from what I remember of that box you had at the fest, there were a whole bunch of kernel messages about /dev/hda timing out
prepare-all is complete.
Got EMC2 and ./configure says it completed.
but get errors with make.
what sort ?
Ah it still things 2.6.9
says 2.6.10 in src/Makefile
you're running 2.6.9, so the configure script will try the 2.6.9 module dirs, I think
should report 2.6.9
This install was from the bdi4.20 disk. I guess I need to go all the way to a new install if I'm running the wrong kernel
4.20 is using a 2.6.10 kernel
Odd. I've got both 9 and 10 in /boot. I'll fix menu.lst and reboot.
and don't forget to check you have the 2.6.10 kernel sources installed & built.
I did the prepare-all
Is that it.
prepare-all is all that is needed in the kernel src dir, yes.
And... configure/make in the emc2 tree should still have run.
Now no initrd named 2.6.10
I've been compiling emc2 HEAD for a 2.4.25 kernel on this box, and I'm running a 2.6.11 kernel.
Comment out the initrd line in your grub menu.lst
03paul_c 07bdi-4 * 10emc2/src/emc/rs274ngc/interp_convert.cc: Moved convert_axis_offsets() call down the order in convert_stop().
paul_c: What do you make of the kde vulnerabilities?
how do you mean "vulnerabilities" ?
They have released new fixes for some sort of problem with security.
two weeks in a row.
looks like pretty minor stuff to me...
I think only the PCX parsing bug is remote
Kommander is low risk, and as cradek says, PCX is remote.
Cups & kpdf would be a higher concern.
But let's be realistic about it..... How many unresolved critical security holes are there in that other OS ?
and how many exploits are doing the rounds that are targeted at *nix ?
I think that's irrelevant. I don't use that other OS and it is not a metric against which anything serious should be measured.
the good thing to see is that the k* bugs were fixed pretty quickly.
At the end of the day, it boils down to risk assesment.
How secure is the operating system, and how good is the firewall.
balanced against the risk of successful attacks.
and how many porn sites do you surf
All right. So I don't have to just run out and get the newest.
oh, you probably should, but probably nobody's going to target KDE.
1. windows/ie 2. firefox 3. sendmail/bind/ssh ...... N. kde
You don't know how many enemies I've acquired over the years.
I need deadband on my mouse.
g92 grumble grumble....
users, grumble, grumble..
Yes, there was a bug in the M02/M30 routine that affected G92 operation.
g92 should be stricken from the interperter
Doesn't everybody use G92?
* SWPadnos ducks
and that is now fixed, certainly in the bdi-4 branch.
* paul_c aims low.
* SWPadnos runs
g92 is a great way to crash a machine
* paul_c tells Mr Bush where to find the missing WMD's
up his ....
with a amd
paul_c: He knows they have to be SOMEWHERE.. He's got the receipts from where his daddy sold them to him. :)
If G92 isn't used correctly, and you make wild assumptions about it, then, yes. G92 can be dangerous.
how's it going Ray?
and after an abort, it is safe to assume the machine will be in an undefined state.
yes it will
Me good. screw g92
I wonder if Tom still reads the emc-users list?
All the posts did some good though. Weeded out a bunch of list members who really didn't want to be there.
so I get the idea that some people think there's a company that is programming this stuff
and they should get whatever they want
Open source does not mean free.
catch you all tomorrow.
better late than ... really late.
rayh I got an email from Hal May the other day
he was still having problems with his rotary axis
I helped him get the computer set up and talked him through a bit of it
but I don't think I can do a whole lot more for him
said he wasn't getting predictable moves out of the A axis
did he ever get in touch with you?
A while back he did. I pointed him toward the Sherline inch ini on his install.
That one uses some better numbers for rotary than he had.
how's the continuous lookahead in the BDI 4.20?
continuous speed/feedrate in BDI 4.20?
* A-L-P-H-A feels ignored.
I'm just setting up 4.20 now.
Was running 4.18 on the sherline. worked fine there.
But that is not real high speed
well, my system is only like 85ipm.
or something like that.
