#linuxcnc | Logs for 2013-01-11

[00:38:49] <Loetmichel> mornin'
[00:40:38] <tjb1> Morning.
[02:22:19] <DJ9DJ> moin
[03:09:00] <tjb1> Morning DJ9DJ and Loetmichel
[03:09:13] <DJ9DJ> good morning tjb1
[03:09:38] <tjb1> 4 am :)
[03:11:42] <DJ9DJ> hmm, 10 am here
[03:12:01] <DJ9DJ> where are you, east coast?
[03:32:40] <tjb1> yes
[03:33:29] <DJ9DJ> :)
[03:50:36] <mrsun> hmm, 55 degrees i get the dovetails to on the sieg x1 ...
[03:51:18] <mrsun> ahh seems that some other bloke has come to the same conclution :P
[03:53:10] <mrsun> wouldnt using a set screw with an angled top be the best to tighten gibs ?
[03:53:21] <mrsun> that is having the angle in the end the same as the dovetail angle?
[04:00:53] <cncbasher> mrsun:use a small ball bearing between the screw and gib plate
[04:01:07] <mrsun> hmm that is an idea =)
[04:01:25] <mrsun> but would leave just a very small contact area to begin with and it would dig in ?=
[04:01:35] <mrsun> or do i drill a small dimple in the gib ?
[04:01:42] <cncbasher> this self adjusts to whatever angle is needed
[04:02:07] <cncbasher> most have a small dimple in the gib yes
[04:02:55] <mrsun> http://lh5.ggpht.com/_ozT1G3WqXg4/TUqsPuzQqeI/AAAAAAAALKM/HweM0rgDKQU/s800/gib.jpg http://lh5.ggpht.com/_ozT1G3WqXg4/TUqsPuzQqeI/AAAAAAAALKM/HweM0rgDKQU/s800/gib.jpg
[04:03:05] <mrsun> cncbasher, and you sure that works good? :)
[04:03:14] <mrsun> so i dont bite myself in the arse doing that :P
[04:03:18] <archivist> the taper in the bottom of the dimple helps tighten the gib if you are lucky/unlucky
[04:03:47] <cncbasher> iv'e done it like that for years , not had any problems
[04:04:24] <archivist> I have had directional friction differences on some slides
[04:04:40] <cncbasher> dont like the idea of the lower edge of the gib not being angled ,
[04:05:05] <cncbasher> i can see that being a problem and jamming up
[04:05:08] <mrsun> gonna try and make new gibs for the mill as the people who made them 1 .. made them to thin, 2 the "height" of the gib differs 2mm over its length :P
[04:05:25] <cncbasher> needs to be angled to fit properly , similar to the top
[04:05:55] <mrsun> cncbasher, its the angle of the set screw i mean from those pictures, that they havent angled the bottom of the gib to make it fit i find kinda stupid =)
[04:06:09] <mrsun> the set screws on my mill is like the first picture
[04:06:12] <mrsun> round nosed
[04:07:19] <cncbasher> yes i agree , hence the reason to use a ball bearing , far much better than a round end on the scree
[04:07:28] <cncbasher> screw
[04:08:07] <mrsun> ye a ball bearing wouldnt twist either
[04:08:14] <archivist> I rather like taper gibs because of the endwise restraint
[04:08:17] <mrsun> as the set screw would just rotate against the ball bearing
[04:08:37] <mrsun> archivist, ye but flat gibs in this mill and i do not have the tools to modify it for tapered :P
[04:09:30] <archivist> you need some form of restraint (dimples) else the gib can slide (stick to the stationary part)
[04:10:22] <mrsun> archivist, there are in the original gibs, but looks mostly like they have just tightened the gib screws hard to make dimples in it :P
[04:10:24] <archivist> that emco drawing just looks wrong from that perspective
[04:10:28] <mrsun> kinda trashed :P
[04:11:04] <cncbasher> yea that drawing is amateurish and not good practice
[04:12:26] <mrsun> but so, get some ball bearings then that fit inside the gib screw holes and new gib screws that fit agains the bearings better, and new gibs
[04:12:28] <mrsun> :P
[04:12:38] <mrsun> and drill a small dimple for each bearing ball
[04:12:52] <cncbasher> yes
[04:12:54] <archivist> I did come across a new (to me) taper gib, tape is not length but across and pushed in sideways
[04:13:31] <mrsun> i guess one could make "tapered" gibs by having two gibs
[04:13:37] <mrsun> and pushing them against eachother
[04:13:59] <mrsun> gapered a couple of degrees each
[04:14:07] <mrsun> hard to lock them in place tho :P
[04:14:20] <mrsun> prone to jamming i guess
[04:14:36] <archivist> dunno how you would arrange the push out screws
[04:15:04] <mrsun> hehe would only be push in screws :P
[04:15:06] <mrsun> on both sides :P
[04:16:31] <mrsun> if there was more space one could just make a stationary tapered shim also
[04:16:41] <mrsun> and add in a tapered gib
[04:16:51] <mrsun> but at 3.75mm maximum space i think that would be hard =)
[04:17:31] <archivist> tapers gibs self jam if the push out screw is not hard up against the end
[04:18:11] <mrsun> mm
[04:20:10] <mrsun> nah gonna just try and replace the gibs, gonna start with the X axis and see where it takes me =)
[04:20:13] <mrsun> making new ones
[04:20:27] <mrsun> just need to find some 4mm stock to mill down to 3.5
[04:21:21] <mrsun> good project to do while sick =)
[04:22:47] <mrsun> and gonna try cncbasher ball bearing trick =)
[04:22:56] <mrsun> time to burns ome more money ... sigh :P
[04:27:15] <mrsun> and noticed i need to straighten the leadscrew on X also ... removed the gib and ran it from the motor and the table wobbles like hell :P
[04:29:19] <mrsun> the never ending mill project ... :P
[05:14:55] <tjb1> im out
[05:14:56] <tjb1> t6am
[05:15:00] <tjb1> peace fellas
[05:29:49] <HarryHood> I used to use 5I20.7I33 setup, now im using 5I25 - 7I77, my encoders have only ever been used on A, B, Z, however, now my machien does not stay in time, with the 7I77 do I need to also use A/ B/ Z/ ?
[05:30:47] <jthornton> HarryHood, you can use single ended on the encoder
[05:31:14] <HarryHood> is it less accurate this way?
[05:31:37] <jthornton> less resistant to noise
[05:31:47] <HarryHood> I could just connect the rest of the pins, is that the reason for my machine stepping away from true 0?
[05:32:27] <jthornton> if you have them I would connect them, the machine may not be tuned as well
[05:32:50] <HarryHood> thats my question really, is it something in the tuning or encodeR?
[05:33:54] <HarryHood> if I tell the machine to go x5, then x0 and repeat. its like, 2 steps left one step right
[05:33:59] <jthornton> did you tune the axes with halscope?
[05:34:19] <jthornton> is this a servo machine?
[05:34:20] <HarryHood> I used my old tuning from the 5i20
[05:34:51] <HarryHood> just plugged in numbers
[05:35:30] <HarryHood> i just dont see how the encoder is offsetting
[05:35:59] <jthornton> does the DRO agree with the position?
