Aha! Got 3 Minarik PCM4 signal isolators coming off Ebay !...That solves tht problem!
G night all
Jacky^game is now known as Jacky^afk
djb_rh: you can say that again
ooh my fc5t1 torrent is done
dare I install it and ruin my perfectly functioning laptop?
jepler Got Ghost?
Jymmm: is that some commercial software? Uh, no, I don't have Ghost
actually I have a separate partition where I'd install fc5t1 anyway
unless my finger slips during installation
speaking of fc4 or fc5, is there a real-time kernel that will compile with gcc4? I seem to remember big warnings the last time I looked saying I had to use gcc 2.95 or something like that...
do you really mean gcc4?
that's the only packaged version of gcc for fc4
I'm a little bit surprised there's not a gcc 3.3 packaged in extras or something
oh, here it is. compat-gcc-3.3
so is there a RT kernel that will build with gcc33 or gcc4?
well I apparently built my 2.4.20 kernel with gcc3.2.2
probably never thought twice about it
well maybe I just dreamed the bit about gcc 2.95, or confused it with something else, or it was a piece of folklore someone told me
jepler: Turkey Hallucinations
nah, couldn't be.
oh I know... Cranberry Hallucinations!
maybe it would just work. https://mail.rtai.org/pipermail/rtai/2005-March/010787.html
this may be where I got the impression that gcc 2.95 was required: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl/emcinfo.pl?BDI-4_Compile_EMC2
sometimes a problem in the past makes wariness hang on for much longer than necessary
in some cases it can even turn into perpetual superstition
mmm, my pies turned out great
or I guess in this case.... cool pies =)
cradek did cutting the SP help?
03paul_c * 10emc2-auto/wiki/ (9 files in 7 dirs): "Auto update wiki from a cron job. Fri Nov 25 05:30:01 GMT 2005 "
Hello, is anyone using emc2?
martin is now known as chinamill
I just installed emc2 - ran the machine around a bit
No one is out shopping? - I forget some of you are not in the US.
This is the big day after thanksgiving shopping frenzy
I try to stay far far away
With emc2 hal and rewrite - does a faster computer perform faster motion? I am playing on a 700mhz amd and get following errors at around 20 ipm for one of my axises - would a faster proccessor help?
check PERIOD in the ini file too
choose an appropriate valuce for your cpu
This is emc 2 - there are 3 periods if I remember right - not like emc1
I know only one
check the max feedrate for any axis too
Hi Skunkworks...for steppers a faster computer definately helps...I had a 100 MHz PC on my small stepper mill initially, even with a PCI parallel card
you can find it tryng and retryng ..
And switching to a 700 MHz pc made a world of difference in the max speed ... Its been a while but I would say a factor of 2 ....
The 700 MHz PC is now on the servo system, and never tried the 100 MHz PC on the servo system, so cannot comment there....
I run 10 microsteps on the stepper system....
wb9mjn: wich drivers ?
It has Gecko 201 A s...which are excellent...never miss a step...
wb9mjn: really good drives ;)
Ive to try my G340 yet :(
get no motors and encoders yet
For servo's I went with the MotenC-100 and Gallil amplifiers - brushed servos...
Like the idea of having the PID loop going throug the computer....
seems nice, ive not much knowledge about PID
i eas thinking the PID is already did by the driver
thanks - there are 3 periods in emc2. base_period, servo_period and traj_period
skunkworks: base_period is the one you want to mess with
The machine I buil the stepper system out of was really cheap...I did it for the learning experience...Initially I tried to use some locally made full step
traj_period is not really used
and servo_period should be an integer multiple of base
I think I used defauls values with my 800 mhz celeron
drivers, and had all sorts of missed step problems, and stalling, even with the motors on the bench....
the olny thing, i found optimal value for axis speed, and it work pretty good
wb9mjn: where you from ?
I m near Chicago...
skunkworks: no problem
oh .. too far
wb9mjn: you on the map yet?
Yep....I m on the Map -- Don L ...
i was looking for cheap used motors
im not able to find any here around..
Sorry, no spares I cannot risk not having available here...
