Back
[00:21:29] <yoyoek1> hi all
[00:32:03] <frallzor> hello mah baby
[01:18:30] <HRyba> Where can I find documentation on using the uln2003 chip and EMC2? I have tried google already, and searched on the wiki but no luck.
[01:20:52] <HRyba> I have a circuit already, I really just need to learn how to program the hal configuration file, I guess.
[01:21:14] <oPless> oh
[01:21:20] <frallzor> isnt it step/dir?
[01:21:25] <oPless> I was going to tell you to look at
http://www.mathworks.com/matlabcentral/fileexchange/authors/54516
[01:21:26] <oPless> :)
[01:21:59] <HRyba> frallzor, no
[01:22:10] <HRyba> I thought it was, too, at first.
[01:23:08] <andypugh> ULN2003 is just a buffer. It sinks 8 lines of currrent to ground at half an amp when the corresponding 8 inputs go high.
[01:24:15] <andypugh> You need a quadrature stepgen set up in HAL and 4 p-port pins to your motor. The motor must be 5,6 or 8 wire so that it can be wired unipolar
[01:24:15] <HRyba> Yep :) How can I configure the .hal to use a buffer chip like the ULN2003
[01:24:48] <HRyba> Wait, isn't the quadrature stepgen set up as 5,5,5 or something?
[01:25:06] <HRyba> I have 5 wire motors on my machine, so I'm set there
[01:26:03] <andypugh> http://www.8051projects.net/stepper-motor-interfacing/uln2003-stepper-big.gif
[01:27:36] <HRyba> thanks
[01:27:41] <andypugh> http://www.linuxcnc.org/docs/2.1/html/man/man9/stepgen.9.html
[01:28:26] <andypugh> But apart from all that, I wouldn't do it that way. You will find the performance disappointing.
[01:29:26] <andypugh> Mainly because you need to set the voltage low enough for the winding resistance to limit the current, whereas with a proper PWM controller you can use 10 - 20x rated voltage to get more speed.
[01:29:29] <HRyba> I'm really just doing this as a proof of concept so that we can feel comfortable buying a bigger better controller
[01:29:55] <HRyba> I've already heard online about the poor preformance
[01:33:10] <andypugh> OK, then set up a new config with the Stepconf wizard, then edit the HAL file to use a type 5 (probably) stepgen and hook up the extra pins by copy-pasting the ones that Stepconf creates and editing the indexes.
[01:33:58] <HRyba> I don't have to worry about editing anything else in the hal?
[01:34:03] <HRyba> That's easy enough
[01:34:48] <andypugh> That ought to be pretty much enough.
[01:36:36] <HRyba> cool, let me set up the circuit and hook it up to the port
[01:47:50] <ries2> any mechmate owners here?
[01:49:17] <HRyba> andypugh: I set up the circuit and have modified the step_type; what do I need to copy and paste?
[01:51:24] <andypugh> You need to double up the pins. Conventional stepgen only sets up 2 pins per stepgen, so you will need to copy the lines that link the stepgen outputs to the parallel port pins. (And probably rename the signals from X-step, X-Dir to X-A, X-B etc. However the signal names are arbitrary.
[01:52:39] <andypugh> The pin names (stepgen.0...) etc are not arbitrary and have to be correct. You can see them in Hal Config as long as your HAL isn't too broken to prevent AMC starting
[02:11:00] <HRyba> Apparently my hal is too borked to load it in EMC
[02:11:28] <andypugh> Scroll to the bottom of the error log and it will give you a clue, or type dmesg
[02:14:30] <HRyba> "stepgen.0.stepspace not found"
[02:15:37] <andypugh> Comment it out and try again. Then once you get into EMC you can look at the HAL config and see what is there
[02:15:56] <HRyba> is // comment?
[02:16:23] <andypugh> Actually, delete that line. There is no .stepspace except in a type 0
[02:16:44] <andypugh> I think comment is #
[02:17:24] <andypugh> http://www.linuxcnc.org/docs/2.1/html/man/man9/stepgen.9.html
[02:17:33] <HRyba> okay, now I get the error that stepgen.0.step not found
[02:17:37] <HRyba> that can't be right
[02:17:51] <andypugh> Lists a lot of parameters that don't exist except in type 0
[02:18:06] <andypugh> No, there is no step parameter either.
[02:18:34] <HRyba> dir?
[02:18:45] <andypugh> You need stepgen.0.phase-A, phase-B, phase-C and .phase-D instead
[02:19:12] <andypugh> No, no .dir, .steplen, dirsetup or dirhold
[02:19:55] <HRyba> okay, let me code that in
[02:20:34] <HRyba> If I am not going to use some of the axis, can I remove the parport definitions?
[02:20:55] <andypugh> Yes
[02:21:02] <andypugh> In fact, you probably should
[02:21:56] <andypugh> Also, for simplicity, use pins 2 to 9. Avoid the hardware-inverted ones at this point (pin 1 is inverted)
[02:22:08] <HRyba> I have 2-5
[02:22:12] <HRyba> for the one axis
[02:22:17] <andypugh> That's good
[02:22:20] <HRyba> :)
[02:23:07] <HRyba> Do you know what the setp paraport ... reset 1 does?
[02:23:20] <HRyba> and if so is it necessary?
[02:23:56] <andypugh> No, I keep meaning to find out. I think it is a dodge to get shorter pulses in ster/dir mode. Delete it.
[02:25:18] <HRyba> * HRyba crosses fingers
[02:25:34] <HRyba> O_O EMC loads the config
[02:30:08] <andypugh> Time to fire up Halscope and watch the paralell port pins for activity
[02:30:29] <HRyba> Halscope?
[02:30:43] <andypugh> Machine-Halscope in EMC
[02:31:14] <andypugh> It's like an oscilloscope that lets you watch any signal in the software
[02:34:34] <HRyba> The motor is trying to run...
[02:34:43] <HRyba> I think I need to modify the timings
[02:35:04] <andypugh> It should work at low speed
[02:35:09] <HRyba> andypugh: Thank you for all your help so far!
[02:35:31] <HRyba> Haha, I changed the speed and it worked
[02:37:09] <HRyba> I can get 10 in/mi out of it
[02:37:23] <HRyba> not a lot of torque though
[02:37:44] <HRyba> thanks again
[02:38:12] <andypugh> Aye, not surprising at 500mA
[02:38:46] <HRyba> Think I could hook up some FETs to up the current?
