#emc | Logs for 2009-11-27

[00:38:28] <DaViruz> i think i'm in love with pc104
[00:47:14] <MarkusBec> MarkusBec is now known as MarkusBec_away
[00:51:16] <tom3p> did you know you can put "apt:viking' in the url bar of firefox to use apt to get the packages for viking? ( or whatever you want to install, just like apt-get install blah ) i didnt know that :o
[00:52:06] <tom3p> cool, it shows up in menu bar under education ( its a gps app )
[01:24:33] <tom3p> sounds like a great deal, but will the mobo run emc2 well? http://www.tigerdirect.com/applications/searchtools/item-Details.asp?EdpNo=5530098&sku=B69-0154
[01:26:54] <tom3p> woohoo! its on the list http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Latency-Test
[01:29:15] <tom3p> the nvidia pci card may not play well
[01:31:30] <tom3p> haha lets see what google find for nvidia on emc ... "01:44:55 <cradek> * cradek spits on the nvidia driver" oh, ok, i spose thats bad
[03:56:51] <fenn> is there something like the screw compensation code as a hal module for more arbitrary stuff? piecewise linear interpolation
[03:59:09] <cradek> I don't think so
[03:59:27] <cradek> getting the table into it would be a pain
[03:59:47] <fenn> did anyone ever figure out how to do software axis alignment, like if your cross slide is not perpendicular
[03:59:59] <fenn> or straightening a bent axis
[04:01:13] <MattyMob> fenn that's easy if you want linear interpolation. the trick would be putting the layer in emc. would you e.g. like to see the transformed gcode?
[04:02:34] <fenn> no, i want the machine to pretend it's perfect
[04:04:19] <MattyMob> emc must co-operate with that, so it'd need to go down below the gcode level
[04:05:14] <cradek> seems like you could write kins for that
[04:05:41] <MattyMob> could be
[04:09:12] <fenn> i'm thinking hacking the blend module to allow for multiple linear functions
[04:09:23] <fenn> then set the values with a shell script
[04:09:53] <fenn> and add that offset to the joint value
[04:11:38] <tom3p> fenn: then it wouldnt jog straight with a bent screw, or jog square with x mislligned to y. (cuz its in blend )
[04:12:13] <tom3p> in kins it would
[04:12:55] <cradek> fenn: do you have a machine with an error like this that's hard to just fix?
[06:36:02] <awallin> there was a blend paper quite recently
[06:36:19] <awallin> I asked for the code, naturally they refused.
[06:36:39] <awallin> university research sponsored by heidenhain etc
[06:38:18] <awallin> http://dx.doi.org/10.1016/j.ijmachtools.2009.11.005
[06:39:04] <awallin> I think it is going to be hard to make progress on blending/lookahead if the requirement is that it be very general 6-axis motion and applicable to all kinds of kinematics
[06:39:26] <awallin> 3-axis trivial kins is probably doable, even for amateurs like us...
[06:59:18] <tom3p> awallin: commercial cncs expect trivkins , never complex (cnc controls not cnc machines) , so that (non-triv) may be a nice thing but a real complication
[07:00:59] <awallin> if someone sends me a HSM machine I promise to work on it :)
[07:01:32] <tom3p> :) well it is near xmas, you been good?
[07:01:55] <dimas> :)
[07:02:35] <tom3p> domani !
[07:47:49] <Engineer> good morning all :)
[07:51:09] <dimas> hi
[07:53:19] <Engineer> hi dimas
[08:07:53] <MarkusBec_away> MarkusBec_away is now known as MarkusBec
[08:39:06] <numen> good morning
[08:39:27] <micges_work> hi
[08:51:33] <numen> why did i dranke some alk yesterday?
[09:08:59] <numen> re
[10:50:17] <numen> has anyone here an lasercutter?
[10:52:24] <micges_work> yes
[10:53:24] <numen> i search one, who could cut me 2parts for my mill
[10:54:32] <numen> steel 10mm
[11:10:10] <awalli1> we have an old 25W CO2, I wonder what could be done with it...
[11:10:56] <numen> awalli1 could you cut steel with this?
[11:12:21] <awalli1> no idea, it burns wood, I think for steel with a thickness of more than 1-2mm you will want 50 or 100W or much more...
[11:15:15] <numen> i need the 2 sidewalls for my mill ^^
[11:16:56] <archivist> frallzor found someone in the UK to cut thick sheet and ship to him
[11:17:14] <archivist> but you must have a local to you
[11:32:37] <tarzan> why not saw it?
[12:03:39] <numen> i will show you an pic
[12:12:28] <numen_> http://imagebin.org/73178
[12:12:31] <numen_> numen_ is now known as numen
[12:16:50] <tarzan> small bottom holes meant for laser too?
