who makes the best allen wrenches?
I have some mother effin tight small screws (2.5 and 4mm wrenches) and I'm twisting wrenches
jmkasunich: snap on
thanks - now I just need to find some
have you tried boudhouse
I've had my set for years
no, all I have on hand at the moment is crapsman
Actually sears might seel bondhaus
my inch set are Elkund (or something like that, good ones), but the metrics, dunno
when I think of Bondhaus, I think of ball end, which would _not_ work here - the ball would snap off
they make micro tools for jewerly and electronics... not one .ca screw I can't remove using them
these ain't china setsecrews
they have the oddest heads and usually get stripped the most
they're in what I sure was a very expensive machine tool spindle
I'm not stripping the head, I'm twisting the shank of the allen-wrench
it doesn't help that the screw is 2" down a hole, and I have to use the long end of the wrench in the screw
I'm just syaing that they make the end of the tool actualyl fit the head so that doens't occure
how large is the opening?
jmkasunich: what about something like this... http://www.quinntoolsonline.com/catalog/TH-412S.jpg
then use a ratchet extension
and an impact wrench =)
that might work
but I need a 4mm blade that can get down a hole about 1-1/2" deep and maybe 3/8" diameter
right now I'm trying to figure out how to apply a bit of heat - I think they used locktite, and heat should soften it
toss a lil debatured alcohol doen the hole and light a match
burns with a white flame, so be careful - almost invisable
[00:48:12] <JymmmEMC> http://en.wikipedia.org/wiki/Denatured_alcohol#Uses
interesting use for sanding wood
got one out
a soldering iron fit nicely down the hole
it was coated with locktite
what are you crackign open?
disassembling some spindles
speaking of cracking, how is the safe?
I've be helping someone doing some hdd recovery so havent' been on it. But I called the mfg and they faxed over the manual.
I need to pinout the PCB to see what they're doing.
But at least I now know that there is NOT a button inside the safe, all programming is done on the ouside, so I still need to figure out why there are three wires going into the safe.
One is common to GND, not sure about the other two
They're using an Atmel uC and a EEPROM
both are off the shelf components
I look at the footer of the fax I got in, says I can now change my prerences to receive PDF's. okey cool. Goto my account, and if I want DPf I have to pay a monthly fee.
No thanks, I'll stick with the fee account tyvm - I don't recv faxes that often
hi is anyone in here?
woo, I changed tools
and I have most of the estop chain and most of the limits/homes
sounds like lots of progress for one weekend
yes very much so
you still don't have the resolver-to-encoder boards right?
right, no motion for a while
plenty of other stuff to do though.
there are four signals coming from the turret position sensor - one doesn't do anything
but it only needs three, so maybe it's on purpose. I haven't deciphered the pattern yet.
cradek: you mentioned running a coolant pump from single-phase. I cobbled together a static phase converter and idler motor a while back.
jtr: I currently have a big rotary converter, but I hope to get everything running on single phase to avoid the noise and complexity
hm, .5 HP new vfd is $110...
The idler is 1 HP, two terminls to the two legs of the power line, 35 UF run cap between one leg and the third terminal. nothing else. so you may be able to run your coolant pump with a single cap.
starts in less than a second.
woo, that sure sounds simple.
does it always start (and always in the same direction?)
my phase convert has a small 1ph starter motor, so no caps
I just looked - this motor is marked 1/6 HP.
yes and yes. I had plans to do it up nice with extra balancing caps and a start circuit, but it works too well to bother with. Maybe when I get the mill hooked up...
is it a small mill? 1hp idler motor sounds pretty small. mine is 8hp.
The lathe it was built for has a 1/2 hp motor; the mill will probably need a better setup, but I don't remember the size offhand.
[03:06:45] <cradek> http://www.metalwebnews.com/howto/ph-conv/fig1.html
they show an extra cap for starting only... I haven't found a diagram of run-cap only
I had thought I might have to kick-start until I got the better setup, but it fired right off.
aha, "self starting method" sounds like what you have
I was going to use these plans eventually: http://www.metalworking.com/dropbox/_1998_retired_files/FRW-n.txt
Fitch helped a number of people tune them on rec.crafts.metalworking
neat. I will read
thanks for the pointers, I'm encouraged to try it
welcome - and with Surplus Center nearby, caps shouldn't be expensive.
I built a rotary phase converter with an old three phase motor for an idler and some caps and a potential relay
and a magnetic starter.
it starts with a push button and when it gets up to speed the starting cap drops out.
if you're interested I can dig up the articles I used to build it...
thanks but I think I'm set. I'm browsing run caps now...
jtr: how many MFD per HP do you have on yours? I see numbers from 30-50
the Carlson article mentions 70 UF/HP starting cap - I'm actually using 35 to start and run the unloaded idler. So maybe 6UF for the 1/6 HP, but it won't be completely unloaded, so you might go up a little.