Its like the 4 th largest city in Illinois now...was only 26K people when we moved here....
Its the city inthe US with the most number of old houses being torn down, and replaced with "McMansions" ....
alex - what does that mean interger multiple? if I divide the sever by the base number I should get an interager?
servo by the base period - sorrry
skunkworks: if you have .00020 then 0.00100 or .00080 or 0.0040
you get the point
makes sense now. I read it in the ini file but didn't click until now.
I have not been back to the shop to check the feed issue with emc2 - (seems to be using the max velocity instead of the individual axis max velocity) and pull the ini file off of the box.
wb9mjn - what kind of feeds are you getting out of your system?
Been a while...Let me turn it on and see....probalby about 60 in/min with the 100 MHz system, and 120 in/min with the 700 MHz system...
what is your input scale? for your axises?
The little Proxon MF70A CNC only has XYZ of 128, 46 by 57 ....
Input and output scales are all 2000 ... Its a 1 mm pitch lead screw on all axis...with 2000 microsteps per revolution...
The XY and Z are travels are in mm ...
sounds about right - that is similar to what I am getting. My input scale is around 2540 and can get around 100ipm
my z has an input scale of 20000 and i have trouble getting 25 ipm
emc2 seems to put out a smoother pulse train than emc. from the little I have played with it
Looking at my old ini file for the 700 MHz stepper system, it has max speed set to 4.0 (240) on the three axis, and the default set to 3.0 (180)...
The 100 MHz system is still booting, hi..!
But unless all the ways and everything were perfect, I remember I would get stalling.....After some machining, and some fine swarf from the 15K RPM spindle,
the ways would start to get gunked up....and steppers start to stall....
With everything clean, and lubricated, it could do that....except possibly the Z axis which has a nut that is badly worn, and is lifting allot more
weight, due to a replacement of the spindle motor with something bigger...
Ok, have it on now...Its doing 60 in/min on the 100 MHz system, with the PCI parallel card....
I was getting almost 200 ipm with emc put that was with the period set to .000016
100 in/min is set as the max ...
I have not played with the period with emc2 yet
In the 700 MHz config, I had period set to .000020 ..
The 100 MHz pc mini.ini has it set to .00003 , and the OS system is quite sluggish ...
Its one of the original EMC BDI disks sent out by Ray , most likely...I rebuilt the 100 MHz stepper system like 6 mo ago, trying the various BDI CD's
I have, and probably got the latest one that would load in the old CD ROM it has...
I tried .00002 on the emc2 box and the system stalled ;) set it back to the default .00005
I had screwed it up trying to a newer version of Red Hat on it, which the CD just would not work right with it...
Yes, its EMC1 ....
The CD rom is so old, it would not load any newer EMC bdi's, let alone EMC2 ...
some of the older cdroms will not see burned cds
you could try burning the cd at the slowed speed your buner will do - 2X or 4X - it might help
So, apparently the 700 MHz system sped up the stepper mill by a factor of 3 ...
is the 700 a amd or intel?
Does not help, as the width of the bits is smaller, I believe, and fixed....
Its an Intel Pentium ...
Its too slow for Linux anyway, and only bareable as an every day computer with W95 ....
Its OK for a EMC computer, if that s all that is going to be on it...
The default in the 700 MHz ini is 3.0 (180 in/min) , versus 1.0 (60 in/min) ...
yes - The computers I am playing with are ones that are just too slow to do anything usefull (other than machine control)
Actually..I wrong...That should be 180 mm/min, versus 60 mm/min ...
Yes, it takes about 2 min to traverse the X axis !
In the servo mill, which is set in inches as its native dimension, it does 360 inches/min peak ....
But, its ball screws and servos, and linear beary slides ....
running my 6 foot gantry - it takes about 25 seconds to run from one end to the other ;) (200 ipm emc1)
used to be a film step and repeat machine. Just had to raise it to add a z
what are you using for a servo driver board? and servos
playing with emc to maybe use it to retrofit one of our bigger mills with servos
MotenC-100, with clifton precision motors ....clifton precision is now owned by another company...
right now it has hydrolic servos (eww)
is the 100 the 4 axis board?
