#emc | Logs for 2008-08-08

[00:13:35] <Guest179> Guest179 is now known as skunkworks
[01:51:02] <robin> meep
[01:54:36] <jmkasunich> peem
[01:56:22] <robin> so
[01:56:28] <robin> once I got the kb to work
[01:56:37] <robin> doing a plain install from CD without the live thing ..
[01:56:41] <robin> worked perfectly
[01:56:55] <robin> grub installed and stuck right
[01:57:00] <jmkasunich> I've never been particularly fond of live CDs
[01:57:18] <robin> it even seems to like my Mesa card :)
[01:57:51] <jmkasunich> you sound surprized
[01:57:59] <robin> well
[01:58:24] <robin> given in the past how anyting EMC related seemed to be based on goats blood scarifices and deep magic .. i was a bit
[01:58:36] <robin> very impressed
[01:58:40] <SWPLinux> heh
[01:59:14] <robin> i love the "copy this setup over, create desktop icon" functionality
[02:00:41] <robin> oops. closed the terminal
[02:00:55] <SWPLinux> as a user, do you think the relatively flat list of sample configs is good? do you think a hierarchical tree would be better?
[02:01:17] <robin> there is some hierarchy
[02:01:38] <SWPLinux> one level deep for each root dir
[02:01:45] <robin> maybe
[02:01:50] <robin> was not a problem
[02:01:53] <robin> I found what I needed
[02:02:11] <robin> maybe a good idea to have "3 axis stepper mill" really close to the top though
[02:02:30] <SWPLinux> ok - just wondering. I'm working on a tree now
[02:04:35] <jmkasunich> actually, 3-axis stepper mill should be a symlink to stepconf these days
[02:05:46] <jmkasunich> hmm, thats a thought
[02:06:16] <robin> is it still easy to mix stuff .. like 3 axis of servo, plus an axis of stepper for some random thing
[02:06:18] <jmkasunich> perhaps we should have one or more stepconf generated configs, complete with whatever trick files stepconf uses when you go back and tweak the config
[02:06:25] <jmkasunich> robin: of course
[02:06:42] <robin> just a bit of HAL tweaking?
[02:07:07] <robin> sorry, itas ben a long time :)
[02:07:13] <robin> brain has forgotten much
[02:09:15] <robin> ok off to bed
[02:09:18] <robin> see you around
[02:10:08] <SWPLinux> it's not easy with stepconf, unless the servo drives are step-to-servo like geckos
[02:10:16] <SWPLinux> sinece it's meant for stepper machines
[02:10:18] <SWPLinux> since
[02:10:40] <jmkasunich> whats not? mixing axes?
[02:11:05] <SWPLinux> using a combo of steppers and servos, with stepconf taking care of the details for you
[02:11:10] <jmkasunich> anything like that involves either skipping stepconf, or using it only to get a start, and then doing the rest like a real man
[02:11:15] <SWPLinux> doing it is easy with a text editor and a brain
[02:11:19] <SWPLinux> yep
[02:12:43] <SWPLinux> there's a change I'd love to see in the various HAL drivers and components, which would do two things for us: 1) allow a sim version to be used to do real configurations and 2) throw errors when the correct hardware isn't found
[02:12:54] <SWPLinux> actually, it's two changes
[02:13:17] <SWPLinux> 1 is to have a sim mode in the drivers, or have some pseudo-auto-generated pin/func/param list for each driver
[02:13:37] <SWPLinux> the second change would be to explicitly list every piece of hardware that's expected to be found
[02:13:54] <SWPLinux> (thinking largely of the PPMC here, but it's applicable to others as well)
[02:13:57] <jmkasunich> I don't get it
[02:14:12] <SWPLinux> having drivers that auto-detect hardware is kinda wrong
[02:14:18] <jmkasunich> why?
[02:14:29] <SWPLinux> what if a piece of hardware dies?
[02:14:43] <jmkasunich> then the pins that it exports aren't there, and the config fails
[02:14:47] <SWPLinux> the system won't load, but the reason will be hidden somewhere
[02:15:21] <SWPLinux> what if you add hardware, like another DAC card, to the PPMC, which changes the names of all the pins you have now?
[02:15:24] <SWPLinux> (or does it?)
[02:15:43] <jmkasunich> that depends on EPP bus addressing I think
[02:16:04] <jmkasunich> your original statement was very generic
[02:16:09] <SWPLinux> add a Mesa card - which is which
[02:16:11] <SWPLinux> yes :)
[02:16:49] <jmkasunich> I am adamantly opposed to anything that doesn't let you choose to load the driver and then see what it gives you
[02:16:56] <jmkasunich> discoverability is VERY important
[02:17:09] <SWPLinux> well, that could also be an option - load --discover or something like that
[02:17:22] <SWPLinux> load my_module config=DETECT
[02:17:24] <jmkasunich> heh - I was about to suggest that the other case should be the option
[02:17:27] <SWPLinux> heh
[02:17:43] <jmkasunich> loadrt some_driver better_find=far,bar,andblat
[02:17:48] <SWPLinux> exactly
[02:17:59] <SWPLinux> well, that should be a required parameter for normal operation
[02:18:07] <jmkasunich> why?
[02:18:10] <SWPLinux> you really want to get an error if the hardware isn't what you think it should be
[02:18:42] <SWPLinux> auto-detection is good, error detection is also good
[02:19:12] <jmkasunich> but the definition of an "error" depends on the user
[02:19:20] <jmkasunich> if I have a foo card, then abscence of foo is an error
[02:19:29] <SWPLinux> yes, it depends on the expected hardware
[02:19:32] <jmkasunich> if you have a bar card, abscence of bar is an error
[02:19:43] <SWPLinux> if I load the bar driver, it ignores foos
[02:19:52] <jmkasunich> so the _user_ should be the one specifiying (explicitly) what he wants to see
[02:19:58] <SWPLinux> yes
[02:20:03] <jmkasunich> I'm talking about ppmc, where one driver support foo and bar
[02:20:07] <SWPLinux> ok
[02:20:27] <SWPLinux> I see it similarly to the way you see setting up a system
[02:20:34] <jmkasunich> if you load a foo driver, and it doesn't find a foo, it prints to dmesg and exits (for those that can do that)
[02:20:44] <SWPLinux> you start with bare HAL and tune servos (for instance), then put the numbers you found into the EMC config
[02:21:02] <SWPLinux> I'm thinking that hardware detection should be that way as well
[02:21:23] <SWPLinux> first make sure you've got what you think you should, by using the detect funciton
[02:21:44] <SWPLinux> then put that information into your config, and hope to hell you get an error if anything changes that you didn't do on purpose
[02:21:47] <jmkasunich> we're arguing over a detail I think
[02:21:55] <SWPLinux> could be
[02:22:20] <jmkasunich> I want the detect to be the default - why should a newbie have to figure that out
[02:22:31] <SWPLinux> the added advantage is that the driver is told exactly what should be there at load time, so there can be a "fake this setup" option :)
[02:22:43] <SWPLinux> ie, export pins but don't twiddle with hardware
[02:22:44] <jmkasunich> the ONLY time you want that checking is once the config is done
[02:22:55] <SWPLinux> sure, could be
[02:23:32] <SWPLinux> but at this point, I don't think you can explicitly tell a driver that there should be two of X (STG, Mesa, Motenc ...) - they detect the number of cards
[02:23:54] <jmkasunich> that could be changed of course
[02:23:57] <SWPLinux> yes
[02:24:05] <jmkasunich> although I have grown to absolutely hate insmod params
[02:24:13] <SWPLinux> actually, it should be possible to write a "hardware detector" program
[02:24:15] <SWPLinux> heh
[02:24:21] <jmkasunich> discoverability again
[02:24:36] <SWPLinux> something that looks at lsmod, lspci, etc. and compares to a list of known hardware
[02:24:56] <SWPLinux> won't work for everything though - I think ISA cards (and parports, really) are about impossible to detect
[02:25:12] <jmkasunich> you write that.....
[02:25:20] <SWPLinux> heh - eventually ;)
[02:25:30] <SWPLinux> I'm just learning python and tcl, remember?
[02:25:31] <jmkasunich> personally, I think trying to write things that replace a human brain is hard
[02:25:38] <jmkasunich> regardless of the language
[02:25:57] <SWPLinux> well, there are always better idiots. eventually my programming skill will catch up with the bottom of the lot
[02:26:21] <SWPLinux> oh hmm. I guess I should pack up and get to my plane
[02:26:31] <jmkasunich> on your way to tormach?
[02:26:34] <SWPLinux> see - start complaining and the time just flies
[02:26:38] <SWPLinux> nope, close though
[02:26:50] <SWPLinux> going to the In-Laws, headed for Green Bay from Chicago right now
[02:26:55] <jmkasunich> ah
[02:27:18] <SWPLinux> maybe I'll get a chance to play with the lathe setup some day
[02:27:29] <SWPLinux> gotta run. thanks for the chat - see you
[02:28:05] <jmkasunich> goodnight
[02:44:00] <cradek> well I'm committed... it's sure easier to start a retrofit when the machine won't run as-is.
[02:44:23] <cradek> I mapped out pretty much all the wiring that'll stay
[02:44:30] <jmkasunich> did you start disconnecting stuff?
[02:44:40] <cradek> yes it's pretty well torn apart
[02:44:52] <cradek> I figured out why the spindle brake doesn't work -- it's missing
[02:44:58] <jmkasunich> heh
[02:45:03] <cradek> no other big surprises though
[02:45:19] <jmkasunich> thats good
[02:45:28] <jmkasunich> a VFD should be able to handle braking
[02:45:36] <cradek> yeah that'll be nice
[02:45:53] <cradek> but, I'll have to leave one of the clutches engaged for it to work
[02:46:20] <jmkasunich> twh clutches for two different drive ratios?
[02:46:27] <jmkasunich> two
[02:46:31] <cradek> lyes
[02:46:40] <jmkasunich> dog clutches, or friction ones?
[02:46:51] <cradek> looks like friction
[02:46:57] <cradek> like an AC compressor
[02:47:03] <jmkasunich> engage both at the same time - instant brake
[02:47:35] <cradek> I thought about that. I might try it.
[02:48:02] <cradek> might be best to make sure the motor is off before doing that.
