#emc | Logs for 2010-09-10

[00:04:08] <theorb> theorb is now known as theorbtwo
[00:31:59] <mike> mike is now known as Guest63860
[03:19:43] <cradek> ichudov_: http://timeguy.com/cradek-files/emc/truetype-tracer-and-heeks-pocketing.png
[03:20:31] <Jymmm> cradek: Can you give a 3D (isometric) view?
[03:21:55] <cradek> not much to see: http://timeguy.com/cradek-files/emc/truetype-tracer-and-heeks-pocketing-perspective.png
[03:22:31] <Jymmm> you took the output of TTT and ....
[03:22:59] <cradek> ttt generates dxf - import dxf in heekscad - draw a rectangle around it - tell heeks to pocket it - post and save gcode
[03:23:18] <cradek> the paths are excellent - arcs where arcs should be
[03:23:21] <Jymmm> Oh, I thought TTT generated gcode, not DXF
[03:23:27] <cradek> it will do either
[03:23:35] <Jymmm> ah
[03:23:39] <Jymmm> didn't know that
[03:23:42] <cradek> it can generate dxf with nice smooth biarcs for the splines
[03:24:20] <Jymmm> I asked for the perspective because I couldn't see any depth
[03:24:57] <cradek> yeah the cuts are all one depth - the idea is you'd leave just the letters sticking up
[03:25:21] <Jymmm> I dont see any z moves in the gcoe either
[03:25:41] <cradek> look at the ends of the cyan lines (rapids)
[03:25:57] <cradek> it's a feature that it doesn't have to move up and down much to do the pocket
[03:26:15] <Jymmm> OH! Very shallow
[03:26:34] <Jymmm> very cool =)
[03:27:03] <cradek> yeah it's at -0.25 I think
[03:29:32] <Jymmm> Hmmm, I wonder if asking for a bevel aorund the profile of each char is asking too much?
[03:29:59] <cradek> with a chamfer tool? you can get that in heeks
[03:30:13] <Jymmm> In aluminum, it's no big deal. In wood, it helps prevent chipout
[03:30:59] <Jymmm> Or just a V-Groove bit
[03:31:00] <cradek> in that situation you could draw first *on the path* with a pointy tool, and then do this pocketing with a regular cutter
[03:31:11] <cradek> yep
[03:31:21] <cradek> would work fine
[03:32:08] <Jymmm> who devedoloped heeks?
[03:33:07] <cradek> I don't know the history
[03:33:38] <Jymmm> Is it purely CAD, or CAD/CAM ?
[03:34:04] <cradek> more cam than cad, imo
[03:34:22] <Jymmm> Oh danielfalck wrote it
[03:34:39] <Jymmm> ooops, sorry
[03:34:44] <Jymmm> Daniel Heeks
[03:34:52] <cradek> et al
[03:35:21] <Jymmm> http://heekscnc.blogspot.com/
[03:37:17] <Jymmm> Oh, it looks like you and seb have been commiting to it =)
[03:37:39] <cradek> yes I've fixed a few things lately
[03:38:11] <cradek> but it's funny - I haven't even talked to Dan H yet
[03:38:34] <Jymmm> Heh, doens't mena he doens't know who you are though =)
[03:38:52] <cradek> uh-oh, hope I don't have a reputation
[03:39:32] <Jymmm> cradek: Yep, you do.... More of an 'Olsen Twins' than a 'Paris Hilton' one though
[03:45:48] <pcw_home> Had an IGBT explode in my face today, that was kind of exciting
[03:46:05] <cradek> yikes. do you wear glasses?
[03:47:14] <pcw_home> Yes I do, not too many big pieces flying but nice bang and flash
[03:49:25] <pcw_home> Fellow engineer got the clever idea of using as desk light to test the brake threshold voltage on some 8I20s we were testing
[03:49:27] <pcw_home> worked fine when we checked the threshold (100 and 120V), not so fine with 360V VBUS
[03:50:55] <cradek> they don't fail just a little bit, do they
[03:51:13] <pcw_home> No, kinda sudden...
[03:58:50] <pcw_home> It was a triple banger first the brake IGBT failed (after a really bright flash from the desk light), then a cable blew at a spade lug crimp
[03:58:51] <pcw_home> dumping the motor energy who knows where (looks like the card VBUS got to ~1000V before the Hbridge module shorted out)
[03:58:53] <pcw_home> nothing on the low side (serial interface, FPGA card, PC) was damaged so thats a good thing
[03:59:31] <cradek> wow, that seems surprising
[04:00:06] <cradek> goodnight, hope you avoid explosions tomorrow
[04:01:27] <pcw_home> Nite, we are now putting a plastic shield over HV stuff we test :-)
[04:01:49] <cradek> better go with polycarbonate :-)
[04:05:00] <pcw_home> Yep and review test procedures...
[04:10:34] <Jymmm> I built a PS in electronics class, just got finished soldering up every thing and was bench testing it. I noticed the voltage was dropping consistantly. I left to grab the instructor and before had a chance to come back I heard a BOOM!
[04:11:13] <Jymmm> The big old cap exploded with one guy sitting right next to it.
[04:11:20] <Jymmm> poor bastard.
[04:15:12] <pcw_home> Where I used to work years (and years) ago a radial capacitor blew up in the face of a test technician who proceeded to chase and threaten grave bodily harm to the
[04:15:14] <pcw_home> assembler person who and hand assembled the card. (this was in the days of through hole parts, hand poked into PCB sitting on foam)
[04:15:48] <Valen> had the odd fet catch fiew
[04:15:50] <Valen> fire
[04:15:58] <Valen> Fire Emitting Transistor they be called
[04:16:14] <Valen> they actually seem to work for a while when thy are on fire
[04:18:06] <Valen> best explosion I've had was a tantalum cap
[04:18:14] <Valen> was making a prototype
[04:18:31] <Valen> and it had a short somewhere, I couldn't work out what was causing the problem though
[04:19:09] <Valen> eventually dad had a solution, he says "its not pretty but it'll work", take all the expensive chips off the board and hook the 5v rail direct to a car battery
[04:19:21] <Valen> the thing that explodes is the short
[04:19:48] <Valen> thankfully when i was just about to do it he said to turn the board over just in case
[04:20:04] <Valen> thing cooked off like a firework red hot crap blasting across the desk
[04:20:19] <Valen> burnt a bunch of quater inch holes into the surface of the desk
[04:20:37] <pcw_home> Tantalums are responsible for a fair number of holes burned clean through PCBs...
[04:22:15] <pcw_home> You can do that test a little more gently with a lab supply (set to say 1V and 10A current limit)
[04:22:17] <pcw_home> then just feel around for the hot part (or use a voltmeter and follow the drop_
[04:23:13] <Valen> it was a very good short, it was sucking around 1.5A without getting warm
[04:23:20] <EricKeller> I guess nobody from Italy would be on her now
[04:23:24] <EricKeller> here
[04:24:30] <EricKeller> is axis called axis in the process list?