I got one of the original ones, which is 8 axis I believe...He now sells a 4 axis one too...
Using Gallil PWM servo amps, which work well ...
ah. it looks like a nice board - heard nothing but good things about them so far
There were some initial issues....Mostly if you do not get your PID close enough to something that works, it can do wierd things...But once you get
the PID tuned up so that the axis stay in control the wierdness goes away ...
now - is there any feed back from the motors to the drives? or is it only the posision loop though emc
emc > drive > motor > encoder > emc?
The loop is throug EMC ... On the X and Y I have 5000 line encoders, and 250 on the Z...Once I am cutting metal, the Z axis will be upgraded to a 5000 line
encoder as well ...
The encoders are on the X and Y screws, and the Z motor ...
The encoders feed into the MotenC-100 board, read by EMC, computed, and an error voltage comes out (+/- 10 V) out the MotenC-100 which then
that is neet. Our big old mill with hydrolic servos had an inner and outer feed back loop. outer posisioning andd an inner velocity iirc
is fed to the Gallil amp, and to the motors, through a shaft coupler, back to the screw...
simple and sweet - I like that
how long did it take you to tune the pid? did you use any of the utilities that has been written for emc?
Besides the PID values, there are 3 feed forward factors position, velocity and acceleration....One needs to tune up at least 5 factors per axis to get
it right ...
I would say 2 weeks total...When I get the heavy machine table fastened down to the motion table, I will probably have to do the X and Y again...
Which should take a day ...
Yes...I used the plotting capabilities in EMC ... to plot following error ....
thanks - I read about it but it is nice to ask questions to someone that has actually done it
PID is actually somewhat sloppy to tune...Its the VelFF and AccFF that fine tune in the following error, and allow one to move the P up so that you get
stiff control action ...
don't have much experiance with that. The time will come though.
Moving the P up without the two FF's adjusted will result in oscillation ...
But initially, the P needs to be kept down...Kinda scary to have a 40 pound milling head jumping up and down, hi...
I have heard stories of bridgports rocking back and forth on its base while x was oscolating
Yep...best to bolt it down...My D+M 6 is somewhat smaller than a Bridgeport, and it was scary enough for me ...
Might even be prudent on a Bridgeport to put in some out-riggers to avoid tipping, our yanking out the mounting bolts...
scary - I have to do some running - It was nice talking to you. will be back later
who asked me how economy is going here ?
may les_w ??
[16:02:45] <Jacky^> http://www.economist.com/surveys/displaystory.cfm?story_id=5164061
Goodbay 'sweet life' ..
Nov 24th 2005
alex_joni_away is now known as alex_joni
indeed it works
when I connect, the window starts freaking out, like it's scrolling thru many pages of stuff, over and over
once I send something (like "test") it stops
* jepler *heart* irssi
last time I didn't send anything, and after about 30 secs, ksirc crashed
* Jacky^ agreed with jepler
jmkasunich: all ok?
seems ok now
great. and yourself?
same as always, so much to do, so little time
trying to focus on the ppmc driver, get the damn thing done so I can do other things
* alex_joni is chillin for a while
playing some notpron ;)
was looking at the trackers on wednesday, looking for easy ones to close out and improve our "average age"
you fixed a lot a couple weeks ago, my turn now ;-)
* alex_joni closed quite a few for the same reason :)
was unsure about a few of them
of the remaining ones I mean
one of them was about inconsistent use of '-' and '_' in hal names
I've checked, all pins and params now use '-'
but many (most) functions still use '_'
seen your update on that one
debating with myself whether to just say "that is the standard, functs use '_'" or to convert all to '-'
I kinda used '_' all over too
seems more common for function definitions
you can't define a function called foo-bar
but you can foo_bar ;)
not in C
not in C, not in C++
but you cant define a variable called foo-bar either
not sure about Ada or Fortran though
unlikely in any language
yes, but you don't have HAL variables ;)
you have pins, which don't exist in any language
so that's ok
if you allow math operators, you tend to interpret '-' as "minus", not as a word separator
in LISP it's customary to use - as word separators
alex_joni is now known as aj_65
in list, subtraction is like "( foo bar - )", right?