[05:36:27] <HarryHood> it shouldnt really matter how accurately the machine moves, so long as it the encoder reads dro reads 0 the orientation of the motor shafter should be the same
[05:36:30] <HarryHood> yes
[05:37:36] <jthornton> does the DRO move back and forth while your sitting still?
[05:37:43] <HarryHood> no
[05:38:25] <jthornton> I've only been awake 5 min so brain not fully functional yet
[05:38:37] <jthornton> not that it ever is fully functional
[05:40:21] <HarryHood> if I start the software and assume all motor shafts are at 0 and the keyway of the motor shafts are in the 12O'clock position. then send motor to x1 and back to x0 repeated for example 10 times, when I sendit back to zero, the shaft sits at 4 O,clock
[05:40:28] <HarryHood> lol i know that feeling
[05:41:08] <pcw_home> If you are using single ended encoders, you should make sure
[05:41:10] <pcw_home> that the 7I77 is jumpered for single ended encoders
[05:41:11] <jthornton> ok, so the position and the DRO do not agree after moving, this does sound like noise on the encoder lines
[05:41:29] <HarryHood> both axis are equally offsetting
[05:41:34] <jthornton> and wait for Peter to magically show up at 3 am
[05:41:45] <HarryHood> I assume the jumper is wrongly set for only using a b z
[05:42:21] <HarryHood> I need to either hook up a/ b/ z/ OR relocate jumper
[05:42:29] <pcw_home> I think default is differential
[05:43:08] <HarryHood> differential refers to a a/ b b/ z z/? and single ended is a b z?
[05:43:21] <pcw_home> yes
[05:43:26] <HarryHood> ughhh
[05:43:30] <HarryHood> one jumper
[05:43:36] <HarryHood> I knew it was simple
[05:43:38] <pcw_home> three
[05:43:56] <pcw_home> three per encoder
[05:43:57] <HarryHood> 1 for each encoder? no?
[05:44:04] <HarryHood> three per encoder?
[05:44:12] <pcw_home> 1 for each signal
[05:44:22] <jthornton> if I had differential I'd use it
[05:44:41] <HarryHood> got ya
[05:44:49] <HarryHood> yeahhh im sure this is the problem
[05:44:50] <pcw_home> jthornton couldn't sleep
[05:45:28] <jthornton> yuck
[05:45:41] <jthornton> I slept in to 5:30
[05:45:48] <pcw_home> might not be _the_ problem but certainly can be a problem
[05:46:18] <HarryHood> I am almost 100% sure this will solve it
[05:46:50] <HarryHood> Its defaulted to have all six signals and im only giving it three
[05:47:26] <pcw_home> diff encoders will usually work in TTL mode but the reverse is not true
[05:47:40] <HarryHood> right
[05:47:59] <pcw_home> back to bed...
[05:48:07] <HarryHood> I try to avoid coming directly to the chat room for help
[05:48:15] <HarryHood> its been about a year now
[05:48:33] <HarryHood> everytime though you guys have been incredibly helpful
[05:48:48] <HarryHood> thanks again! sleep well
[05:58:22] <jthornton> anyone remember where to put isolcpus= in 8.04? I can't seem to find anything with my googlefu
[07:08:34] <V0idExp> Hello everybody!
[07:11:23] <V0idExp> Using linuxcnc 2.5.1 (shipped with ubuntu, got it from linuxcnc.org). I have the following configuration: XYZ stepper + encoder machine, which basically does "painting". Briefly, it's a machine that draws figures on a XY plane, and Z axis raises/lowers the drawing head
[07:12:34] <V0idExp> The problem is that my figures are very detailed, I mean, there are commands G01 X.. Y.. sent in some places every millimeter or sometimes even more.
[07:12:58] <V0idExp> So, the machine does not move at full speed (the one specified in G01 command with F parameter)
[07:13:54] <V0idExp> Because it accelerates and slows at each point, and can't manage to reach full speed. All points are commanded to be reached at speed F2000, but rarely I can reach those speeds.
[07:14:47] <V0idExp> I noticed that with less detail, or said in other terms, with more space between points, actual speeds increase
[07:15:24] <V0idExp> Is there any way to tell LinuxCNC to don't slow down at each point but keep going at commanded speed?
[07:21:17] <DJ9DJ> G64 may help
[07:21:26] <DJ9DJ> http://linuxcnc.org/docs/html/gcode/gcode.html#_g64_path_blending_a_id_sec_g64_path_blending_a
[07:22:29] <V0idExp> Thanks! Reading it now
[07:22:57] <DJ9DJ> not sure about that, but thats at least what I understood ;)
[07:23:23] <V0idExp> Already have a question... :) Does it throws points away?
[07:24:55] <DJ9DJ> i think it does not throw points completely away, but yes, it decreases accuracy
[07:24:55] <V0idExp> Imagine a triangle. What my NGC program generator does it is adding very much detail at angles and almost no points at straight lines. (ideally only 2, but being the figure got from machine vision software, sometimes it does not see "ideally straight" lines)
[07:26:00] <DJ9DJ> i think linuxcnc cannot optimize your gcode...
[07:26:30] <V0idExp> No, no! My GCODE is already optimized this way :)
[07:26:46] <V0idExp> The only problem is that linuxcnc doesn't go at nominal speeds, and that's what I need
[07:27:38] <DJ9DJ> yeah, if you have very many points next to each other, the machine will slow down to "hit" each point exactly
[07:27:45] <V0idExp> Is there a way to force it to don't try to correct speeds based on geometry, but just to try to go through commanded points at speeds that I programmed?
[07:27:58] <V0idExp> Ok, understood
[07:28:04] <V0idExp> Well, this explains many things :(
[07:28:23] <Tom_itx> sounds like a crappy gcode optimizer
[07:28:32] <DJ9DJ> but perhaps some of the old linuxcnc experts might help you better than me, i am just a noob without much experience
[07:28:53] <V0idExp> So, what you suggest, is to add more tollerance to G64 commands?
[07:28:55] <DJ9DJ> nor in cnc stuff, neither with linuxcnc ;)
[07:29:27] <DJ9DJ> yes, that was my intentional idea. set something like G64 Pxxxx
[07:29:35] <DJ9DJ> where xxxx i the tolerance
[07:29:40] <DJ9DJ> *is
[07:30:15] <V0idExp> thank you!
[07:30:31] <DJ9DJ> no prob :)
[07:35:09] <jthornton> V0idExp, have you read this? http://linuxcnc.org/docs/html/common/User_Concepts.html
[07:50:43] <V0idExp> Reading it now ;)
[08:46:15] <Mr_Wolfsl> Hi
[08:47:14] <Mr_Wolfsl> anyone know the kinematics 5axiskins.c?
[09:07:32] <cradek> hi Mr, go ahead and ask your question about it
[09:07:45] <cradek> then if someone can, they will help
[09:33:59] <Mr_Wolfsl> hi cradek how can I import the tool-length in the kinematics?
[09:34:04] <Mr_Wolfsl> my haldata is empty
[09:35:16] <cradek> that's kind of the wrong approach. you probably want to put your tool lengths along W and set up your kinematics so a change in W is always along the tool vector
[09:35:59] <cradek> otherwise you will get jumps in position when the offset changes, which will make your machine disagree with you
[09:36:10] <Mr_Wolfsl> I put the length in the colum W in my tool table
[09:36:52] <cradek> ok, good
[09:37:48] <cradek> so in your gcode you do G43 and you'll see W change. then you can issue G0 W0 and you'll get a nice move that applies the tool length
[09:38:02] <Mr_Wolfsl> when I put the pivot-length + tool-length manually in the .hal it's OK
[09:38:13] <Mr_Wolfsl> yes I have the G43
[09:38:57] <Mr_Wolfsl> I see the W change in the GUI not with the kins
[09:39:14] <cradek> well you don't want the tool length in the hal file, because it changes! like all the other offsets, that is a part of the joint position your kins gets.