[12:18:22] <numen> nope
[12:18:58] <numen> only the slot for the ipa steel and the big hole
[12:21:14] <numen> why?
[12:22:51] <tarzan> i thought small holes were hard to do on laser
[12:23:29] <archivist> that size is not small
[12:24:27] <numen> 5mm
[12:24:31] <numen> in 10mm steel ^^
[12:25:06] <archivist> 5mm bolt is too mall for 10mm plate
[12:25:40] <numen> nope, but these, i dont want to laser
[12:26:16] <numen> there i will have to cut some threads
[12:39:21] <numen_> re
[12:39:27] <numen_> numen_ is now known as numen
[13:36:22] <ries> can I run talency-test on the command line? Or is this a X utulity only?
[13:37:24] <jthornton> yes
[13:38:28] <ries> jthornton: I am getting TclError: no display name .........
[13:38:34] <jthornton> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Latency-Test
[13:47:30] <ries> jthornton: I found it here actually : http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?TroubleShooting
[14:19:14] <tarzan> numen, does laser cut hardens machining surface ?
[14:27:15] <numen> tarzan not realy
[14:27:18] <numen> why do you mean?
[14:32:42] <tarzan> the laser cutting may harden surface for thread cutting being difficult?
[14:33:18] <numen> i dont want to laser the holes, where i want to cut the threads
[14:33:34] <numen> just the outlines and the H with its big hole
[14:37:59] <ries> funny... my latency test shows far worse number in a terminal then in X
[14:54:20] <jthornton> archivist turns out the control on the Hardinge is alive... robh showed me how to turn it on with the secret buttons and handshake
[14:54:35] <archivist> hehe
[14:54:49] <archivist> all running then?
[14:56:58] <jthornton> dunno, I have to load the PLC file back into the controller first
[14:57:31] <jthornton> if the power is off the battery goes flat and when you turn it on you have no PLC program so nothing happens
[14:59:24] <archivist> nasty
[14:59:59] <celeron55> that's why they invented flash memory 8)
[15:00:04] <archivist> and someone gets to charge for reloading
[15:00:32] <jthornton> yep, that's the game they played
[15:15:37] <MattyMob> have you looked at the tech? is the ram & rom 28 pin dips?
[15:18:04] <MattyMob> 686 mobos have 29C020 flash chip which fits in an eprom circuit
[15:20:23] <jt-plasma> no, I have not popped the lid off of it yet
[15:20:36] <tarzan> jthornton, secret buttons and handshake?
[15:21:57] <jt-plasma> tarzan: yes, if the battery is dead you have to hold the diag button down while you press the control on button to get to the setupscreen
[15:26:25] <numen> bbl
[15:26:41] <christel> [Global Notice] Hi all, We're about to perform some memleak related maintenance on a handful of the ircd servers, at the same time we'll be rehubbing a bit. It will be noisy for the next ten minutes or so. Apologies for the inconvenience.
[15:27:28] <christel> [Server Notice] Hi all, the server you are connected to will be going down shortly for maintenance. Apologies for the inconvenience.
[18:23:59] <ChanServ> [#emc] "This is the #emc channel - talk related to the Enhanced Machine Controller and general machining. Website: http://www.linuxcnc.org/, wiki at http://wiki.linuxcnc.org/"
[18:25:04] <ChanServ> [#emc] "This is the #emc channel - talk related to the Enhanced Machine Controller and general machining. Website: http://www.linuxcnc.org/, wiki at http://wiki.linuxcnc.org/"
[19:22:59] <numen> can anyone check an shematic?
[19:23:24] <awallin> is it very advanced?
[19:24:58] <tarzan> just curious, pls post
[19:26:08] <numen> eagle ok?
[19:26:16] <numen> or what kind do you need?
[19:26:26] <celeron55> a .png is good
[19:26:30] <awallin> put a picture on imagebin
[19:26:31] <numen> ok
[19:26:33] <celeron55> i'd like to see it too
[19:28:02] <numen> mom
[19:30:17] <numen> http://imagebin.org/73215
[19:30:22] <numen> but no values for r and c
[19:32:25] <tom3p> irf drivers, irf fets, but wtf is the main chip?
[19:32:49] <numen> a3986
[19:35:02] <tarzan> MOSFET Driver?
[19:35:41] <numen> a3986 is i microstep stepper controller with internel highside drivers
[19:35:53] <numen> but i want to use with externel drivers for a higher motor voltage
[19:42:08] <tom3p> do you common the motor - to the 3986 gnd ?