[04:26:32] <EricKeller> you guys are going to make me go all the way downstairs and crank up a computer just so my answer to the Italian guy about axis starting in the background is correct?
[04:27:46] <Valen> i would have to drive for 20 minutes
[04:32:12] <EricKeller> a likely story
[04:33:35] <Valen> well it should be
[04:33:51] <Valen> if your feeling lazy you could download it and fire it up inside a VM
[04:39:43] <EricKeller> I have a sim version, but apparently the axis crashes
[04:40:00] <EricKeller> I can get the tk interface to run
[04:40:45] <EricKeller> probably have to build it again, too lazy
[04:41:37] <Valen> good excuse to try out 10.04
[04:41:52] <EricKeller> this is my desktop running 10.04
[04:42:06] <EricKeller> maybe I should try to figure out how to install a vm
[04:42:07] <Valen> I C
[04:42:17] <Valen> vmware player is pretty simple
[04:42:23] <Valen> kvm is open source
[04:42:31] <Valen> (and built in)
[04:42:32] <EricKeller> the thought of vmware player gives me hives
[04:42:40] <Valen> eh "it works"
[04:43:07] <EricKeller> "it didn't last time I tried and I gave myself nightmares trying"
[04:43:08] <morfic> i like virtualbox, but use it more at work on windows host and plubuntu guest
[04:45:01] <morfic> just don't install from the ISO, install plain ubuntu and build emc2 (SIM performance was better compared to 8.04 installed into VM), don't think i have space or time to try 10.04 iso into vm
[04:45:36] <Valen> morfic whats wrong with the livecd?
[04:45:59] <morfic> nothing
[04:46:44] <Valen> I installed it into a vmware player instance and it seemed to work ok, whined about RT-delays when starting the sim but eh
[04:46:49] <morfic> just a run in place sim ran better than trying to use what you find in VM after installing from the old iso
[04:47:18] <EricKeller> sim whining about rt-delays makes me wonder
[04:47:19] <morfic> yeah, those dealys i could ignore if OS was usable, which it was not
[04:47:34] <morfic> EricKeller: in a VM, i'd expect ot have BAD RT :)
[04:47:56] <Valen> why is the sim using RT anyway?
[04:47:57] <EricKeller> bad RT, bad!
[04:48:08] <morfic> Valen: clicking menu to open some app and it taking seconds per menu that pops up is what i refer to
[04:49:29] <Valen> didn't see that on my install
[04:49:38] <morfic> Valen: all i know is following suggestions from here, i installed 10.04, built sim as run in place and i can now work on it fine, just didn't want to send EricKeller off to try VB w/o pointing out that if 10.04 iso gives similar results, he has a good option of plain 10.04 + building emc to run in place
[04:49:59] <morfic> Valen: i suppose your desktop and my desktop are not likely the same ;P
[04:50:30] <Valen> this was on an old laptop
[04:50:39] <Valen> dual core ~2ghz i think
[04:50:45] <Valen> vista host
[04:50:56] <Valen> vmware player 3
[04:50:56] <morfic> what kind of dual core? intel core1?
[04:51:20] <EricKeller> downloading the iso is no big deal, I was going to install it anyway
[04:51:26] <Valen> dunno, i *think* it might be core2 but it is a few years old
[04:51:37] <Valen> its dads machine
[04:51:51] <Valen> he wanted something so he could test his mastercam outputs on
[04:52:38] <morfic> Valen: single P4 era Xeon+HT, 3.x GB of ram (what ever it is XP gets you, think it is 3.2, while i thought PAE should get you 3.6 in XP?) and i can't shut down other apps while VB is running
[04:53:26] <EricKeller> I've never su'd to my own account before
[04:53:27] <Valen> PAE lets you get full memory, but no process can hit more than ~2gb
[04:53:29] <morfic> the emc2 is for that? why does he have mastercam if he built a machine just to test the code? usually people have the machine or need to use mastercam with.for
[04:54:23] <Valen> i stuck an EMC virtual machine onto his desktop so he can make the toolpath in mastercam then stick it onto the VM, open it in EMC and make sure it runs ok
[04:55:58] <Valen> we are still learning mastercam
[04:56:06] <morfic> Valen: you can't get 4gb, and i think 3.2 is showing something on this machine is wrong, *IF* i am right PAE is typically around 3.6GB, with a 4GB address space and other things than main memory to address, you could not get all of 4GB ram
[04:56:24] <Valen> oh if you happen to have a postprocessor tuned for EMC it'd be really handy ;->
[04:56:47] <morfic> i wish i had the cash to buy mastercam for my garage workshop :)
[04:56:50] <Valen> I dunno, I havent bothered too much with XP and 4gb ram
[05:02:27] <EricKeller> mastercam and solidworks would be nice :)
[05:02:34] <EricKeller> I could rule the world
[05:03:18] <Valen> if you have solidworks why not use solidcam?
[05:03:32] <EricKeller> is it any good?
[05:04:06] <EricKeller> mastercam easily takes solidworks
[05:04:39] <EricKeller> anyway, I don't have $20-30k for software, so world domination is right out
[05:07:01] <morfic> keep on eye on cradek and heeksCA[D|M]
[05:07:40] <morfic> most cam is happy if you get it a nice solid in one of many formats, there does not need to be a real relationship between both apps
[05:08:16] <morfic> CAD Guy uses Inventor, i take it, smash it or slice it and put toolpaths on, and if it's an acme thread, i may forget to change infeed from 60deg to 29deg :/
[05:08:59] <EricKeller> the cam software companies have worked very hard to make the connection easy
[05:11:00] <Valen> looking at what I use in rhino, the actual CAD software itself if you start off from a 3D base shouldn't be that hard
[05:11:10] <Valen> the maths is beyond my level though
[05:11:28] <Valen> (by hard, i mean there isn't a great need for lots of features)
[05:11:45] <Valen> its mostly extruding curves, capping and subtracting
[05:19:41] <morfic> this is the last time i mentioned the install iso btw, always turnes into way too long a talk after, and i thinks it's always valen asking me too :P
[05:20:24] <Valen> lol
[05:20:49] <Valen> well its out now so its all good as far as i'm concerned
[05:21:49] <EricKeller> what's the name of 10.04?
[05:22:14] <Valen> lucid
[05:22:39] <EricKeller> if Hardy is capitalized, is lucid?