[09:40:21] <cradek> changing an offset will NEVER cause a change that kins can see, because changing an offset does not move the machine. on your next programmed move you'll get the motion with respect to the new offset. that's why I said you then issue G0 W0
[09:40:45] <cradek> are you following that? it's critical you understand it
[09:41:00] <Mr_Wolfsl> ok i will try
[09:43:17] <Mr_Wolfsl> I tried but I have joint 8 following error
[09:44:11] <cradek> tried what?
[09:44:25] <Mr_Wolfsl> put G0 W0 in the Gcode
[09:44:59] <cradek> you'll have to loopback the W motion pins in the hal
[09:45:24] <cradek> since W isn't a real joint with real feedback
[09:48:34] <Mr_Wolfsl> it's already done
[09:51:35] <cradek> if you got a ferror, you don't have it looped correctly. pasetbin your hal file?
[09:51:46] <cradek> er pastebin
[09:52:37] <Mr_Wolfsl> # core HAL config file for simulation - 5 axis
[09:52:37] <Mr_Wolfsl> # load RT modules: first one the 5-axis kinematics
[09:52:37] <Mr_Wolfsl> loadrt 5axiskins
[09:52:37] <Mr_Wolfsl> loadrt [EMCMOT]EMCMOT base_period_nsec=[EMCMOT]BASE_PERIOD servo_period_nsec=[EMCMOT]SERVO_PERIOD num_joints=[TRAJ]AXES
[09:52:37] <Mr_Wolfsl> loadrt ddt count=9
[09:52:39] <Mr_Wolfsl> loadrt probe_parport
[09:52:41] <Mr_Wolfsl> loadrt pwmgen output_type=0
[09:52:44] <Mr_Wolfsl> # add motion controller functions to servo thread
[09:52:46] <Mr_Wolfsl> addf motion-command-handler servo-thread
[09:52:48] <Mr_Wolfsl> addf motion-controller servo-thread
[09:52:50] <Mr_Wolfsl> addf ddt.0 servo-thread
[09:52:52] <Mr_Wolfsl> addf ddt.1 servo-thread
[09:52:54] <Mr_Wolfsl> addf ddt.2 servo-thread
[09:52:56] <Mr_Wolfsl> addf ddt.3 servo-thread
[09:52:58] <Mr_Wolfsl> addf ddt.4 servo-thread
[09:53:00] <cradek> no no no
[09:53:00] <Mr_Wolfsl> addf ddt.5 servo-thread
[09:53:03] <Mr_Wolfsl> addf ddt.6 servo-thread
[09:53:05] <Mr_Wolfsl> addf ddt.7 servo-thread
[09:53:05] <cradek> use a pastebin website
[09:53:06] <Mr_Wolfsl> addf ddt.8 servo-thread
[09:53:08] <Mr_Wolfsl> # create HAL signals for position commands from motion module
[09:53:10] <Mr_Wolfsl> newsig J0pos float
[09:53:12] <Mr_Wolfsl> linksp J0pos <= axis.0.motor-pos-cmd
[09:53:17] <Mr_Wolfsl> newsig J1pos float
[09:53:18] <Mr_Wolfsl> linksp J1pos <= axis.1.motor-pos-cmd
[09:53:21] <Mr_Wolfsl> newsig J2pos float
[09:53:23] <Mr_Wolfsl> linksp J2pos <= axis.2.motor-pos-cmd
[09:53:25] <Mr_Wolfsl> newsig J3pos float
[09:53:26] <Mr_Wolfsl> linksp J3pos <= axis.3.motor-pos-cmd
[09:53:28] <Mr_Wolfsl> newsig J5pos float
[09:53:30] <Mr_Wolfsl> linksp J5pos <= axis.5.motor-pos-cmd
[09:53:32] <Mr_Wolfsl> # loop position commands back to motion module feedback
[09:53:34] <Mr_Wolfsl> linksp J0pos => axis.0.motor-pos-fb
[09:53:36] <Mr_Wolfsl> linksp J1pos => axis.1.motor-pos-fb
[09:53:38] <Mr_Wolfsl> linksp J2pos => axis.2.motor-pos-fb
[09:53:40] <Mr_Wolfsl> linksp J3pos => axis.3.motor-pos-fb
[09:53:43] <Mr_Wolfsl> linksp J5pos => axis.5.motor-pos-fb
[09:53:46] <Mr_Wolfsl> # send the position commands thru differentiators to
[09:53:48] <Mr_Wolfsl> # generate velocity and accel signals
[09:53:50] <Mr_Wolfsl> # define the signals, and hook them up
[09:53:52] <Mr_Wolfsl> newsig J0vel float
[09:53:54] <Mr_Wolfsl> newsig J0acc float
[09:53:56] <Mr_Wolfsl> linksp J0pos => ddt.0.in
[09:53:58] <Mr_Wolfsl> linkps ddt.0.out => J0vel
[09:54:00] <Mr_Wolfsl> linksp J0vel => ddt.1.in
[09:54:03] <Mr_Wolfsl> linkps ddt.1.out => J0acc
[09:54:05] <Mr_Wolfsl> newsig J1vel float
[09:54:07] <Mr_Wolfsl> newsig J1acc float
[09:54:08] <Mr_Wolfsl> linksp J1pos => ddt.2.in
[09:54:10] <Mr_Wolfsl> linkps ddt.2.out => J1vel
[09:54:12] <Mr_Wolfsl> linksp J1vel => ddt.3.in
[09:54:17] <Mr_Wolfsl> linkps ddt.3.out => J1acc
[09:54:18] <Mr_Wolfsl> newsig J2vel float
[09:54:19] <DJ9DJ> muha
[09:54:20] <Mr_Wolfsl> newsig J2acc float
[09:54:22] <Mr_Wolfsl> linksp J2pos => ddt.4.in
[09:54:24] <Mr_Wolfsl> linkps ddt.4.out => J2vel
[09:54:26] <Mr_Wolfsl> linksp J2vel => ddt.5.in
[09:54:29] <Mr_Wolfsl> linkps ddt.5.out => J2acc
[09:54:31] <Mr_Wolfsl> newsig J3vel float
[09:54:33] <Mr_Wolfsl> newsig J3acc float
[09:54:35] <Mr_Wolfsl> linksp J3pos => ddt.6.in
[09:54:37] <Mr_Wolfsl> linkps ddt.6.out => J3vel
[09:54:37] <cradek> http://i94.photobucket.com/albums/l112/nbbruin26/Derp-Muppets.gif
[09:54:38] <Mr_Wolfsl> linksp J3vel => ddt.7.in
[09:54:40] <Mr_Wolfsl> linkps ddt.7.out => J3acc
[09:54:43] <Mr_Wolfsl> # amp control
[09:54:44] <Mr_Wolfsl> newsig xena bit
[09:54:48] <Mr_Wolfsl> linksp xena axis.0.amp-enable-out
[09:54:50] <Mr_Wolfsl> newsig yena bit
[09:54:53] <Mr_Wolfsl> linksp yena axis.1.amp-enable-out
[09:54:55] <Mr_Wolfsl> newsig
[09:54:57] <Mr_Wolfsl> oops sorry
[09:54:59] <Mr_Wolfsl> I don't know what a pastebin website is
[09:55:07] <cradek> I should've been more specific
[09:55:19] <cradek> go to pastebin.com or pastebin.ca and paste your file there, then copy the url here
[09:55:33] <cradek> it's a common way people share large blobs of text
[09:55:55] <Mr_Wolfsl> ok
[09:57:31] <Mr_Wolfsl> http://pastebin.com/6pEsvcL4
[09:57:40] <Mr_Wolfsl> my hal is for the simulation
[09:58:02] <Mr_Wolfsl> cause I have tried with my machine
[09:58:11] <Mr_Wolfsl> not in simulation
[10:01:21] <Mr_Wolfsl> http://pastebin.com/XYN9bbpv
[10:01:32] <Mr_Wolfsl> this is my real machine
[10:02:07] <cradek> http://pastebin.com/6pEsvcL4 does not contain anything about axis.8, so you don't have it looped
[10:03:06] <Mr_Wolfsl> ok how can I loop it
[10:03:49] <cradek> net whatever axis.8.motor-pos-cmd axis.8.motor-pos-fb
[10:05:22] <Mr_Wolfsl> and for the simulation?