gimp's procedural database lets languages that can use "-" and "_" interchangably (scheme) and others use "_" (python)
with whitespace around the '-'
(- 3 2) => 1
(- foo bar) -> foo minus bar
(let ((a 3) (b 2)) (- a b)) => 1
(- foo-bar barf) => foo-bar minus barf
it must distinguish between - by itself and between words
anyway, we're trying to decide on a convention to use for hal
should components/drivers/etc export names with -, or _? for pins and sigs the convention is -
mostly because its easier to type, when writing a hal file, or useing halcmd from the command line
english and LISP speakers would tend to prefer to type -
same argument exists for functs
C and python speakers would tend to prefer to type _
I don't see a compelling reason to use _.
I prefer typing _ for functions
keeps the name the same one as the real c function
I know that's not always the case..
but it would be nice if, like gimp, I could write foo_bar to refer to the signal foo-bar if I'm using a language that doesn't support dashes in identifiers
so you're suggesting to just replace _, - with a HAL internal sep.
so both will work
the internal one can be either - or _, or some other char
or replace _ with - (ie use - as the hal internal sep)
yes, something like that
right, but allow a .hal file to contain _ in the names
even if they actually refer to -
the internal sep will be the one that appears in listings, so it shouldn't be some arbitrary thing
it'd be nice if any self-documenting facility was consistent, not mixing - an _.
the list and show functs will display the actual name as stored internally
I don't know how gimp implements it, but for users it is consistent and convenient
we have two separate issues here
code that translates _ to - might be a nice addition to halcmd for user convenience
but there is also the issue of the names exported by existing modules
ok.. have it all with '-'
and the translation besides that
* aj_65 can do the latter if you want..
do i smell lisp here?
I smell dead fish
note that we can't implemnt the translator until after all modules are fixed
because with the translator in place, anything with _ becomes unusable
ahh.. anything that now is _ is unusable..
modile exports funct "read_all"
ok, got it
aj_65 is now known as aj_66
nope.. getting more advanced on notpron :P
[17:59:22] <aj_66> http://deathball.net/notpron/
too much of a time sink right now
damn nice :)
I agree on that..
but it's addictive...
most time sinks are
anyway, after I get this driver working (or maybe just partly working and committed) I'm gonna go thru the HAL moduels and change funct names from _ to -
or maybe I'll just do it now
* jmkasunich is procrastinating on the driver code
aj_66 is now known as aj_67
aj_67 is now known as aj_68
there is a HAL component called encoder_ratio
so _ is used in the prefix as well, not just in the function name
aj_68 is now known as aj_69
aj_69 is now known as aj_70
03jmkasunich * 10emc2/ (20 files in 4 dirs):
changed all hal components to use '-' instead of a mixture of '-' and '_' in
identifiers. For example, changed pid.0.do_pid_calcs to pid.0.do-pid-calcs.
Also changed all standard .hal configs to use the modified names. Users with
customized .hal configs will need to revise them, the changes are limited to
aj_70 is now known as aj_71
aj_71 is now known as aj_72
ok .. enough for today ..
aj_72 is now known as alex_joni
* alex_joni returned :)
my head hurts
alex_joni, stop smacking it between the door and the doorframe then.
I would've stopped at aj_69
happy turkey day you people
or is it turkey day?
maybe you people
I heard there are veggie turkeys for thanks giving :)
for vegeterians that is..
on the radio they were talking about the new trend... organic turkeys... like green turkeys, no steroids, or anti-biotics.
a freak'n turkey is $70!!!! CDN.
for a 20lb turkey.
around here they are between $0.49 and $1.49 per lb
The farmed ones are cheap
those are probably at the high end of the range
organic turkeys are the fuck'n hippy birdies
you want organic, get a gun and go hunting
jmkasunich, I'm in Canada.
turkeys aren't up here... I think
my brother said that a wild turkey chased his children
in their own yard
I've seen a swan chase people.
and my inlaws hit one with their mini-van while driving on the PA turnpike, wrecked the windshield
we called the swan henry. Henry was a giant swan, that lived on the beach when I was lifeguarding there
I've seen canada geese chase folks too
bah, geese are allllll over the place here.
you got water, you got a lawn, you'll get geese.
so, pool + green grass ==> geese.