[05:22:41] <morfic> EricKeller: cam software companies working hard? the Esprit guys did not get the memo (talking what i consider lack of features, imports are fine, but every time i touch a mfc/.net mix app i get squeezy) i understand a need to change over eventually, but do not take me from 100%/0% mfc to 90/10. 80/20, 60/40 and so on, too much UI stuff moving around, option pulled off one tab and put on another, and something that changes the code is not accessible beca
[05:22:41] <morfic> use it is on the "roughing tab" which is disabled when you work on a finish op on the same tab
[05:23:14] <EricKeller> morfic: CAM companies other than mastercam seem to be mostly con artists
[05:23:16] <morfic> all proper nouns are, no? except in what morfic types, he does use caps more casually
[05:23:22] <EricKeller> and I'm not sure about the guys from mastercam
[05:23:45] <morfic> EricKeller: i loved VirtualGibbs, it has better stuff years ago than Esprit does now
[05:24:32] <EricKeller> I'm not really a consumer of Esprit, the machinist at work has mastercam, it works for what we do
[05:24:56] <morfic> Mastercam i am sure appeals to the CAD guy who was told to go learn how to program the new CNC machine, Gibbs back then felt like it was to appeal to the machinist who was supposed to program more efficiently than in a text editor, stuff there used to make sense to me
[05:25:28] <EricKeller> I've seen more rants about Gibbs than any other company
[05:25:30] <morfic> if Esprit comes up, find out who said it, look him/her in the eye, and slap, HARD
[05:25:59] <morfic> EricKeller: i have not used it in years, if it has changed this much it would be sad, then i have nothing left to hope for
[05:26:42] <morfic> surfcam gave out cds to install demos for learning.testing, i should try to find if gibbs does this
[05:26:47] <morfic> you got me curious now
[05:27:51] <EricKeller> vm of the liveCD running finally
[05:30:51] <EricKeller> that was a very long-winded way of finding out that the process name of Axis is, indeed, "axis"
[05:31:31] <EricKeller> I got two errors about real time delay, really makes me wonder how that is generated
[05:32:38] <morfic> that's all you wanted?
[05:33:01] <EricKeller> yeah, nobody would answer the question and I was too lazy to walk down to the basement
[05:33:04] <EricKeller> :)
[05:33:15] <morfic> isn't it booting the RT kernel? then even sim is relaying a kernel message i bet
[05:33:24] <morfic> does dmesg have the same delay message?
[05:33:55] <morfic> EricKeller: then make it worthwile, is it running well at least? not like you are on a old P60?
[05:34:12] <EricKeller> no, it's a pig
[05:34:46] <EricKeller> you have to very carefully move the mouse
[05:35:31] <morfic> if you plan to use this again, kill it, install plain ubuntu and build it yourself there, based on wiki instructions
[05:35:54] <EricKeller> I didn't install the livecd, just running a vm from iso
[05:35:56] <morfic> EricKeller: thanks for keeping me from beeing the only one with a dog slow sim off iso
[05:36:31] <EricKeller> the livecd is a real time kernel, so I'm not sure how that plays with a sim config
[05:36:43] <morfic> oh that should be even better, i tried to avoid the squashfs off the "CD"
[05:37:04] <morfic> i think you have the answer there :)
[05:38:22] <EricKeller> I wanted to see how a vm worked anyway, I've used them, but only on the server at work
[05:38:38] <EricKeller> which I wasn't responsible for, but I didn't want the guy that was responsible for it to get fired
[05:39:45] <EricKeller> system administration in the real world of bosses that don't want to pay for real IT people
[05:44:08] <Valen> EricKeller: your not using vmware either are you?
[05:44:17] <EricKeller> no, kvm
[05:44:31] <Valen> which has the same heart as virtualbox
[05:44:36] <Valen> I wonder if thats the issue
[05:45:28] <EricKeller> I could try vmware player and see, but I'm too lazy
[05:45:56] <EricKeller> kvm was a couple of apt-gets away
[05:46:10] <Valen> vmplayer wasn't too hard
[05:46:22] <Valen> hardest part is getting through their download page lol
[05:46:29] <EricKeller> yeah
[05:46:46] <EricKeller> I was also trying to virtualize an installed version of windows
[05:47:02] <EricKeller> which it seemed like it would do, but it got all strange
[05:47:03] <Valen> you need to do work on that before you try the move
[05:47:18] <Valen> change the ide drivers
[05:51:55] <morfic> if you have drivers available it should do a couple "woah, hw changed, installing stuff now" else it freaks out
[05:53:30] <Valen> if you don't set the ide back to "standard ide" then shutdown, then move it into the VM it'll barf on inaccessable boot device generally
[06:00:37] <Jymmm> What is the name of those clamps you would install on an enclosure to give it a tight seal?
[06:01:01] <EricKeller> clecos?
[06:01:58] <EricKeller> clamps for sheet metal work, or permanent installation?
[06:02:18] <Jymmm> like you see on band gear
[06:02:37] <EricKeller> butterfly clasps?
[06:02:42] <Jymmm> or test equipment containers.
[06:03:07] <EricKeller> casket latches?
[06:03:15] <EricKeller> draw latch?
[06:03:21] <EricKeller> work with me here
[06:03:33] <Jymmm> looking for pic
[06:03:46] <EricKeller> look up draw latch on mcmaster
[06:04:37] <Jymmm> sorta like this http://www.amazon.com/Closed-Length-Capacity-Southco-Latches/dp/B002IB9QR2/ref=sr_1_9?s=industrial&ie=UTF8&qid=1284098657&sr=1-9
[06:05:03] <EricKeller> that's what I was calling a butterfly latch
[06:05:27] <Jymmm> but nothing came up for that on google
[06:06:29] <EricKeller> mcmaster calls them turn-to-open draw latches
[06:06:43] <EricKeller> page 2951
[06:07:56] <EricKeller> I've had one of the flush mounted ones like they use on music cases for the last 20 years and haven't figured out a good use for it yet
[06:10:05] <Jymmm> I was hoping that I could use it on a 3/4" thick front panel of MDF for hte air scrubber I'm making, but I'm not sure If I can mount the hook or not
[08:12:41] <Valen> mcmaster doesnt want my money
[08:12:50] <Valen> its cos i'm an evil furrrener
[08:26:07] <Endeavour> Anyone here ever use grbl?
[13:22:04] <Jymmm> Heh, SolidWorks 2011 even has 'Environmental Impact Analysis'. But these ppl are so full of themselves, the student edition is limited to 12 months.
[13:22:59] <Jymmm> JT-Work: Yo
[13:23:08] <JT-Work> yo
[13:23:28] <Jymmm> JT-Work: Heh, SolidWorks 2011 even has 'Environmental Impact Analysis'. But these ppl are so full of themselves, the student edition is limited to 12 months.
[13:23:46] <JT-Work> just buy a copy like I did
[13:24:16] <Jymmm> JT-Work: Insert Credit Card here: [ ]
[13:24:33] <Jymmm> JT-Work: what version and how much?
[13:25:00] <skunkworks> does this spell out emc2-axis? http://www.youtube.com/watch?v=EwfhjVV7D7Q
[13:26:16] <skunkworks> ^ that is pretty nice work!
[13:28:05] <Jymmm> skunkworks: I don't think it does.
[13:28:46] <Jymmm> There wee accents that didn't look commonplace.
[13:28:52] <Jymmm> s/wee/were/
[13:32:48] <Jymmm> Does HeehsCAD support assemblies?