[02:48:15] <jmkasunich> friction brakes (or clutches) wear, VFDs don't, so using the VFD would probably be better
[02:49:06] <cradek> good point.
[02:49:19] <cradek> there would be plenty of room for a heater under the big cabinet
[02:49:38] <cradek> if I want to rigid tap, I'll probably need a brake heater
[02:49:42] <jmkasunich> however, engaging both clutches after it is stopped might be a convenient lock when changing chucks, etc
[02:49:52] <cradek> it has a locking pin for that
[02:49:57] <jmkasunich> you mean a DB resistor?
[02:50:00] <cradek> yes
[02:50:22] <jmkasunich> I saw a retrofit where a guy used an electric stove element as the DB resistor - simple, rugged, cheap
[02:50:43] <cradek> smart. I bet they have about the right resistance.
[02:51:18] <jmkasunich> and they can take 240V continuously
[02:51:31] <jmkasunich> put it in a screen box on the top
[02:57:50] <jmkasunich> I'm gonna do something novel and try to be in bed before midnight
[02:57:53] <jmkasunich> goodnight all
[03:02:37] <cradek> darn, he left
[03:04:40] <cradek> hi stuart!
[03:05:00] <stustev> hi chris how is the lathe coming?
[03:05:12] <cradek> great. I started tearing it apart in earnest tonight.
[03:05:29] <stustev> it is fun taking things apart
[03:05:32] <cradek> I mapped out all the wiring, proxes, clutches, relays, etc. that I will keep.
[03:05:52] <stustev> can you keep the majority?
[03:05:56] <cradek> I have not found any big surprises yet. I will be able to pull it off.
[03:06:07] <stustev> I had no doubt
[03:06:08] <cradek> yes all the magnetics and isolation is pretty much there and working
[03:06:14] <stustev> great
[03:06:22] <cradek> it will mostly be a matter of wiring it up and writing ladders
[03:06:33] <stustev> just a little time
[03:06:48] <cradek> homing is unusual on it - it will take some finesse
[03:07:04] <stustev> is there anything you need - I will try to find something here
[03:07:27] <stustev> how could homing be unusual?
[03:07:36] <cradek> jon elson's resolver boards? :-)
[03:07:48] <stustev> I haven't heard much from him
[03:07:57] <cradek> I think I have everything, but I will keep your offer in mind, thank you
[03:08:09] <cradek> last I heard, he is waiting for his pcbs to be made.
[03:08:21] <stustev> that's all I know too
[03:08:55] <cradek> homing is unusual because there are many indexes per ballscrew revolution since the resolver is geared up. so they have a coarse home switch (slide) and a fine home switch (prox on the ballscrew pulley)
[03:09:20] <cradek> I think I will have to AND the fine home switch and the actual index and feed that to my index
[03:09:29] <cradek> this will make it so there is only one apparent index per ballscrew rev
[03:09:44] <stustev> yes - unusual
[03:09:59] <cradek> another problem is the Z home is somewhere in the middle of travel. if I start on the wrong side homing will fail
[03:10:26] <cradek> same for X but the 'wrong side' is very short
[03:10:54] <cradek> I could home to the outer limit switches except then I would not be able to hardwire them to the amp disables.
[03:11:07] <stustev> move the home switch
[03:11:12] <cradek> so I have to make some tradeoffs, or redo some hardware
[03:11:33] <cradek> it is very hard to do that on Z - the prox is in a hole in the main casting
[03:12:16] <cradek> I will figure something out. it will be easy to make it work 'ok'
[03:12:47] <cradek> jepler suggests we could have ping-pong homing -- if on the wrong side of the home switch, bounce off the limit and home the other direction -- I think/hope he was kidding.
[03:13:20] <stustev> not a bad idea
[03:14:04] <stustev> if you hit the limit before the home switch move the other direction until you see the home switch
[03:14:32] <cradek> then go past it, then home in the original way
[03:14:53] <cradek> but if limits don't disable the amps, you may as well home to the limit switch and have it much simpler
[03:15:31] <stustev> you would have to have amps with directional disable
[03:15:53] <cradek> true (these don't have it unfortunately)
[03:16:04] <cradek> that is a great feature
[03:16:10] <stustev> so far my amps have + disable and - disable connections
[03:16:27] <cradek> these have one 'inhibit' terminal
[03:16:42] <stustev> scratch the ping/pong home
[03:16:45] <stustev> :)
[03:17:13] <cradek> both screws have allens on the free end! it was a very regular occurrence to get stuck on a limit I think. homing is very flaky as-is.
[03:18:02] <cradek> they recommended you always jog both axes "in" before homing so you would not be on the wrong side
[03:19:10] <stustev> the cinci has the X homing in the middle - you have to jog the cam past the switch in the negative direction before you home the machine
[03:19:31] <cradek> bleh
[03:19:42] <cradek> can you make a new cam that's half the table length?
[03:19:56] <cradek> with emc, you could fix it right, by doing that
[03:20:12] <stustev> yes I agree - I should have moved the cam and switch a long time ago
[03:20:30] <stustev> I could have done it right with OpenCNC
[03:20:41] <cradek> if it is slow I can see wanting to home in the middle - you just need a very long cam
[03:21:16] <stustev> the cam is in a slot on the underside of the table - very easy to move
[03:21:37] <stustev> the microswitch would be more difficult
[03:21:47] <cradek> ah, nice and simple
[03:21:55] <cradek> is the cinci running now?
[03:22:16] <stustev> I will check - I may be able to do it by just moving the switch - we are ready to reinstall the spindle again
[03:22:53] <stustev> I will not be in the shop tomorrow - maybe it will be together when I get back Monday
[03:23:01] <cradek> neat
[03:23:06] <cradek> I'm anxious to see it
[03:23:29] <stustev> It is NOT possible for you to be any more anxious than I am :)
[03:23:39] <cradek> haha I bet
[03:23:57] <a-l-p-h-a> wuzzup folks.
[03:23:59] <stustev> I think my teeth are almost worn down from gritting them
[03:24:07] <cradek> it has been a much longer project than you expected. I hope I have better luck on mine but I will not bet on it.
[03:24:38] <stustev> you don't have a shop full of old machines with old controls to eat up all the time
[03:25:04] <cradek> but I do :-)
[03:25:21] <cradek> not as big a shop but stuff is just as old!
[03:25:23] <stustev> yes - I can't wait to see your shop
[03:25:55] <cradek> it is getting full
[03:26:46] <stustev> we had a lighting strike very near the shop last week - it seems as if every control has a quirk now
[03:27:13] <cradek> ouch
[03:27:43] <stustev> we checked the ground potential on the incoming service - we got 0.8 volts - it should be 0.0
[03:28:46] <stustev> we looked outside on the switch gear - the ground wires on every pole have been removed - copper thieves are unbelievable here
[03:28:57] <cradek> oh man
[03:29:09] <cradek> they are just asking to be killed doing that
[03:29:30] <stustev> I will call the power company but I don't expect much response - the thieves have hit all over town
[03:30:21] <stustev> I saw on the news - not in Wichita - someone had climbed a pole to harvest wire - 7200 volts got him
[03:30:48] <cradek> should I feel guilty if that doesn't make me very sad?
[03:32:07] <cradek> john k says people in cleveland are going around to abandoned (repo'd) homes and tearing out all the plumbing and leaving them to flood
[03:32:19] <cradek> essentially destroying the houses
[03:32:47] <stustev> it is sad but I don't feel very guilty - the darwin awards will maybe pick that up for an award
[03:33:33] <stustev> the copper thieves are doing that here also - sometimes someone will come home and their AC will not work - the unit has been stripped during the day - wow
[03:33:50] <cradek> good grief.
[03:34:03] <cradek> you'd think that can't be the easiest way to 'make a living'
[03:34:57] <stustev> this isn't a living - it is for drugs - these people work harder at not working than if the got a job
[03:35:14] <stustev> bbl - I am trying something with brlcad
[03:35:41] <cradek> goodnight
[03:38:02] <fenn> they should just switch to aluminum
[03:40:34] <stustev> good night
[08:54:26] <Poincare> stustev: here they steal the copper overhead cables from the electric trains. And with a few kV it sometimes DOES go wrong :-D
[09:47:58] <fragalot> Sweet, found a workshop that has all the tools & parts i'll need to make a CNC machine :) he said he MIGHT have some steppers laying about, so all i need now is the stepper controller :p
[09:49:48] <fragalot> which brings me to a lil' question, would i go for bearings, or make plastic guide bushings that slide over the pipes
[09:55:50] <archivist_ub> pipes? solid bars I hope
[09:56:11] <fragalot> yedah
[09:56:14] <fragalot> yeah*
[09:56:53] <archivist_ub> * archivist_ub looking for mass, strength and stiffness but not in that order
[09:56:58] <fragalot> actually it's a pretty thick diameter and short distance, so pipes would suffice
[09:57:22] <fragalot> thick walls, keeps the weight down and doesn't make that much of a difference in stiffness
[09:57:47] <archivist_ub> what accuracy are you looking for
[09:58:11] <fragalot> .5mm would already be nice for a first one, more is a plus
[09:58:33] <archivist_ub> ok pipes should achieve that
[09:58:45] <fragalot> esp. pipes with a diameter of 30m >.>
[09:58:46] <fragalot> mm
[09:59:21] <fragalot> i'd grab my calculator and tell you exactly how much it'd flex but i can't find it :p
[09:59:38] <archivist_ub> remember how cheap pipe is made and what specs they dont care about
[09:59:51] <fragalot> who said "cheap pipe"
[10:00:08] <fragalot> cheap pipes have a weld running along the length, which i don't want
[10:00:29] <archivist_ub> others are extruded
[10:00:37] <fragalot> yeah
[10:00:53] <fragalot> Anyways, pipes of full rods,.. bearings or bushings ;)
[10:00:59] <fragalot> s/of/or/
[10:01:02] <archivist_ub> so are not straight if they sag out of the die hot
[10:01:35] <fragalot> true
[10:01:53] <archivist_ub> bearings/bushings is friction v cost
[10:02:05] <fragalot> and simplicity of design
[10:02:51] <archivist_ub> both about the same unless there is added play reduction
[10:03:36] <fragalot> well bushings can just pe pressed into the MDF, for bearings i need to cut 2 strips to fit the L profile in, with 4 bearings on that, who'm are all 4 perfectly aligned
[10:04:38] <fragalot> just wondering if theres a big advantage of the bearings system over the bushings, thats all
[10:05:32] <archivist_ub> could design for very low friction and 0 play with bearings
[10:06:31] <fragalot> true.. bushings would be a balance between perfect fit, and friction..