+ gun ==> fun
I'd like a silenced bb gun.
geese + green grass = green goose poop everywhere
I would love to mount a rear firing bb gun to my CAR.
fuck'n stupid mofo's tailing me while the ground is slippery from snow.
WTF are they thinking?
morons, first snow falls and they forget how to drive
I got snows on, and I'm doing 60 in a 60... while people are crawling like at 20.
it took me anhour to just drive around people.
I hate snow for the fact other people can't drive.
A-L-P-H-A: I know exactly what you mean
same thing around here ;)
* alex_joni drives a 4x4 usually
so I kinda don't care about the snow..
seriously... good damn retards...
it's a GIANT convoy going down the highway... it's 3 lanes... and stupid mofos are all driving in the middle lane.
they should move south
okay, forget it... jump in the passing lane, zooom... off I go at 100km/h... while they mofos can all peddle at 40
I was like, wth,.... seems like everyone is a like "OMG, it's WHITE!!! what the hell is this stuff... better super doper extra careful... it may attack me back"
A-L-P-H-A: and it's cold, and wet
damn... I think this massage I have booked at 4:30, will put me to sleep
jeez.... I didn't realize that Jon E connected his FPGA pins directly to the outside world
no, just commenting on his design (or lack therof)
right.. I know how it is :(
step and dir go straight from the FPGA to the off-board connectors
what? you wanted to super duper flux capacitor optoisolators?
was helping a guy once to solder a pin
encoder sigs go straight from the connector to the FPGA
and he didn't remove the stuff from voltage
no noise filterine, no buffering, no nuttin
so when the solder was placed on the pin .. bang ;)
bang? how many volts was it?
jmkasunich, wait... serious question, what would you have liked? a hex inverter 7404? or an optisolator?
100V because of bad earthing
I would expect pfft
it closed through ground
not much current though, but enough to spoil a very expensive DSP
alex_joni, how much magic smoke got let out?
soldering iron was live? or ckt was live and iron grounded?
no smoke.. just it never worked again
I've seen magic smoke a few times... I put a chip in backwards to be programmed... I couldn't figure it out for the life of me, why it wasn't working. hahaha
APLHA: encoder ins, some kind of buffer, maybe a schmidt trigger
iron wasn't separated from 220V
and the board was connected to a PC through GND/parport
jmkasunich, 7404 is a schmidt trigger. [I believe]
soldering a live circuit?!
live iron is bad, bad, bad
and the pc was pluged the other way around in the socket
7414 is schmidt, 7404 is plain
we can do that over here :/
ok, who's the numbnuts soldering a live circuit?!
Jymmm: I know.. I asked twice if it's removed from voltage
alex_joni asked who?!
he said yes, and I only checked Vcc
alex_joni, you should have checked it yourself.
APLHA: at work, there would be ESD protection, and probably RC noise filtering, upstream of the schmidt
resistor and cap
one resistor, one cap
series R, cap to gnd
RC, opposed to RLC ;)
alex_joni dont confuse the boy
Jymmm: how so?
diodes to ground and supply, to clamp spikes
I do that right before the chip's VCC, and touch it to ground with a 0.1uF cap...
RLC is RC with an additional inductance
run away, run away!!! rayh is coming.
too late... we've been infected.
* jmkasunich hides under the desk
A-L-P-H-A: yes, but it's safest to have an most pins that go outside
alex_joni if he doesn't know what a RC circuit is, you think he's gonna about LC, RCL circuits?
alex_joni, say again.