[13:40:23] <JT-Work> Jymmm: I'm using 2010 but when 2011 comes out I get upgrades (maintenance contract)
[13:40:57] <Jymmm> JT-Work: Ok, cool. When you get 2011, you can sell me 2010 for $1
[13:41:16] <Jymmm> JT-Work: 2010 wasn't a very good year
[13:44:16] <JT-Work> yea, I had to do quarterly payments for the maintenance contract in 2009/10 to keep it
[13:44:52] <Jymmm> JT-Work: Did they let you keep at least one arm?
[13:45:33] <skunkworks> we do that here at work.
[13:45:41] <skunkworks> I think we have 2 seats
[13:46:12] <JT-Work> just my mouse hand
[13:46:51] <JT-Work> heh 5.1 is out
[13:47:06] <JT-Work> sp5.1
[13:55:01] <JT-Work> skunkworks: do you generate any cam from SW?
[13:56:14] <Jymmm> JT-Work: cradek just did using his TTT to create the DXF file brought into HeeksCAD to generate the gcode
[13:57:09] <Jymmm> http://timeguy.com/cradek-files/emc/truetype-tracer-and-heeks-pocketing.png
[13:57:18] <Jymmm> http://timeguy.com/cradek-files/emc/truetype-tracer-and-heeks-pocketing-perspective.png
[14:00:37] <JT-Work> LOL that is a better path than I get from OneCNC
[14:00:55] <cradek> the pocketing really seems quite good
[14:01:27] <cradek> the arcs are arcs
[14:02:21] <Jymmm> Is Alibre PE work it?
[14:02:25] <Jymmm> worth
[14:14:58] <JT-Work> I need to try and get HeeksCad to work for me again
[14:27:57] <Jymmm> Fuuuuuuuuck... Never thought it would be so difficult just to find 4" smooth wall duct hose with wire reinforcement that doesn't produce ridiges when bent
[14:28:41] <Jymmm> Found what I want on the saws at HD. Just got off the phone wiht the service company and they no longer carry THAT hose *sigh*
[14:45:26] <Jymmm> Oh great... the isolation mount I want is not in stock.
[14:45:37] <Jymmm> and is uncommon *sigh*
[15:58:11] <skunkworks> JT-Work: (it isn't my department - but they bought bobcam. They have used it about twice so far ;))
[15:58:46] <JT-Work> I feel sorry for them, that is what I used at first with the Anilam
[15:59:17] <Jymmm> skunkworks: Ew
[16:00:15] <Jymmm> skunkworks: They are probably one of "THOSE" ppl that read cnc-zone too =)
[16:01:47] <tom3p> anilam animal
[16:02:22] <JT-Work> I can run it with my eyes closed now
[16:11:28] <Jymmm> http://tmux.sourceforge.net/
[16:16:46] <jepler> meanwhile, I hacked pcb2gcode to convert its "stairstepped" gcode with short moves into larger primitives, including arcs. http://emergent.unpy.net/files/sandbox/pcb2gcode-with-arcs-preview.png
[16:17:38] <cradek> jepler: my zoomy wheel doesn't work in your screenshot
[16:18:04] <jepler> try this instead: http://emergent.unpy.net/files/sandbox/pcb2gcode-with-arcs.ngc
[16:18:13] <cradek> looks good, but did those start out round?
[16:18:20] <cradek> not that it matters much
[16:18:40] <jepler> yes, that was a round pad in the gerber file
[16:19:37] <EricKeller> the error cascade when I try to build heekscnc from svn is so long I can't tell what happened
[16:20:37] <elmo40> can someone paste that ubuntu heeks svn link again? i lost it when i closed the chat.
[16:21:17] <jepler> it uses a recursive algorithm based on Ramer-Douglas-Peucker but after having found the "most outlying" point it can either fit a line or an arc to it; it was mostly a translation from my existing Python code in rs274/author.py.
[16:21:42] <EricKeller> elmo40, you mean the build instructions?
[16:22:03] <elmo40> no. someone pasted a link with a built .deb file of the latest heeks svn
[16:22:10] <elmo40> might have been in #cam, though
[17:06:30] <EricKeller> mini-box.com is out of stock on all barebones systems because they all use the pico-80 power supply
[17:09:44] <gene> Hey guys!
[17:10:21] <elmo40> hi gene
[17:11:09] <gene> I just managed to get 10.04 & emc-2.4.3 up and running, not too bad once I got it to use the vesa driver.
[17:11:51] <elmo40> why vesa?
[17:13:29] <gene> But I have a SCALE problem on the Z axis. the calculations for it look good, but it moves about 8% (SWAG) too far. And I had to stick a - sign in front of the X SCAL value. Has anyone else noticed this?
[17:14:20] <gene> vesa shares IRQs, and for a realitime display is about 2 seconds faster than the radeon driver in serviceing the backplot.
[17:14:40] <elmo40> ah, ATI. yes, that could been an issue
[17:15:16] <elmo40> and why a scale problem? adjust the settings with stepconf. 8% is a lot!
[17:15:20] <gene> Issue? More like a PIMA
[17:17:02] <gene> Ok, xyloytex 4 channel, running in 8 step microstep, 10 tpi acme screw, 17 tooth pulley on the motor, 42 teeth on the acme nut carrier
[17:18:18] <gene> this should return a scale in the 39,527.xxxxx range, but to get it to move an inch (after the backlash is taken up) I had to lower it to about 34,299.xxxx
[17:19:57] <gene> And I had to add a - sign in front of the x scale to get it to job the right direction. All this from an .ini file that worked perfectly for 2.3.x.
[17:20:23] <gene> s/b jog above
[17:21:51] <gene> And I just found my saitek game controller is dead with 2.4.3 too.
[17:22:59] <elmo40> dead?
[17:23:28] <gene> So that's not terribly important, but the scale error is.
[17:24:04] <gene> Yup, no response from it, for any button or stick.
[17:24:14] <elmo40> same .ini file brought over from 2.3 does not give same results in 2.4?
[17:24:26] <gene> Correct
[17:24:49] <elmo40> interesting. i wonder why tht would be.
[17:25:25] <gene> Restored from backups. But let me try another config justr for grins. brb.
[17:28:01] <morficmobile> gene: vesa vs open source radeon?
[17:28:55] <gene> Yes
[17:29:12] <gene> radeon is a friggin disaster, as it always has been here.
[17:31:44] <gene> And I just recopied my old 2.3.6 configs from the backup, and killed emc. Humm.
[17:31:53] <elmo40> lol
[17:32:01] <morficmobile> i love my radeon on my home desktop, but i run most libs and drivers from git
[17:32:33] <gene> looks like I ned to go look at a few logs again, but I have other things to do this afternoon, so it will be later.
[17:32:38] <morficmobile> i think i will see if the drm stuff ports to the 2.6.34 magma and how it affects to have working kms color tiling with RTAI
[17:33:21] <morficmobile> gene: and gnome as desktop environment?