[10:06:36] <fragalot> bearings it is! :)
[10:07:48] <archivist_ub> with bearings you can have one side fixed and strong springs the other
[10:08:26] <archivist_ub> as long as springs are above cutting forces seen by the springs
[10:08:36] <fragalot> mm yeah
[10:10:02] <archivist_ub> and track is straight and smooth enough that steppers can push bearings against springs during moves
[10:11:07] <fragalot> bbl food
[10:43:14] <micges> good morning
[11:08:52] <fragalot> back
[11:08:54] <fragalot> morning micges
[11:35:36] <fragalot> "I god had meant us to us the metric system, we would have been born with ten fingers" .. now you may call me a liar, but the last time I checked, I have 10 fingers and 10 toes...
[11:37:01] <archivist_ub> 8 fingers, two thumbs ....
[11:37:10] <fragalot> the thumb is also a finger afaik
[11:42:34] <fragalot> bbl
[11:55:27] <lyll> is it possible to use emc with a cnc-milling machine connected to a serial port (ttyS0)
[11:58:19] <archivist_ub> lyll, doesnt make sense as emc is designed to have full control although I have seen the odd request
[12:09:59] <lyll> i have a small milling machine that only havs a serial port for comunication
[12:10:22] <lyll> an older version of this: http://www.max-computer.de/x5e/small_format_a5a3.html
[12:11:07] <lyll> so there is no way to use this wirh emc?
[12:12:15] <archivist_ub> probably with some modification depends on your skill level
[12:13:58] <jepler> it won't be "plug & play" -- you would have to have full knowledge of the serial port communication, and figure out for yourself how to make that fit with the way emc works.
[12:15:41] <archivist_ub> lyll, if it has get at able stepper drivers internally, adapt and connect to the parport via optos
[12:16:53] <jepler> yes, getting direct access to the steppers or servos is another option
[12:16:55] <lyll> * lyll has no parallel ports :(
[12:18:17] <jepler> you could obtain a PCI parallel port or a smarter PCI card like those from mesanet.com
[12:24:26] <lyll> ill try to adapt it.. thx for the information
[12:25:23] <archivist_ub> someone here most of the time to answer questions, american timezone is best though
[14:26:04] <fragalot> jeez the chineese are CRAZY.. did you guys see the opening ceremony?
[14:26:09] <fragalot> it's like O.O
[14:41:58] <henley> hi all, I have a little problem... stupid enough. I registered to the emc-user mailing list and I keep receiving all the post in my mailbox. Where can I change the option to not receive all the post in my mail box?
[14:42:40] <henley> I checked the website, but I only find the subscibe section...
[14:42:57] <cradek> do you mean you want to unsubscribe?
[14:43:25] <henley> no, just don't receive all the post
[14:43:58] <cradek> do you mean you want to receive a digest instead of separate messages?
[14:44:18] <cradek> I'm having trouble understanding what you want, because to subscribe to a list generally means you want to receive the messages
[14:44:39] <henley> how does the digest present itself, an archive?
[14:44:49] <henley> yeah...
[14:45:10] <henley> once I wanted to post a question... but now
[14:45:26] <cradek> if you follow the last URL at the bottom of every message, you can enter your email address there and edit your preferences
[14:45:55] <henley> cool, thanks, I'll check now
[14:46:20] <cradek> follow the URL that's after any advertising that gets added
[14:46:46] <henley> It is not that messages from others don't interest me, its just most of the time, I am of no help...
[14:47:16] <henley> I keep deleting them to make room in my mail box.
[14:48:39] <archivist_ub> use gmail its free and they allow gigabytes
[14:49:03] <archivist_ub> no need to delete then
[14:54:25] <henley> Thanks, but personaly, when I have gigs of stuff lost around, I feel like my whole life is a mess.
[14:54:45] <henley> Thats fine, I found how to edit my profile.
[14:55:01] <henley> Thanks to all. ciao
[15:12:45] <fragalot> greaaaaat
[15:13:01] <fragalot> instead of using a market, I just used tipp-ex to point out an important part of my coursebook >.>
[15:13:06] <fragalot> marker*
[15:41:31] <jymm> hola
[15:42:07] <archivist_ub> holy
[15:44:29] <jymm> I need some ideas.... I'm getting a digital datasafe w/o the combo. So I need to come up with a "device" that can punch the buttons for me and pull in the handle. any suggetions on solenoids? need 12 of them (but cheap)
[15:45:13] <fragalot> cheap solenoids.. o.0
[15:45:25] <archivist_ub> make them
[15:45:31] <fragalot> this is EMC, here you use 2 steppers and a solenoid
[15:45:33] <jymm> archivist: ha
[15:45:47] <fragalot> bbl
[15:45:58] <archivist_ub> 3 steppers! no solenoid
[15:46:11] <fragalot> :/
[15:46:20] <fragalot> if you always hav eto do the same distance, there is no point in having 3 steppers
[15:46:25] <fragalot> too much added complexity
[15:46:39] <fragalot> real men would hook the solenoid up to the water pump function
[15:46:47] <archivist_ub> actually a 4th needed to test open the door
[15:46:54] <fragalot> haha
[15:47:23] <fragalot> why not just weld a big bolt onto the door, and screw a hydraulic press onto that?
[15:47:24] <fragalot> :)
[15:47:38] <fragalot> or a water jet
[15:47:54] <fragalot> bbl
[15:48:01] <archivist_ub> * archivist_ub imagines it wont need that much effort
[15:48:12] <jymm> I can get into the safe anytime I want, but this is a data safe, I want to maintain the integrity
[15:51:36] <tomp> "w/o the combo"? is this just to not leave fingerprints? ;)
[15:52:05] <jymm> tomp: say that again?
[15:52:26] <tomp> "w/o the combo"? is this just to not leave fingerprints? ;) ;)
[15:52:30] <jymm> Well, found the cheap part... just not mountable http://www.electronicsurplus.com/ccp73696-6vdc-solenoids-intermittant-duty-small-a420-066410-00-7678.htm
[15:52:40] <jymm> tomp: I meant rephrase it =)
[15:53:25] <tomp> it sounds like you didnt want to touch it , explaining jokes is useless
[15:53:45] <jymm> ah, they don't know the combination to the safe,
[15:55:36] <jymm> this has potential... http://www.electronicsurplus.com/objects/catalog/product/image/img73691.jpg
[16:39:00] <fragalot> if you don't want to touch it, use a breadstick >.>
[18:17:49] <robin_sz> hey
[18:18:01] <robin_sz> so, encoder and mesa card all wired up ...
[18:18:15] <robin_sz> im single ended on the encoder, but the mesa is diff in,
[18:18:27] <anonimasu_> hm.. then you are fscked :/
[18:18:33] <robin_sz> I presume I just ground the /A and /B inputs?
[18:18:37] <anonimasu_> (unless your mesa can do either)
[18:18:39] <anonimasu_> :p
[18:18:53] <anonimasu_> what mesa board?
[18:19:01] <robin_sz> well, assuming its just a diff amp ..
[18:19:03] <robin_sz> 7137
[18:19:13] <archivist_ub> conversion to diff is also possible
[18:19:18] <robin_sz> well, yeah
[18:19:26] <robin_sz> but i ouwl have thought not needed
[18:19:41] <robin_sz> surely just grounding the /A and /B inputs will do it?
[18:19:44] <robin_sz> no?
[18:20:02] <anonimasu_> is that 7i37?
[18:20:07] <robin_sz> yeah that
[18:20:37] <anonimasu_> but that's just the isolator board
[18:20:42] <robin_sz> yes
[18:21:04] <anonimasu_> it dosent have a encoder input
[18:21:21] <robin_sz> thnaks for playing
[18:21:25] <anonimasu_> :P
[18:21:38] <anonimasu_> what's your other mesa board?
[18:21:59] <robin_sz> oh
[18:22:01] <robin_sz> errm
[18:22:12] <robin_sz> 7i33
[18:22:15] <robin_sz> oopsie
[18:22:17] <robin_sz> my bad
[18:23:24] <anonimasu_> they can do ttl also..
[18:23:32] <robin_sz> yes
[18:23:37] <anonimasu_> page 5 of the datasheet..
[18:23:40] <robin_sz> but I need to ground the spare input
[18:23:42] <robin_sz> ??
[18:23:44] <robin_sz> or not
[18:24:03] <anonimasu_> no, I dont think that matters
[18:24:05] <robin_sz> ok
[18:24:12] <robin_sz> in that case it should work ;)
[18:24:20] <robin_sz> * robin_sz scratches his head
[18:24:42] <jepler> robin_sz: which board? 7i30? 7i33?
[18:24:43] <anonimasu_> differential pins are other inputs no?
[18:24:56] <robin_sz> err
[18:24:58] <robin_sz> i guess
[18:25:01] <archivist_ub> * archivist_ub might expect to set it at mid rail not gnd
[18:25:17] <anonimasu_> then ignore them..
[18:25:30] <robin_sz> k
[18:25:35] <anonimasu_> it dosent make sense for it to take both the differential and the normal one in consideration..
[18:25:39] <robin_sz> so in TTL mode it just uses the A and B pins
[18:25:42] <anonimasu_> not after you've set what you want to care about..
[18:25:56] <robin_sz> in diff mode it uses the A and /A, B and /B
[18:26:11] <jepler> yes I'm pretty sure that you do not connec the /ENCA and /ENCB pins when you use the TTL jumper setting
[18:26:17] <anonimasu_> * anonimasu_ nods
[18:26:20] <robin_sz> ok
[18:27:28] <robin_sz> might need a harder pull-up resistor then
[18:28:03] <anonimasu_> hm.. why do you need a pull up?
[18:28:13] <anonimasu_> dosent it work for you?
[18:28:17] <robin_sz> open collector outputs on encoder?