[02:39:03] <andypugh> What's the motor rated at?
[02:39:47] <HRyba> I don't know, the motor isn't labeled with anything
[02:40:10] <andypugh> You could double-up the ULN pins. 2 channels for each phase
[02:43:05] <HRyba> I would only be able to get 2 axis out of the machine XD
[02:43:05] <andypugh> No, each parport pin can drive at least 2 ULN2003 channels
[02:43:54] <HRyba> that just requires some extra wires?
[02:44:24] <andypugh> Aye
[02:44:40] <HRyba> okay, I will do that
[02:44:46] <HRyba> I cannot thank you enough!
[02:44:50] <HRyba> I have to go now
[02:44:52] <HRyba> bye
[02:50:56] <andypugh> That was quite a short departure
[02:51:02] <HRyba> haha
[02:51:17] <HRyba> I decided just to keep this channel on for the electronics season.
[02:51:44] <HRyba> :)
[02:53:23] <andypugh> OK, coming up to 3am here. Time to not be here any more.
[03:11:11] <MarkusBec> MarkusBec is now known as MarkusBec_away
[07:21:59] <tomp> got heekcnc to mill pockets, nice. need to get cutter comp on it tho, or cad it with that allowed for.
[07:24:47] <tomp> i noticed the default tool table for sim-mm is in odd place /usr/share/doc/emc2/examples/sample-configs/sim/sim_mm.tbl
[07:25:17] <tomp> nobody would ever think the file was there
[07:25:24] <tomp> emc 2.3.4
[07:38:27] <tomp> heekscnc pocketing knows tool dia, and desired pocket dimensions, it doesnt use cutter comp, and left me .2mm wall stock for finishing
[07:39:20] <tomp> heekscnc gcode output gave me a 'tool table' in the gcode eg: N50G10L1P5 R1.500 Z15.000
[07:42:37] <tomp> i had to remove the 8 lines like that, emc2 would not execute them , and gave no error, never got the pgm to execute
[07:43:15] <tomp> the prev versions of emc had an emc err log, dont see it now
[07:43:22] <tomp> gnite
[08:07:37] <MrSunshine> a
[08:15:33] <roh> tomp that .2 sounds like you didnt set clearance to 0 for the endmill
[08:17:32] <roh> think of it like 'you need 4 pockets, 3mm corners, 20x20' .. then you use a 'bigger mill' to do the 'raw work' with a clearance of 0.2 and then do the same pocket function with e.g. smaller dia endmill and clearance 0 to make it nice and shiny
[08:17:52] <roh> atleast that is how i understood it
[08:27:18] <MarkusBec_away> MarkusBec_away is now known as MarkusBec
[08:29:23] <mIreland> say, what does it mean when when a bunch of people exit and then reenter a chatroom? can you tell anything about the nature of the failure by the flurry of msgs?
[08:30:42] <Jymmm> mIreland: It's a net-split.
[08:33:20] <toastydeath_> mIreland: a netsplit is when one server loses connection to another server. Since IRC networks are trees, the users on each side of the split can no longer receive messages from one another.
[08:37:08] <MarkusBec_away> MarkusBec_away is now known as MarkusBec
[08:39:12] <mIreland> looking at coding under linux with the opencv library. anybody here familiar w that?
[08:41:38] <mIreland> I got swept into this silly sentry gun project... I got it working in 2 days for under a hundred bux but now i see that i must write better software
[08:41:53] <mIreland> cv as in computer vision...
[08:42:25] <MattyMatt> can it distinguish the neighbour's dog from a ninja? :)
[08:43:37] <mIreland> my two rcservos ziptied together w laserpointer? no
[08:44:27] <mIreland> in phact, it tracks in revers in x and is useless. cat toy, in the words of the writer of the software
[08:45:35] <mIreland> not that im complaining. he admits he did it w little knowledge. plus i think its in polish originally.
[08:46:07] <mIreland> but poking around i learn of opencv library and get curious.
[09:36:46] <MattyMatt> yeah steppers and crappy thread
[09:37:29] <MattyMatt> with one used ballscrew unit which will make a nice X so I'll be able to scan reasonably fast
[09:37:58] <mIreland> sounds like you've skipped to something that can work
[09:38:04] <mIreland> we started w allthread
[09:38:09] <mIreland> and drillpress
[09:38:37] <mIreland> then i got job in machine shop. now i have bridgeport, engine lathe, chucker, and grinders.
[09:38:53] <MattyMatt> I'm going to try the webcam-watching-vernier idea for Y, and if it works well I can use any servo
[09:39:26] <mIreland> wait, is the wwv idea a thing i should be familiar with?
[09:39:45] <MattyMatt> my 38mm drill is packed in the 44mm ring of MY drillpress :)
[09:40:18] <MattyMatt> I get better holes with the brace & bit, generally
[09:40:29] <mIreland> what do you make?
[09:40:42] <MattyMatt> so far, routing machine(s)
[09:41:17] <mIreland> do you lie awake at night thinkin of machines that can manufacture themselves?
[09:41:22] <MattyMatt> I'm going to do some artistic/furniture carving to raise some cash for metalwork
[09:42:15] <MattyMatt> those machines on the moon, yeah
[09:42:29] <mIreland> where are u? i think i asked b4, and i think uk, but i am dim and must ask again
[09:42:35] <MattyMatt> yep
[09:43:11] <MattyMatt> nr Liverpool
[09:43:15] <mIreland> i am in killafornia. suckramento. i will help any person with metalfab. if they can reach me.
[09:44:09] <mIreland> long way for a visit
[09:44:24] <MattyMatt> yep
[09:45:23] <Jymmm> only 90 minutes away
[09:46:27] <mIreland> i marvel all the time that i have retrofitted a production lathe; added new functionality to an already venerated machine, and have an irc line to a global network of wizards and fellow fools on the control and it is my plaything in my garage... you cannot pay enough to get this.
[09:47:03] <mIreland> jymmm... 90min rly?
[09:47:12] <MattyMatt> I have a grungy idea for a nut. tap an m8 hole in the knob on the handle of an old brass spoon, and squash the bowl flat & drill for mounting
[09:47:24] <Jymmm> mIreland: yep
[09:49:07] <mIreland> i am interested in a closer association with capable persons. I have things to offer. i know you (jym)are here all the time.
[09:49:21] <MattyMatt> yeah teamwork rocks
[09:49:35] <mIreland> but for fun we can play hot and cold... are you 90 min due south?