[17:33:40] <gene> good luck, I need to goto the landfill and see if I can find an ignition module for my SP mower. Later.
[17:33:59] <morficmobile> wow, those are some searching skills
[17:34:17] <morficmobile> landfill and specific part sounds like hay stack and needle
[17:34:40] <morficmobile> i think you need the luck more than me
[17:36:13] <elmo40> SP mower? do you mean POS mower? ;)
[17:38:14] <gene> self propelled, and the landfill is a separate stop. Cub Cadet parts for a Kawasaki motor are hens teeth. I expect the dealer will have to order, and about $60
[17:39:42] <Jymmm> gene: Tried ebay for parts yet?
[17:48:57] <elmo40> or craiglist?
[17:49:02] <elmo40> or mold your own ;)
[18:52:44] <q8istud> hello everyone
[18:53:36] <fossrox> hi :)
[18:53:45] <q8istud> i have an error when i want to run the machine
[18:54:25] <elmo40> well... what is the error? ;)
[18:54:46] <q8istud> RTAPI: ERROR: Unexpected realtime delay on task 1 This message will only display once per session. Run the Latency Test and resolve before continuing.
[18:55:01] <q8istud> that's an error
[18:55:20] <EricKeller> there is quite a bit of info about that on linuxcnc.org
[18:55:28] <EricKeller> what is the computer?
[18:57:09] <q8istud> its my made :)
[18:57:31] <EricKeller> you might have a loose solder joint then :)
[18:57:32] <elmo40> specs...
[18:58:34] <EricKeller> is it a stepper motor system?
[18:59:47] <q8istud> no, the machine work when i test in in the wizard .. yes stepper motor sys.
[19:02:36] <celeron55> my friend just told me emc2 calculates the maximum step rate wrongly as 1/(base period), as it should be 2/(base period) because it needs two processings for one pulse
[19:02:55] <celeron55> is he wrong?
[19:03:45] <q8istud> how can i know that ??
[19:04:16] <q8istud> how can i know if its in maximum ??
[19:04:22] <EricKeller> q8istud, many conversations may happen at once
[19:04:29] <celeron55> q8istud: this is not related to your problem
[19:04:31] <EricKeller> on irc
[19:04:53] <celeron55> just writing random thoughts in here for everyone to read
[19:04:59] <EricKeller> q8istud, did you run the latency tests?
[19:05:44] <q8istud> yes, and its work very well
[19:06:18] <Jymmm> q8istud: what were the values of the latency test?
[19:10:08] <EricKeller> I always wondered about the warning on startup, but when I've used a computer that was fast enough for my settings it never happens
[19:14:24] <q8istud> motor steps: 1000, Drive microstepping: 2, pulley teath: 1:1, leadscrew pitch: 20, max. velocity: 40, max. acceleration: 30, home location: 0.0, table travel: 0.0 to 0.8
[19:14:38] <q8istud> is that is ??
[19:14:44] <elmo40> https://m21.hyte.de/wiki/SpindleMeter
[19:14:54] <EricKeller> no, it's the thread timings that we need
[19:15:59] <q8istud> what is the word in the programs??
[19:16:00] <elmo40> my links are all random. not related to any specific conversation... sorry for the confusion ;)
[19:17:46] <q8istud> are you want the driver timming sitting ??
[19:18:00] <JT-Work> cradek: your collets and rack are on the way finally
[19:19:34] <roberth__> what freebies do i get
[19:20:28] <q8istud> my machine from hobbycnc .. :)
[19:20:29] <EricKeller> q8istud, read this page and tell us your numbers after running the latency testhttp://wiki.linuxcnc.org/emcinfo.pl?Latency-Test
[19:21:15] <q8istud> ok thank you
[19:21:57] <cradek> JT-Work: thanks! I forgot to send you the check - I will get it in tomorrow's mail
[19:22:05] <EricKeller> q8istud, you'll probably have to change some numbers as shown on this page:
[19:22:09] <EricKeller> http://wiki.linuxcnc.org/emcinfo.pl?TweakingSoftwareStepGeneration
[19:26:47] <q8istud> thank you again, i will try it soon.
[19:37:26] <elmo40> Jymmm: you still in need of a 1/2" collet with 'A' drill hole? i got it made. Plus a few extras ;)
[20:17:27] <elmo40> great... error in running EMC
[20:18:26] <elmo40> http://pastebin.ca/1937515
[20:21:57] <roh> elmo40 you need reasonable graphic drivers
[20:22:16] <roh> no opengl no axis
[20:23:05] <cradek> the most common cause of this is futzing with the proprietary nvidia drivers, especially installing and then trying to uninstall them
[20:23:15] <roh> *sigh* i just wish somebody would pay me for such stuff, not only boring it work.
[20:23:40] <roh> irgh.. wrong window
[20:24:10] <cradek> if it's a new install, the easiest fix is to install again and then DO NOT use the proprietary drivers when it offers them
[20:24:53] <roh> i can recommend older ati hw for that. radeon 7000-9000 agp stuff. fits well to the machines used for emc now.
[20:25:16] <roh> also avail in passively cooled :) good for dusty workshops
[20:25:27] <cradek> I like older matrox stuff for the same reasons
[20:25:40] <roh> true. but thei gl is all sw and slow
[20:25:41] <cradek> I think all my real emc machines have matrox pci video cards
[20:25:55] <cradek> it's perfectly fine for emc in my experience
[20:26:14] <roh> and they got no dvi output. we got 1680x1050 or hullhd panels on all the emc machines
[20:26:24] <roh> s/hull/full/
[20:27:18] <cradek> perhaps our requirements are different
[20:27:55] <roh> i guess so :) i am just happy as soon as we got rid of the win32 on the laser
[20:29:14] <roh> heh.. now i know why we didnt use the matrox ones. (just checked the old hw box) ... the matrox ones all got no dvi OR have a sticker on it that dvi is fried the rest is ok
[20:29:46] <roh> evil... but i can remember someting about them was weird. whatever.
[20:41:40] <andypugh> I have been boring and turning a 7" x 2" OD bit of steel to various diameters and 3mm wall thickness. Very boring and chattery, mainly because all the diameters are bigger than my chuck bore, and I don't have a fixed steady. 6" protrusion and a 5" boring bar, non-ideal.
[21:06:33] <Connor1> Hey Guys, I'm ready to test my driver board and stepper motors.. and have never used EMC, can someone help me? I'm using one of those controllers that has the stepper drivers built in.. (4 axis version).
[21:07:48] <andypugh> Stepconf should hand-hold you through the setup.
[21:08:12] <andypugh> You just need to know what pin on your cable is which signal.
[21:08:17] <Connor1> what about com port ID's ?
[21:08:30] <Connor1> yea, that's what I'm not sure about.
[21:08:35] <andypugh> (And whether the "enable" pins are active-high or active-low)
[21:08:39] <alex_joni> com port?