[18:28:20] <archivist_ub> some encoders are open collector
[18:28:24] <robin_sz> mine are
[18:28:34] <anonimasu_> ah
[18:28:47] <robin_sz> i was expecting the 'joint' to display the position when I moved it by hand
[18:28:54] <robin_sz> even in estop
[18:29:28] <archivist_ub> do you have a scope to check levels on that pin
[18:29:37] <robin_sz> twiddling the axis by hand does not seem to register in emc
[18:29:40] <robin_sz> yes, I do
[18:30:44] <jepler> I'm 100% sure that in "machine off" state, emc tracks encoder inputs to display on the DRO. I'm 99% sure that it also does in estop.
[18:33:01] <anonimasu_> it would make sense for it to track them always
[18:36:30] <jepler> yes I agree
[18:36:31] <robin_sz> ok
[18:36:38] <robin_sz> so I added astronger pullup
[18:36:39] <fragalot> microstep or halfstep?
[18:36:43] <robin_sz> I hav 200 ohms now
[18:36:45] <fragalot> (which should i program into my PIC)µ
[18:37:10] <robin_sz> and about 2.5v square wave
[18:37:14] <robin_sz> should be enough
[18:37:20] <robin_sz> hmm
[18:37:30] <jepler> fragalot: what task is your PIC performing?
[18:38:09] <robin_sz> the mesa has IO 1-24, 25-48, 49-72
[18:38:27] <robin_sz> how do Iknow which ribbon cabel is servos , which is limit switches?
[18:38:35] <fragalot> jepler: converting step & direction inputs to stepper output
[18:38:58] <fragalot> also, does EMC step on the down or up flank of a step signal
[18:39:05] <jepler> robin_sz: "P2" is encoder/dac 0 through 4. see http://linuxcnc.org/docs/html/hal_drivers.html#r1_5_4
[18:39:23] <robin_sz> ok, im using P2
[18:40:13] <jepler> fragalot: stepgen uses a TRUE pulse on the step output for a step, but you can invert the outputs that appear on the parport or other digital output if your device requires it different
[18:40:26] <robin_sz> im assuming they come out as per the Mesa manual once they have gone through the 7i33
[18:40:38] <fragalot> jepler: so it steps on the rising flank of the true pulse
[18:40:41] <fragalot> k
[18:41:10] <robin_sz> jepler: he could also get stepper outputs direct from EMC, without the step/direction conversion too
[18:41:23] <robin_sz> or could once anyway, I assume we stil lcna do that
[18:41:33] <fragalot> jepler: so, microstepping or halfstep for the steppers ?
[18:42:42] <archivist_ub> I use halfstep, I dont expect great speed from them
[18:42:59] <jepler> fragalot: yeah, half step will be easier to implement and it's perfectly adequate for a small mill
[18:43:05] <fragalot> k
[18:43:07] <fragalot> thanks
[18:43:09] <jepler> you don't get much more positional accuracy from microstepping
[18:43:26] <fragalot> no, but it might have had another benefit
[18:43:28] <fragalot> :)
[18:48:22] <archivist_ub> fragalot, reduced resonance, so easier to accelerate though the critical speed
[18:48:32] <fragalot> I see
[18:48:46] <jepler> IMO look at l297 for an idea of what a good half-step translator & constant current controller will do. when I looked at doing something like that in a microcontroller (I considered AVR rather than PIC) the main barrier was having two analog comparators or DACs that were fast enough to give good chopping performance like l297.
[18:49:04] <fragalot> jepler: doing so right now
[18:49:15] <jepler> er, not DACs but ADCs
[18:49:33] <archivist_ub> yup they do the current limiting for you
[18:49:44] <fragalot> I'm also looking at alternatives to the µC
[18:49:57] <fragalot> like dedicated circuits, such as the l297
[18:50:20] <archivist_ub> I would use dedicated or bought in
[18:53:19] <maddash> sweet.
[18:57:32] <fragalot> jepler: http://www.electronics-lab.com/projects/motor_light/045/schematic_th.gif how is this
[18:57:36] <fragalot> archivist_ub: ^
[18:57:49] <fragalot> http://www.pchilton.co.uk/images/stepperdriver.jpg or this one.
[18:58:37] <fragalot> $26 .. can't really go wrong on the first one (premade!)
[19:00:16] <robin> well, coinfusing, I have nice 3V square waves going into the 7i33
[19:00:26] <robin> 0 to 3v,
[19:00:29] <robin> pleanty for TTL
[19:00:36] <robin> jumpers in TTL posiotion
[19:00:48] <robin> 7i33 connected to P2 of the 5i20
[19:01:24] <robin> now ..
[19:01:38] <robin> now is it a good time to start with the goats blos and chicken bones?
[19:02:04] <anonimasu_> it has to be polarbear blood. for it to be of any use
[19:02:06] <jepler> fragalot: schematic_th.gif looks like a pretty typical L297+L298 board for bipolar motors.
[19:02:33] <robin> oh pooh
[19:02:41] <robin> so difficult to get polarbear blood
[19:02:51] <jepler> robin_sz: transfer it to connector P3 and use halscope/halmeter/whatever to see whether looking at those signals as digital inputs you see any changing input values
[19:03:15] <robin> hmm
[19:03:22] <robin> sigh
[19:03:26] <robin> opening the PC again ...
[19:03:32] <robin> moving the monitors again
[19:03:35] <robin> oh welll ...
[19:03:58] <robin> ah wait, I havbe a cable ready for P3 .. woo hoo
[19:04:13] <fragalot> jepler: cool, I found it for $26, looks like it's a reasonable price (and i don't have to do any design work ^_^ yay)
[19:08:01] <jepler> fragalot: oh, the maximum current for L298 is 2A
[19:08:07] <jepler> make sure that's appropriate for your motors
[19:08:55] <robin> wow
[19:09:00] <robin> suddenly
[19:09:03] <robin> it works
[19:09:05] <robin> blink
[19:09:12] <jepler> robin: what did you change?
[19:09:34] <archivist_ub> he switched it on
[19:09:46] <robin> not a thing
[19:09:47] <fragalot> jepler: 2A is /MORE/ than enough for me.
[19:09:54] <robin> i read the PDF
[19:09:57] <robin> and it worked
[19:10:00] <robin> is it linked?
[19:10:02] <jepler> robin: ah so documentation is useful
[19:10:06] <anonimasu_> lol
[19:10:09] <robin> seemingly
[19:12:56] <jepler> fragalot: and your motor is capable of bipolar operation? (4, 6, or 8 wires)
[19:13:17] <fragalot> jepler: 4
[19:14:05] <fragalot> just figuring out which transformer i'll need to get 240->24VAC (eg. how much amps it 'l all pull)
[19:15:17] <jepler> fragalot: beware -- their design is limited to 30VDC by the capacitor C3's voltage rating. 24VAC rectified is more like 34V.
[19:15:28] <archivist_ub> I got lucky and have a 28volt switchmode psu which adjusted to 30v nicely
[19:15:37] <fragalot> 18V it is
[19:15:38] <jepler> L298 is good to 40V+ but C3 isn't
[19:15:51] <fragalot> jepler: C3 is easilly replaced..
[19:16:30] <jepler> hm and 1N5822 is limited to 40V
[19:16:38] <archivist_ub> c3 on diag is 35v
[19:16:52] <fragalot> archivist_ub: voltage spikes occur.
[19:17:02] <fragalot> owell, i'll just use an 18V toroid
[19:17:03] <archivist_ub> electrolytics run close to voltage
[19:17:16] <archivist_ub> unlike other devices
[19:17:20] <fragalot> just figuring out which VA i'll require (could do "as big as i can find", but tahts expensive :p
[19:17:22] <jepler> archivist_ub: yeah but that doesn't leave any margin or allow for V+ to rise due to back EMF
[19:17:24] <anonimasu_> switchmode sucks for cnc
[19:18:08] <archivist_ub> working fine for me
[19:18:41] <fragalot> mm, 18V 160VA -> 48euro
[19:19:06] <fragalot> and a big capacitor.
[19:19:17] <anonimasu_> :)
[19:19:44] <fragalot> and badass rectifier.. ... meh i'll just make one with the pile of 6A6 diodes i have laying arround
[19:20:51] <anonimasu_> switchmodes work too.. it's just that if you have back emf.. they usually shut down :)
[19:21:13] <jepler> huh I haven't seen that problem
[19:21:28] <alex_joni> hi guys
[19:21:28] <robin_sz> meep
[19:21:39] <jepler> in my piddly little machine the switchmode supply works fine so far
[19:21:43] <anonimasu_> :)
[19:21:47] <archivist_ub> Im using a Coutant Lambda one not a chinese cheapo
[19:21:55] <robin_sz> so
[19:21:55] <jepler> but ask me again after 1000 hours of use
[19:22:03] <anonimasu_> it might just be the one I used in my app..
[19:22:07] <robin_sz> thats the encoder hooked up and scaled easy peasy
[19:22:08] <anonimasu_> it was a DC to DC converter..
[19:22:36] <archivist_ub> anonimasu_, maybe an extra bit of decoupling was needed
[19:22:38] <robin_sz> i guess an input scale of 100 is a bit coarse
[19:22:52] <alex_joni> robin_sz: / mm?
[19:23:02] <alex_joni> sounds good to me :)
[19:23:08] <fragalot> now, 160VA is just overkill (i hope 3 stepper motors of 1A ), but who cares. -> how big of a capacitor would i put on :p
[19:23:17] <robin_sz> yeah, mm is hte only way
[19:23:33] <alex_joni> finished installing a new bot today :)
[19:23:42] <robin_sz> has it killed anyone yet?
[19:23:59] <alex_joni> nope .. not a rogue one :)
[19:24:04] <robin_sz> ah well
[19:24:06] <robin_sz> give it time
[19:24:10] <alex_joni> :P
[19:24:24] <alex_joni> it's a mid-sized one .. so definately the power is there
[19:24:31] <robin_sz> yes for sure
[19:24:34] <alex_joni> ~ 9m long
[19:24:36] <robin_sz> 50kg?
[19:24:40] <robin_sz> wow
[19:24:43] <robin_sz> 9m
[19:24:47] <robin_sz> bigggggg
[19:25:06] <alex_joni> nope.. small
[19:25:13] <robin_sz> that is envelope?
