anyone got a mesa board working on a mill with servos?
yes (not me though)
I think petev has one, and possibly roguish
SWPadnos: I thought you had one too, in the bulk purchase?
I missed a bulk purchase?
few months back
that must have been a big secret
not really, just quick I think
course, I don't know what I would do with two
haven't hooked the first one up yet
I'm trying to decide if I really need any extra relays/etc hanging off of the 7i37 board
seems like it should be able to do almost everything I need without external hardware
I have 3 5i20s and 5 5i22s at the moment, but none of them is running a mill
I think the manual for the 7i37 says that you should use catch diodes if you switch inductive loads (like coils). even smallish relay coils can cause large-ish negative spikes at turn-off
a-l-p-h-a2 is now known as a-l-p-h-a
"Matlab has two problems: 1) it's expensive, and 2) it costs a lot." http://www.dspguru.com/sw/opendsp/mathclo2.htm
( regarding the simple TP files written in 'm' )
I get matlab from work
but the toolboxes are limited
just installed octave ( clone ) and will try to see if the .m code will load there ( and/or read the octave docs to see what the code meant )
tomp, what code you talking about?
no, what m files?
oh, the .m code code for simple tp is found at ...
[02:26:17] <tomp> http://wiki.linuxcnc.org/uploads/runcycle.m
the associated wiki page is http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Simple_Tp_Notes
makes two plots that they were too lazy to label
I could post them somewhere as jpeg
thank you maybe http://imagebin.org/
interesting, i think the code develops a new velocity, but is called a trajectory planner
it generates positions, actually
well, it generates positions. correct ones are preferable ;)
[02:34:59] <eric_u> http://imagebin.org/10279
[02:35:28] <eric_u> http://imagebin.org/10280
i was reading the web page just now, it describes generating a new velocity "3.2.1 simple tp math"
eric_u: thank you
tha lower "curve" could sure use some smoothing
well, it is acceleration
jerky accel ...
who wrote simple tp?
hey there. does anyone know the status of gantrykins?
I'm not sure what the exact status is, but I understand that there are some changes in CVS (which will be in EMC2.2) that make "interesting" machines easier to set up
things like XXYZ, XZ, XYUV should work better in 2.2
I love that part of the code where the units of one of the variables change, can't afford to waste those bits now, can we?
i looked into cvs and found a few recent files by Cradek.
(this is only from what I've seen on IRC and the mailing lists - I didn't have anything to do with writing the code, and I don't know exactly what's been done)
it's in simple tp, and I remember that from when I was looking at the trajectory planner in the production emc code
it's on line 41 of runcycle.m
oh. me not know matlab
well, cradek put it in a comment anyway
in cvs i did not find a plain 'TRUNK' in the tag drop down. what is the correct tag to view the latest and greatest code?
whoever did it in the emc code just did it without comment
Roguish, no tag
cvs -d:ext:firstname.lastname@example.org:/cvs co emc2
what? that's it?
well, you could add compression and a target dir if you want
cvs -z5 -d:ext:email@example.com:/cvs co -d emc2-TRUNK emc2
i'm just view right now. the linux box is in the shop, shut down until Friday when the day job is done for the week.
hmmm. I bet tortoise could be used to get an anonymous checkout, but strangely enough, I've never tried
day 2: http://jmkasunich.dyndns.org/cgi-bin/blosxom/sewers-09-05-07.html
my house must have been built by the same people dig within 10' of the place and you're hitting rocks and chunks of cement
the bedrock is shallow here - the sanitary sewer line is setting in a trench in the bedrock
and everything they dut out making that trench was used for fill
built in 1928, and I'm guessing without much power assistance
the trench in the rock is exactly wide enough to hold one sewer pipe, with the storm stacked on top, and the supply on top of that
wider would have meant more rock breaking
eric_u: did you find any calculations that resulted in positions in the runcycle.m code?
was that what I was looking for?
its what i'd like to ask if you dont mind, thanks
what they call progress appears to be position
at first blush
right T - P are differnces of position, but i didnt see any calculation of a new position, only saw old data being tested/compared
T-P is like heidenhains 'way to go'
distance not position
thanks, & goodnite
T is your target position
P is progress
T-P is target minus progress = work left to do
a distance, not a position, np
I think T is a target position
P is something like the position along the path
what other position you want?