[09:49:41] <Jymmm> mIreland: I don't know nuttin, I just troll here.
[09:50:03] <MattyMatt> even coding works better if you're in the same room tho. teleworking isn't everything
[09:51:04] <mIreland> i cant accomplish anything except grpwing unabomer hair without company. too bad i know so few linux geex
[09:51:40] <mIreland> i am an inspiration to those who know me tho. free as in speech.
[09:52:13] <Jymmm> Well, go unibomb Da Terminator already!
[09:52:15] <MattyMatt> I code for money, or I try :)
[09:52:31] <mIreland> i dont bomb nothin. i read.
[09:52:45] <mIreland> but i am easy to pick out of a crowd
[09:54:39] <archivist_emc> so am I being a hairy git
[09:55:06] <mIreland> i code for inner peace. my best wrk is a tightknit algorithm for factoring primes... not to par with hi math but a serious coding version of meditating on your navel...
[09:55:52] <mIreland> pascals triangle looks neat rendered in prime factors
[09:57:09] <mIreland> plus it is a map for figuring permutations
[10:00:35] <mIreland> lucky for me, my prime number obsession has abated for now. it was bad, ask my family. Now we are playing in the shop which is much better.
[10:01:21] <mIreland> one proj is a robot arm, made simply to take up some of the excess pneumatic tubes laying about
[10:02:51] <mIreland> the other is the cat shooter, our response to the 30 semi feral cats next door who own our front lawn. rcservos and a hose, by way of the valve from a washing machine.
[10:03:03] <mIreland> Highly effective and humane.
[10:03:34] <mIreland> sentry gun is obvious extension and has proceeded very quickly
[10:04:08] <mIreland> oh the satisfaction
[10:06:17] <mIreland> alas, the coder lacked my drive for complete dominion and i must outdo him with better code.
[10:07:10] <mIreland> i have always believed that this sort of machine had important application... i will confine myself to pest abatement for now.
[10:20:21] <anonimasn> * anonimasn yawns
[10:49:08] <EbiDK> EbiDK is now known as EbiDK|AWAY
[11:00:05] <mIreland> yawns
[11:55:38] <anonimasn> anonimasn is now known as anonimasu
[12:00:47] <MrSunshine> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?UpdatingTo2.3 full of dead links!
[12:04:19] <jthornton_> fix them
[12:12:16] <micges> only 5 broken links :)
[13:30:45] <tomp> roh: i think the program intentionally left .2 per side and it is intended to be for finishing. its not wrong, just surprising.
[13:40:34] <tomp> mireland: opencv will be very useful for pick&place with emc2, or automated inspection. why is there 1 'l' in useful ?
[13:46:01] <MarkusBec> MarkusBec is now known as MarkusBec_away
[15:11:29] <roh> http://yamato.hyte.de/packages/heeks/ubuntu_9.10_2009-11-08/ got karmic packages for heekscad and cam. please test
[15:17:45] <tom3p> roh, still testing hardy :), pretty cool too
[15:18:32] <roh> will do a fresh hardy build soon
[15:18:34] <tom3p> running more pocketing, wish there were kreiss rather than n-gons
[15:19:07] <roh> had a party here and the milling room is still full of crap so i dont want to enable the power there not yet
[15:19:17] <roh> party on friday->sat.
[15:19:18] <tom3p> hahaha
[15:19:42] <roh> crap as in 'put the stuff you want to preserve in there and lock the door'
[15:21:05] <roh> now we also know that one needs to put in an effort to meet cops. much more than a mill or powertools can do in this workspace
[15:21:19] <roh> eh.. when i comes to noise
[15:36:47] <frallzor> anyone here tried topsolid?
[15:37:46] <awallin> http://www.youtube.com/watch?v=aCVVW34wNM8 enjoy :)
[15:50:43] <tom3p> awallin: nice
[15:51:05] <MarkusBec_away> MarkusBec_away is now known as MarkusBec
[15:52:03] <roh> hrhr
[15:53:30] <roh> nice chamber
[15:56:33] <MarkusBec> MarkusBec is now known as MarkusBec_away
[15:58:58] <MarkusBec_away> MarkusBec_away is now known as MarkusBec
[16:14:15] <EbiDK> EbiDK is now known as EbiDK|AWAY
[16:15:27] <frallzor> http://www.topsolid.fr/resources/images/pr/175.jpg wouldnt mind having one of these
[16:19:15] <tom3p> crank grinder, worked on one for ship engine, the machine wrapped around the crank and spun. the cranks were too large to put inside a reasonable machine ( > 5 meters ) so, move the mountain to mohammed
[16:19:45] <tom3p> yoyoek1: good morning
[16:19:47] <yoyoek1> hi
[16:20:52] <frallzor> its a cute machince nonetheless
[16:21:04] <tom3p> oh yes!
[16:21:16] <frallzor> id like one in my garage
[16:21:28] <frallzor> and then extend my garage by like 10x to fit it :P
[16:23:55] <celeron55> you could use the machine as a garage
[16:24:23] <frallzor> good point
[16:24:32] <frallzor> mill a car when needed
[16:25:04] <celeron55> sure you could fit a bed in there too, so no need for the house either
[16:25:55] <frallzor> I could mill a Alu-wife too
[16:32:43] <oPless> I seem to have missed a url :)
[16:46:44] <MattyMatt> one crank grinder per town. vote communist and have it installed in the People's Shed
[16:47:35] <MattyMatt> I miss the local college. that was the same thing once you'd signed up for a course
[16:47:57] <MattyMatt> or looked like you had :)
[16:49:44] <MattyMatt> angle grinder on a puma would achieve the same thing, if it had good enough eyesight
[17:47:23] <motioncontrol> ggoe evening. I have on maschine the A axis circular, but for move this axis i first move up the table indexet gear.if i write g1 a180 f100 , the axis normaly turn,but is necessari first up the table index and after turn.i have thing create a custom M fuction for move up the gear , when the movement up is complete reset the hold axis and move the A axis.But this solution not is very good. Tha operator maschine want simp
[17:47:23] <motioncontrol> le write g1 a180 f100.exist a pin i ceck for a axis , change status when i command it?
[17:52:40] <Jymmm> For all you snow-bound people, a nice little shop heater with good reviews....
http://www.northerntool.com/webapp/wcs/stores/servlet/product_6970_503599_503599
[18:06:54] <MattyMatt> 59 kW toasty
[18:15:18] <MattyMatt> can you fit a crucible inside it?