[19:25:16] <robin_sz> or track length?
[19:25:19] <alex_joni> heh.. no
[19:25:19] <anonimasu_> linear track?
[19:25:23] <alex_joni> 9m linear track
[19:25:25] <jepler> fragalot: http://geckodrive.com/upload/Step_motor_basics.pdf suggests C=80000*I/V where I is current and V is supply voltage
[19:25:25] <alex_joni> then the robot
[19:25:27] <robin_sz> ah eight
[19:25:38] <alex_joni> and 2 positioners
[19:25:40] <fragalot> jepler: sweet, thanks
[19:25:49] <jepler> fragalot: in uF
[19:25:50] <alex_joni> one for 5t parts, one for 2t parts (lathe type)
[19:25:56] <robin_sz> fragalot, you could just buy geckos, will be cheaper
[19:26:16] <alex_joni> robin_sz: the next one (I start working on on monday) is 15m long
[19:26:22] <alex_joni> but the big one is ~20m long
[19:26:23] <robin_sz> heh nice
[19:26:31] <robin_sz> truck chassis?
[19:26:41] <alex_joni> nope.. beams
[19:26:44] <fragalot> robin_sz: do geckos take 240V in? (and last time i checked, a single gecko starts off at a higher price than 3 of these other circuits)
[19:26:45] <robin_sz> right
[19:27:01] <anonimasu_> fragalot: a gecko will outperform it for sure.. too :)
[19:27:33] <robin_sz> of course they dotn take 240v in, neither does an L297, and 3 geckos costs lots less than building 3 1A drives
[19:27:35] <anonimasu_> fragalot: no, they take 48v in max..
[19:27:35] <fragalot> anonimasu_: i'm not after performance that badly ;)
[19:27:47] <anonimasu_> yet..
[19:28:00] <fragalot> robin_sz: actually,.. no
[19:28:08] <alex_joni> robin_sz: sent you a pic :)
[19:28:30] <alex_joni> fragalot: robin_sz counts the labour time too
[19:28:42] <fragalot> robin_sz: atm one stepper driver costs $26, the coil i just found in one of my catalogs costs $35, the rectifiers, just diodes that i have laying arround, capacitors, i have laying arround,...
[19:28:49] <archivist_ub> we cheapskates dont
[19:28:52] <fragalot> alex_joni: I like a bit of work :)
[19:29:01] <fragalot> besides, it's just for in my spare time
[19:29:05] <alex_joni> fragalot: if you can spare the time :)
[19:29:06] <fragalot> beats beeing on IRC
[19:29:09] <alex_joni> * alex_joni just got home..
[19:29:15] <alex_joni> and it's 10PM :(
[19:29:18] <robin_sz> fragalot, well, weve only been on this channel 6 or 7 years, and, as yet, weve seen a few hundred people try to build their own drives because it was "cheaper" .. as yet, none have saved money, most have not succeded, and most have bought geckos in the end anyway, but, feel free to prove us wrong.
[19:29:43] <alex_joni> robin_sz: I have 2 geckos and one 297/298 on my machine :P
[19:29:56] <robin_sz> heh, i had 207s as well
[19:30:00] <alex_joni> I didn't quite build it myself, I had one of my employees do it :D
[19:30:00] <robin_sz> horrible things :)
[19:30:01] <fragalot> robin_sz: it's a kit with the 297/298 combo on it
[19:30:08] <fragalot> lol alex_joni
[19:30:25] <alex_joni> fragalot: well.. in the end I ended up debugging/tuning it until it ran
[19:30:28] <robin_sz> alex_joni, i started out with L297, but they are pretty bad
[19:30:29] <fragalot> I hate FANUC.. I really REALLY hate them machines
[19:30:39] <fragalot> alex_joni: explain
[19:30:49] <fragalot> if it runs, and it runs without stalling, i'm happy
[19:30:56] <cradek> robin_sz: I built four L297/8 driver boards and am still using them. geckos are awfully expensive.
[19:31:00] <alex_joni> fragalot: an L297/298 is quite more than 2-3 chips on a board
[19:31:10] <alex_joni> you still need to tune current limits, and so on
[19:31:23] <fragalot> alex_joni: don't call that debugging, call that finetuning.
[19:31:31] <jepler> and my experience is that xylotex also works quite reliably, though I'm not using it near the current or voltage limits
[19:31:56] <fragalot> I also considered the hobbyCNC board
[19:32:21] <alex_joni> I can surely feel quite a difference from the geckos to the 297/298
[19:32:32] <robin_sz> cradek, ive run both, the geckos make the 297s look like .. well, a ferrari compared to a trabant. the performance is so different, it makes the 297s look broken,
[19:32:34] <alex_joni> but I didn't justify it yet to buy another gecko for the Z axis too
[19:32:36] <fragalot> all in all, i think the price will be relatively the same, and if it breaks, in this case by using 3 seperate drivers, i can replace the one that broke instead of the entire board
[19:32:48] <fragalot> robin_sz: won't argue on that
[19:32:51] <alex_joni> but I have tiny steppers.. so it doesn't really matter
[19:32:55] <anonimasu_> hm.. also geckos dont break..
[19:32:59] <anonimasu_> except mine.. :p
[19:33:02] <fragalot> anonimasu_: haha
[19:33:03] <alex_joni> as the steppper get bigger, it matters
[19:33:06] <anonimasu_> dont run them without current limit set.
[19:33:12] <alex_joni> anonimasu_: I also have some 201's
[19:33:18] <alex_joni> those break easily :)
[19:33:23] <anonimasu_> hm.. I dont think so
[19:33:37] <fragalot> anonimasu_: that's what they say about FANUC drives too... guess what, wher eI work, those are the most common things that break. the cheap taiwaneese relay-based boards hardly ever break.. sofar had to replace 2 relays,.. in total, on 7 machines
[19:33:37] <anonimasu_> I mean obviously you cant run them without motors..
[19:33:39] <robin_sz> ive had a bridgeport running on 201s :)
[19:33:45] <fragalot> anonimasu_: oh, and their PSU's, but thats because they underpowered it :p
[19:33:55] <alex_joni> bbl
[19:34:05] <archivist_ub> I had a duff driver from my source (seems an intermittant)
[19:34:16] <anonimasu_> hehe
[19:34:20] <robin_sz> so, now to hook up the analogue out :)
[19:34:38] <robin_sz> and pray I get it the right polarity :)
[19:34:45] <anonimasu_> test them first..
[19:35:08] <robin_sz> comanded poitin at zero
[19:35:18] <robin_sz> move axis to +10
[19:35:36] <archivist_ub> will the bits fly this far north? should I duck
[19:35:40] <jepler> robin_sz: you hooked up estop already right?
[19:35:41] <robin_sz> yes
[19:35:59] <jepler> and tested it to work?
[19:36:06] <robin_sz> jepler, what is the "estop" of which you speak, is it a custom amongst your people?
[19:36:13] <anonimasu_> lol
[19:36:33] <jepler> robin_sz: yes. it is the custom among my people that once you've pressed a woman's estop button, you are married to her.
[19:37:04] <anonimasu_> better be careful with thoose rapids
[19:37:47] <cradek> robin_sz: I was just letting you know that all of "as yet, none have saved money, most have not succeded, and most have bought geckos in the end anyway" did not apply in my case
[19:37:51] <alex_joni> * alex_joni hates safety regulations
[19:38:18] <anonimasu_> cradek: well, you are special :p
[19:38:50] <alex_joni> things get more and more extreme these days
[19:39:07] <alex_joni> first they started using double contacts on all safety switches and estops
[19:39:37] <alex_joni> now you need additional PLC's connected to limit switches and workcell selection
[19:40:32] <cradek> my lathe has some air powered stuff, like a special slide for a parting tool. I am trying to figure out what it should do during estop.
[19:40:55] <alex_joni> for the stuff I'm talking about you actually feed the axis position into a PLC
[19:41:02] <alex_joni> then you teach it the location fo the switches
[19:41:18] <alex_joni> and if it reaches the switch position without tripping the switch it estops the whole machine
[19:41:29] <alex_joni> (it assumes the switch is broken)
[19:43:03] <tomp> i need a new programmer for a new technician, i do PICs PEELS 8051's not too fancy. I used to use Needham's. they're outta bizness. Any suggestions?
[19:43:46] <alex_joni> upgrade? :D
[19:43:49] <lerman> We typically have our technicians pre-programmed by their parents. :-)
[19:44:07] <alex_joni> lol @ ken
[19:44:10] <archivist_ub> I use one of Microchips own for PIC
[19:44:16] <alex_joni> usually genetically programmed ?
[19:44:19] <tomp> outta business (Needhams & Grandparents;)
[19:45:08] <tomp> archivist yes, got that with PicLab, but these guys wanta swiss army knife that does everything
[19:45:40] <archivist_ub> tomp my other one is also an out of business one :(
[19:46:27] <anonimasu_> alex_joni: insane
[19:46:32] <alex_joni> tomp: http://www.progstar.com.pl/combo3/
[19:46:35] <alex_joni> anonimasu_: I know..
[19:46:52] <anonimasu_> alex_joni: why not rely on a linear resistive encoder anyway..
[19:46:53] <alex_joni> whenever something changes, you need to reteach everything.. it's .. &^*(&(& to say the least
[19:47:03] <anonimasu_> alex_joni: if it breaks you know it's broken..
[19:47:13] <alex_joni> anonimasu_: this is for workspace separation
[19:47:25] <anonimasu_> hm.. like robot cells?
[19:47:33] <alex_joni> if you have 2-3 work places along the same linear track for example
[19:47:37] <anonimasu_> yeah
[19:47:56] <alex_joni> but it can also be 2 tables left and right of a robot
[19:47:58] <tomp> alex_joni: thanks, thats pretty much the items i burn, so how much is a Brutto?
[19:48:21] <alex_joni> brutto is the price with taxes
[19:48:38] <tomp> ooops, is the unit Euro then?