Les was saying there were problems with 60 +
oh... well, my feedrate is only 25ipm - 30ipm. :) so I should still be okay
I'm not doing any highspeed machining. :)
But Dave ran the same spiral program on his Mazak up to 100 with only loss of feedrate during the final small moves.
that's the interesting thing to me - the control knew it couldn't go at the programmed rate, so it slowed down
(either that or it just couldn't keep up, and the cycle time slowed :) )
I was wondering if it was an accel issue or starvation or what.
jmk mentioned a test to see if it's mechanical or software
just run the output from the motion planner directly into halscope - no hardware at all
sorry - the servo controller
Anyone got experience with emc2 in a recent build?
then you see if it's the software screwing things up, or if it has to do with PID / machine parameters
not me - I suspect Paul, Alex, and John are the only ones :)
I have an emc2 splash and tkemc. There is a dog.
dyslexics - Untie!
what's the dog look like?
Hey I'm outa estop.
Old guy, really old guy.
I love it when e-stop turns off
(using a USC card and BridgeportIO)
I was pretty impressed when the same combo pulled the estop relay at Matt's.
cool - that was a trick :)
It was your code that did it.
Jon loaded a live rc46 and his special stuff.
ah - the commits I did for Paul when he was at Les' place?
Must be. It was a copy of sf from about a week after devFest
OK - that's what it was.
my only contribution was the rescan speedup suggested by Fred.
To this old, late night mind, it all runs together.
I've got halcmd reporting things like pins and such.
yes indeed. I should run as well - it's even later here
I hope to try EMC2 one day soon - but I hope more that I can pull off this camera project (and thus have income this year :) )
jmk and I were working on a way to use tickle to show and set halcmd stuff.
You up with microsoldering?
you could say :)
Can you say eyestrain!
I'm quite familiar with it
I've been probing a connector with 0.010 inch pitch contacts, under a microscope
I had to make special probes with 0.003 shim stock and thin sticky plastic sheet for insulation
Now to write the first tickle screens for halcmd.
well, good luck with that - I'm off for the night.
SWPadnos is now known as SWP_Away
asdfqwega: I was out of Crivens, I substituted plastic peanuts. :)
anyone still awake?!
* Phydbleep is digging for his BDI-4.20 disk.
I'm trying to decide the length of my Z axis
I WANT 18", but I'm not sure if that'll casue issues
The bearing blocks I'm using are 6" long, so that'll leave about 12" of travel.
I've gotten everything else cut today, except the Z axis stuff.
Alright, be back in the mornin... G'Night Folks!
Does Debian read the ide contrllers in reverse order or something?
* mshaver is away: mshaver
* Phydbleep cant get 4.18 or 4.20 to work with an msi 741 board.
Morning anonimasu :)
* Phydbleep has been fighting with bdi and an msi 731 board.
4.18 and 4.20 both barf when it comes time to partition the drive. :\
I hope the QT book lands today
so I can begin to play with my gui..
But, both 98 and XP are doing fine.
* Phydbleep hands anonimasu a paper-towel.
* Phydbleep is waiting for the package that ups is holding hostage.
heh damn them
Tom sent it ups before I sent him one usps.. He'll get his today or Fri and they say mine will be here tuesday.
And for <$1 more he could have sent it usps and I'd have it by Fri.. :)
perhaps I should call the customer from yesterday and ask how it'd go
anonimasu: Was it you what got the turret-spindle mill cube?
no, it got sold.
Bummer..What was the final price on it?
same as I said..
* Phydbleep goes "*WAHHHH!*" :(
That would have looked so good through the window of my shop as it sank out of sight in the backyard. :)
Note to self: Even in a desert, if you want to keep it afloat, put it on pontoons.
I would have loved it
[08:58:58] <anonimasu> http://www.gnosjomaskin.se/detail.asp?Id=1817
but well, I could buy it instead of a car ;)
Buy it and some billets and make a motorcycle. :)
Road taxes will be cheaper. :)
Morning alex_joni :)
Hm, seems like the brand new matsuura high speed machines have rapids 50m/min
for the fastest one.
all other models have 25m/min
and the ones with linear motors, 90m/min
* Phydbleep would be happy with 1m/min if it would fit in the shop and handle .3m x .6m with a .2m Z
hm, 1m/min isnt that fast :)
anonimasu: I can be very patient if it means I can walk away from a high-pitched cutter whine. :)
* Phydbleep hates noisy tools..
