say a developer do things in a certain way
he could try 1000 times and get no issues
a 'stupid' user can help , instead
just because he think different
and act different ,,
but they dont contribute much to an open source project
in this sense they can
open source projects tend to go where the delopers want them to go
commercial projects go where users want them to go
but I think a big projecy needs donation
donations in money or HW
thats the difference
i.e I'm a free software supporter
sometimes OS projects do things no commercial project would do, because users dont want (or think they want) that function
I donate 250 euro/year just for that
also 100 euro on wikipedia too
this really help
but you need to become popoular before..
I do not pay for licenses
in OS projects, I dont think users are the power, its developers
as many peoples instead have to do
and in commercial projects, its the sales team
you should ask that to who develop night and day ..
OS developers develop what they feel motivated to do
paid dvelopers do what they are told/paid for, no matter how dull
I didnt say to pay developer, but help
right .. and moving swiftly on ...
and there are many ways to do it
this is a NICE curry ...
mushroom biryani. mm nice
btw, I think its quite simple today find a software to run a cnc under doze
I don't know much, because I just used emc since I started
yeah, there is stuff around
not much as flexible or reliabe as EMC though
But I see lots of peoples in cnc forum they do not talk about linux or free software
just using what they have on theyr hands ..
but the front ends are nicer, and easier to install / set up if you are a doze person
good or bad
robin_sz: I know the difference between 'easier to use' and 'good to use'
learning curve is very different
tried mach2/3 yet?
but many peoples don't think that
what's that ?
a good rule when developing code, or anyting really, machines, whatever, is to research what the market leaders are doing, get what ideas you can amd select the best features of them
its what artists and musicians have doen for centuries
have to look at 360°
so, tomorow, when I ask "have you tried mach2/3 yet" maybe the answer will be different
I hope so for you :)
oh I tried it many times ...
its not that great to use seriously
but, as you say, you need to look around 360 degrees
dmessier, I think I'm gonna try out Astoria tonight.
Jymm is now known as Jymmm
anyway .. as I was saying, that was a NICE curry :)
and the beer is helping wash it down
the problem is the time..
darn time, stop it
Whenever I think of curry, a secondary thought soon follows - many white people can't handle spice...
primary thought of, YUMMY.
strangely, of all the restaurants in the UK, curry restaurants are proably the most numerous
robin_sz... yeah... soo many of my friends are now thinking indian food is good... though their stomachs can't handle the spice, and their systems do a little purging.
I love this http://en.wikipedia.org/wiki/Spaghetti
cant stay without, any 2-3 days I have to eat it
<shrug> spaghetti is yummy... but I like vermicilli better.
or however you spell it.
we do homemade tomato sauce every year in august
with tomatoes from my garden hehe
that's the big secret
for a good quality pasta you should try barilla brand
the best !
we pay it 1/3 more of other here ..
exported should be not cheap too
but its damned good
[22:13:10] <giacus> http://www.annamariavolpi.com/how_to_cook_pasta.html
and finally http://www.annamariavolpi.com/how_to_eat_spaghetti.html
doh, the video too :D
spaghetti is one of the few good things to come from Italy :)
few ? :)
Ducati and Limon Cello being the other two
in the past they were many !
indeed, all change ..
oh? is barilla good?
Lerneaen_Hydra: absolutely the best
I sayd isnt cheap too ..
compared to others
I think they are quite easy to find in sweden
it comes from Parma
I suppose Sweden must import all its spaghetti,
same place of Prosciutto do parma (bacon)
too cold to grow it I assume?
or cheese (parmigiano reggiano)
hmm, I'm not sure. it depends on how sensitive it is
I think it needs good sun to ripen on the trees
here where I live it's not all too cold, we've got most standard fruit trees (plum, pear, apple and so on) and most plants that live in germany & france, so it's not all too cold
probably around 20-25°C during the summer
there was a very famous programme on the BBC in 1960, about the Swiss spaghetti harvest
where I live near 0
maybe 20-30 meters above, at most
Spaghetti is not a widely-eaten food in the UK and is considered by many as an exotic delicacy.
Mr Dimbleby explained how each year the end of March is a very anxious time for Spaghetti harvesters all over Europe as severe frost can impair the flavour of the spaghetti.
He also explained how each strand of spaghetti always grows to the same length thanks to years of hard work by generations of growers.
[22:24:38] <robin_sz> http://news.bbc.co.uk/olmedia/70000/video/_70980_aprilfool_vi.ram
oh, april fools, right
it was the first april fool ever to be done on television
until then the BBC was always VERY serious
wait, 1957 ...
long time ago
robin_sz: I bought one of these for my mother http://www.vicentia.net/eng/products.htm
as gift, years ago
have to add some stepper :D
wanna try to build one like that ?
nah, I kn ow it grows on trees ;)
Produce over the 600 different shapes of pasta.
where is that pesky les_w?
so .. this week, ther was a company that closed locally, but hey still had one very good product, and they still have orders
but now, they have no factory ... so the old boss came to see me, maybe I can make them and ship them, he can still make some money
did they close les_w's factory?
nah, not in the USA
watching the video .. http://www.vicentia.net/filmato.wmv
here in the UK
if we get the contract it will be very good
its like twice our current turnover :)
not bad ;)
all in "chromweld" too,
not quite stainless steel
1.4003is the euro spec, or 3CR12
sounds like a PITA to machine
its ok to laser and weld
oh, but cutting machine and so on (mill/lathe) is not?
or well, like stainless
aaaaanyway, it's late here now
cya all. (g'night)
Is there any instance where the GUI (TkEMC) would display different Z axis position than the stepgen.x.position-fb would display ?
I am seeing the wierdest thing, when I run in manual mode I can move Z axis and I get the same readout from TkEMC as I get from halcmd but when I start a gcode TkEMC displays something way different
I change back into manual after running in auto and the same thing as auto ...
you probably have offsets active
in MDI mode program G90x0y0z0 and G10L2P1x0y0z0
and maybe a G54 for good measure
I implemented the feed-hold a while back
that might be the cause
RS274NGC_STARTUP_CODE = G50
that won't make offsets
but does it successfully turn on adaptive feed by default? I never tested that
yup it does :D
I use it for my feed-hold function
I have not implemented any offsets to the best of my knowledge, can they be included in the gcode or ?
yes gcode can set them
if you program MDI g28 does it show 0,0,0?
I am checking
switch to MDI, type in g28 I get work offsets 0 0 0
ok then you don't have active offsets
and in the auto mode there are no offfsets
The actual distance traveled should be 300 but is 11.8111 which gives a ratio of 25,4
I know that number .... :D
ok back up, what's the problem again?
If the gcode is in inches do the "DROs" also display inches ?
is the machine configure inches or mm?
so the gcode program has a g20 in it?
it'll probably show you mm still, but I'm not sure in tkemc
is there a problem with the distance it moves, or just the display?
just the display
isn't there a button to change the display?
sorry figured out what I was doing wrong I think
(now I feel really stupid)
There is a menu to select mm/inches and it seems after loading a file with g20 you have to change back to mm
the g20 will continue to be in effect until you change it
you can go back to mm with g21
*every* gcode program should have g20 or g21 at the beginning
I will file this one as pebkac (problem exists between keyboard and chair)