[10:26:26] <Mr_Wolfsl> Do you have any idea for the simulation?
[10:53:21] <mrsun> hmm new gib made huge difference, but i think i got to small ball bearing balls, got the presure point for the gib to low even lower then the original gibscrews .. might be fixable by making a 3mm "flat" in the gibs at the center of the gibscrews .. hmm
[10:54:03] <mrsun> should be 3.5 or 3.6mm balls insted but cant get that anywhere here :/
[10:54:48] <cncbasher> yea come in through the screw holes with a long drill and get the angle in the gib
[10:55:59] <cncbasher> yes 3.5mm balls or there about works for me
[10:56:03] <mrsun> yeah but there is so little clearance that i cant get a drill in there with any kind of power thingie :/
[11:01:27] <mrsun> 3.5 doesnt seem to be a very common size :P
[11:01:54] <mrsun> more then on ebay
[11:01:56] <mrsun> but then i have to wait
[11:01:59] <mrsun> a long time :P
[11:02:12] <mrsun> but damn i could tighten the new gib ALOT more then the old before binding :P
[11:02:24] <mrsun> beforei had like .5mm rattle of the table at the ends
[11:03:04] <mrsun> now i got about 0.01 in one end and like 0.1 at the other most extreme ... dont know why i cant tighten it up more in that end tho might be the dovetails that are out that much also :/
[11:04:01] <mrsun> using 3.5 insted of 3 i guess would lift the center about 0.25mm dont know if that is enough... :/
[11:04:20] <mrsun> the dovetail of the table is a little to small .. should have been 1mm longer and it would be alot better
[11:09:48] <mrsun> one of the problems with small stuff i guess :P
[11:12:15] <r00t4rd3d> http://imgur.com/a/urmuz
[11:23:12] <mrsun> hmm might have to go and make a go two directly if im lycky i havent thrown away the "flex" thingie i had .. it wasnt very flexy but might work to get into those crammed placed :P
[11:27:43] <Loetmichel> re @ home
[11:40:27] <ReadError> hey yall
[11:40:37] <ReadError> was wondering if any gcode wizards could look at something
[11:40:39] <ReadError> http://pastebin.com/QKA7CUjR
[11:40:59] <ReadError> im getting an error "Near 3132"
[11:41:18] <ReadError> Cannot use axis values without gcode that uses them
[11:41:41] <IchGuckLive> hi
[11:41:57] <ReadError> morning sir
[11:42:21] <DJ9DJ> hi there
[11:43:55] <cradek> ReadError: didn't we answer this question already? maybe it was someone else... after g80 you must specify a new motion code. http://linuxcnc.org/docs/html/gcode/gcode.html#sec:G80-Cancel-Modal
[11:44:19] <ReadError> oh wasnt me, but thank you sir
[11:44:35] <cradek> heh, no problem, sorry for assuming
[11:45:00] <cradek> it was just so familiar I knew the answer before looking at your gcode, haha
[11:45:36] <ReadError> N15640 G81 X6.7669 Y-1.0756 Z-0.25 R0.2 F4.
[11:45:36] <ReadError> N15645 G80
[11:45:37] <ReadError> N15650 Z0.6
[11:45:47] <ReadError> so that needs to be G0 Z0.6
[11:45:58] <cradek> yes, if you want a rapid there
[11:46:02] <ReadError> its throwing an arbitrary Z value i see
[11:46:12] <ReadError> without prefacing it with a G code
[11:51:13] <ReadError> ahh i see, seems its drilling algorithm is broke in this software
[12:03:44] <ReadError> hsmexpress seems pretty decent for a free cam application
[12:07:38] <IchGuckLive> use heekscad on drilling O.O B)
[12:08:13] <ReadError> its not that many lines
[12:08:22] <ReadError> i should learn some gcode and fix it
[12:13:34] <tjb1> Smart alarm on android is amazing...
[12:14:04] <ReadError> does it use the accel to tell when you are almost awake?
[12:14:26] <IchGuckLive> ReadError: or just klick the center and drill then hit Go and get the gcode
[12:15:03] <tjb1> No it just starts real low
[12:15:27] <tjb1> I tried that fancy one that watches your movement and about melted the phone
[12:23:27] <IchGuckLive> ReadError: do yu got apattern to share with us ?
[12:24:33] <ReadError> http://pastie.org/private/vsy5zcqkjynoqd5yitzk1w
[12:24:37] <ReadError> thats the bad gcode
[12:24:45] <ReadError> i assume the pattern is good but the syntax is wrong
[12:27:18] <IchGuckLive> why dident you get the 3rd drill also in
[12:27:47] <IchGuckLive> ok other depth
[12:27:52] <IchGuckLive> O.O
[12:28:03] <ReadError> well the depth should be the same
[12:28:14] <ReadError> i think its an error in the generation maybe too
[12:28:23] <IchGuckLive> Z-0.25 ! Z0
[12:29:27] <IchGuckLive> so Z0 shoudt go Z-0.25 ß
[12:30:41] <IchGuckLive> is this just a snippsel as i see high linenumbers
[12:31:20] <ReadError> nah thats the broken part
[12:31:28] <ReadError> http://pastebin.com/QKA7CUjR
[12:31:31] <ReadError> thats the full code
[12:32:02] <IchGuckLive> G98 and R is bad
[12:32:21] <IchGuckLive> as G98 Retracts to the prior (Started) T
[12:32:24] <IchGuckLive> Z
[12:32:58] <IchGuckLive> get the G98 off
[12:33:24] <IchGuckLive> N15635 no Gcode
[12:33:35] <IchGuckLive> yust number
[12:33:54] <IchGuckLive> is there a hole YES
[12:34:17] <IchGuckLive> so no N15625 G80
[12:34:19] <IchGuckLive> N15635 X6.7669 Y-1.0756 Z0.2
[12:34:27] <ReadError> let me grab the gcode from my other generation
[12:34:47] <IchGuckLive> N15640 G81 and here the G81 is also a fail
[12:35:12] <IchGuckLive> or just get N15635 G0 X6.7669 Y-1.0756 Z0.2 and you are done
[12:35:22] <ReadError> http://pastie.org/5669617
[12:35:32] <ReadError> thats some working gcode from another cam app
[12:37:17] <IchGuckLive> are you doing a postprocessor rewrite ?