Noise is inefficient. :)
Phydbleep: I hope you know that walking away can be expensive..
anonimasu: That's why I'm going to wire all the limits to the e-stop. :)
oh, they wont trip if you try to mill your vise off
Nope.. I'll mill the pin off the fixture and have to hammer out the stub and hammer in a new one. :)
* Phydbleep has a whole box of big roll-pins just for that. :)
anonimasu: Ever seen a "Tesla Valvular Conduit" ?
I see how you are.. Join the channel and wander off for a coffee. ..
* alex_joni ?
had a hard time waking up this morning
alex_joni: We said "Hi" 10 minutes ago when you joined. :)
I cant seem to find any good place to buy a broken cnc mill
Phydbleep: I know
but.. I was busy ;)
was doing my 30 minutes of work for today
brb fetching food
anonimasu: I know where there's a 10,000 kilo turret mill... "Acropolis" is the name on it.
I ordered food
I could keep it at work, work wouldnt sink.
anonimasu: ~3m cube
not a problem :)
but the shipping %"!�
anonimasu: You could rent a shop here and I could maintain the machine for you. :)
Phydbleep: yeah, just need somthing to manufacture..
somthing "TOO" heavy for the other shops ;)
make smaller mills with it
Forget it... This is a small one for the shops around here.
I've seen some nice big mills myself
not that big.. but still about 10-15m high
I saw some in finland.. on a tour.. to the factory where they make forest machines
they were neat
The shop I go to for lathe parts has a hole in the roof with a vent duct sticking up for clearance for a mil.
brb, fetching my junk food
16 foot ceiling in there and the duct goes up another 12? 14?
They added a new door in the wall they open to get the full travel on the X.
The door is 12 feet from the machine when the table is homed.
don't you guys got decent measurement units?
Just one if you keep that up. <Jk> :)
Finger that is. :)
what's your unit for microns?
for measuring small things that is
Well.. Lets see.. A cubic millimeter is a "gnats ass" so a micro must be "fsck all". :)
alex_joni: Depends on what I'm doing/working on.
alex_joni: welcome to the world of american precision
* Phydbleep goes to nano-meters or angstroms for true precision.
you'd better go to nano-feet
I found a 10" x 20" slab of polished natural granite for a flat spot.. it looks nice and flat if you use it like a mirror and examine a piece of graph paper.
I fotgot to check the mail
there's a shop that does 0.001 work over here..
or even with more precision
edm shop :)
* anonimasu digs in a book about 3d graphics
trying to find somthing about splines.
nothing in there :/
was digging for algoritms
[10:17:18] <Phydbleep> http://www.genlogic.com/rel27.html
[10:18:53] <Phydbleep> http://www.lighthouse3d.com/opengl/maths/index.php?catmullrom
[10:21:06] <Phydbleep> http://www.crbond.com/graph_app.htm
anonimasu: You want the math? http://news.povray.org/povray.binaries.tutorials/attachment/%3CXns91B880592482seed7@povray.org%3E/Splines.bas.txt
That ought to make your eyes bleed. :)
Phydbleep: I am more curious how you blend a couple of splines togther
that catmullrom site was good
[10:29:17] <Phydbleep> http://www.cs.wpi.edu/~matt/courses/cs563/talks/curves.html
De Nada.. :)
fenn_afk is now known as fenn
what does a little hat sitting on top of a letter mean in mathematics? i'm looking at http://graphics.cs.ucdavis.edu/CAGDNotes/Catmull-Rom-Spline/Catmull-Rom-Spline.html
right after "We define the curve on each segment"
fenn: segment between the two points
thanks, i eventually got it
that page is pretty relevant to what anon wants to do
well it's a spline all right
but you don't want to do that on the whole trajectory
only at the blending parts
trying to replicate cradek's following error prob...
No joy with tkemc
* paul_c suspects PERIOD is too low, and/or max_accel to high.
fenn is now known as fenn_afk
paul_c: good morning
good morning steve
paul_c: got any idea what a sensible accel is?
.25g or so
hm ok, what does that end up like in the config?
I keep mine set way too high so I get ferrors at max speed..
depend on units..but I caution that at .25g the current program will tear a machine apart
for the current program prob 5-15 in/sec^2
that is about 1/40 g
what speed is it maybe I shouldnt keep it at 450
emc should be able to cope with that accel right=
well you are using freqmod right?
what's a sensible default accel then?
half or 2/3 of it?
the figure may need to be lower to avoid slipping steps
now the ferror is just for the virtual servo
or at the servos?
too large accel and the steppers will slip
they wont slip
if that is not happening don't worry about it
that sounds reasonalbe..
what are your Ferror values in the .ini?
and 1 min ferror..
otherwise it did error on rapids..