[18:16:06] <MattyMatt> if it's cold, melt more metal
[18:41:14] <MrSunshine> * MrSunshine milled his first round pocket today .. got some latchback it looks like =)
[20:06:50] <ds3> 8
[20:19:41] <Tecan> http://www.youtube.com/watch?v=VebOTc-7shU << fall of the republic full HQ
[20:34:24] <frallzor> backlash ffs MrSunshine :P
[20:43:59] <gweepprefect> hello.. anyone alive?
[20:45:44] <MrSunshine> frallzor, yeye :)
[20:45:48] <MrSunshine> once again :P
[20:47:43] <gweepprefect> I'm having a somewhat baffling issue with accuracy on my Y mill axis, I figured I'd poke my head in and see if anyone had any advice.
[20:48:46] <MrSunshine> backlatch ... *pkes frallzor*
[20:48:52] <MrSunshine> latchback :P
[20:48:54] <MrSunshine> lashback? :)
[20:48:59] <MrSunshine> backlash ? :P
[20:49:07] <MrSunshine> so many ways and only one right
[20:49:32] <gweepprefect> It may be a backlash issue, but I'm not sure. My X and Z don't behave the same way, and they're very similar mechanically.
[20:50:51] <micges> gweepprefect: any more info?
[20:51:19] <gweepprefect> Sure, I can give you a brain dump or if you have specific questions, shoot.
[20:51:43] <celeron55> a brain dump would be a good start
[20:52:23] <gweepprefect> OK.. Basics.. latest live CD. PCI parallel port, connected to Gecko 203V, to NEMA23 steppers.
[20:53:13] <gweepprefect> Mill is an X3 with a PROMiCA conversion kit, 5mm ball screws, single nut, on all axes
[20:54:41] <gweepprefect> it's a 1:2 motor:leadscrew ratio with this kit
[20:54:52] <gweepprefect> I work in inches, so I think my math is right that it's 5.08 rev/in
[20:55:10] <gweepprefect> and I entered 10:20 for the ratios.
[20:55:34] <gweepprefect> Driver microstepping appears to be 10, and it's the usual 200 steps/rev on the motors.
[20:57:03] <gweepprefect> So, on to what I'm seeing: if I put a dial indicator on the bed against the head and run the X back and forth 0.2", it behaves as I expect.. If I zero the dial indicator, it will bang back and forth very close to 0 on either end.
[20:57:15] <gweepprefect> I'm doing the "back and forth" bit with the stepconf "test this axis" option
[20:57:32] <gweepprefect> Same deal with Z - it works as I'd expect.
[20:58:06] <gweepprefect> Unfortunately, Y is not working so well. If I do 0.2, it'll usually come up 5 thou short, and then drive right back to 0.
[20:58:36] <gweepprefect> If I bump it to 0.3 it's only a few thou short, and at 0.4 it seems to cross a zero and it's right on. Then beyond 0.4, it seems to run a little long.
[20:59:05] <gweepprefect> So it sort of feels like the pitch calculation is screwy for some reason, but I doubt that's really what's going on.
[20:59:22] <gweepprefect> And here I am, looking for advice. :)
[20:59:48] <celeron55> have you tried swapping motors and drives
[21:00:16] <gweepprefect> Not yet.. I will confess I just got the mechanicals installed on the mill yesterday, so I'm pretty early on in debugging.
[21:02:25] <celeron55> is that what you just said about Y perfectly repeatable?
[21:03:25] <gweepprefect> It seems to be, for at least a handful of tests I've done
[21:04:10] <celeron55> that seems quite mechanical
[21:04:10] <gweepprefect> That's why I don't suspect lost steps or anything, it seems to be very repeatable
[21:04:20] <tom3p> did you try differnt places? maybe its leadscrew in one area, if same result in diff areas, then it could be math
[21:04:36] <gweepprefect> I tried it at 2 different spots anyway
[21:04:40] <tom3p> k
[21:04:50] <gweepprefect> Very similar results at each
[21:05:04] <celeron55> was the motor and the leadscrew at different starting positions then
[21:05:28] <gweepprefect> Most likely, I just jogged the table to a different spot and tried the same test again
[21:05:52] <celeron55> you should do a few more tests on that
[21:06:05] <gweepprefect> Let me try swapping drives, since that takes only a couple minutes..
[21:06:05] <celeron55> i think
[21:06:07] <LawrenceG> loose coupling/pulley, play is thrust bearing block, loose nut mounting???
[21:06:40] <gweepprefect> Everything seems solid mechanically, I just tore that axis down and rebuilt
[21:06:55] <celeron55> i'm pretty sure swapping drives won't make any difference
[21:07:09] <gweepprefect> Me too
[21:07:12] <celeron55> because it's repeatable
[21:07:42] <celeron55> and the drives just copy the step pulses directly to steps of the motor
[21:08:16] <tom3p> the graph you speak of ( crossing a zero), i cant think of what causes it. i think i understand what you see. but my guess would always be long or short, not varying from short to 0 to long
[21:09:10] <gweepprefect> The only thing that makes sense to me there is the rev/in calculation being off slightly, but I really don't think a bug that obvious would have made it past testing
[21:09:22] <gweepprefect> Not to mention the other axes work just fine...
[21:09:23] <tom3p> unles the motion was actually circular ( doubt that you couldnt see that in .2" to .6" )
[21:09:37] <MrSunshine> when cutting soft aluminium .. what do i have to think of ?
[21:09:52] <MrSunshine> got a maximum of 2000rpm on my mill .. and i wouldnt be suprised if that is only like 1500rpm
[21:09:55] <celeron55> swapping motors would maybe help track the problem
[21:10:07] <celeron55> if the motors is bad, it might rotate a bit nonlinearly
[21:10:17] <gweepprefect> I can swap my X and Y fairly easily, maybe 10-15 mins.
[21:11:10] <tom3p> and full swap, meaning check the suspect axis with another motor AND a non-suspect axis with the suspect motor
[21:11:18] <celeron55> yes
[21:11:36] <gweepprefect> OK, give me a few and I'll do that
[21:12:57] <tom3p> while the motor's off, 'feel' the slide
[21:13:00] <MrSunshine> how many flutes for milling alu ? :)
[21:13:07] <MrSunshine> only got 4 flute mills atm i think it is :/
[21:18:20] <tom3p> MrSunshine: cutting alum tip #1 cheap alum cuts crappy, but heres some better info
http://www.cnczone.com/forums/showthread.php?t=48642
[21:19:04] <MrSunshine> i like my own casted aluminium .. or whtever it is .. melting down engine stuff and puring it into blocks, hard and nice to cut :)
[21:19:19] <MrSunshine> can feed as fast as i like and no clogging like it is with soft alu
[21:20:36] <gweepprefect> Half swap done, my X motor is on the Y.. It's doing the same thing.