[12:38:08] <ReadError> oh i was just trying this hsmworks software
[12:39:15] <IchGuckLive> ok
[12:39:49] <IchGuckLive> do you got a dxf ?
[12:40:47] <ReadError> i can export one
[12:42:07] <IchGuckLive> nice if you do so
[12:43:15] <ReadError> grr i think this HD is about dead
[13:03:33] <mrsun> http://imagebin.org/242524 point A and point B is like 150mm from the center, in each direction when the table is at point A and i rock it i get about 0.1mm of "rocking" at point A, at point B i get about 0.03mm of rocking, im guessing in relation to the center point (the end mill) this rocking is very amplified am i right? at these points what would be acceptable values of "rocking" of the table? :)
[13:06:59] <IchGuckLive> optical sisaorder
[13:09:03] <mrsun> the green and red represents the table in the far extremes of the travel to left and right
[13:09:18] <andypugh> No, movement at the tool will be slightly less than at points A and B
[13:09:30] <andypugh> (About half)
[13:09:37] <mrsun> andypugh, huh ?
[13:10:03] <mrsun> cant measure the movement at the tool and if i rock the table at the far extremes the movement at the tool (center black point) should be very small shouldnt it? :)
[13:10:10] <andypugh> The ditance from the pivot (the edge of the black block) is less for the tool than for your measuement point
[13:10:45] <andypugh> We are vehemently agreeing by using different words.
[13:11:01] <mrsun> oh :P
[13:11:19] <mrsun> cant tighten the gibs any more for point A tho, table binds. ..
[13:11:28] <mrsun> might be better if i scrape the gibs real flat tho
[13:11:36] <mrsun> as they can differ some in height as its just flatstock ive used
[13:12:42] <andypugh> Scraping is traditional :-)
[13:14:05] <mrsun> the dovetails might need some touchup also ... ive scraped the flat surfaces of everything so its very very flat for being a sieg X1 mill :P
[13:15:15] <mrsun> but the dovetails are sooo small its hard to even get a very small scraping tool into them :P
[13:15:47] <mrsun> but anyways what would be acceptable values in rocking of table on a mill :/
[13:15:56] <mrsun> there has to be some "slop" else everything just binds up
[13:15:57] <mrsun> but how much
[13:16:52] <tjb1> Tom_itx: I was sent to infest this channel of 3d printers and the like.
[13:24:11] * mrsun has also learned today not to work on the mill table with an endmill in the spindle
[13:24:16] <mrsun> damn those things are sharp
[13:25:38] <andypugh> Hole in the back of your hand now? I have never done that.
[13:27:10] <mrsun> http://www.engineeringtoolbox.com/friction-coefficients-d_778.html hmm why is some values missing .. like cast iron - brass lubricated
[13:27:42] <mrsun> andypugh, nah .. shaved of some skin from the side of the finger ... mill wasnt running luckily :P
[13:27:59] <mrsun> hmm maybe should make gibs inc ast iron :P
[13:28:07] <mrsun> cast iron - cast iron .. 0.07 lubricated
[14:03:57] <ReadError> dang this HD is practically dead
[14:04:05] <ReadError> i need to migrate data off asap
[14:05:20] <Jymmm> ReadError: Freeze it if you can
[14:05:34] <ReadError> Jymmm: its still chugging
[14:05:40] <ReadError> just 10-20MB/s
[14:05:49] <ReadError> fortunately i have a 2TB installed already
[14:05:55] <ReadError> just need to change over
[14:06:03] <Jymmm> ReadError: You probably shouldn't wait to image it
[14:06:19] <ReadError> Jymmm: its on my hackintosh
[14:06:37] <ReadError> im just going to re-install with mountain lion and migrate my home directory over
[15:22:21] <tjtr33> jthornton, how goes the touchscreen? just got KeyTec MagicTouch running on 10.04, will now try touchy out
[16:17:08] <DJ9DJ> gn8
[16:38:05] <andypugh> Feeling smug :-) https://picasaweb.google.com/lh/photo/0gdDpGWrW8U1R6aLwPmnVNMTjNZETYmyPJy0liipFm0?feat=directlink
[16:39:58] <andypugh> It's the worm from a dividing head. I was a bit more relaxed when I found that a replacement was "only" £70
[16:39:59] <andypugh> https://picasaweb.google.com/lh/photo/MF2HxVmb9Z_tlVzJl0IFwtMTjNZETYmyPJy0liipFm0?feat=directlink
[16:40:32] <cradek> beautifully done
[16:41:32] <andypugh> I was worried that it might be too hard, but once CBN-ed through the case it was fine.
[16:42:04] <andypugh> (I didn't make the worm part, I hasten to add, just machined down the shaft, shortened it, and bored the hex in the end)
[16:45:33] <andypugh> The plan is to have a hex collar around the motor shaft, and then I can have a neat installation with the moto r shaft inside the worm shaft, rather than a really long extension tube with a coupling in it, like this one: http://divisionmaster.co.uk/examples_files/divisionmaster_data/vertex-bs0.jpg
[16:47:40] <cradek> neato
[16:48:44] <andypugh> Of course this relies on me managing to get a hex exactly coaxial with a bore.
[16:49:30] <andypugh> In theory I could make the hex spigot on the lathe too :-)
[16:51:43] <cradek> do you want a slightly tapered hex? that'd be no harder with your setup...
[16:51:52] <andypugh> I considered making the socket pentagonal, just for fun.
[16:52:06] <cradek> ha
[16:52:14] <andypugh> No, the whole point is to have no end thrust on the motor shaft.
[16:52:26] <cradek> oh, gotcha
[16:52:39] <andypugh> Though I accept that your plan means that I can have zero backlash.
[16:52:48] <cradek> sure
[16:53:11] <cradek> hard to meet both goals
[16:53:20] <cradek> gotta run, bbl.
[16:58:12] <ReadError> andypugh: is that for a rotary table?
[17:07:29] <andypugh> ReadError: Yes
[17:08:01] <andypugh> Well, technically a dividing head. Shall we call it a "4th axis"
[17:29:17] <andypugh> Any welders here? This is remarkable: http://youtu.be/ygcm0AQXX9k
[17:44:07] <ReadError> wow thats pretty awesome
[17:45:11] <andypugh> No more perfect beads 1/2" to the left of where the panels meet :-)
[17:53:40] <tjtr33> and you get to be a cyberman too http://en.wikipedia.org/wiki/Cyberman
[17:57:36] <ReadError> HSMXpress is pretty nice, most of the features i need atleast
[17:57:57] <ReadError> besides bunk drilling gcode
[18:02:40] <andypugh> I imagine I will someday start making things, instead of making things to make things. Or software to make things to make things.