I don't run freqmod but it would seem that you could just set these way high...since they are only virtual
check with others
I had the accel set too high..
that was the trouble with it.
I can probably lower them now
3 more days then I can strip my mill into molecules..
finally replace the gearbox
and the motor with a 2800 rpm one.
and 2-3 hp more
going to gear it 6500/2800
that wont take away too much torque
what do you think of that gearing?
ow burned my tongue..my grits are too hot
on the spindle?
down to 2800 from 6500?
the other way
ok for some plasics and small cutters
depends on depth of cut..
the spindle I have now can I push to 4000 but without torque..
just watch the bearing heat
should be ok
typical 15-20mm bore bearings are speed rated to about 10krpm
tapered roller bearings would not like that though
I don't know what is on there
the bearings are going to be replaced..
so I can pick some that matches my specs..
the trouble is getting the right SFM with smaller dia cutters
did you see that spindle design that guy that wrote your book made?
[14:33:27] <anonimasu> http://pergatory.mit.edu/perg/awards/TURBOTOO.html
well...busy day starting new projects and such
I will continue to write down TP maths as time permits
But I have one question for others....
emc velocity adaptation obviously breaks down at some point
higher servo rates do give higher performance.
question: of the user/RT comms that limit servo rate....
are any not needed?
or can data structure br rearanged for more efficient (fewer) comms?
The data structure could be reduced to just enough for end points & velocity
and leaving setup data and non-rt commands to be sent via an alternative mechanism.
are time effects of comms cumulative? i.e. more comms =slower?
well you cant send more data at a time then "this"
yes I would think
so you have to reduce the ammount of data until you always can submit your data fast enough..
it's probably cumulative
paul_c: feel free to correct me :)
Well anyway that's the thought....
at SOME arbitrarily high servo rate performance would be "acceptable" if arc blending were fixed-right?
and if input data were prefiltered to stay out of the velociy adaptation brekdown point
anyway I will leave you with that....off to get some other work done
do any of you use the g54 offset like has been discussed recently on the list?
it's been a while since I've had a machine working, but yes
I use all offsets , several tool length comp, and MDI g92 on the turkey call production code
All works fine
ever seen the problem I described last night in email?
let me read it
it seems crazy but it's really there
I use numerous g10 in the turkey code and I can say for sure that all G0 moves are fine
I predicted you would say that.
I am using a 1+ year old version
mine's somewhat old too. I don't think yours is significantly different...
cradek: A couple of things you could try...
and/or reduce the max accel.
I get ferrors when I run accel high
I've run many hours with this setup - it only breaks when I have that offset programmed
unique to freqmod I guess
well off to do some work...new project starting today
paul_c: I'll try those things - I'll also turn on debugging and see if I can figure out what it thinks is going on.
* anonimasu yawns
I got more parts to make for work.. tonight
DEBUG=0x00000010 will get you the TASK commands as they get sent out
I got my QT book now :9
Working hard today?
got two days vacation
I don't think that's right.
but Ben is sick, so I'm hanging here with him today
Ah. How's the shop coming?
he got the flu this week and missed 3 days school
good. I've been turning things on the lathe for a couple weeks
about to get the mill set up I think
did you get on IRC the other day with a Mac?
Yes. It was a dual g5 that my kid Joe got hold of.
was it pretty fast?
Had a bitch of a time finding an irc client.
I use X-chatAqua
Awesome speed. Renders stuff beautifully.
I'll let him know about x-chat.
I just added a 120 Gig hard drive to this little cube last weekend
Now clue what we wound up using.
something off the wall.
Now you have several weeks worth of storage.
I still love using this machine. Silent, small, does what I need for surfing,listening to music
yep. Ben wants to put a bunch of his music on it, so now he can
Joe is quite a mac user. Airport, laptops, ipods...
whatcha working on today?
writing a tcl front end to halcmd.
Compiled emc2 last evening with bdi 4.20 and started looking around.
are there many people using emc2 on actual machines?
Not tried it with a machine yet.