You were infected long before today!
not infected, infested
A-L-P-H-A: on connection lines too
not only on Vcc
are these pull up resistors, and pulldown resistors that we're talking about?
borne resistors and such?
outside ---resistor----*---- pin
burne, borne, whatever.
what's the *?
bourne is a brand name?
that's where the cap goes to ground
omg... why am I learning this.... I'm studying to be a forensic accountant.
you're the one who asked
A-L-P-H-A: you need to know this for forensics
thanks... but I really must be moving...
hang around with engineers, and you might get infected
kk... chat with you all later.
he might get awoken
rayh: whats up in the up?
the temp is not up
I bet not... it's pretty damn cold even here right now
jmkasunich: soon it will get up, if you use abs()
sunny tho, it's pretty outside, as long as I am inside
Right. 40's 40 right.
The lake froze over just about the same date it always does.
going skating? ;-)
Not for a few days. It is still creaking and cracking.
Have others had a chance to test the Lerman interpreter modifications?
err.. bon soir
any carbide grinding expereince around?? soir... apre midi... ; )
dmess: excuse moi, mais je suis un peux fatigue..
too busy on the ppmc driver (or avoiding working on the ppmc driver, like now ;-)
vas te couchez
il n'est pas tard
I understand exactly, jmkasunich.
[22:59] <alex_joni> il n'est pas tard
at least I get to practice on my french :P
Is that driver going to be a problem?
no, I just need to get my ass in gear
the docs leave a bit to be desired
sa fait pas de differance... va te couchez si tu est fatigez
I was wondering if he had done something irreversible with the estop.
dmess: je ne sais pas tres fatigue, jus'que un peux
he did things annoying with the estop, but nothing irreversible that I'm aware of
dmess: and feel free to correct my errors ;)
Those boards will be nice for mid sized systems with the high power outs.
you mean the SSRs?
solid state relays
I was hoping that I could test emc2 here with the system before I have to deliver it.
you are working on job that uses his boards?
i got opto's too
Yes. I've got pwm and step versions
oh, then I guess I better get going
Matt's got a pwm though.
planning on using both for the same system?
variations on a theme.
I'm trying to write the driver so that it will go out on the parport/EPP bus and discover what is there, then export HAL pins for the lot
IMO the pwm makes more sense for few more dollars.
so you could mix and match any of his boards
That will be fantastic.
hmm just found a fused board with 10 opto22 ssr ' kickin about
be easier to test if I actually had more than one board
re: the opto-22's on jons board - have you actually installed the SSRs yet?
Ship me stuff and I can backup your tests.
I've got 4 on one board running.
I found that my SSRs obstruct the fuse sockets - either mine are too wide, or his layout is for something narrower
send me an addy ray i'll get the board to ya...
I've got a coffee can full of 120 volt when the time comes.
ray has board(s) I think, he needs the code ;-)
dmess I never refuse hardware though.
im not sure if this whole board works even.. i got it for components only
well, knowing you have boards and a job for them makes this easier
although you probably can't connect both to the same parport at the same time anyway
would need some kind of Y connector
jmkasunich: probably would mess up EPP real bad
not if the stub length is short enough
I mean the timings in the protocol
the boards are intended to work that way, each can have a different EPP addy
ssr' should be activatable as m codes or special g code options
then maybe he might consider another output port from the board
like you get on scanners :)
to connect your printer further down the path..
fat chance, if he isn't even willing to spring for RC filters
Crydom ssr's work fine.
like the commadore 64 used...
Let me look for the other style, rectangular.
he has a 26 pin (2x16) header that might be connected in parallel with the DB25
these are opto22 model oac5p ray
not on the schematic he sent me tho
and I'm too lazy to follow traces
Tiy're right he's got the offset on the wrong side for motorola MP120D2 modules
with 2.5 A at 240v fuses
and you don't have a second board I imagine..
alex: no, only one
but ray has 2
jmkasunich: then it probably won't matter to you ;)
if I had 2 I could test the driver to see if it discovers and exports for both
but jone can test aswell
he should have a lot of the boards :)
btw (OT) did you see the g-rex'es?
was a pic on the gecko list..
what about them?
a pic with a few hundred of them
I have one, not planning to do anything with it right away, but someday....
yeah. mariss manages to cram a lot onto not very much pcb space
and he has a machine to build them
not sure if it will go through..
that is impressive
pictures make all the difference
if you see that you know that it's not a little enterprise doing garage business :)
my boards are from Barderair pneumatic logic controller
31 per panel, 12 panels.... 372 units
something like $40,000 selling price
now imagine finding a bug in the design :D
I think Jon has done less than one panels worth in 5 years
imagine calling in a FLEEET of airplanes cause the l/g is facked....
alex_joni_ is now known as alex_joni
stacks like in that pic are why Mariss can afford to give out hardware to folks who write drivers or otherwise contribute
well.. actually the capital that allowed him to produce that stack
pick and place is fun to watch
yeah.. seen some at a local factory
Quick question....I have a 90 v DC spindle drive motor...