[18:04:41] <ReadError> only dumb thing is, it goes to home, which it thinks is 0,0,0 after the job
[18:04:48] <ReadError> and that would crash my spindle in to the table
[18:04:59] <ReadError> so i need to change the postprocessor
[18:06:07] <tjtr33> sounds like a static postamble is used
[18:07:59] <jdh> or that's the default and he didn't change it.
[18:11:32] <andypugh> Des it return to G33 perhaps?
[18:15:36] <ReadError> nope
[18:15:41] <ReadError> it has some custom stuff
[18:15:45] <ReadError> let me paste in pastie
[18:17:12] <ReadError> http://pastebin.com/bBG8HYbZ
[18:17:39] <ReadError> but theres no way I can tell that lets the software know where home really is
[18:18:05] <ReadError> so i just commented it out, it alread does Z0.6 which is good enough
[18:19:14] <andypugh> There are a couple of places that you can define in machine goordinated. G33 is one
[18:19:51] <ReadError> http://pastebin.com/Qzh6wScZ
[18:20:00] <ReadError> thats the "official" emc post processor
[18:20:55] <ReadError> so using that, it basically tries to send the machine to 0,0,0
[18:21:40] <tjtr33> this is C? but interpreted? (homeX, homeY looks like the bugger)
[18:22:05] <ReadError> javascript i think
[18:22:43] <tjtr33> and they are assigned values by smarter-than-you ( aka stupid) code
[18:22:54] <ReadError> apparently
[18:22:55] <tjtr33> the code is stupid
[18:23:08] <ReadError> lol
[18:23:11] <tjtr33> xyzFormat.format(machineConfiguration.getHomePositionX());
[18:23:12] <andypugh> Do I see G28 on line 1206?
[18:23:16] <tjtr33> chg that ^^^ cfg
[18:23:45] <ReadError> i just commented out the lines
[18:23:56] <ReadError> brings Z up 0.6" which is good enough for me
[18:24:21] <andypugh> I think it might be going to G28: http://www.linuxcnc.org/docs/2.4/html/gcode_main.html#sub:G28,-G30:-Return
[18:25:30] <ReadError> N16800 G0 Z0.6
[18:25:30] <ReadError> N16810 M9
[18:25:30] <ReadError> N16815 G28 G91 Z0.
[18:25:30] <ReadError> N16820 G28 X0. Y0.
[18:25:30] <ReadError> N16825 M30
[18:25:37] <ReadError> thats what it was doing
[18:25:44] <andypugh> Yes, G28.
[18:25:58] <ReadError> so rapid to Z0 ?
[18:26:10] <andypugh> Define your G28 and I think it will be OK
[18:26:57] <ReadError> theres no where in the postprocessor or configuration to denote that
[18:27:50] <andypugh> It's a machine setting
[18:27:59] <andypugh> See that link I posted
[18:28:14] <tjtr33> in Linuxcnc define g28 http://www.linuxcnc.org/docs/2.4/html/gcode_main.html
[18:28:26] <tjtr33> like andy said
[18:30:07] <tjtr33> we didnt see your gcode, just the postprocesser source
[18:31:35] <ReadError> so is G28 defined within my actual linuxcnc configuration or in each gcode file?
[18:32:00] <andypugh> It is stored in persistent variables in your config
[18:32:26] <andypugh> Jog somewhere safe, MDI a G28.1 and then see what happens
[18:33:42] <tjtr33> stored in 5161-5166? 6 or 9 parms?
[18:34:30] <tjtr33> eh, an nc program should not rely on external params... thats ... dangerous
[18:35:24] <andypugh> G28 is meant to be the safe position.
[18:35:37] <ReadError> oh neat
[18:35:42] <ReadError> i set G28.1
[18:35:47] <andypugh> The #params are just to set and view the position.
[18:35:52] <ReadError> then moved it, ran G28 and it returned back to that position
[18:36:22] <tjtr33> yeh, i've used similar, but what we saw was, it wasnt really safe, it was 'assumed' safe
[18:36:37] <ReadError> im too absent minded sometimes to remember to set this
[18:37:14] <tjtr33> make your code stand alone then, fix it a the post source
[18:39:06] <ReadError> when i first built my mill, i was like oh this will be easy, draw some stuff, click some buttons and im done
[18:39:09] <ReadError> booooy was i wrong
[18:40:30] <tjtr33> its responsibility, the g28 is ok IF the position is ok ( catch22) and stand alone code is ok. its still big metals and sharp edges following specific commands (pretty unforgiving )
[18:40:42] <andypugh> Well, you _can_ do that. But you get scrap and broken tools. mainly
[18:45:34] <tjtr33> this is intersting in that post "setWorkPlane(new Vector(0, 0, 0)); // reset working plane" can they tilt the working plane? ( circles outside of g17/18/19)
[18:46:10] <andypugh> I am not sure.
[18:46:42] <tjtr33> doesnt amtter,c ant afford SW anyway :P
[18:46:47] <tjtr33> gnite
[18:46:54] <andypugh> You can certainly rotate X and Y about Z
[18:47:15] <tjtr33> i meant it looked like a normal to a surface
[18:47:45] <tjtr33> thx again
[18:48:27] <ReadError> bahhh
[18:48:31] <ReadError> even in this version
[18:48:36] <ReadError> holes break gcode
[18:50:15] <andypugh> Or does G-code break holes?
[18:50:37] <ReadError> something so simple too
[18:59:15] <ReadError> actually this might be something that can fixed in the post processor
[19:05:57] <Jymmm> When a fuel pump goes out, will it show a open? (not even a lil resistance?)
[19:08:23] <jdh> depends on failure cause
[19:15:14] <ReadError> hm ok, so
[19:15:37] <ReadError> it was generating: G80
[19:15:37] <ReadError> Z0.6
[19:15:46] <ReadError> i changed that to G0 Z0.6
[19:15:48] <ReadError> works that way
[19:18:52] <andypugh> Jymmm: A fuel pump can fail mechanically but be electrically perfect.
[19:19:21] <Jymmm> andypugh: Well, unless I missed something... I tested the fuel pump from the last connector directly to the battery - nothing, not even a noise. Then I smacked the gas tank with a dead blow hammer while powered directly to the battery - still nothing.
[19:19:29] <ReadError> ahhh, i think i need to do G80 to stop motion, then issue the G0 Z0.6
[19:19:49] <andypugh> G80 cancels canned cycle mode, as far as I know
[19:19:58] <Jymmm> andypugh: I then ohm'ed it out, showing completely open.
[19:19:59] <ReadError> yea
[19:20:20] <ReadError> G80 Cancel Modal Motion
[19:21:14] <andypugh> What sort of pump? The diaphragm pumps will show open until the diaphragm moves back far enough cause the next pulse.
[19:21:57] <Jymmm> andypugh: http://www.autozone.com/autozone/parts/Airtex-Master-Fuel-Pump/2001-Ford-Expedition-2WD/_/N-jfarqZ8vcz2?itemIdentifier=491824_29353_0_4568
[19:24:27] <andypugh> I think that's a rotary pump. If it shows open-circuit then it's dead.
[19:24:51] <Jymmm> andypugh: Sad thing... KBB Private party sale "Very Good" is only $3500
[19:25:27] <andypugh> I have no idea what those phrases mean in that sequence
[19:25:32] <Jymmm> andypugh: go ahead, say it....