[19:49:19] <anonimasu_> no
[19:49:23] <anonimasu_> pln
[19:49:31] <tomp> zlotny = half a buck
[19:49:34] <tomp> thanks
[19:49:50] <alex_joni> 1 Polish zloty = 0.460723 U.S. dollars
[19:49:59] <anonimasu_> 217.2024 US$
[19:50:12] <alex_joni> not really cheap, but acceptable
[19:50:46] <anonimasu_> holy crap.. zloty better then sek -_-
[19:51:11] <anonimasu_> uh no :p
[19:51:11] <alex_joni> anonimasu_: lol
[19:51:15] <anonimasu_> yes..
[19:51:21] <anonimasu_> 1 Swedish krona = 0.33671313 Polish zloty
[19:51:46] <alex_joni> well.. two of the people I worked with from sweden, were actually polish
[19:53:01] <tomp> heh, trying to find a Linux compatible device programmer, so i dont have to boot M$ to show these guys
[19:58:48] <tomp> wow, that Polish unit is the least expensive i've found
[19:59:13] <tomp> i've seen/alex found
[20:04:50] <alex_joni> tomp: it was the first google hit
[20:05:13] <alex_joni> * alex_joni heads to bed.. been a long week
[20:05:22] <alex_joni> and it's not over yet
[20:06:57] <jepler> anonimasu_: geez why can't everyone just use metric?
[20:07:02] <jepler> how many mm in one zloty?
[20:08:50] <anonimasu_> jepler: 65535
[20:14:15] <robin_sz> everyone does use metric
[20:14:36] <archivist_ub> no we dont
[20:14:55] <archivist_ub> well not all the time
[20:15:34] <archivist_ub> * archivist_ub hugs the mm/in button on the guessing stick
[20:16:08] <robin_sz> can't say weve got anyting that works in inches ...
[20:16:13] <robin_sz> hmm
[20:16:17] <robin_sz> nope
[20:17:03] <archivist_ub> I have imperial jobs on metric machines and metric on imperial
[20:17:17] <archivist_ub> always swapping
[20:17:39] <robin_sz> we just dotn see imperial jobs at all really
[20:17:47] <robin_sz> a few, very few, enquiries
[20:17:53] <robin_sz> but we tend to decline them
[20:18:25] <archivist_ub> we do restoration work, so see old stuff
[20:18:40] <robin_sz> you just know if its in inches they wont have electronic drawings ...
[20:19:12] <archivist_ub> drawings!!!!, I get worn parts no drawings
[20:19:23] <robin_sz> heh, we run a mile at that point
[20:19:26] <archivist_ub> make it fit
[20:19:30] <robin_sz> mile ..
[20:19:33] <robin_sz> oh never mind
[20:19:48] <archivist_ub> send them here, we need more work
[20:20:16] <robin_sz> farmers are the worst
[20:20:26] <archivist_ub> hehe yes
[20:20:33] <robin_sz> "can you make someting to medn thei bag of rust?"
[20:20:45] <archivist_ub> gas axe
[20:21:01] <robin_sz> "no, not the whole thing, just this bit, dont want to spend more than I have to"
[20:21:38] <archivist_ub> dont damage the twine holding it together
[20:21:39] <robin_sz> its easier to just send them on their way
[20:22:05] <robin_sz> my favourite is "im sorry, we dont do welding" ..
[20:22:30] <robin_sz> always funny as there are 6 welding bays behind me with bright blue light coming out of each one
[20:22:39] <archivist_ub> hehe
[20:23:13] <archivist_ub> production welding and repair welding are in a different world
[20:23:20] <robin_sz> yes indeed
[20:23:50] <robin_sz> repair welding is best left to guys in garages
[20:23:51] <archivist_ub> I did production welding for a while
[20:24:08] <robin_sz> its a different skill
[20:24:24] <robin_sz> our guys are all coded
[20:24:36] <robin_sz> perfect welds first time everytime
[20:24:49] <robin_sz> you see them do a 2m long run sometimes in one hit
[20:24:50] <archivist_ub> yes have done own car repair, chasing paper
[20:25:17] <archivist_ub> thats a long run
[20:25:44] <archivist_ub> we had machines for long runs
[20:26:00] <robin_sz> best on watertight work to not stop and start
[20:26:15] <archivist_ub> boats?
[20:26:17] <robin_sz> has to be dye-pen tested afterwards
[20:26:42] <robin_sz> mmm ... coolant tanks and chip conveyor systems
[20:26:52] <archivist_ub> ah ok
[20:27:18] <robin_sz> we have a couple of small clients ;)
[20:38:40] <tomp> gnite alex
[20:44:22] <owad> My router is (finally!) all wired up and working. I need to make a cutting table, now, and am considering a few pieces of angle iron running the lenth, with a sheet of MDF bolted to it. Any opinions?
[20:46:22] <jepler> I originally had a plywood table 1/2" thick, and about 1' between the supported ends. I found it was too uneven and flexed too much for trace isolation milling, but it would probably be OK for many other tasks. I replaced it with an aluminum table, also 1/2" thick and it's much better.
[20:46:46] <jepler> obviously flexing/bowing will increase with the distance between supported points
[20:47:28] <jmkasunich> owad: so, how big is your machine?
[20:47:56] <owad> work area is 2' x 4'
[20:48:31] <jmkasunich> and what level of accuracy are you aiming for? router work, +/- 0.010" or so? or PC and milling, 0.002" or better?
[20:48:36] <owad> jepler: yeah, that's why I was thinking about the angle iron underneath\
[20:49:11] <owad> right now, just router work. Eventually I want to get better leadscrews and nuts and try some fancier stuff, but not yet
[20:49:46] <jmkasunich> holding 0.002" over a 2' x 4' table with anything other than serious amounts of steel would be tough, so thats a wise decision
[20:50:02] <owad> that'd be an awfully large circuit board :-)
[20:50:48] <owad> jepler: what sort of table do you have now?
[20:50:55] <jepler> owad: aluminum 1/2"
[20:51:27] <jepler> http://axis.unpy.net/files/01215190154/table-medium.jpg
[20:51:31] <jmkasunich> note that is work area is about 1' x 1' (maybe a bit longer and narrower)
[20:52:54] <tomp> good point ( whatever 'beam deflection' is for a plate )
[20:53:29] <jmkasunich> angle iron is far from the stiffest possible shape
[20:53:34] <owad> just ordinary bolts, typically, to hold the work down?
[20:53:41] <jmkasunich> square tube or I-beam shapes will be stiffer
[20:53:54] <owad> ok, I'll see if I can find that affordably
[20:54:44] <jmkasunich> what is the frame of your machine made from?
[20:54:51] <owad> MDF
[20:54:58] <jmkasunich> all of it?
[20:55:18] <jmkasunich> wow
[20:55:21] <tomp> what is the stuff zen-bot was made of?
[20:55:26] <owad> http://tinyurl.com/5ahdnm
[20:55:50] <jmkasunich> owad: so the X rails are on "I-beams" made of MDF?
[20:56:01] <owad> yes
[20:56:28] <jmkasunich> the table has to go over the top of that crosspiece, right?
[20:56:39] <owad> yes
[20:56:49] <owad> the table will have to be about 6' long, to fit over the machine
[20:57:00] <jmkasunich> and the crosspiece moves with the gantry, so the table can only be supported from the ends
[20:57:08] <owad> right
[20:57:11] <jmkasunich> ouch
[20:57:38] <jepler> tomp: the white plastic is HDPE if my old blog entry can be trusted
[20:57:42] <jmkasunich> to use MDF alone, you'd need two "I-beam" structures at least as tall as the ones holding the Y rails up
[20:57:45] <jmkasunich> X rails
[20:57:52] <tomp> stand it up so gravity doesnt bow the 6' ft table
[20:57:56] <tomp> jepler: thx
[20:57:57] <jmkasunich> which obviously won't work
[20:58:26] <owad> MDF on top of steel square tube?
[20:58:38] <jmkasunich> how much height do you have to work with?
[20:58:47] <owad> a good bit -- let me measure
[20:58:49] <jmkasunich> (total height of table and sub-structure/frame
[20:59:01] <jmkasunich> I'm not seeing that much
[20:59:18] <jmkasunich> from the ridge across the back of that crosspiece up to the tool is maybe 8"?
[20:59:45] <jmkasunich> I assume you want at least half of that for the workpiece and tooling
[21:00:56] <jmkasunich> a 6' long piece of 2" wide x 3" tall x 1/8" wall steel tube is $40.72 from metal express, weighs 22 lbs
[21:01:29] <jmkasunich> two of those running lengthwise would be a start, although you'd have sag inbetween
[21:01:43] <jmkasunich> maybe thats where you use the angle iron or smaller square tube
[21:02:43] <jmkasunich> for the long support members, taller is better (stiffer), they also have 2" x 4"
[21:02:55] <jmkasunich> both are available in a variety of wall thicknesses
[21:03:25] <owad> 4" for the permanent table (from the farthest the tool will extend down, to the top of the under carraige)
[21:03:38] <jmkasunich> did you do all the MDF cutting or was it a kit or something>?
[21:03:49] <owad> a kit
[21:04:25] <jmkasunich> so you don't have the tools to cut say 6' long 3" wide strips of MDF from a 4'x8' sheet?
[21:05:02] <owad> yes, I can do that. I have a table saw
[21:05:42] <jmkasunich> you might consider cutting a bunch (maybe 10) 3" wide strips, then gluing them on edge to the underside of another sheet of MDF
[21:06:14] <jmkasunich> 10 strips spaced 2.4" apart running lengthwise will provide reasonalby good rigidity in that direction
[21:06:20] <owad> you'd recommend that over angle iron? I guess that'd be cheaper
[21:06:36] <jmkasunich> adding another thinner sheet on the bottom would make it even stiffer
[21:06:44] <jmkasunich> I'm not recommending one over the other
[21:06:59] <jmkasunich> it depends on what you can work with, etc
[21:07:22] <jmkasunich> for example, if you can weld, I'm sure you can make a much stronger table out of steel tube, and just add an MDF top
[21:07:39] <jmkasunich> but if you don't have a welder, that gets hard to do
[21:07:42] <owad> no welder
[21:07:57] <owad> I can trill, tap, and bolt, though
[21:07:58] <jmkasunich> I have no idea what MDF costs
[21:08:07] <owad> pretty cheap
[21:08:29] <jmkasunich> does it glue well? is that thing glued and screwed, or just screwed together?