[21:23:20] <gweepprefect> I just hooked up the whole X drive to the Y and ran the same test using the X axis configuration tester and it's doing the same thing.. Starting to sound more and more mechanical.
[21:24:47] <tom3p> MrSunshine: mcdonald-douglas feed calculator sez "0.010" feed per tooth at 1", so for a .125" diameter cutter, that should be .00125"
[21:25:18] <tom3p> gweepprefect: how did it feel with the motor off?
[21:25:18] <MrSunshine> omg i hate inches ...
[21:25:18] <MrSunshine> :P
[21:25:39] <tom3p> .01" = .25mm
[21:25:49] <tom3p> ~
[21:25:55] <gweepprefect> tom: slight bits of "stiction".. a little bit like the detents in a stepper motor
[21:26:03] <gweepprefect> but nothing that stopped me from easily moving it
[21:26:57] <tom3p> leadscrew nut tight to slide ?( not moving, that'd be nonlinear )
[21:27:36] <gweepprefect> it could be, though i loosened it up from the holder to see if that would have an effect earlier
[21:28:03] <tom3p> not from what you say... look to see the whole drive train is solid, just to eliminate that
[21:28:25] <celeron55> it seems quite obvious it's something related to the leadscrew, anyway
[21:29:02] <tom3p> then yopu should get same result by hand turning the screw ( same as turning with motor )
[21:29:05] <tom3p> you
[21:29:07] <gweepprefect> there was a minor bit of fitment problem with the Y ball nut holder mating to the chinese "slight imprecision".. it could be related.
[21:53:53] <gweepprefect> OK, which one of you guys stole my 3mm T-handle?
[21:55:39] <frallzor> I did
[21:58:52] <gweepprefect> Answer: still sitting in the last gib screw I adjusted.
[22:03:07] <gweepprefect> I took the table off, loosened up the fit on the Y nut holder, the screw feels less "catchy" now
[22:03:39] <gweepprefect> Now to put the motor back on and see if it actually makes any difference
[22:07:40] <gweepprefect> It's "better," but it's still exhibiting the under/zero/over problem
[22:11:47] <anonimasn> anonimasn is now known as anonimasu
[22:13:02] <gweepprefect> .1 under by .004, .2 under by .0025, .3 under by .0015, .4 just a hair over, .5 over by almost .002, .6 over by just about .003
[22:14:16] <anonimasu> hm.. what kind of screw are you using?
[22:15:11] <gweepprefect> pretty sure they're Thomson, that's what the nuts are
[22:16:07] <anonimasu> is this moving in only one direction?
[22:16:41] <gweepprefect> It's bouncing back and forth, using the stepconf tester.. I know backlash should show up, but it shouldn't be acting like that
[22:17:06] <anonimasu> hm..
[22:17:11] <anonimasu> try moving in only one direction
[22:17:27] <celeron55> what kind of gears are between the motors and the leadscrews?
[22:17:44] <gweepprefect> XL pulleys with a belt
[22:17:47] <anonimasu> that's the first thing you need to do..
[22:18:13] <gweepprefect> 10t on the motor, 20t on the screw
[22:18:13] <anonimasu> so you'll see that it's not a issue with not getting steps out
[22:18:29] <archivist_emc> gweepprefect, are you missing steps its the most likely, reduce acceleration a bit
[22:18:32] <gweepprefect> anon: my other axes are working well (not sure if you saw that part of the conversation)
[22:18:42] <anonimasu> hm, my client died
[22:18:45] <gweepprefect> i've also got the speed and accel cranked all the way down for this test
[22:18:45] <anonimasu> so, no
[22:18:51] <anonimasu> well, try one direction first
[22:18:54] <gweepprefect> (mostly so i can read the dial indicator)
[22:19:27] <anonimasu> and then change direction to see if it's backslash
[22:19:47] <gweepprefect> ok, just need to save off this config and test it in emc proper
[22:20:38] <anonimasu> then if that works well enough reverse and check if it's backslash
[22:22:40] <gweepprefect> it's overshooting, even in one direction, going 0.1 at a time and resetting the dial indicator
[22:23:22] <anonimasu> hm..
[22:23:24] <anonimasu> dont reset it..
[22:23:27] <gweepprefect> i backed it all the way out enough to soak up backlash, took it to 0, reset, then went 0.1 a few times in the same direction, it's definitely behaving oddly
[22:24:00] <frallzor> http://www.camcraftsa.com/cnc-routing-textures.htm Is it just me or would block with textures like that look really cool on a wall? =)
[22:24:03] <anonimasu> keep stepping with the indicator unmoved
[22:24:06] <frallzor> *blocks
[22:24:36] <gweepprefect> ok, i'm at zero, here goes.
[22:24:46] <archivist_emc> gweepprefect, the other thing to look at is the timing to the driver
[22:25:14] <gweepprefect> going to -.1 i got .102
[22:25:18] <archivist_emc> some drivers dont see the reverse fast enough
[22:25:39] <gweepprefect> i'm using slightly more conservative timing than the datasheet recommended, but i guess i could try backing it way off
[22:25:48] <anonimasu> frallzor: they do :)
[22:25:56] <gweepprefect> again, my other axes are behaving right, and they all use the same hardware, so it seems like timing is ok
[22:26:33] <anonimasu> that's really weird
[22:26:45] <gweepprefect> going to -0.2 my dial indicator says -0.204
[22:27:02] <frallzor> * frallzor wants to be able to mill blocks like that soon!
[22:27:03] <anonimasu> this is continuing from -0.1 right?
[22:27:07] <archivist_emc> scale error?
[22:27:12] <frallzor> I have a vision for them allready
[22:27:12] <gweepprefect> to -0.3 i'm reading -0.306
[22:27:19] <anonimasu> seems consistently wrong
[22:27:22] <anonimasu> with the same amount
[22:27:22] <gweepprefect> yep, i haven't touched the indicator, i just read it off
[22:27:34] <anonimasu> why dont you nudge the scale just a tiny bit
[22:27:49] <celeron55> gweepprefect: continue it still a bit
[22:27:57] <anonimasu> 0.2 - 0.4 - 0.6...