[21:08:47] <Connor1> err.. parport
[21:08:49] <andypugh> lspci -v isn't it?
[21:08:51] <alex_joni> hopefully it's not a serial port :)
[21:09:06] <andypugh> Or am I misemebering?
[21:09:12] <Connor1> I know.. I was just playing with a Arduino setting up a RPM metter\
[21:09:44] <SWPLinux> you can use numbers like 1, 2, and 3 to use the first, second, and third parallel ports, respectively (up to 16 I think)
[21:10:11] <SWPLinux> as detected by Linux, and also assuming that you're using something recent, like emc 2.4.x
[21:10:18] <SWPLinux> (maybe in 2.3 also, I don't know)
[21:10:35] <andypugh> Ah yes, I had forgotten that bit :-)
[21:10:35] <alex_joni> not 2.3
[21:11:31] <Connor1> I'm running 2.4.3
[21:11:52] <SWPLinux> then use the number 1 if you have only one parallel port
[21:12:07] <SWPLinux> assuming that the Linux kernel sees it
[21:13:10] <Connor1> Okay.. in stepconf... Parallel port settings.. first parport base adderss 0x378
[21:16:20] <SWPLinux> type in 1 instead
[21:16:31] <SWPLinux> if 378 doesn't work (which it probably should)
[21:17:24] <Connor1> okay, what about the motor leads? I have red, black, green, blue.. looks like black & green is a coil, and red blue is a cool
[21:18:03] <andypugh> They go to the driver. What are the terminals on the driver labelled as?
[21:18:24] <Connor1> XA+,XA-,XB+,XB-
[21:18:54] <SWPLinux> connect one coil to XA+ and XA-, the other to XB+ and XB-
[21:19:01] <andypugh> OK, for a start put red and blue to A+ and A- and black and green to B+ and B-
[21:19:17] <andypugh> If you switch a coil you reverse the motor.
[21:19:33] <andypugh> If you connect one coil between two phases, the drivers blow up...
[21:19:33] <SWPLinux> there are several ways to fix it if the motor goes the wrong way, which is all that can go wrong *if* you don't cross-connect the coils
[21:19:42] <SWPLinux> what he said too
[21:20:37] <andypugh> Also never disconnect a motor from the driver with the driver powered up.
[21:21:11] <motioncontrol> good night at all
[21:21:22] <SWPLinux> or connect one
[21:21:27] <andypugh> Easy way to find which wires are pairs is to touch the ends together, and see if they make the motor stiff to turn.
[21:21:36] <SWPLinux> and/or use a multimeter :)
[21:22:28] <andypugh> 4-wire motors are easy. 8-wire can be annoying (and the multimeter can't distinguish between a good and bad permuation with those).
[21:22:55] <andypugh> But that doesn't matter right now :-)
[21:23:38] <Connor1> okay, wired up.. what happens if a driver fires up with NO motor attached?
[21:23:48] <SWPLinux> no problem
[21:23:52] <Connor1> okay.
[21:23:54] <Connor1> here goes..
[21:23:59] <SWPLinux> well, should be no problem anyway :)
[21:24:58] <Connor1> oh crap.. it's asking about pin assignments.. I have no clue.
[21:25:20] <Connor1> http://cgi.ebay.com/4-Axis-Nema-23-stepper-motor-287-oz-in-CNC-Router-KIT-/270631971793?pt=LH_DefaultDomain_0&hash=item3f02ec2bd1
[21:25:27] <Connor1> that has a link to the controller I'm using
[21:27:14] <andypugh> you probably have a USB stick with some badly-written instructions on it?
[21:27:30] <Connor1> cd, and wrong stuff.
[21:27:39] <SWPLinux> if you have no other manual or information, then you're out of luck, unless you're good at reverse-engineering stuff
[21:28:03] <SWPLinux> since the "reference wiring diagram" only says that you should connect the LPT connector to a PC
[21:28:40] <andypugh> Well, Xylotex and Sherline presets exist, one of those is actually very likely to work.
[21:29:03] <Connor1> I'm looking for a manual from someone else.. they're ALL over ebay.
[21:30:27] <andypugh> Worst case you should be able to find the Driver IC datasheet and trace the tracks.
[21:31:03] <andypugh> The CD is probably right, even if it is the wrong board.
[21:31:31] <andypugh> And getting the P-Port pinout wrong is not a problem, it just won't work.
[21:32:28] <andypugh> (My advice about crossing coils and disconnecting motors cost me £150 to discover, you get the knowledge for free, isn't the internet marvellous?)
[21:32:45] <Connor1> :) Thanks..
[21:33:00] <Connor1> Why does disconnecting the motors cause issues?
[21:36:30] <Connor1> okay.. this thing has XYZA enables...
[21:37:50] <andypugh> On the P-Port connector, or on a seperate header?
[21:37:59] <Connor1> p-port
[21:38:11] <andypugh> Stepconf supports that.
[21:38:39] <Connor1> Amplifier enable ?
[21:38:48] <andypugh> (Waste of 4 precious pins, but there you are)
[21:38:59] <andypugh> The Stepper drivers are also called amplifiers
[21:39:17] <andypugh> They convert 5v mA to 24V 2.5A....
[21:46:40] <Connor1> Subsystem: Device a000:2000
[21:46:40] <Connor1> Flags: bus master, fast Back2Back, medium devsel, latency 32, IRQ 10
[21:46:40] <Connor1> I/O ports at 1018 [size=8]
[21:46:40] <Connor1> I/O ports at 1010 [size=8]
[21:47:50] <andypugh> Nurse! Nurse! He's talking gibberish again!
[21:47:59] <Connor1> I have a onboard parallel port, and 2 on a PCI card.
[21:48:12] <Connor1> I'm using one on the PCI card at the moment..
[21:48:18] <moopy> anyone know about jacobian stuff in genserkins??
[21:48:26] <Connor1> I don't think linux is seeing the onboard one..
[21:49:46] <moopy> what chipset does the board use connor?
[21:50:07] <Connor1> The pci card or the mobo?
[21:50:24] <moopy> which ever one is not being recognised
[21:50:35] <moopy> pastebin the output of lspci -v
[21:51:30] <andypugh> moopy: Have you read http://en.wikipedia.org/wiki/Jacobian_matrix_and_determinant
[21:51:44] <moopy> no i have not andy
[21:52:03] <Connor1> http://pastebin.com/HStKEFHH
[21:52:04] <moopy> but i just want to kow why it is disabled
[21:52:13] <andypugh> It probably won't help at all with the practice, but might hint what it is for.
[21:52:57] <moopy> i know what it is for (i think) but i want to use what has been disabled and want to know if there is some deeper reason it has been commened out
[21:53:39] <andypugh> Connor: Perhaps it is disabled in the BIOS?
[21:54:28] <andypugh> moopy: I guess if it is not used it would have been commented out to save thread time.