teh new position, how th target is calculate, the thing a trajectory planner does
there must be a higher level
this is effectively a path planner for a line segment
really, it just takes your current position and a target and tells you how fast to go for the next sample period
what i saw in 'this' is a velocity planner, based on T-P, acc & desired (programmed) velocity. and that's my frustration ( its called an orange but its really a buick )
right it calculates velocity, not position
the velocity comes from PID
I'm pretty sure this tells you what velocity you need to go to stay on your path, the commanded velocity comes lower down
the coed we just read seems to calculate a velocity, maybe some other code also does that, but this code seemed to be touted as another/alternate trajectory planner
ok - could be
I haven't looked at it, so definitely don't listen to me
I looked at the c code this is based on over a year ago
i aggree that in emc the velocity can come from PID, but i was trying to see the trajectory get planned in this code
but this seems like a very straight forward translation
are you guys talking about the simple_tp branch in cvs?
the matlab code
* cradek goes back under his rock
didn't you write it under there?
thanks again & its goodnite from me :)
after one year I remove all traces, so nobody can blame me anymore
so, I think you're too late
I never blame people that write code
people that don't write code much more evil
Think I'll do a step by step of how to adapt those Surplus Center Ametek servos... For my use they will need mods.
are they currently available servos?
they are not water tight
yes - cheap
I bought 3
had one from a different lot
do you have a link?
had one from a different lot - allready ( total of 4)
[04:54:08] <Skullworks-PGA1> http://www.surpluscenter.com/item.asp?UID=2007062621461319&item=10-2351&catname=
that's nice it has a dual shaft
the wiring diagram is not currect - it says brown wire is not used - it is a chassis ground
yeah - but the shaft is too short...
I am going to make some press-on extensions
they are about 600 oz-in peak
motors could have been done up as 4 brush - these only have 2 installed
also bearing in rear cover is like an 8mm skate bearing
my other Ametek has 5/8 shafts at each end
they sould do just fine for my Seig X3
also these are 37V nominal with a 12A stall
the other motor I have is a 40V winding - and I think is a 4 brush
they do come as new old stock - in bags with cardboard sleevs on the shafts.
you gotta love surplus
I will prolly make replacement rear covers that have a compartment for the encoder and conduit outs
yeah - Camtronics sells a version of these for $165/ea.
but he does fit some simple U.S. Digital encoders (no index)
those are probably surplus too
a quality brushed motor will go for more than that
these really are sweet - have the skewed armature so there is no cogging
most motors out there have the exposed leads or spade connectors - I hate that!
with the replacement cap and use of silicone and some extra seals these will be clean and long term collant proof.
a program of a rect in xy plane is tilted 45deg to ZX and 45degree to xy ( like a compound sine). done in APT, sent to EMC, displayed in Axis
[05:22:43] <tomp> http://imagebin.org/10285
i purposely had the post make the circles into tiny line segment rather than helical arcs ( another agenda )
kudos to crochetyguy
Pretty interesting, what was the code generated with?
nuclear code reactor
is anyone planning on porting aptos to windows
my sources say no
aptos? is that what it was? I just logged on probably missed something
[05:56:42] <Jymmm> http://www.editthis.info/opencam/Main_Page
Now there's some cool stuff on there . thanks Jymmm
skinnypuppy1334: np, looked useful
toast: aptos is in python isnt it? so porting to windows would be a non-issue
fenn: is it?
i don't know the first thing about python
or how to run it on windows
python has most of its libraries built in, so the windows support is pretty good (~100% compatible)
[06:04:27] <fenn> http://www.python.org/download/windows/
fenn: not python, the APT language
i will investigate this, thanks
anyway aptos is not even close to finished, and i expect apt360 has sapped some of his motivation to get it done
apt360 works but it's in fortran and assembly for the ibm 360, so its not exactly code you can hack on
fenn: Why, doens't everyone have access to a IBM 360?
i dont even know what an ibm 360 is
[06:07:04] <Jymmm> http://en.wikipedia.org/wiki/IBM_360
so i guess apt360 isn't uh
going to run on windows
hm i dunno, it runs on linux, dont see why it wouldnt run on winderz
I know where I can get an account on one if that helps
no, you seem to be missing the point
nobody knows the language it's written in, and code conversion tools dont make it easier to work on
who said anything about converting, I'm talking running baby!!!