[22:28:05] <anonimasu> err 006
[22:28:07] <celeron55> gweepprefect: you were getting different kind of results to the other direction
[22:28:31] <gweepprefect> -.4 = -.408
[22:28:48] <anonimasu> what if you go to .6 ?
[22:28:51] <gweepprefect> .5 = -.5095
[22:29:26] <gweepprefect> -0.6 = -.611
[22:29:53] <celeron55> well, more, it might start getting a repeated pattern
[22:29:55] <gweepprefect> hm, i have an idea.. I wonder if it's step polarity
[22:30:04] <archivist_emc> scale setting error
[22:30:24] <anonimasu> I agree with archivist..
[22:30:24] <celeron55> the scale is supposed to be and is the same on all axis?
[22:30:26] <gweepprefect> I was wondering that, but all the screws are the same, the other axes were behaving right
[22:30:40] <gweepprefect> as far as i know - all the same nuts and screws
[22:30:58] <anonimasu> what kind of pitch accuracy does your screws have?
[22:30:59] <gweepprefect> it sure feels like a scale error
[22:31:06] <celeron55> what does it show at -1.0
[22:31:20] <anonimasu> (it's another thing) it might be a screw that's just slightly different
[22:31:37] <gweepprefect> -1.020
[22:31:53] <celeron55> that's fairly linear then
[22:32:18] <celeron55> try some other values and see if it is at them too
[22:32:25] <gweepprefect> i'm not sure about pitch accuracy.. they came as part of a kit. it's a very high quality conversion kit though, i expect good specs.
[22:33:04] <archivist_emc> kit from?
[22:33:06] <gweepprefect> i'm out of dial indicator travel at the moment
[22:33:18] <gweepprefect> PROMiCA - cnckits.com.au
[22:33:29] <archivist_emc> dont know them
[22:33:38] <tom3p> get rid of the reversal in the test, go up against a 123 block on 3" side, set 0, pull block , replac eso its 2 " go SAME dir, see 0, flip to 1" go SAME dir see 0 again ( never reverse dir during test ) see wha that tells you
[22:33:39] <gweepprefect> they make a nice kit for the Sieg X3 mini mill
[22:34:17] <gweepprefect> this last test had no reversal
[22:34:23] <gweepprefect> it was all one way
[22:34:34] <archivist_emc> I wonder if the kits come from the same source as the UK sieg kit
[22:35:09] <gweepprefect> as far as i know they make them themselves, on converted X3's.. there's a US distributor, but i don't think there's a UK one
[22:35:14] <tom3p> didt read that, but, a scaling eror is linear
[22:35:48] <anonimasu> Thomson specifications on the leadscrew and ball nuts we use state backlash of 0.002" to 0.013" ... we've measured our prototypes and they average 0.003"
[22:35:49] <HRyba> andypugh: the ULN2003 preforms awesome on the machine
[22:36:15] <HRyba> The motors must be rated for less current than the one I tested last night
[22:36:19] <andypugh> For suitably conservative definitions of "Awesome" ?
[22:36:44] <HRyba> good torque and speed
[22:37:06] <anonimasu> thoose kits seem nice
[22:37:08] <andypugh> Glad to hear it.
[22:37:29] <gweepprefect> Yeah, I had a hard time coughing up the extra bucks versus the CNC Fusion kits, but I like their design a lot better.
[22:37:35] <anonimasu> gweepprefect: the idea of them supplying better bearings and stuff is a _very_ nice ting
[22:38:00] <gweepprefect> Yeah.. The Sieg bearings are kind of weak.
[22:38:27] <gweepprefect> It's a lot of machine for the money, but it has its quality limitations from the factory.
[22:39:14] <gweepprefect> You can see pictures from my conversion (and some youtube video) at
http://www.tangentaudio.com/
[22:39:26] <HRyba> andypugh: I have some videos and pictures if you want to see it
[22:39:40] <andypugh> Yeah, post a link if you have one
[22:40:12] <HRyba> okay, let me find them
[22:41:02] <_Poincare> _Poincare is now known as Poincare
[22:41:14] <anonimasu> gweepprefect: one thing I've got to say
[22:41:17] <anonimasu> nice videos!
[22:41:31] <anonimasu> err two
[22:41:44] <anonimasu> and dont mill with a drill holder
[22:41:55] <anonimasu> :)
[22:41:58] <gweepprefect> yeah, i know.. i didn't have a collet small enough for that dremel bit
[22:42:17] <anonimasu> I like the look of that kit
[22:42:33] <gweepprefect> for a quick test in plastic, i figured the cutting forces weren't even remotely a problem
[22:43:32] <archivist_emc> gweepprefect, I was helping a new user on one of those on friday night, have you had the spindle apart yet
[22:43:59] <gweepprefect> no, i haven't had the nerve to pull that apart yet
[22:44:25] <archivist_emc> heh, feel for play, if any it needs doing
[22:44:32] <gweepprefect> there's a nice guide on tearing it down and rebuilding it, but it looks somewhat daunting
[22:44:47] <tom3p> that test must have used a long stroke indicator if it moved .6" w/o moving indicator, still i cant see cosine error being non-linear
[22:45:03] <archivist_emc> not too hard, but the bearings are far too tight on the shaft
[22:45:18] <gweepprefect> tom3: yeah, it's got a little over an inch of travel
[22:46:00] <tom3p> use you calipers over 4 or 5 lobes of the screw, compare a sequence of 5, you should have no variance, else screw is wonky
[22:46:31] <archivist_emc> gweepprefect, two of us did the spindle in 2 hours, we were not working too hard
[22:46:46] <gweepprefect> let me see if i can do that from underneath without pulling it apart again
[22:47:29] <tom3p> cosine err is when the measurement is not parallel to the thingy being measured ( but its a linear error )
[22:47:30] <archivist_emc> we did the spindle only without any other recommended job
[22:48:47] <gweepprefect> i set my calipers to 25mm and locked them down, the jaws drop right into the grooves across about 3-4" of the screw that i can reach
[22:49:14] <gweepprefect> archivist: did you use the Arc Euro Trade guide?