Several prototype machines. Weyland just put a page on the wiki with his.
ray lies (where?)
jymm did you find the bearings you were looking for?
I found it by clicking recent changes.
rayh oh that. nm
dan_falck : on that page you gave me, yes. But I'll just use what I have and replace later
rayh I thought you meant like today or yesterday =)
jymm what kind of bearings do you have now?
The link is under case studies.
The most recent change to the machine was running emc2
I did a lot of work with angular contact bearings over the last month. Radial could be a problem
rayh lol kinda funny that to find somethign you have to look at last changes instead of a TOC =)
TOC would be too easy
dan_falck really mucks em up huh?
the thrust on radial bearings?
Jymm:if you want to set your preload high and the bearings aren't right- it feels like gravel
if you have angular contact bearings and they aren't set up right-same thing
we were working on some Setco slides that came in stock, and felt terrible
Well, the mounts will be replacable, so I could use anything later on.
if you have problems with slop (preload not related to leadscrew/ballscrew) you'll know were to look
I have all parts except screw/bolts right now. I want this up and at least running by EOM.
router table right?
Come June, I can get the AC bearings
those surplus bearings at $4 each would be good
even though they are "surplus" they might be new in box
I'm still trying to decide the length of Z, I want 18" but I'm not sure if that'll cause issues. The slides are 6" so that leaves me with 12" of travel.
what kind of parts do you want to make?
surplus is fine, it's RFE/used that I have a problem with.
they might be ok. $4 is cheap
I will order some and let you know what they are
oh, ok. cool.
I need two sets for my mill/drill
1/2" ID, you remember the OD ?
1.875? I can't remember for sure
I checked a couple weeks ago and they would fit the end plates on the mill
Ok, well I haven't cut the leadscrew bearing mounts yet, I make em 3 or 4" then.
Anyone have any thoughts on a 18" Z axis length?
with a 12" travel
what kind of parts are you wanting to make?
If you want to do inlay work on flat pieces, you might only need 4" of travel.
well the extra travel is really to be able to engrave on 6" high jewlery boxes
you might not need all the travel, if you could just bolt the slide up higher, just to clear the part
are you doing a gantry with the table moving in one axis and a fixed bridge over top>
moving gantry, not table
I've got a CNC router like that. I can bolt the Z axis in different positions to accomidate different part heights.
and iwas trying to avoid remounting the Z axis if the extra height was needed.
usually that means re aligning too
I did a lot of inlay work a couple years ago. Usually the mother of pearl was only .045" thick.
so Z movement wasn't a big deal
I didn't worry about alignment much either. It was mostly one off stuff. I set home at the time that I turned on machine.
Murphy's law says that if I make the length 12" giving me 6" of travel, the 2nd job I'll do will need 12.125" length.
Jymmm: that applies to everything..
be it 5 cm or 5meter
anonimasu shush you! =)
Im sorry :D
so, you think 18" length on the Z would be too much?
18" woul be okay if you can do it
I just wanted to give you some options, if it's a problem
sometimes it hard to make the Z axis compact enough
for the Y or X to push around, without stalling
depends on everything else as usual :)
I understand. I just hate limitations. But I dont want to fsck myself either.
heh, I hate limitations too.
but my budget does like them.
I was gonna make x and Z bearing mounting blocks 4", then I decided to make them 6" instead. I have the extra berings, and it'll had a LOT more support.
in what material?
I need to re-machine mine
they are good, but I dont want tem in iron, I want them in aly and with a support bearing for the gear..
I was gonna make the the bas of the Y axis only 3/4" thick, but decided to double that to 1.5" as well.
err the outside gear..
work a bit..
by the time I get done, this sucker is gonna weight 150 pounds!
thats w/o motors/router too =)
alright, bbl. gots more cutting to do today + LOTS of drilling!!! =)
dan_falck let me know how those AC bearings are =)
paul_c: I sent another message with the debug output. It's very apparently wrong (a bogus EMC_TRAJ_SET_VELOCITY)
That will help to trac the problem down.
I would sure like to know if anybody else can reproduce it. It's really easy.
Can you email me both your ini and var files please.
oops, I meant to attach the ini
OK... You have different max velocities for each axis
so the velocity along a vector can be higher than any one axs
You need to reduce the velocity in [TRAJ] to approx 0.167092
but that's true even if they're all the same
....correction (dropped a digit)... 0.8167092
that'll screw me when I hook up the rotary axis again.