I will get 5, 6, 7.5 - they are very cheap
sounds like one of them should balance it ok.
I used a clamp-on ammeter to check the currents in the three leads to the motor, and measured the voltages across the windings to see that I wasn't generating real high voltages.
I don't have one of those ammeters, but I will check the voltages and should be able to pick the best cap value
[03:29:48] <cradek> http://surpluscenter.com/item.asp?UID=2008081022193269&item=11-3201
these sure look nice
Of course, I was "tuning" with the lathe running as a load. That is a nice switch - you thinking for the E-stop chain?
yes I need to put one or two on it
Sounds like you've made great progress tonight. Good luck, and take pictures! The cats have all gone to bed, so i guess it's time for me, too. Good night.
This is on sale at the retail outlets for $39.99 http://www.harborfreight.com/cpi/ctaf/displayitem.taf?Itemnumber=3377
They are pretty decent too, just don't exceed rated weight =)
handle folds down easily
is it ok that pressing F1 while running program change path mode from G61 to G64P1 ?
Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2008-08-11.txt
micges: as far as we know, the problem you reported was fixed in 2.2.6. Are you sure you have the problem on 2.2.6 like your report says?
I think this may be my first time in the room this early
I saw some of you guys were on early and not as much later
I will try it for a while at work
that's where I am right now
I figured that - maybe the responses will be slow as work interferes?
jymm is now known as jymmm
* archivist_ub regards work as an interference
jymmm is now known as jymm
[14:44:03] <skunkworks_> http://fantasticcontraption.com/?designId=356856
I could get adicted to this game..
skunkworks_: it's fun :)
I finished it 2 times, and that was enough :D
brain lock: how do I compare two s32 for equality?
seems like a messy way to do it.. http://fantasticcontraption.com/?designId=35744
damn, that fantasticcontraption game is addictive, i've been wasting a fair bit of time on it!
was wondering if anyone might now where some "Arm" plans are
Open Source Preferably
thanx, after more searching, lots found something i can use
I have a pinout question for the stg and the hal
does hal label the pins different from the stg card's manual?
i cannot get the emc out of estop.
what is stg?
servo to go
[17:27:51] <mr_jimjames> http://www.servotogo.com/dwnld/hardman2.pdf
ok, out of my guesspertice level
how about a different question
i can get the machine to activate my relays by hitting estop
the estop button doesn't come out of estop
what's a i733? (just kidding)
are you using the axis interface?
i use mini, but tried tkemc too.
i could try axis
anyway you might have the estop input on or off as the case might be
I think that would prevent you from changing the state of the e stop button in axis
I've never used mini or tkemc...
are you talking about the hal file, or the actual pins?
well if you have the e stop pin configured in your hal file and don't have the input hooked up it might act like you describe.
I'm not on a linux computer so I can't look and see but if you pastebin your hal file perhaps someone can spot it
i don't have it on me right now.
are you using one of the sample configs that came with EMC?
it said in 12 ... out 7
the hal file
i think i have it connected wrong.
I would ask which config but I can't look at them AFAIK from here
actually I can look, what config are you using?
what version of EMC?
the new umbuto with emc2 i believe
one moment customer on phone
there is a section that starts out connect e-stop... If you don't have your e-stop wired up yet you will have to comment out the lines indicated
I have a transistor/relay setup to connect the machine to the computer using +5v from the card.
in my last install i used the estop button to do this.
i am guessing i don't have the right pins connected.
i should be on pin 12 and 7?
page 12 connector p1 http://www.servotogo.com/dwnld/hardman2.pdf
i don't think that is labeled the same.
kinda looks like the e stop is connected to 12 and 7 is connected to the pc
I've never seen a stg so I'm guessing a bit
connected to the pc?
I have to leave now, perhaps someone can answer your question better
c'è qualcuno che parla italiano?
so, let me see .. pid tuning ..
if FERROR is 1.0 (mm) ....
then, P must be a MINIMUM of 10?
otherwise it will never get to +10 on Aout?
mine seem to oscillate like anyting with a P term over 5
I was hoping to tighten up FERROR ... but that would impl a larger P term no?
c'è qualcuno che parla italiano?
luca: Provi a parlare inglese (babelfish translation, don't hurt me)
excused if I write badly, use a translator babelfish
I wanted to have information on the compensation g41/42
luca: here are some links: http://linuxcnc.org/docs/html/gcode_main.html#sub:G40,-G41,-G42: http://linuxcnc.org/docs/html/gcode_main.html#sec:Cutter-Radius-Compensation
because of the "concave corner" limitation (220.127.116.11 in that link), it is generally not possible to make CAM software generate code that works with emc's G41/G42. However, it can be useful for hand code when you add appropriate arcs on all corners
I have not understood well as an example here, in order to make a square
g1g41 d2 x2 y2
why it cannot be compensated?