[21:54:50] <moopy> is was thinking that might be the reason
[21:55:05] <andypugh> You could see how long the servo thread takes to run with it commented out, then re-enable it and look again. Perhaps it is that simple?
[21:55:54] <alex_joni> moopy: it's not used at the moment
[21:56:01] <alex_joni> jacobian I mean
[21:56:06] <moopy> yeah i think the jacobian will be useful if i want to branch out more joints from a linkage
[21:57:10] <andypugh> I have been reprimanded for using the same tangent on two lines of code, rather than deriving it from some sins and cos already calculated. Saving execution time is important in realtime modules.
[21:58:05] <Connor1> looks like my parport should be 0x1008
[21:58:18] <andypugh> Try "2"?
[21:58:39] <Connor1> Realtime already running. Use 'halrun -U' to stop existing realtime session.
[21:59:00] <andypugh> Is EMC running at the same time as stepconf?
[21:59:12] <Connor1> no
[21:59:17] <andypugh> Odd.
[21:59:49] <moopy> just type 'halrun -U' at a console then
[21:59:50] <Connor1> brb
[21:59:52] <andypugh> in the terminal, type (I think) halshow. Perhaps halcmd show all.
[22:00:15] <andypugh> Or if you don't care about the "why" just do that :-)
[22:00:31] <moopy> if that dont work try 'ps -ef' and then kill any running emc processes
[22:00:57] <Connor1> non running.
[22:01:06] <Connor1> I have this chat window running on same machine..
[22:01:10] <Connor1> wouldn't think that would be a issue.
[22:01:38] <moopy> type dmesg and pastebin the output
[22:01:49] <andypugh> No, the IRC does not use Realtime
[22:02:47] <andypugh> Or restart. (I think he maybe just did?)
[22:04:20] <Connor1> OKay, rebooted
[22:07:12] <Connor1> raceback (most recent call last):
[22:07:12] <Connor1> File "/usr/bin/stepconf", line 1552, in on_xaxistest_clicked
[22:07:12] <Connor1> def on_xaxistest_clicked(self, *args): self.test_axis('x')
[22:07:12] <Connor1> File "/usr/bin/stepconf", line 2050, in test_axis
[22:07:12] <Connor1> 'scale': data[axis + "scale"],
[22:07:12] <Connor1> TypeError: %d format: a number is required, not NoneType
[22:32:22] <MattyMatt> I run firefox on my emc machine, and it's only the FF that suffers :)
[22:32:28] <andypugh> Connor1: Looks like we are all stumped by that one
[22:32:48] <Connor1> I think it had to do with needing the enables..
[22:33:07] <Connor1> I just set the X to Amplifier enable.. and left the Y and Z and A to Unused ATM
[22:33:15] <Connor1> but, I still can't get it to do anything
[22:33:45] <andypugh> Almost certainly not the real issue. EMC2 doesn't know if the amps are working or not, or if the motors are moving or not.
[22:34:08] <Connor1> Yea, but, it was using the same config option for all 4.. I think it confused it.
[22:34:09] <andypugh> Steppers are very much "Send pulses, and hope"
[22:35:29] <andypugh> Aye, that could cause a bit of bother.
[22:36:02] <andypugh> Any noise from the motors? LEDs on the driver board?
[22:36:14] <Connor1> none.. just the fan running on the heat sink.
[22:36:32] <Connor1> what about the step times and stuff at the very begining?
[22:36:44] <andypugh> Try inverting the "enable" pin?
[22:36:49] <Connor1> yup.
[22:37:19] <andypugh> Put big numbers in, say 100,000
[22:37:36] <Connor1> I have 5000,5000,20000,2000
[22:38:48] <andypugh> Is the motor stiff to turn?
[22:39:18] <Connor1> no more than it normally is.
[22:40:48] <andypugh> Sounds like amp-enable is not active then.
[22:41:08] <andypugh> You _have_ connected the main power and the logic power?
[22:41:14] <Connor1> yes.
[22:43:26] <Connor1> okay, I powered it down, and put it on the Y axis.. and it's hard to turn.
[22:43:32] <Connor1> err.. I can't turn it.. rather.
[22:44:04] <andypugh> Interesting.
[22:44:13] <andypugh> And a good sign.
[22:45:10] <andypugh> You have no idea how frustrating this is, if just about any of us were there (who have done this, with one of those boards) you would be spinning motors now.
[22:45:44] <Connor1> Yea, I know.
[22:47:22] <andypugh> What did you switch (from what to what)
[22:47:49] <Connor1> I'm unplugging power from it.. and moving it from X to Y to Z.. testing..
[22:48:05] <Connor1> it looks like Y has breaking on the motor... dunno why.
[22:48:08] <Connor1> let me try something..
[22:48:37] <Connor1> okay,.. That's odd.
[22:49:02] <Connor1> This board has 4 LED's (1 for each AXIS).
[22:49:10] <Connor1> two were lit, two were not.
[22:49:21] <Connor1> unplug parport.. they all go dark..
[22:51:00] <andypugh> I wonder if the board is hardware-inverting the hardware-inverted p-port pins that emc is software-inverting?
[22:51:50] <Connor1> I'm about to mark them all as unused..and walk the invert down the pins..
[22:55:44] <andypugh> You might find this useful: http://wiki.linuxcnc.org/emcinfo.pl?Parallel_Port_Tester
[22:56:21] <archivist> forgetting to pull up pins is a common error
[23:02:50] <andypugh> Shouldn't be an issue with this board, which has an actual DB25 on it
[23:03:46] <Connor1> I'm at a loss...
[23:03:59] <Connor1> I'm --><--- to loading up XP and Mach3.. just to test the hardware..
[23:04:17] <andypugh> Did you try the p-port tester I linked to?
[23:04:19] <Connor1> 0x1008 or 0x1000 has to be the par-port.
[23:04:21] <Connor1> Yea..
[23:04:35] <Connor1> Nothing changed on the LEDS.. motor never twitched.
[23:06:08] <andypugh> Did you edit the HAL file to point to your suspected base address?
[23:06:19] <Connor1> yup. tried both
[23:06:26] <andypugh> (and it might be worth trying 1, 2 and 3)
[23:10:33] <andypugh> I am not clear if it is the hal_parport module that uses parport_pc to find parports, or whether Stepconf does it.
[23:12:22] <andypugh> You have an "active" Y axis, but not an "X". This either indicates that the p-port is working, or that the amp-enables are serendipitously connected to hardware-inverted pins.
[23:12:45] <Connor1> X and Y were lit..
[23:12:50] <Connor1> only Y held the motor.
[23:13:08] <moop> what driver board ae you using?
[23:13:17] <andypugh> That's _very_ odd
[23:13:39] <Connor1> one from ebay.. like this one.. http://cgi.ebay.com/4-AXIS-CNC-Router-Mill-Stepper-Motor-complete-kit-/230493424213?pt=LH_DefaultDomain_0&hash=item35aa7a8a55
[23:13:39] <andypugh> Connor1: Which continent are you on?