The max velocity should still be clamped in the RT code....
but I bet you have figured out the bug. It picks the velocity appropriate for a diagonal move (0,0)-(1,1) when in fact it's only moving (0,0)-(0,1)
lemme run some tests after tea.
certainly, something is a little screwy there.
thanks, I'll look for the bug too.
dan_falck : Heh, cool.
Is the about of force needed to lift an object the same when lifted using a leadscrew?
Is the amount of force needed to lift an object the same when lifted using a leadscrew?
vertically that is
'nother new face...
I was reading Waylands Wiki page at lunch this afternoon and I am having the same problem he had with my emc not handling the backlash compensation. Any thoughts?
jmk is working on a solution
(didn't know Wayland had a wiki page)
I thought that was the name! Any way he said he upgraded to EMC2 and problem solved. IS there a detailed description availaible to help a new to linux user do this?
only what is on http://wiki.linuxcnc.org
Ok I'll take a look and see if I can find it. THANKS
paul_c : http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Weyland's
paul_c did you see my Q above?
Could you repeat the Q
paul_c : Bascially, I'm wanting to know what if any conpensation I'll need for the Z axis. 269 inoz steppers.
backlash comp ?
paul_c : Actually I just dont wnat to overwork the motors
direct drive, 1/2"-10 ACME
Jymmm: you might want to ask machine questions on one of the yahoo groups forums
down motion isn't a problem (gravity), it's the UP/HOLD I'm worried about
[18:48:02] <dan_falck> http://groups.yahoo.com/group/CAD_CAM_EDM_DRO/
[18:48:22] <dan_falck> http://groups.yahoo.com/group/DIY-CNC/
* paul_c is looking at cradek's problem....
and thinking, much of emccanon.cc is 'riggin daft
the problem is GET_EXTERNAL_POSITION
it has several issues
it nukes canonEndPoint (switches it to relative coords, not absolute)
it subtracts currentToolLengthOffset twice before returning position
I know how to fix it, but I'm trying to figure out the *right* fix
got a function() name where that happens ?
(don't ask me why it's uppercase)
blame messrs Proctor & Shackleford
hm, I should turn on some debugging and see if I can find why emc2 stops responding..
or well, why it refuses to switch modes..
paul_c: testing the fix now...
EMC_TRAJ_SET_ORIGIN should be setting the canon end point....
paul_c: bug is fixed in my tree
The whole method of handling unit conversion in emccanon.cc sucks...
should be using one unit system frm the bottom up.
and get rid of those legacy units that no one uses in the real world.
yeah like "
millimetres & degrees from the bottom up would solve a lot of troubles with unit conversion at the GUI level
Yes they would.
Anything other is a display thing.
How would the interp handle this?
if (mode == G20) convert_to_metric();
That would easily work for me.
Any problems for the step interp?
03cradek 07cradek_stable * 10emc/src/emctask/emccanon.cc: Fixes wrong max velocity when G54 offsets are in effect
Any coordinates fed in could be converted to the base units at the very first stage.
That would work for any interp that we have control of
Or are you saying at the task level.
Do it at the interp level.
evrything has to go through the interp before it gets to task.
I wonder if Tom would be up for that?
03cradek 07cradek_stable * 10emc/src/emcio/ (Makefile extbridgeportio.c):
Support emulated IO for some bits while the other bits are real -
this lets you specify the same port for rt and non-rt IO. Just
don't touch the lower 8 bits in non-rt.
To specify that a bit is to be emulated, negate its number.
Kramer or Hubin ?
Mandating everything uses millimetres at task and below might meet some resistance....
03cradek 07cradek_stable * 10emc/src/emcmot/ (emcmot.c emcsegmot.c): it's my branch, and I have no switches.
and angles need to be in Radians as possible really - Those are the native units for C/C++
from folk like josh hall that wants angstroms
hm, why do they need angstroms?
one angstrom = 1.0 *10^-10 metres
so they are in metric anyway...
I dont get why they need them though
I saw an actuator that could move 1 angstrom
the trick was watching it and proving that it moved.
Many of the pizzo crystal things will move in fairly small increments.
hi all I was trying to read a learn a little about RTAI
but I got sleepy.
but I do have questions
does emc use FIFOs, mbuff or poorman?
just trying to learn.