[19:25:42] <Jymmm> andypugh: you KNOW you want to =)
[19:26:23] <andypugh> I am not sure what sort of "Private Parties" cost $3500
[19:26:40] <Jymmm> vs tradein value (which is always lower)
[19:27:13] <andypugh> Sorry, no, still lost. Is the KBB like the KKK?
[19:27:50] <Jymmm> Kelley Blue Book... A standard used for pricing cars http://kbb.com/
[19:28:26] <andypugh> Ah, OK, so you are still talking about your car.
[19:28:58] <Jymmm> Yeah, $1100 for them to change the pump, but the car is only worth $3500,
[19:29:11] <andypugh> The most I have ever paid for a car was $350.
[19:29:11] <Jymmm> It's not even a frickin blown engine!
[19:30:10] <andypugh> $1100 to change the pump is insane. I have done it at work in 10 minutes in the carp-ark, because I couldn't be bothered booking the job into the workshop.
[19:30:52] <jdh> in-tank pump?
[19:31:15] <andypugh> Yes
[19:32:24] <andypugh> It helped that there was a velcroed-on panel under the rear seat, that the rear seat was just the foam, with no uphoulstery and not held on in any way.
[19:33:14] <andypugh> I suppose if they have to drop the rear suspension to get the tank off then it's no fun.
[19:33:15] <ReadError> hey yall, quick question, i seem to have trouble using a 1-2-3 block to get my touchoff for Z perfect
[19:33:24] <ReadError> lately I have been turning on the mill
[19:33:27] <andypugh> Bit there normally is an access hatch.
[19:33:42] <ReadError> and going down in values of 0.001 until i can see it start to chip
[19:33:52] <ReadError> then going up a little
[19:33:59] <ReadError> is that a viable touchoff method?
[19:34:18] <jdh> get a gauge pin
[19:34:27] <andypugh> How much mechanical hysteresis do you have?
[19:35:35] <ReadError> oh interesting
[19:36:02] <ReadError> me andypugh?
[19:36:19] <andypugh> Yes
[19:36:40] <ReadError> oh im 100% keyboard jockey
[19:36:54] <ReadError> so im basically forced to learn as I go
[19:37:37] <andypugh> If you put a DTI on the head, or the tool. How far do you have to jog back up after jogging down before the needle moves?
[19:38:10] <ReadError> well im doing backlash compensation on x/y
[19:38:19] <ReadError> but nothing on z, have not measured it TBH
[19:38:20] <andypugh> My machine is 0.05mm, I am hoping I can improve that.
[19:43:58] <ReadError> ill never be skilled enough to get anywhere near that
[19:44:49] <jdh> my Z is 0.054 but I currently have a delrin ballnut mount
[19:46:43] <Jymmm> andypugh: No suspenion to deal with, just everything else
[19:47:08] <Jymmm> andypugh: I just dont have the tools compressor, air ratchet, jacks, etc to do it
[19:47:30] <Jymmm> andypugh: I might see if there is a access panel, but I doubt it
[19:47:40] <andypugh> You sure there is no access from under the seat?
[19:47:59] <Jymmm> andypugh: I'm going to check in a moment
[19:48:06] <andypugh> If there is a panel, then you can do the job with a full tank
[19:48:43] <Jymmm> I'd have to do it with a full tank anyway, I have no place to put 20Gallons of gas
[19:51:21] <andypugh> You could go shopping. Your US barge probably only does 3mpg
[19:51:46] <jdh> I know a guy that used a 9v battery in series to jump the pump past the dead spot
[19:52:25] <jdh> once a month or so for a year
[19:57:51] <Jymmm> andypugh: Nope, not unless I grab the angle grinder
[19:58:24] <ReadError> doesnt andy work for ford? ;p
[19:58:39] <Jymmm> jdh: you mean to 21V?
[19:59:34] <Jymmm> jdh: When I ohm'ed it out, it's showing a full open. Not sure if 24VDC would make a difference
[20:00:01] <jdh> dunno. he had a chevy
[20:00:31] <Jymmm> Id be far easier to find a bolt on pump
[20:00:42] <Jymmm> but I dont think they exist
[20:01:00] <Jymmm> min psi 85
[20:01:08] <jdh> I had to go pick him up at the gate once. Then he got it fixed.
[20:01:08] <Jymmm> min gph 40
[20:05:45] <jdh> submerged for cooling and noise
[20:07:14] <andypugh> Jymmm: If I were you I think I might look quite seriously at the "making an access panel" idea.
[20:07:29] <andypugh> I have a nibbler.
[20:07:51] <Jymmm> andypugh: If I knew EXACTLY where to cut, I would.
[20:08:42] <jdh> is that not googleable?
[20:10:13] <Jymmm> jdh: If it is, my GoogleFu is no good
[20:10:56] <andypugh> Jymmm: http://www.ebay.co.uk/itm/Trick-Flow-High-Flow-Electric-Fuel-Pump-43-GPH-85-PSI-25004P-/370724001947?pt=Motors_Car_Truck_Parts_Accessories&hash=item5650df589b
[20:12:13] <Jymmm> andypugh: piggy back it?
[20:12:50] <andypugh> There are more on eBay at 100psi, I didn't try 90psi
[20:13:16] <andypugh> You ought to be able to suck through the existing pumo
[20:14:06] <Jymmm> Yeah, I was thinking that. I even saved the old fuel filter that I could use to cut apart to make fittings to tap into the existing fuel lines
[20:14:48] <andypugh> I am not advocating that course of action.
[20:14:57] <Jymmm> which?
[20:15:01] <andypugh> Any
[20:15:03] <andypugh> :-)
[20:15:10] <Jymmm> yes you are dmanit =)
[20:15:49] <Jymmm> But.... what I can do is disconnected the filter and try to suck from there to see if I get any flow
[20:16:56] <Jymmm> andypugh: can you think of any reason why it wouldn't work?
[20:17:14] <Jymmm> emissions, whatever
[20:17:21] <andypugh> No, none.
[20:17:55] <Jymmm> there is a fuel tank pressure sensor in the tank, but I think an external pump would still trigger that correctly
[20:19:03] <andypugh> It's my job to make sure that the diesel HP pump has enough pressure, but the only reason not to run the pump full-bore is to avoid wasting electrical power.
[20:19:19] <Jymmm> heh
[20:19:39] <Jymmm> I think I can waste a few amps on the laternator
[20:19:42] <Jymmm> alternator
[20:20:05] <andypugh> Ah, no, if the sensor for "enough pressure to crank" is in the tank then an external pump might have a problem.
[20:20:42] <Jymmm> I dont know what it's for, it does goto the PCM
[20:21:14] <Jymmm> just says "Fuel Tank Pressure Sensor"
[20:21:20] <andypugh> We have 800W of heated windscreen and 1000W of cabin heater, plus fans and power steeering, even the 240A alternator is a bit close to the edge.
[20:22:21] <Jymmm> andypugh: fwiw I have the FORD (tm) Wiring Diagram manual
[20:22:44] <Jymmm> now some 3rd party crap
[20:22:48] <Jymmm> not
[20:26:31] <skunkworks> andypugh: did you try your hex yet?
[20:28:09] <jdh> I paid like $250 for fuel pump replacement in a suburban with a 40 gallon tank
[20:31:01] <andypugh> skunkworks: Not yet. It's an 8.45mm hex. Exactly right for the job (mid-point between the shaft and the thread root. I stopped working in round numbers a long time ago, and my designs got better.