[23:13:41] <moop> moop is now known as moopy
[23:13:45] <Connor1> US
[23:14:06] <andypugh> There goes that idea.
[23:14:15] <Connor1> what was the idea?
[23:14:36] <andypugh> If you were < 20 miles away, I would pop round :-)
[23:14:41] <Connor1> oh.
[23:15:20] <moopy> i have heard people have problems with the chinese import stuff
[23:15:43] <moopy> have you contacted the seller with your problems?
[23:16:00] <Connor1> not yet.. I would have to try it on the Mach3 first.
[23:16:46] <moopy> have you checked the board serial number against the manual you are using?
[23:17:14] <andypugh> Can you see a 555 timer on the board? I see you have a manual interface. My similar board had an onboard 555-based pulse generator which would spin the motors without a PC involved simply with jumpers and twidling a pot
[23:17:16] <moopy> i have heard of people being given the wrong manual for the board they buy
[23:17:36] <Connor1> I looked for the manual that matched.
[23:18:20] <moopy> i think some of the chinese stuff has really awful manuals
[23:19:09] <Connor1> okay. Success at triggering the spindle relay
[23:19:28] <Connor1> 0x1008 is the 2nd port on the card.
[23:19:37] <moopy> manuals so bad that some of the pin assignments are wrong in the manual
[23:20:50] <moopy> so you may have to go through and try every pin by trial and error?
[23:20:51] <andypugh> That sounds like progress
[23:22:26] <moopy> try inverting each parport pin in turn to find which pins enable which motors
[23:22:31] <andypugh> Completely off-topic, but I like this "The farmer wondered whether the wether would weather the weather, or whether the weather the wether would kill"
[23:23:00] <andypugh> The parport tester should help figure that out
[23:23:01] <cradek> is a wether an animal?
[23:23:09] <andypugh> It's male sheep
[23:23:23] <moopy> welsh like sheep
[23:23:53] <cradek> I was thinking about the infamous "known unknowns" speech when you said learning the manual doesn't have the right pin descriptions is progress
[23:24:37] <Connor1> !WooT!
[23:24:40] <Connor1> I got it.. I think.
[23:24:50] <Connor1> Now for tuning..
[23:25:20] <Connor1> okay, it's going in the same direction no matter which way I do it..
[23:25:31] <Connor1> with the Jog controll.
[23:25:43] <andypugh> Odd, but inverting the direction pin ought to switch it
[23:25:52] <moopy> try inverting each unassigned pin in turn until it changes direction
[23:25:57] <archivist> invert the right pin
[23:26:00] <andypugh> Or, your direction pin is the wrong pin
[23:26:27] <moopy> eventually you will find all the right pins
[23:26:35] <Connor1> What about the + / - on the motor ?
[23:26:43] <Connor1> it's kinda jittery..
[23:27:19] <moopy> it is chinese
[23:27:39] <andypugh> My chinese one was fine, just not very robust
[23:28:22] <andypugh> Try longer step and direction times. You can reduce them later
[23:28:39] <Connor1> okay.. I have microstepping turned on 1/2
[23:28:49] <Connor1> 200 steps per
[23:28:55] <andypugh> Also, don't expect too much speed from a stepper
[23:28:56] <Connor1> driver microstepping is set to 2
[23:29:38] <andypugh> 4 might be smoother, then you enter diminishing returns territory.
[23:36:40] <Connor1> Y Z and A work.. something wrong with the X
[23:36:45] <Connor1> Only goes 1 way...
[23:37:03] <MattyMatt> Conner1. the CD I got had all the docs for the whole range. that one in the photo is the HY-TB4DV-M, same as mine
[23:40:20] <Connor1> okay, this is odd.. ony the y,z,a the enable has to be inverted..
[23:40:31] <Connor1> for me to get X to move.. I don't invert it..
[23:40:37] <Connor1> I think something wrong with X.
[23:41:31] <andypugh> Connor1: Could be. Check the traces with a multimeter, look for bridges between tracks, breaks in the tracks etc
[23:41:43] <Connor1> yup.
[23:41:57] <Connor1> I can't figure out WHY it wouldn't work on the other parport..
[23:42:14] <andypugh> Vibrating in the spot is a sign that step and dir are switched, so it isn't that
[23:43:11] <MattyCNC> http://filebin.ca/nyzdy/HY-TB4DV-M4axisdriverboardmanual.doc
[23:43:22] <MattyCNC> pinouts in there
[23:43:31] <andypugh> If you look up the datasheet for the driver IC you will see that all the board you have does is (a) connect the driver ICs to the the motors and (b) cost a lot less than you could by the 4 ICs alone for
[23:43:57] <MattyCNC> optos are nice touch
[23:44:31] <andypugh> Actually, a duff opto could easily be the problem here
[23:44:45] <andypugh> Check that one hasn't got a leg folded under.
[23:44:55] <MattyCNC> and it ain't true. these chips are £2.66 each
[23:45:52] <andypugh> Where from? I was paying £8 from HK, and more from RS during my short-lived driver-IC-exploding hobby
[23:46:20] <MattyMatt> HK via ebay of course. me looks again
[23:47:06] <MattyMatt> my optos were always wonky in the sockets. if it's just arrived press them all down
[23:47:29] <MattyMatt> very cheap DIP sockets
[23:47:34] <andypugh> The ones with the ICs in holes in the board look a lot easier to desolder than mine. (I think that is actually jolly clever)
[23:47:56] <MattyMatt> yeah could potentially put another H/S on the back too
[23:48:50] <Connor1> I noticed that one set of sockets isn't flush with the board...
[23:48:57] <Connor1> but, they're soldered ok..
[23:48:58] <andypugh> That filebin link fails for me, no idea why
[23:49:39] <MattyMatt> worked for me on this machine
[23:49:46] <andypugh> Compare input and output voltages with a working channel. I doubt that they do any actual testing.
[23:51:41] <MattyMatt> http://cgi.ebay.co.uk/TB6560AHQ-TB6560-STEPPER-MOTOR-DRIVER-IC-TOSHIBA-/110535851885 forst hit 2.91+p&p
[23:52:07] <MattyMatt> I like this board. worth cloning IMO
[23:53:08] <andypugh> Professional design - Two stage signal processing with super anti-jamming !
[23:53:39] <MattyMatt> emc>emp
[23:53:45] <andypugh> Whatever that means
[23:54:27] <andypugh> My board was the other common one, which I think is a bit less capable.
[23:55:37] <andypugh> (TA8435H)
[23:57:16] <MattyMatt> yeah that's the 2A version
[23:57:59] <MattyMatt> which would have been perfectly OK for my motors, but I thought I'd give myself some headroom :)
[23:58:25] <andypugh> Yeah, rice puddings didn't fear their skins with that one, and wet brown paper bags were positively rude.
[23:59:31] <MattyMatt> I'm still on 12V, so still stuck ~300rpm