you can makes me an example of compensation for a foreign working and an inner one of a square? thanks
Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2008-08-11.txt
16:29 < cradek> micges: as far as we know, the problem you reported was fixed
in 2.2.6. Are you sure you have the problem on 2.2.6 like your
* alex_joni heads to bed
good night all
cradek: I've installed 2.2.6 from update manager
that pretty sure that command "emc" in console launch 2.2.6
and then I made those two tests
hey guys, got a weird error after some regular updating. any ideas what i screwed up?
[25183.753824] rtapi: no version for "nano2count" found: kernel tainted.
[25184.148047] motmod: Unknown symbol sincos
[25184.762613] RTAI[math]: unloaded.
wrong kernel booted
2.6.24-16-rtai (#1 Sun Apr 13 17:50:16 EEST 2008)
if you installed from normal Ubuntu then added EMC2+the RT kernel, then any normal kernel updates will set themselves to boot
well, maybe it's something else then :)
i'm trying to be really careful with all this updating stuff.
this one is my fault, and it looks like I forgot to fix it on CVS TRUNK when I did in the v2_2_branch: [25184.148047] motmod: Unknown symbol sincos
nano2count is surprising though, that's a kernel symbol
or it should be
oh, is that RTAI?
oops - gotta run
sorry, had a phone call myself.
oh man, missed him by that much
Roguish: a CVS up should fix the 'sincos' error now, and the other is just a warning
heck, i even got my script working to do the update!!!!
BigJohnT: I looked into luca's question & stumbled on a paper by Kramer ( one of the fathers of EMC ) on milling bores http://citeseerx.ist.psu.edu/viewdoc/download;jsessionid=6805026FD5756DF4F8E0D24A5336A711?doi=10.1.1.31.231&rep=rep1&type=pdf
it reminded me of your wizards.
this is awesome. (not done by me) http://fantasticcontraption.com/?designId=192073
regarding luca's question, i do not know if emc's interpreter can compensate discontinuous/'not-smooth' paths
(where subsequent gcode moves are non tangential)
this would require foreshortening the 'inside square' and automatically adding radii to the 'outside square'
so, I set some values in PID based onthe ziegler-nichols suggestions ...
and it tracks ok, but sounds sort of ... mmm .. rough
like running on sand
hm I could have sworn the wiki had a page on jon elson's interactive tuning procedure .. having tried Z-N and automatic tuning before I've never seen them give really good tuning, but Elson has a pretty good step-by-step procedure that he's demoed multiple times at CNC workshop
I think his is more or less the "simple method" described here: http://linuxcnc.org/docs/html/motion_pid_theory.html#r1_1_3
though a big part of his presentation is showing *how* to do this in halscope
well I give up -- I can't find the wiki page
ok, I'll look
i looked in the integrators manual, was expectign to see some procedure involving halscope
but, it didnt mention it
yep, thats Z-N,
I was hoping to see some small description of how to hook up halscope to read Aout and then spply impulses and whatch the response
one last one.. then I am done.. this one I did. http://fantasticcontraption.com/?designId=363142
heh, jon's vs ZN: they have the same start point, then Jon's is art & ZN is formula. I was taught Jon's. I bet most old-school servo tuners use it too.
I guess I'll just hook a real scope up and play
i have some neat baldor drives that tune themselves
whether its optimal or not I dont know
[22:20:35] <jepler> http://emergent.unpy.net/index.cgi-files/sandbox/axis-calibration.png
get the calibration menu with Machine > Calibration and halscope with Machine > Hal Scope
choose at least 2 channels and servo-thread
I only have one channel
in halscope you get up to 16 :-P
oh I hought you menat axes
"position" is used mostly for triggering
so for channel 1 select the position command -- for me it's signal Xpos-cmd and I think that's typical of sample configurations in emc
click the "1" button, then click "signals" and scroll to find Xpos-cmd or something else that looks plausible
or you can choose the PID input signal, pid.0.command, which should be the same thing
then click "2", click "pin", and choose pid.0.error
now go into machine on and mdi a move, then click "force" in halscope to make it capture
(you might have to click 'normal' under 'run mode' first)
* robin_sz scribbles all this down
once you've done this, you can click the "1" or "2" label again, then use the "gain" and "pos" sliders to position and scale the trace you selected
then to set a trigger, you're going to click where my screenshot says "source chan 1", and select channel 1 (position)
P seems to be about 4
I is around 2
D around 0.075
then you can move the "level" slider so that it is in the middle of the move that you are going to use for testing -- I was using "g0 x0" and then "g0 x1" to move 1 inch each time
once you've done that you can repeatedly mdi G0X0 / G0X1 and you'll get a fresh trace whenever you G0X1
* robin_sz nods
I'll go and have a go
so in the calibration window here's what you do:
type in proposed values for P, I, D and so on. Click "test". Then do your test move G0X0/G0X1 and decide if it's better or worse. If it's better (i.e., you want to keep it) click "ok"
if it's worse (i.e., you want your prior numbers back) click "cancel"
when you're done, you can in principle click "save to file" though personally I'd jot down the numbers and edit my inifile by hand. otherwise the values you found will be lost when you restart emc
* robin_sz nods
ok, off to try
thanks for the help
and the fish
oh and it might also be useful to select 3 channels and make pid.0.output be one of them
(actually you have to select 4 channels in that case)
then you can see stuff like the DAC saturating at +10V
hey Jep... can you help with a restoration of a system at all??