[21:08:40] <jmkasunich> wood screws (self tapping) or thru-bolts with nuts?
[21:09:00] <owad> screwed, with cross dowels
[21:09:16] <jmkasunich> how thick is the MDF?
[21:09:26] <owad> when I use wood screws, it tends to split
[21:09:35] <jmkasunich> ah, that would be a problem
[21:09:50] <owad> 3/4" thick
[21:10:02] <jmkasunich> I was thinking of a top surface made of 3/4" MDF
[21:10:25] <jmkasunich> on the bottom of that sheet, route 3/4" wide x 1/8 to 1/4 deep grooves the long way, about 8-10 of them
[21:10:44] <jmkasunich> glue 3" wide, 3/4" thick strips into the slots as ribs
[21:11:10] <jmkasunich> then if you can get say 3/8" MDF, route 1/8" deep grooves in that, and glue it to the ribs as a bottom sheet
[21:11:25] <owad> that sounds pretty sturdy
[21:11:30] <jmkasunich> result is a 4" thick "slab" that is sort of like corrugated cardboard
[21:12:14] <fenn> and then fill it with sand :)
[21:12:31] <owad> as far as mounting things to the table -- just drill holes in it, or is there a cheap way to do something approachign T-slots?
[21:12:35] <jmkasunich> fenn: one nice thing about MDF is that it is well damped compared to steel
[21:13:17] <jmkasunich> owad: I think I've seen an aluminum extrusion that was intended to be set into slots in MDF to make a t-slot
[21:13:38] <jmkasunich> personally I can't get too excited about that, but I'm a mill/metalcutting guy
[21:13:42] <jmkasunich> might be OK for wood
[21:13:50] <ds2> think woodcraft sells those
[21:14:30] <jmkasunich> hmm, can you get thinner MDF? like 1/2 or 3/8?
[21:14:57] <jmkasunich> the ribs don't have to be 3/4" thick, the strength comes from their height - if the ribs _and_ the bottom were 3/8" thick, it would be lighter
[21:15:01] <owad> found it. T-Track: http://www.woodcraft.com/family.aspx?familyid=3782
[21:15:06] <jmkasunich> the top needs to be thick of course
[21:15:24] <owad> I'm not sure what thicknesses I can get
[21:15:42] <jmkasunich> that t-slot stuff needs a slot over a half-inch deep
[21:15:44] <owad> what about plywood?
[21:15:55] <jmkasunich> 2/3 of the way thru your top - that would suck
[21:16:25] <jmkasunich> plywood for the top would suck unless it is very high-grade ply
[21:16:39] <fenn> if you ran the ribs the other way it wouldn't really matter
[21:16:42] <jmkasunich> I suppose you could use it on the bottom and for the ribs
[21:16:59] <jmkasunich> the ribs need to run lengthwise - the table is only supported at the ends
[21:17:13] <jmkasunich> you could run the t-slots crosswise I suppose
[21:17:44] <jmkasunich> looking at those t-slot extrusions, I'm about 99% sure they are NOT designed for clamping stuff down
[21:19:18] <jmkasunich> I suppose they could work if you drilled them every 6" or so, countersunk them, and screwed them down
[21:20:20] <jmkasunich> have to be very carefull using traditional clamping approaches tho - normal metalworking strap clamps pull up on the t-nuts and push down on the work and the step block, neither of which will be on top of the track
[21:20:43] <jmkasunich> the clamps they make for that track push down directly on the lips of the track, so they aren't trying to rip the track out of the table
[21:22:36] <owad> ok, I'm going to go to the hardware store and see what I can find
[21:22:43] <jmkasunich> owad - just a sec
[21:22:43] <owad> I really appreciate the help!
[21:22:46] <owad> ok
[21:23:01] <jmkasunich> go to mcmaster.com, and look at page 3211 in the catalog
[21:23:21] <jmkasunich> they have 1/4-20 threaded inserts that screw into drilled holes in particleboard (and probably MDF)
[21:23:29] <jmkasunich> 100 inserts for $12.51
[21:24:02] <jmkasunich> you could drill a grid of holes and run them in with an electric drill
[21:24:10] <owad> as a matter of fact, I already even have some of those
[21:24:32] <ds2> or just get the ones you pound in from the back
[21:24:33] <jmkasunich> a 2" x 2" grid over a 24" x 48" table would take 288 of them
[21:24:36] <owad> haven't tried them in MDF, but they work great in wood
[21:25:02] <jmkasunich> you could probably use the machine to drill the holes for you ;-)
[21:25:23] <owad> that's the only way they'd end up in a straight line :-)
[21:25:51] <jmkasunich> I'd be reluctant to use the ones you pound in from the back
[21:26:16] <ds2> why?
[21:26:34] <tomp> fwiw a 24 oz bottle of upsidedownium lets it sit above a simple static work table http://imagebin.org/index.php?page=add
[21:26:39] <jmkasunich> my experience is that the prongs are not very sharp, and rather wedge shaped
[21:27:04] <jmkasunich> in wood they're OK, but in particleboard or MDF, they displace material and cause crumbling or cracking - not very strong
[21:27:20] <ds2> ah but having a full flange is better then thread forming stuff onto MDF
[21:28:01] <jmkasunich> maybe - the ones I was looking at have "knife threads", not 60 degree threads
[21:28:31] <ds2> still threads in MDF just sounds scary
[21:28:50] <ds2> I much rather drill oversize and embed in epoxy
[21:28:54] <jmkasunich> I've seen the back-insert ones that had a round flange with holes for small nails (brads) - those might work nicely
[21:29:03] <jmkasunich> the formed metal prongs scare me tho
[21:29:14] <tomp> bbl/bebacknextweek
[21:29:33] <jmkasunich> in any case, I'd start with a scrap of MDF and several candidates, and try them, before buying nearly 300 pieces of anything
[21:29:55] <ds2> hahaha, unless you are doing a first article for production, then 300 peices ain't too bad ;)
[21:30:38] <jmkasunich> this is for hold-down holes on a router table - the 288 pieces _are_ the production run
[21:31:14] <ds2> now if you were manufacturing router tables....
[21:31:18] <owad> I've got plenty of scrap, so I'll give it all a try
[21:31:44] <jmkasunich> owad: I think threaded inserts will cost less and probably be stronger than the track thing
[21:32:02] <owad> the track is definitley out
[21:32:22] <owad> I'm concerned the threaded inserts will split the table, but we'll find out with some scrap pieces
[21:32:36] <ds2> of these MDF, nylon slab, plexiglass, or HDPE/UHMW - which one am I likely to find the cheapest for a 6x8x0.375 sheet?
[21:32:50] <jmkasunich> MDF by far I suspect
[21:32:59] <jmkasunich> 6 x 8 feet?
[21:33:03] <ds2> inch
[21:33:08] <jmkasunich> oh
[21:33:16] <ds2> making a base to display a circuit board
[21:33:23] <jmkasunich> in that case, you won't be able to buy MDF that small
[21:33:41] <ds2> *nod*
[21:33:55] <ds2> plexi cut offs used to be cheap but they are going up in price
[21:33:58] <owad> Lowes sells 2'x4' sheets and they cut it free, though
[21:34:14] <jmkasunich> 6x6x3/8 nylon at mcmaster is $9.40
[21:34:16] <ds2> HDPE/UHMW is basically a cutting board if I can find them cheap
[21:34:37] <ds2> owad: the local Lowe's panel saw is perpetually broken :(
[21:34:40] <robin_sz> * robin_sz prepares to fire up the servo drive
[21:34:42] <jmkasunich> 12x12 is $27
[21:34:56] <owad> I'm off to the hardware store. Thanks for the help, jmkasunich.
[21:34:59] <owad> Thanks jepler, ds2, fenn.
[21:35:12] <ds2> McMaster needs shipping unfortunately :(
[21:35:21] <jmkasunich> 12x12 acrylic (plexiglass) is more - $44
[21:35:25] <ds2> guess I am back to pondering a way to make this cheaper
[21:35:42] <ds2> I can get 12x12 locally for $20 cut rounded off
[21:35:51] <ds2> trying to see if I can do it for sub $5
[21:36:26] <jmkasunich> ah, that will rule out most plastics, unless you can get creative - like actually using a cutting board
[21:36:37] <jmkasunich> you only need the one piece, right?
[21:36:53] <ds2> yep, one time thing
[21:37:01] <jmkasunich> has to be 3/8" thick?
[21:37:05] <ds2> needs to be hard enough that a 4-40 nut won't dent it
[21:37:13] <ds2> nope, 1/4-1/2 is all good
[21:37:40] <ds2> even 0.75 if it is cheap enough, but that's starting to get heavy and would complicate getting 4-40 fasteners long enough
[21:37:51] <jmkasunich> what about thinner?
[21:38:09] <ds2> I don't want to go below 1/4 as it would flex too much as a display panel
[21:39:24] <jmkasunich> gotta be 8" long?
[21:39:33] <ds2> nope, that's flexible too
[21:39:40] <jmkasunich> 6x6?
[21:39:49] <ds2> I think 6x6 might work if I am careful with the board placement
[21:40:04] <jmkasunich> mcmaster has 6x6x1/2" ABS plastic, beige color, $6.11
[21:40:22] <jmkasunich> they have 12x12 too, but that is 4x the square inches, 4x the price
[21:40:23] <ds2> * ds2 looks
[21:40:37] <ds2> still over my target price but it is getting close
[21:40:53] <jmkasunich> they have a lot of things in 6x6 that aren't too pricy, but as soon as you go to 12x12 it gets high
[21:41:11] <jmkasunich> ABS is probably among the cheapest
[21:41:29] <ds2> I really would like to get it under $5... it is partially a volunteer effort
[21:41:46] <jmkasunich> 6x6x1/2 PVC, gray, $5.54
[21:42:00] <jmkasunich> but as you say, shipping is significant on small orders like this
[21:42:07] <ds2> well, there is shipping on top of them
[21:42:29] <ds2> is there anything common like a cutting board that would be a minisheet ?
[21:42:58] <jmkasunich> if shipping is an issue, we really can't help you - you gotta go to a local store/supplier and see what they have
[21:43:22] <ds2> I was more of looking for suggestions as to things that are flat and roughly that dimension that I can cut up
[21:43:38] <lerman> wood?