are there any other apt compilers
toast: yes, i believe dmessier has the source to a different one
its probably also in some ancient language
i guess i'm asking a different question - there is no ready to run apt compiler, anywhere, for windows
didnt you major in computer science?
fenn: if i wanted to dick around with things, i'd take up embroidery
alright, so, what seems to be the problem?
the problem is i'm asking a simple question with a simple answer
"no" is the answer, but it took entirely too long to get there
"no, but it wouldnt be that hard to port it"
if you had the inclination
i'm sure it is quite simple, but it's beyond what I want to invest in my own time
but thank you for the suggestion.
this is informative, to me at least: http://ometer.com/features.html
sorry, i'm not really commenting on your "feature request" its just something i happen to be reading. linked from here http://morgul.no-ip.com/dokuwiki/doku.php?id=feature_requests
note the "apt360" at the bottom of the page
so maybe i'm confused
and if i am, I apologize
i was under the impression that AptOS was in python, and that Apt360 is in fortran
Apt360 - more stable, but hard to port?
whereas aptos is not really ready for prime time, but, being python, should be simple to run on windows
very stable.. been in use for 50 years now :P
he converts the fortran to C, then compiles the C i think
i dunno how the asm gets compiled
join #cam and ask crotchetyguy
the python is about as far as i'm willing to go on this
I'd like to use this at work and school, and I don't feel the cost/benifit to doing it that way is substantial
versus doing straight g-code or using mastercam
if you have mastercam at work/school then use it
unless you think apt is better
i think each has their own benifits and I'd like to see what kind of things apt lends itself to
i think of apt as more of a geometry library
back when nerds knew how to do math
i have to do a lot of math anyway, so that part doesn't bother me
checking drawings, et cetera
just something i'd like to try out.
one of the more interesting parts of my job is figuring out missing dimensions
that the engineers left off
and can't be bothered to add on
you could be surfing for pr0n
unfortunately, unlike EMC, Fanuc does not allow you to run Firefox on the machine
[14:17:05] <cradek> http://willepadnos.net/jmkasunich/trench-from-front-0727.jpg
though my office is arranged similarly at the moment - a trench from the door to the chair
but no boards supported by the jacks
vapt creates a tilted path for emc the apt code http://pastebin.ca/684041
the gcode output http://pastebin.ca/684069
the axis screenshot http://imagebin.org/10288
the vapt screen shot http://imagebin.org/10289
i used the heidenhain post, just to get the arcs output as short line segments, converted L XYZ to G01 XYZ, then wrapped a couple lines around it for emc :0
the emc post would output helixes, and i wanted segments for an experiment in halstreamer
there is no way to get those arcs in ngc, the lines are the best you can do
vapt has 2 terms, intol & outtol, intol governs gouging (set to 0 ) outtol governs stock left ( set to .0001 ) this is all concerning chord vs true circumference path. so i've told it to stay outside by as much as .0001"
oh, i had to open up the limits on sim axis
a bit slow in here today.
SWPadnos: how is the board building coming?
should be OK. the PCBs are in process, and the assembly house got the parts list yesterday
I found an error, but I figured out an easy fix while I was in the shower this morning
amazing what a nice hot shower will do for you :)
I'm just lucky the op-amp is rotationally symmetric :)
oh - the boards are being (or have just been) silkscreened now
I hope so ;)
now to get bfload to work with the 5i22
and to make an FPGA config for the 5i22
and a HAL driver for it
then it should all be downhill :)
easy as pie
mmmm - pie
atleast you don't have to design the IC's
or - I guess programming the fpga is almost like designing an IC
in the abstract
sort of, but much lower turnaround time
that's the nice thing about VHDL (and possibly verilog as well) - you can take a design, change the underlying library (say from standard logic to 0.09uM SOI process), recompile, and have a more or less identical IC design from your FPGA design
at least that's my understanding
(which could be wrong, but I'll just live in my little fantasy world for a while if so)
[19:33:14] <JymmmEMC> http://www.mercurynews.com/opinion/ci_6814729
Another reason I dont like rebates.