Similar to what s on a chinese mini-mill....
Will there be any problem running that style of motor in reverse ?
why should it?
Its a brushed DC motor....made in england...EMD is the name of the company that made it ...
I was thinking about brush phasing....or is that something that s more a of a problem (comutation timing) with brushless motors ?
brushes might not be located exactly in the neutral position, they could be offset for slightly better performance one way
I m debating buying a 4 quadrant drive....
Or maybe should just save my money until the right brushless servo drive and motor come along for it...
wb9mjn: how often do you plan on changing directions?
or you want to use it on an axis..?
Right now, I really do not have an application for reversing...but I am sure there is one...for example those little through-hole deburing things...
if you want to use it as a spindle motor, and you don't change directions very often.. maybe a relay is better
to just reverse the wirings :)
I have a solid state drive on there now....not sure how it would survive with the relay plugging...
And of course, one can slow down a spindle quite quickly with a 4 quadrant drive....
but also with a big resistor :)
I ve also had some trouble with this drive that is on there now..Keeps blowing this fuse or that.....could be the motor is too small for the R8 Bearings...
Have some isolation devices coming so I can hook up the drive to an extra axis output....
hmmm....might just wait...
till I find a good deal on a brushless spindle drive/motor...
Have had excellent success with a 2 HP (running at 1 HP) on the lathe here...Got a drive and motor as a package for $200....
The drive can take 115 or 220, just drops down the HP ....
Its much to big and heavy for the mill though...about 50 pound motor....versus the 180 pound motor and variable speed pulley set that was originally on
EMC I have seen in the newsgroup can use an axis 10 volt control output for the speed control....
Has anybody done this that is on here ?
that has been done on the mazak
connect the spindle speed to a free DAC, and you're set
* alex_joni goes away for a while..
Ok...using EMC 1 here....guess I will need to get a tach and treat it as an axis?
Found Dave Engvall's comments in the Newsgroup...will try that....
There are quite a few EMC1 machines running full servo.
I know of at least one that runs 8 hour days for the last four years.
what is the preferred way to set spindle speed on those machines (assuming you have a VFD)?
Spindle is the next analog after the axes.
wb9mjn: there you have it...
The one I just mentioned is a bpt and uses the snowmobile belt speed up speed down.
thanks jmk and ray....have some emails from Dave on that...will try that tonite...
Supposedly you can setup a feedback less spindle output, i think...
But one can certainly set the spindle to control a DAC output on the servo board....
I m working on my servo mill k4ts...trying to get the spindle speed control setup...
I have a scr 90 v DC spindle on this machine Ray...trying to get the spindle speed control from program...
Have a signal isolator coming...so, can hook the motenC-100 up to the 0-10 V pot input on the scr drive...
without smoking to many chips, hi!
what s the QTH there, K4TS, i m near Chicago...
saw a neet deal on EBAY...brushless DC motor/contoller in one, for 24 V DC, two hookups, power and pot...intended for motor scooters...
There are a bunch of em...made by Kolmorgen....anybody want to make their own SEGWAY ?
Get a gyroscope and EMC ....he he...
Wonder if there is a gearbox that is like geodesic dome shapped, and you could plug in all the motors, and have it come out on one shaft ?
Been looking at the Prius drive train too much lately...can you tell ?
did I miss anything?
Not much...figuring out how to do my spindle speed control from EMC1 ....
Dave Engvall sent me some notes, and I m going to try it on the MotenC-100 later...at least the output side of things,,,
why not go with emc2?