what's the trouble?
i blew the power supply.. ( thought it was the battery - replaced it ) and MO-BO and HD dont like each other GRUB 21 error
replaced the p/s too
same mobo and hard drive as before the trouble?
or you mean you replaced the mobo because of the battery (soldered battery?)
I would try following these instructions: http://ubuntuforums.org/showthread.php?t=224351
same mobo and drive.. i replaced the battery on the mobo
" probing .... this may take a log time"
maybe you had made some changes in the bios -- boot order, enable/disable ide ports?
changing cmos battery shouldn't affect the bootloader
but if there's a bios setting that the bootloader was depending on
that looked like it might have worked... gimme a few reboots and i'll be back
some recommended feeds/speeds for working in FR4 with small drills -- don't think I'll be going at 200krpm or 125inch/minute anytime soon, though: http://www.kyoceratycom.com/pdfs/200K_FS_FR4_Multilayer_High_Tg_Multiple_CLs.pdf
it's interesting that the chip load curve they found is not linear, but more S-shaped
(what's "max hits"?)
know anything about rtai errors
can you be more specific?
yes, I need to put up the dump on pastebin, but I did an install on another computer of mine it won't initialize rtai
[23:02:16] <pminmo> http://pastebin.com/m9b844dc
jepler were you at the cncfest this year?
pminmo: yes I was
did I meet you?
I don't think so
but I met so many people and I'm lousy at remembering so it's entirely possible
I understand, I have the same problem
I'll remember RayH because of the classes
in your log I notice that before you had the problem with rtai, the kernel had already encountered a problem with the sound driver saa7134_alsa (around line 344 of your paste, 85 seconds after boot)
this suggests a couple of things -- maybe the machine is just flaky. run memtest on it or boot the official ubuntu CD (not the emc2 ubuntu CD) and see if sound works there.
or maybe the version of the saa7134_alsa driver that got packaged with emc is buggy
Gawd, I swear I'm gonna kill the insurance companies!
in that case you can add the line "blacklist saa7134_alsa" to the file /etc/modprobe.d/blacklist with a text editor and try restarting
interesting, it had the previous version of Ubuntu and EMC and runs winxp about every other day
could a bad cd cause this?
I don't know -- you could run the media check from the cd boot menu to try to eliminate that possibility
hey Jepler.. can you drop me that site again
17:49:24 <jepler> I would try following these instructions: http://ubuntuforums.org/showthread.php?t=224351
thx... i think i foun a defective drive...
how did you relate line 344 to the saa7134_also driver ?
pminmo: line 344..378 is all from one error
I guess the saa... driver because it's mentioned on line 377 as containing the actual instruction that caused the error
[ 85.023636] EIP: [<e0a024d5>] snd_card_saa7134_capture_open+0xf/0x101 [saa7134_alsa] SS:ESP 0068:df523e64
down for another try
I really don't need sound I could disable in the bios
well that might work too
anything else you notice as a problem
well there's the error message printed when rtai fails but I don't see anything useful there
i also tried to do latency testing but it wouldn't initialize either
rati i mean
and both situations leave my keyboard non responsive
shoot, I don't have any other ideas offhand
and life is calling .. bbl
what device are you having issues with?
jepler thinks saa7134_alsa sound driver
i just tried disabling in the bios which didn't help so I'm going to try to blacklist it
what is saa7134?
well I was just doing a google search as I thought it was my onboard sound but its not
its a tv card that is plugged in but I don't use, so I'll pull it
It's probably confliciting with the base address of another on-board device.
back in a few minutes
blacklisting the saa7134 did the trick...........:-)