[22:51:05] <archivist_emc> yes as the machine came from there
[22:51:39] <archivist_emc> im 30 miles from them
[22:51:56] <gweepprefect> that's a great guide
[22:52:28] <archivist_emc> we didnt have a press, so block of wood and "effort"
[22:52:29] <gweepprefect> is it helpful to have photos or a video of my setup or the test?
[22:52:50] <tom3p> if screw is uniform, then maybe the ball nut isnt rolling correctly ( the ball tension is varying because the ball pressure isnt uniform, like some backlash in the ball chain itself )
[22:53:32] <gweepprefect> it feels OK by hand now, it was a little "sticky" before i just tore it down and filed down the mounting holes for the Y nut mount to give it some more play
[22:53:53] <gweepprefect> doesn't seem like that would ever amount to *overshooting* though, unless i'm missing something
[22:54:20] <tom3p> i'm thinking the train of balls is not a uniform pressure on the screw, like a gap that is on the screw now, and up in the nut later
[22:56:07] <HRyba> andypugh:
http://www.flickr.com/photos/hryba/sets/72157622639037125/
[22:56:24] <andypugh> Does anyone in the UK have a use for a full set of the big Marlco collets? I would be looking for a donation to the students' fire engine upkeep fund.
[22:57:32] <tom3p> is that a clever british euphemism for beer money
[22:58:05] <HRyba> lol
[22:58:16] <andypugh> Not in this case. They are saving up for new tyres, which will have to me made specially...
[22:59:10] <andypugh> 1000mm x 75mm x 75mm solid rubber tyres. For some reason nobody makes them any more.
[23:00:05] <archivist_emc> I have one collet somewhere and we have a base mount chuck for them andypugh
[23:00:07] <andypugh> HRyba: looking at the motors 1A is probably plenty.
[23:00:43] <yoyoek1> helo
[23:03:57] <andypugh> There is a chuck with no collets on sale on that eBay, I have considered asking him if he wants to buy the collets and re-list
[23:04:06] <archivist_emc> gweepprefect, btw we also didnt remove the bearing outers from the spindle tube, we just did the inner fit
[23:04:29] <yoyoek1> some rc modelers presents ?
[23:05:05] <archivist_emc> yoyoek1, more in ##electronics
[23:05:45] <gweepprefect> archivist: did you upgrade them or just rebuild it?
[23:05:54] <yoyoek1> archivist_emc: I make some example about mGcodeGenerator for rc modeler
[23:06:51] <archivist_emc> gweepprefect, adjusted the fit so the upper bearing slides on the spindle
[23:07:09] <gweepprefect> lathe and emery cloth?
[23:07:24] <archivist_emc> 600 grit, use a very fine
[23:07:51] <archivist_emc> you dont need to take much at all
[23:09:04] <andypugh> Odd, the ones we have are not like this:
http://cgi.ebay.co.uk/MARLCO-INDEXABLE-COLLET-CHUCK-FIXTURE-5-COLLETS_W0QQitemZ310169020110QQcategoryZ12584QQcmdZViewItemQQ_trksidZp4340.m263QQ_trkparmsZalgo%3DSIC%26its%3DI%26itu%3DUCI%252BIA%252BUA%252BIEW%252BFICS%252BUFI%26otn%3D11%26ps%3D63#ht_631wt_1114
[23:10:32] <gweepprefect> well, i tried un-inverting the motor step signals, it behaves the same
[23:12:50] <andypugh> They seem to be Crawford Multibore type collets made by Marlco.
[23:13:04] <andypugh> No idea if they are any use to anyone.
[23:13:55] <archivist_emc> if you find someone in the next week or so point them at me if they need a chuck
[23:14:34] <andypugh> Does the chuck take the big single-taper ones, or the ones that look like a dumpy R8?
[23:15:08] <andypugh> These are like an ER65 if such a thing existed, but with seperate segments spring apart.
[23:15:09] <archivist_emc> single taper with rubber between the blades
[23:15:47] <andypugh> Might be different. I think this is the ones we have.
[23:15:49] <andypugh> http://cgi.ebay.co.uk/Crawford-Multibore-Collet_W0QQitemZ350270237815QQ
[23:16:14] <archivist_emc> not like that one
[23:16:39] <andypugh> There is also a D3 nose collet chuck + full set of collets. But that will just go straight on eBay
[23:24:46] <anonimasu> hm
[23:26:15] <andypugh> Not D3. D1-4
[23:26:28] <anonimasu> I wonder if I can machine a pneumatic chuck for the lathe
[23:26:53] <anonimasu> :]
[23:27:13] <andypugh> All things are possible given time, skill and money
[23:27:44] <anonimasu> it's not so many parts within one.. a tapered inner piece
[23:27:56] <anonimasu> the jaws.. which needs one side tapered..
[23:28:18] <anonimasu> and a open center cylinder..
[23:28:27] <anonimasu> (well the whole body of the chuck ends up as that
[23:28:35] <gweepprefect> i have a video which demos the issue i'm having
[23:28:39] <gweepprefect> let me get it uploaded
[23:29:43] <andypugh> Spindle encoder, ladder logic and a pneumatic wrench operating a conventional chuck?
[23:30:07] <anonimasu> hm..
[23:30:23] <archivist_emc> anonimasu, some automatic chucks have a pendulum arrangement to tighten the grip with rpm
[23:30:54] <anonimasu> that's not a bad design
[23:31:02] <anonimasu> I were thinking machine springs
[23:31:28] <anonimasu> they make them _very_ stiff
[23:31:53] <anonimasu> or make it so preload can be applied with air..
[23:32:30] <andypugh> There's something. When I was working in the materials department at Leeds, they CNC machined tensile test specimens between centres. Just like woodworking, but for metal. I haven't seen that done anywhere else since.
[23:33:23] <MattyMatt> before I run this up, is it anybody's?
http://cgi.ebay.co.uk/ws/eBayISAPI.dll?ViewItem&item=220505249850
[23:33:25] <archivist_emc> andypugh, clock trade does hand turning on metal with a graver
[23:33:50] <archivist_emc> not me
[23:34:03] <andypugh> Nor me
[23:34:37] <andypugh> archivist_emc: Aye, I have read about it, but this was 25mm dia high tensile steel.
[23:34:44] <MarkusBec> MarkusBec is now known as MarkusBec_away
[23:35:09] <andypugh> working with a drive centre you are assured concentricity (critically important with tensile test specimens)
[23:36:12] <archivist_emc> true if and only if the centres are good
[23:36:52] <andypugh> ? How can you lose concentricity?