"The BBB considers [2 complaints out of 51 being responded to] an unsatisfactory record." .. no shit, sherlock
[19:56:44] <jepler> http://www.cs.helsinki.fi/u/hprajani/phun/haha.jpg
-- looks like his PC had some overheating problems?
did he fill in the square in the center too? that's pretty funny
must be a pastry chef
at least that stuff is (usually) non-conductive
I'm just wondering how much in on the heat-sink side of the cpu
some high-end thermal pastes call themselves "silver", I wonder if they're electrically conductive..
arctic silver specifically says that it's not electrically conductive
yeah I see that
I'm pretty sure silver is conductive
which seems weird considering that it has silver in it
yes, it is, which is wnat seems weird to me
electrical insulator in the case of that photo I'd suspect.
"Not Electrically Conductive:
Arctic Silver 5 was formulated to conduct heat, not electricity.
(While much safer than electrically conductive silver and copper greases, Arctic Silver 5 should be kept away from electrical traces, pins, and leads. While it is not electrically conductive, the compound is very slightly capacitive and could potentially cause problems if it bridges two close-proximity electrical paths.)"
I always make sure to buy the cheap stuff or use whatever comes in the processor box
I bought a tube of AS5 or similar, once. So far it looks like it may be a lifetime supply
I just dig in the closet or desk until I find an old tube
then keep squeezing until more white stuff than water (?) comes out
I think NewEgg had an instant rebate or something that made it <$10 for the tube
it actually does work better
* JymmmEMC tosses SWPadnos rebate in the trash unopened.
hah I'm just glad I don't work on audio amps any more. I got really tired of idiots thinking the heat sink grease leaked out of the semiconductors... hence they were bad
don't you know that's the magic sauce that makes them work?
skinnypuppy1334: Just tell them it'll be $100/ea to replace
JymmmEMC, *instant* rebate :)
I'd heard when the smoke escapes the magic is gone... but jeez at the percentage that thought the white grease was the magic
skinnypuppy1334: how do they even know, did they open up the amp and expect to fix it themselves?
Sure enough jymmm their amps and stereos were about as nice as their shitboxxes
when I was in a 'car audio' store a couple years ago, I saw two guys come in and buy a subwoofer box thing for their car, which they called "that piece of shit out there", that had a broken out window and had plastic taped over the hole
That reminds me of the guy that called last night about his email complaining it hasn't worked right for a year... Well DUH... you have to CREATE the mailbox, not just label it in your emial client.
cradek: It's called a portable stereo... duh! lol
GHETTO BLASTER, a blast from duh past
I saw a strange thing today. a "young person" was walking down the street - very short hair, sweatshirt draped over one shoulder, too big T-shirt and sweatpants, holding their crotch and walking like a gangsta
I had a 76 Buick Riviera.... Leaked oil, leaked power steering fluid, leaked water... But damn if the heavy guage steel didn't make is sound good =)
that's strange enough in these parts, but I wanted to tell her that no matter how hard she tries, she's not gonna look like a boy from the 'hood
at least, not without radical surgery
SWPadnos: maybe she had crabs
maybe she did. I didn't ask
SWPadnos: And thus why I always carry a camera with me... for those type of "kodak moments"
I was in a hurry, so I didn't stop (and the light turned green)
SWPadnos: Stop? No, no, this is a drive by shooting
I had other things on my mind - I was headed for a Dept. of labor audit
SWPadnos: as complaintant or defendant?
random verification audit
SWPadnos: We told you to stop hiring those 8yo's and to close the sweat shop you've been running.
oh - they didn't ask about that part of the business
state or federal Dept of Labor?
Anyone know how long gecko usually takes to deliver? hopefully not as long as 800watt...
4 to 18 months
a few days
they shipped my stuff same day both times
That's cool. I'm guessing Jymmm is refering to 800watt
oh, could be
84" ballscrew and servo I'm not bidding on if someone is interested...
[21:12:32] <skinnypuppy1334> http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=290156134389&ssPageName=ADME:B:WNA:US:12
SWPadnos: Heh, my gf said it looked like fristing too when I showed her the cpu pic
that one is on my watch list, but it's a little too expensive now since I have no specific use for it
the seller has another one that's 54" long
or close to that anyway
It appears to be a multi-start ballscrew. Probably NOT what you want if you are building a CNC machine.
it's got a fairly hefty motor on it - that's an 800 W Yaskawa AC servo
Yup. Might be worth buying for the servo.
that was my thought
what sets the jog rate used for axis.N.jog-counts jogging?
lerman_ is now known as lerman