[20:32:09] <andypugh> I have a 14.75mm x 0.9mm pitch thread somewhere, to exactly split a differnce.
[20:33:02] <Connor> FRACK! My KBCC-125 has a issue..
[20:33:20] <Connor> I think on of the terminals on the speed control POT grounded out... now it just wants to run a FULL speed...
[20:33:26] <Connor> anyone seen that before?
[20:33:29] <andypugh> I am considering making the nut a different pitch to the thread, by just enough to be self-locking.
[20:33:40] <andypugh> Connor: Yes
[20:33:57] <Connor> What did I burn out on the controller ?
[20:34:12] <andypugh> In my case the motor shed a commutator segment and killed the controller
[20:34:51] <Connor> I'm pretty sure in this case it was caused by a short to the speed control pot..
[20:35:01] <jdh> that sucks.
[20:35:14] <Connor> I had it setting on the mill with the pot just dangling..
[20:35:19] <Connor> I think it touched my vise.
[20:35:41] <Connor> I had it unplugged.. and when I plugged it up.. poof.. and then it started running full speed.
[20:37:05] <ReadError> im kinda torn about using cutting oil with my setup
[20:37:11] <ReadError> it seems to hold the chips in
[20:37:25] <ReadError> where as the vacuum atleast removes them
[20:37:38] <ReadError> i need one of those cooling setups i suppose
[20:37:49] <jdh> heat goes with the chips
[20:38:36] <jdh> connor: there's a $60 225R on ebay
[20:38:57] <andypugh> Gene keeps mentiioning a HexFet. I just gave up, I fitted a 3-phase motor and VFD: http://www.youtube.com/watch?v=47y6RgAK--8
[20:44:54] <Connor> jdh: That's the 220v model.
[20:44:57] <ReadError> jdh, well ideally one of those flood tables would be nice
[20:45:05] <ReadError> but i can see myself making a huge mess
[20:46:25] <Connor> HexFet ?
[20:46:44] <andypugh> It's a word. I don't know what it means.
[20:47:23] <Connor> Well.. Possible a Mosfet with 8 in/out ?
[20:48:13] <Tom_itx> http://www.irf.com/product-info/hexfet/
[20:48:54] <Connor> okay, yea.. I don't think that's the problem on mine.. I think it's something to do with the control side, not the driver side..
[20:49:51] <jdh> nothing obviously toasted?
[20:49:59] <Connor> Not that I can see.
[20:50:36] <Connor> Shouldn't their be voltage on the + of the speed pot when checked against the gnd ?
[20:50:41] <ReadError> oh crap
[20:50:44] <ReadError> this was almost bad....
[20:50:50] <ReadError> i touch off, start it
[20:50:58] <ReadError> its trying to go to Z0
[20:51:06] <ReadError> almost went straight in to my bed
[20:51:14] <Connor> ouch.
[20:51:56] <ReadError> G90 G94 G17 G91.1
[20:51:56] <ReadError> G20
[20:51:56] <ReadError> G28 G91 Z0.
[20:52:00] <ReadError> first 3 lines ;/
[20:52:35] <andypugh> Connor: If it's a KBIC the pot floats at +100V
[20:53:41] <Connor> 100v ?? How's that possible.. your suppose to be able to use 10V in place of the pot to control it..
[20:54:51] <jdh> guess that's why you need an isolated power supply for the board
[20:54:52] <Connor> It has a schematic of it in the manual online.. but, it's for crap..to fuzzy to read any of the values.
[20:55:32] <Connor> I haven't even gotten to that point yet.. I was still tweaking the spindle..
[20:55:43] <Connor> I just mounted it back on the freaking mill.
[20:56:14] <jdh> 2 down!
[20:56:27] <Connor> ?
[20:56:42] <jdh> didn't you fry your stock controller?
[20:57:00] <Connor> Kinda.
[20:57:43] <Connor> Wonder if it's the Op-Amp that got tosted..
[20:58:34] <jdh> how do you tell it to reverse?
[20:59:00] <jdh> http://solidstatedrives.com/default.aspx?act=faq.aspx&..
[20:59:17] <jdh> some generic troubleshooting stuff for kbic's near the bottom
[21:00:06] <ReadError> http://pastie.org/5672022
[21:00:12] <ReadError> yall dont think that would hurt anything?
[21:00:18] <ReadError> i commented out those first lines
[21:00:27] <ReadError> 3->6
[21:00:46] <ReadError> since its going to G54 and then moving off from that
[21:00:49] <jdh> why get rid fo the g20?
[21:01:13] <ReadError> oh, unit lengths
[21:01:17] <ReadError> yea i should uncomment that
[21:01:43] <jdh> you could set G28 locations and leave them
[21:01:55] <ReadError> i suppose g90 and g91 are okay as well
[21:02:43] <ReadError> so MDI G28.1 to set current location
[21:03:18] <jdh> looks like if you set that, it will be fine as-is?
[21:03:26] <ReadError> yessir that makes sense
[21:03:30] <ReadError> ive never set this before
[21:04:28] <ReadError> hey thats pretty neat, useful
[21:05:10] <jdh> are you homed properly?
[21:06:22] <ReadError> yessir
[21:07:14] <ReadError> ok take 2 on this
[21:07:54] <ReadError> hopefully i dont break anything this time
[21:08:12] <Connor> okay.. looks like maybe the Z2 Diode is toast.
[21:08:22] <jdh> sounds cheap :)
[21:08:26] <Connor> it even looks kinda toasted.. it's silvery looking
[21:10:15] <jdh> I found some 'surplus' 30krpm spindles at work. One is squeaky, the other was working fine when pulled out.
[21:10:29] <Connor> ROFL
[21:11:17] <jdh> I've heard that it is in theory possible to get a pass to get unused materials out. But never found anyone who coudl do so.
[21:11:31] <jdh> someone else suggested writing up the gate pass myself.
[21:12:02] <ReadError> now we are cookin ;)
[21:23:31] <Connor> Its a 1N4748A 22V-1W Zener Diode
[21:29:31] <jdh> costs a dime from mouser
[21:29:42] <andypugh> Connor: Mad as it seems, I am entirely sure about the floating pot. If you look at the manual it says that if you want to use analog voltage control then it is referenced to motor+
[21:30:58] <andypugh> This is actually not a real problem, as generic DC-DC converters isolate their outputs to 5000V, and you only need a tiny cheap one
[21:34:51] <andypugh> Or you can use an opto circuit like: http://www.linuxcnc.org/index.php/forum/30-cnc-machines/25580-grizzly-g1006-milling-machine-cnc-conversion?limitstart=0&lang=english#25670
[21:35:01] <ReadError> increasing my depth of cut/stepover has helped with the finish alot
[21:35:07] <ReadError> i think my stepover was too small before
[21:35:40] <jdh> I think I set mine at about 40%
[21:35:43] <andypugh> However, if the drive is not responding to the pot then replacng the pot with PWM won't help at all
[21:39:10] <ReadError> jdh,40% of endmill diameter ?
[21:39:14] <jdh> yeah
[21:39:28] <ReadError> thats about what im doing now
[21:39:29] <jdh> I think. I haven't set one up in my cam stuff lately
[21:39:34] <ReadError> super nice chip action
[22:01:58] <andypugh> Night all