[23:36:58] <MarkusBec_away> MarkusBec_away is now known as MarkusBec
[23:37:01] <archivist_emc> and you dont have a bad roller in the headstock
[23:37:36] <archivist_emc> center has to be ground on the machine
[23:37:39] <andypugh> I can see you can have taper and out of round, but one thing you should be able to get is concentricity
[23:37:57] <archivist_emc> no dirt in the center
[23:38:05] <archivist_emc> bent center
[23:38:30] <andypugh> Still not seeing it, sorry.
[23:38:32] <archivist_emc> bad balls in a rotating center
[23:39:03] <andypugh> That I can see, but this was a live centre.
[23:39:16] <andypugh> (ie headstock end)
[23:39:38] <toastydeath_> andypugh: there is no guarantee any of the rotating elements in a live center or the point itself is concentric with the axis of rotation of the spindle
[23:40:07] <toastydeath_> that is why the most accurate turning machines rely on vacuum chucks and manual alignment, not centers or any other mechanical system for alignment
[23:40:26] <andypugh> No, but the out-of-true should be consistent as you machine the different diameters of the work. ]
[23:40:40] <archivist_emc> you have to see all the errors, its fun
[23:40:58] <toastydeath_> andypugh: that's also true of a beat up three jaw chuck
[23:41:07] <toastydeath_> i'm not sure what the point is of your statement
[23:41:30] <andypugh> Absolutely. But with a three jaw chuck you can't machine the full length of your specimen.
[23:41:59] <toastydeath_> you can't do it with a drive center either - the dog is in the way, unless you design a fixture that goes on the end of your part that acts as a dog
[23:42:09] <toastydeath_> and then I could just as easily design a fixture to go on a faceplate or anything else
[23:42:09] <archivist_emc> you can because you part off outside the jaws
[23:42:17] <andypugh> Dog? That was my point, no dog
[23:42:42] <gweepprefect> OK, if you check
http://www.tangentaudio.com/ for the top post, that's a link to the video.. Hopefully it's processed and available at your local youtube cache
[23:42:50] <toastydeath_> so you are using a self-driving dog
[23:42:58] <toastydeath_> or a fixtue of some kind?
[23:43:19] <andypugh> You know the drive centres that woodwokers use? A centre and 4 prongs? It was just like that but with hydraulic prongs, made by Sandvik
[23:43:20] <toastydeath_> er, self driving center
[23:43:55] <andypugh> Yes. That was my question. Do they still make them? I have seen no evidence for them the last 20 years.
[23:44:04] <toastydeath_> i have not seen them
[23:45:21] <toastydeath_> sorry, i jumped into the middle there and assumed that was the original topic
[23:45:39] <andypugh> (And in this particular case parting-off the bit held in the chuck wasn't an option, as we were always trying to absolutely maximise the specimen length as it makes for better results, and often we were taking throug-thickness specimens.
[23:46:47] <toastydeath_> understandable
[23:47:25] <MattyMatt> didn't win the metal Z. oh well it was only 4" travel
[23:47:37] <gweepprefect> http://www.youtube.com/watch?v=vBZkyT-3y6g is the debug video which shows the one direction problem
[23:48:23] <toastydeath_> looks like a pitch error
[23:48:29] <andypugh> Daft question, but are you confident in the DTI?
[23:48:29] <MattyMatt> my draper dremel wouldn't work with a 44mm clamp anyway. collar is 19mm, body is 48mm
[23:48:52] <toastydeath_> ooh, clever girl - check it with gage blocks
[23:49:19] <anonimasu> hmmm
[23:49:41] <gweepprefect> well, i measured it with my caliper, it seems to be 5mm as advertised
[23:49:50] <andypugh> OK, just a thought.
[23:50:08] <gweepprefect> it does feel like that sort of problem, but my other axes are behaving properly and they have the same screws/nuts/etc.
[23:50:13] <toastydeath_> if you're sure it's not the dti, and I'm not convinced without gage blocks or something similar, it looks like a pitch error
[23:50:21] <archivist_emc> adjust the scale to that stepper and retest
[23:50:33] <andypugh> It all seems similar to a long thread on the mailing list about one axis on someones machine. Does anyone recall what the upshot of that was?
[23:51:04] <gweepprefect> fwiw, i've completed swapped over motor/drive/output and i'm running the Y with all of the X hardware, just to factor that out, it behaves the same
[23:51:25] <archivist_emc> screw is a constant though
[23:51:46] <archivist_emc> and so is the scale set for that axis
[23:51:46] <gweepprefect> yep, as are the pulleys on the motors... which may factor in
[23:52:16] <gweepprefect> they're timing pulleys, i can't imagine how one would be made wrong and still mate with the belt, but it did cross my mind
[23:52:27] <toastydeath_> i don't see why this isn't fixed by changing the pitch constant in the machine
[23:52:43] <toastydeath_> no screw is going to be dead on pitch
[23:52:50] <toastydeath_> and no pullies/etc are going to be dead on diameter
[23:52:59] <andypugh> You would hope for better than that, though
[23:53:04] <toastydeath_> hope for, yeah
[23:53:09] <MattyMatt> my original design had X & Y same screws, for just this kind of testing
[23:53:46] <andypugh> And with a synchronous drive (ie toothed belt) pulley out-of-round has only a tiny effect.
[23:54:31] <andypugh> How many teeth on the pulleys?
[23:55:31] <gweepprefect> 10:20 motor:screw
[23:56:55] <John_f> gweepprefect: do you have servos or steppers?
[23:57:13] <gweepprefect> steppers
[23:57:24] <gweepprefect> i can recap my basic setup since many of you probably missed it
[23:57:36] <andypugh> If the big pulley is really 19 you would get 5 thou error in each 0.1" step, so it isn't that.
[23:57:56] <gweepprefect> latest live CD, PCI parallel port card, gecko G203V drives, NEMA23 steppers
[23:58:12] <gweepprefect> same motors/screws/nuts on all 3 axes
[23:58:20] <andypugh> Can't be direction hold or something like that because you are only going one way
[23:58:23] <gweepprefect> x and z both work quite well, y is giving me a headache
[23:59:07] <andypugh> Where are you measuring, with respect to the screw?
[23:59:30] <andypugh> It isn't that the table is twisting on the slide is it?
[23:59:36] <gweepprefect> where on the screw is the nut, or where is my indicator?
[23:59:50] <andypugh> Where is your indicator