Found some signal isolators on Ebay...and will need one between the MotenC-100 and the SCR drive first...
I m not compilation competent...and do not have the time to become so...
Maybe next year....
wb9mjn: there is no complicated thing involved with emc2
only very few documented steps.. but it might be wise to wait a bt
I want to see my machine do something first...been a loooooooong project.....
That is, cut metal....
I have all the servo stuff all nice, and I have RC_46 all patched so that the trajectory stuff is now fixed....just do not have a machine table on it yet...
Which should happen next week....
If I had a spare computer and MotenC-100 board...I might risk it now....
wb9mjn: Which way from Chicago?
jmkasunich: still around?
ever heard of ODHL ?
doesn't sound familiar?
[22:03:03] <alex_joni> http://www.opendesign.org/license_models.html
I m in Naperville...due west...
jmkasunich: thought you might be interested
Ah. Not all that far from me.
Where are you tonite ray ?
wb9mjn: you on the map yet?
[22:09:45] <alex_joni> http://www.frappr.com/emctheenhancedmachinecontroller
look for ray henry there
Ray's home is in Crystal Falls, MI, if I remember correctly..in the UP....where my mom's family is from (Calumet) ...
Just about north of you by 400 miles or so.
Yep...about 10 hours by car..hi...
jmkasunich: btw, didn't see you on the map..
Okay. Calumet is not all that far.
damn thing doesn't work for me
jmk I had that problem also.
and I haven't been sufficiently motivated to figure out why
it uses java.. :)
so that might be an problem
why not try firefox?
way better :P
ha... that'll teach the cat to claw my leg
very happy ;)
could be happier though..
grabbed her and trimmed her claws with handy pair of wire cutters
I'm missing some pics
maybe some other time
gots code to write
dammit, who stole my K&R
jmkasunich: how precise is that mapping?
based on zip code, I didn't zoom in really close, but I suspect within 5 miles
you're on Rae Rd
near Mayfield Road
* jmkasunich calls up map
probably that's the center of the area code
nm.. don't let this stop your coding..
I'm about 6 blocks east
just north of oakville
they got pretty darned close
Was just on the phone with DaveE.
He said someone was looking to hook each limit separate on a parport.
How about matrix the signals outside the parport and then sort them out with HAL.
rayh: if you have enough pins you can hook them all to HAL
I was thinking the ordinary 5 in matrixed to 32 signals.
why would you need that?
It isn't need. It's a kludge.
Oh right. Let me think on that definition for a minute.
gotta go away for a bit, dinner time
programming term for a workaround or an inelegant solution to a problem.
a kludge is an awkward or clumsy (but at least
temporarily effective) solution to a programming or hardware design
ok.. I get kludge
so basicly you have 3 axes and 2 limits
and the decoding from 5 signals to 6 (3 x 2 limits) you do in HAL ?
seems really kludgy to me :)
Yes it does.
you only gain 1 signal, and a lot of troubles for that
if you would have 10 axes, then maybe
256 signals in an 8 wide signal
probably a serial HAL reader would be more effective.
Ah here is the correct definition.
kludge n : a
badly assembled collection of parts hastily assembled to serve some
stupid purpose (my addition)
Must get away for a bit.
I double checked the emc2 and yes the max feed in jog is using the traj section instead of the max velocity in the individual axis section
So my z axis only having the max velocity of 25ipm tries to run at the 72ipm of the traj section.
And does it make sense that the lower I make the max accelleration 20 -> 2 the slower I have to run the axis?
I get following errors when I lower the eccelleration for the same speed.
is there a setting for axis g0 axis movement to make them all go at maximum feed instead of them getting to the point at the same time?
skunkworks: I don't believe there is one. But what benefit does this have? Assuming there's not some bug, at least one axis will go at the maximum feed rate during a long G0 move, and the rest will go at whatever speed they need to complete the move in the same length of time.
yes that makes sense. Never thought about it - it will still only go as fast as the slowest axis
one more question - The faster axises seem to run just fine at 72 ipm but the when I jog the axis I get following errors on the de-accelleration.
72 imp when I am doing a g0 move with 1 axis
can't help you there