Gotta go fix a sawmill back later
Doesn't dynamically allocating shared memory take a lot of time?
It can do, but it is only done once at module load time.
Back to spray guns. thanks
Now what's up with jmk I wonder....
What do most people use for generating gcode to machine fairly simple parts?
a text editor & keyboard
I was hoping for something a little more automated. ;)
Some of the stuff I've got, I used some simple C or BASIC programs to generate the files.
Yeah... that's the approach I've been using.
I was hoping to be able to draw simple stuff in a cad program and get it to generate mill paths to fill in outlines.
I guess... not sure what the right term for it is.
yeah, that's pocketing
you need a cam program.
but well, you can pocket the outlines of a contour also..
if you have a specific part or something I can do a cam run for you.
from a dxf
Don't have anything specific.
What would you do your cam run in?
it has a nice pocketing engine...uses lots of G2/G3
[20:35:40] <les> http://members.aol.com/m9685123/M1.htm
Yeah... looking at it.
free functional demo!
That mills the letters A, B and C and some symboles.
or any other shape with islands
umm... the demo does?
oh I was reading..it changes some
might still do dxf shapes
Compiled and ran qcad earlier... looks pretty good.
Doesn't look like it can do anything like generating mill paths though.
here is an example of some of my millwrite parts:
[20:40:11] <les> http://www.lmwatts.com/signwp.html
not all are millwrite though
daryl: Are you using BDI-4.xx ?
QCad is on the CD
Heh... that would have been simpler. :)
Anything else interesting on there?
and there is a Debian package for it ...
Synergy for 3D cad/cam
Either of those make gcode?
or something easily converted to it...
You'd want QCam rather than QCad
and/or a key to use the Synergy package.
Do you have a link for qcam?
Is qcam a linux program?
[20:47:08] <paul_c> http://www.ribbonsoft.com/camexpert.html
Hrmmm.... It would appear that hp2xx will spit out g-code
[20:49:52] <paul_c> http://personal.inet.fi/koti/picnet/cnc/hp2xx.html
I fscking detest UPS.. "Pay us more or we'll store your package for a week."
Or drag it 1600km the wrong direction in the name of 'efficiancy'.
Phydbleep: Sorry - I use UPS mostly due to ease-of-use
asdfqwega: S'okay. :)
* Phydbleep has detested UPS for years. :)
I don't detest UPS - but like many things, you gotta keep your eye on them
It's not like I can actually do anything till I get bdi to install on this machine. :\
Now that I think about it, every package I've ever shipped out west (except those to California), people have asked me to go Fedex or USPS
asdfqwega: The tape was free and I wanted to make sure it stayed shut. :)
asdfqwega: Yeah, If it has to come past Chicago USPS is the way to go.
UPS to Calif goes airfreight so it misses the 3 day layover.
* anonimasu is making parts again
wb rayh :)
How you doing today?
Took the time to read the emc page and realized who you, dan_falck and paul_c are. :)
wohaa... Nothing to do with me....
It all vicious romours
paul_c: Yep, and they keep biting you in the ass. :)
/mode paul_c guru++
Phydbleep: I think you mean "arse" don't you ?
Phydbleep: Twin alien invaders from the planet Mars?
Ass, Arse, Posterior, The back of your front..
See, Paul, you need to disguise yourself so nobody knows you're a guru (among other things)
Just use some reverse psychology:
Take some wire and hook it into your beard and over your ears
that way, everyone thinks you're wearing a fake beard
That doesn't work...you still have the accent
hm I ended up doing somthing strange in my program..
i went to g-50 x-25 or somthing similiar
then I milled the mount into the box..
and when I did the last move before my home rapid, the mill went 90% of the move and then skipped the last part of my g1 y-50
leaving a tiny ridge..
very very strange..
the mill did the same both times..
oh well, dosent matter really..
but it's weird.. y-50 and x-25 should always be the same
* Phydbleep tries turning UDMA mode off in the bios to see if that works better.
I get to the partition the drive bit and it dies. :\
Say's the partition table is invalid, do I want to init the drive, yada yada.. And then it hangs. :\
* Phydbleep switched to vc4 and looked for errors.
hda: lost interrupt < that would do it. :(
I hate the guys that made the T-slot blocks I use for my clamps..
they didnt tap the holes through..
so I cant use them as a fixture,,