[21:43:44] <jmkasunich> cutting boards, plastic storm windows
[21:43:54] <jmkasunich> (dunno if clear is OK)
[21:43:56] <ds2> plastic storm windows? /me googles
[21:44:03] <ds2> I got a can of white spray paint!
[21:44:17] <jmkasunich> why not wood?
[21:44:18] <robin_sz> noooo
[21:44:26] <robin_sz> you cant cut spray cans!
[21:44:29] <ds2> grain
[21:44:30] <robin_sz> think of the mess!
[21:44:50] <ds2> grain I think would be objectionable asthetically
[21:45:00] <ds2> robin_sz: that's what water jet is for ;)
[21:45:07] <robin_sz> ah yes
[21:45:12] <robin_sz> * robin_sz adds that to his list
[21:45:25] <robin_sz> i looked at one this year
[21:45:30] <robin_sz> but decided agaisnt it
[21:46:09] <robin_sz> bought a 2nd laser instead
[21:46:26] <ds2> i wish i had a laser
[21:46:34] <ds2> a little 10watter would suffice
[21:46:37] <jmkasunich> ds2: did you say you have a waterjet?
[21:46:45] <ds2> jmkasunich: no!
[21:46:49] <robin_sz> ds2: mine are 2kw ish
[21:47:03] <ds2> unless you count a water piston used to cut patterns in dry sand ;)
[21:47:07] <jmkasunich> darn
[21:47:07] <anonimasu_> waterjets are f-expensive.
[21:47:09] <ds2> pistol
[21:47:20] <robin_sz> yeah, and waterjet is slow
[21:47:23] <jmkasunich> * jmkasunich looks at 12x12 solid white ceramic floor tile - home depot, $1.09
[21:47:38] <anonimasu_> but well, waterjets cut alot of stuff lasers dont :)
[21:47:41] <robin_sz> the one I looked at as a 4.5axis
[21:47:43] <jmkasunich> like tile
[21:47:45] <robin_sz> 100hp
[21:47:47] <anonimasu_> yeah
[21:47:49] <ds2> robin_sz: different applications... I am happy just to mark stuff
[21:47:58] <robin_sz> laser cuts ceramic tile ok
[21:48:06] <jmkasunich> he doesn't have a laser either
[21:48:12] <robin_sz> poor sod
[21:48:32] <robin_sz> lets club together and buy him onw
[21:48:33] <jmkasunich> ds2: would 12x12 be too big to use as-is?
[21:49:09] <ds2> jmkasunich: yeah, I need to fit it in a box that gets carried on public transportation
[21:49:27] <ds2> I do have a tile saw... Hmmmm
[21:49:56] <jmkasunich> oh, they have 6x6 ones too
[21:49:58] <lerman> just make sure it includes a lantern battery, some wire, a short length of boomstick, and an alarm clock.
[21:50:05] <robin_sz> sub it out for waterjet if you have to
[21:50:28] <lerman> that's always welcome on public transportation :-)
[21:50:32] <jmkasunich> robin_sz: did you miss the part where this is a charity job with a target price of $5 or less
[21:50:37] <robin_sz> yes
[21:50:42] <ds2> lerman: haahahaha
[21:50:45] <anonimasu_> robin: ^_^
[21:51:15] <jmkasunich> ds2: I'd say, get thee to the nearest home depot or similar and look around with an open mind
[21:51:17] <lerman> Multiple layers of thin cardboard, glue, varnish. Make your own plywood.
[21:51:25] <jmkasunich> the 6x6 tiles are $0.42 each
[21:51:58] <ds2> jmkaunich: the killer there is I don't have suitable diamond/carbide drills for the screw holes... regardless, I am planning to visit several dollar stores this weekend
[21:52:13] <jmkasunich> could you glue instead of screwing?
[21:52:27] <jmkasunich> bath tub caulk should stick PCB to tile
[21:52:40] <ds2> no, the PCB needs to come off
[21:52:40] <anonimasu_> lol
[21:52:49] <lerman> I have a mcmaster catalog here. Glue a bunch of pages together with the front and rear covers, and you should be all set.
[21:53:11] <ds2> hahah a used mcmaster catalog is worth more then $5!
[21:53:28] <ds2> I can probally sell that, buy lunch AND some new materials and pay for someone to do it :)
[21:53:29] <lerman> Will aluminum work? I could mail you a 6x8 piece of aluminum sheet.
[21:53:33] <ds2> and still ahve change
[21:53:37] <jmkasunich> what about tempered hardboard, aka masonite
[21:53:51] <robin_sz> mdf ?
[21:53:54] <ds2> hardboard is good, just need to find drops/cut off sources
[21:54:01] <jmkasunich> $7.88 for a full 4x8 sheet, I think they may sell half and quarter sheets
[21:54:27] <lerman> Damn, with all the time we've wasted discussing this, we could all chip in and buy you a block of gold to bolt this too.
[21:54:39] <ds2> hahahahahahahah
[21:54:46] <ds2> anyways, thanks for the ideas
[21:54:54] <jmkasunich> well, if any of us was getting paid for that time
[21:55:00] <lerman> that bolt this 'to'
[21:55:05] <jmkasunich> you're welcome, and good luck
[21:55:27] <ds2> figure it can't hurt to get ideas off other scrongers ;)
[21:55:56] <jmkasunich> the problem with asking scroungers is that we don't know what is in your junkpile
[21:56:05] <pminmo> I just cut a front panel with 1/8" hardboard
[21:56:13] <jmkasunich> (I could probably find a piece of 6x8 lexan or similar in my junkpile)
[21:56:41] <ds2> I go into dollar stores with the idea that everything is raw materials for other projects
[21:57:07] <pminmo> can I ask a quick question?
[21:57:14] <lerman> you just did.
[21:57:35] <jmkasunich> we'll allow you to ask another one though
[21:58:01] <pminmo> ok, the 2nd question is where is the filter for for file extensions? I want to ad .nc to the NC file open in axis
[21:58:19] <lerman> Look in the .ini file.
[21:59:10] <pminmo> i did it has the directory info and the documetation says filters for porcessing
[21:59:19] <pminmo> processing,
[21:59:31] <pminmo> i.e. graphics to ...
[22:00:42] <lerman> It should be self explanatory to add another one.
[22:00:54] <jmkasunich> I think he's actually asking another question
[22:01:24] <jmkasunich> not "how do I run a file thru a filter", but "how do I make the file-open dialog display files with .nc extensions in additon to .ngc"
[22:01:31] <lerman> Yup. He wants .nc to look like ngc.
[22:01:36] <pminmo> yes
[22:01:58] <jmkasunich> I suppose you could try "cat" as a filter for .nc
[22:02:02] <pminmo> i want to see .nc, .ngc .tap files
[22:02:21] <jmkasunich> I don't know how to change the settings for the file open dialog
[22:03:11] <jmkasunich> axis has the ability to use external programs as filters in the unix sense (stuff a file in one end, and a modified file comes out the other) which is what I initially thought you were asking about
[22:03:39] <pminmo> no, I found that in the documentation but it just the load dialog
[22:03:41] <jmkasunich> you can specify filters for various extensions - "if the user opens a .foo file, run it thru filter bar"
[22:04:06] <lerman> John, I think your 'cat' solution will work.
[22:04:11] <jmkasunich> if you have defined a filter for .nc does that make .nc appear in the file open dialog?
[22:04:23] <jmkasunich> if it doesn, then you can define "cat" as the filter for .nc
[22:04:24] <lerman> yes. I believe so.
[22:05:02] <pminmo> I didn't try it as I assumed it would want the following lines to tell it what to invoke
[22:05:23] <jmkasunich> have it invoke cat (which simply echos the original file with no changes)
[22:05:23] <pminmo> I thought it was a quick question, sorry
[22:05:33] <lerman> See the function open_file in axis.py. It adds some extensions to the tuple 'all_extensions'.
[22:06:01] <lerman> As an alternative, just change that function to create the tuple containing ".ngc" and ".nc".
[22:06:10] <pminmo> what directory path
[22:06:21] <pminmo> for axis.py
[22:07:14] <lerman> I wouldn't be surprised if you just add the extension to PROGRAM_EXTENSION in the ini file and did NOT at a line nc = cat, if it did just what you want.
[22:07:46] <lerman> emc2/src/emc/usr_intf/axis/scripts/
[22:08:50] <pminmo> ok I'll go try some things thanks
[22:09:18] <lerman> Looking at the source, just add the PROGRAM_EXTENSIONS = .nc line and try it. Should work.
[22:37:32] <pminmo> fyi adding a [FILTER] section and a PROGRAM_EXTENSIONS = .nc didn't work it recognized PRGROGAM_EXTENSIONS i think because if I changed to PROGRAM_EXTENSION got all kinds of error messages
[22:38:09] <pminmo> thanks anyway
[22:41:49] <robin> amazing
[22:41:50] <robin> emc
[22:41:52] <robin> encoders
[22:42:00] <robin> * robin giggles
[22:59:51] <robin> strange .. when I connect the drive and tie enable high .. the motor oscilates about 20mm +- on the slide
[23:00:09] <robin> even without the +-10 signal connected
[23:02:44] <robin> most odd
[23:16:58] <archivist> archivist is now known as archivist_emc
[23:17:38] <archivist_emc> archivist_emc is now known as archivist
[23:23:40] <archivist_ub> archivist_ub is now known as archivist_emc
[23:24:42] <archivist_emc> archivist_emc is now known as archivist_ub
[23:29:18] <fenn> ds2: go to a sign maker's shop and ask for any cutoffs/trash, or failing that a cabinet maker
[23:30:02] <fenn> the problem with scrounging is that just a little bit of junk isn't enough, you need a whole pile for it to be useful
[23:31:33] <archivist_ub> and once people know you are a repository of old junk...... it keeps coming
[23:39:49] <robin_sz> i think this DC drive amp must have some additional crap in it
[23:40:01] <robin_sz> for armature feedback or some such
[23:40:11] <robin_sz> its got various contants that can be set
[23:41:42] <robin_sz> surely with PID in EMC, I dont need anyting else on the drive do I?
[23:51:08] <ds2> I forgot about sign makers