[19:42:20] <numen> yes
[19:43:00] <numen> i think, i have to, otherwise i would have to use some amp isolators for the current feedback
[19:47:09] <tom3p> http://pminmo.com/phpBB2/viewtopic.php?f=24&t=130&start=15
[19:47:47] <numen> what you want to show me?
[19:48:03] <tarzan> is x1-4 motor gnd an what is mstba6?
[19:48:47] <numen> yes its true
[19:49:08] <numen> mstba6 is the type of connector for the motor^^
[19:53:15] <tarzan> is the a3986 broke on forum?
[19:53:57] <numen> i added the externel fet drivers for a higher motor voltage
[19:58:43] <numen> tom3p i cant understand, what you want to say with this thread?
[20:00:05] <robh> jt-plasma, hi, you around
[20:07:26] <tom3p> numen: the work you are doing avoids the design of the chip (50V max) and the chip has been found to be unreliable according to that thread. so, you are working a lot and may have no reward.
[20:08:31] <numen> why not?
[20:08:47] <numen> sry my english is not very good ^^
[20:09:15] <tom3p> i cannot make you understand whats been printed, so ... i cannot make you understand. best of luck in you efforts
[20:09:35] <numen> so you think, this could not work?
[20:11:18] <tom3p> correct
[20:11:30] <numen> but why?
[20:11:34] <numen> cant you explain me?
[20:15:07] <jt-plasma> robh: yes I am
[20:15:45] <robh> u have email
[20:23:14] <tarzan> numen, why do you microstepp operation?
[20:24:13] <numen> cause like this, the motor works better
[20:24:31] <tarzan> gives you more prescision?
[20:24:47] <numen> it does not turn realy clear
[20:25:26] <tarzan> give a more uniform feed cut?
[20:25:57] <numen> what does this mean? sry for bad english, i speak german ^^
[20:26:44] <tarzan> tool cut is more uniform or gives better finish on piece?
[20:27:41] <numen> motor is more powerfull less missing of steps
[20:31:55] <tarzan> LMD18245 no useful?
[20:32:07] <numen> hmmm
[20:32:14] <numen> i dont linke this one
[20:32:23] <numen> not enough power, not high enough voltage
[20:33:21] <tarzan> voltage cannot increase with mosfet?
[20:33:51] <numen> the a3986 suports only 50v on its gate drivers, so i want to use externel ones
[20:39:52] <celeron55> according to that thread the chip is overally bad
[20:40:09] <numen> what is bad with this chip?
[20:40:14] <celeron55> read the thread
[20:45:23] <tarzan> is the same? http://cuteminds.com/index.php/en/eaglecnc/56-rs274x.html
[20:48:22] <celeron55> that's the basic circuit the chip should be used in, which works with voltages lower than 50V
[20:50:36] <tarzan> it says suitable for big motors
[20:52:12] <numen> tarzan same chip, but max 50v
[20:53:57] <awallin> anyone know alot about THK HSR rails? They have a thin stripe and a thick stripe on the bottom surface. How do I mount them for my X-axis, thin stripes facing in or out?
[21:00:43] <andypugh> awallin: I was looking at servo controllers on eBay, and a sponsoered link came up. http://www.granitedevices.fi/
[21:00:58] <andypugh> Seems you have your own native servo drive manufacturer.
[21:03:35] <tarzan> numen, is you motor big?
[21:03:40] <awallin> yes, he has done very well it seems
[21:04:05] <awallin> andypugh: I think he only started experimenting in 2006-7 when I was playing with brushless drive ideas also
[21:05:11] <awallin> looking for 16x16 left handed lathe tooling today, and a small ~1kW VFD... ebay can become addictive :)
[21:05:11] <andypugh> Sadly I don't see resolvers on the inputs list.
[21:05:29] <awallin> I already ordered an incremental encoder
[21:05:46] <awallin> I need to extend the encoder/hall sensor shaft through the back cover and mount it there
[21:05:51] <awallin> 8mm axis probably
[21:06:14] <andypugh> Sounds like it should be straightforward if you already have a lathe.
[21:06:54] <awallin> there's an M5 or maybe M6 bolt in that shaft which holds it in place... that might be used
[21:08:10] <andypugh> Should work, with a bit of glue, possibly.
[21:11:00] <andypugh> Talking of servos, this pair look quite nice, but I don't know if they would have the power I want for my lathe axes.
[21:11:02] <andypugh> http://cgi.ebay.co.uk/ws/eBayISAPI.dll?ViewItem&item=330378741342
[21:11:21] <awallin> on the manual lathe I don't really know how to cut threads, but I could turn a shaft to the correct diameter, drill it, tap it, and insert a set-screw. This would create a long 'screw' which has a shaft that extends through the back cover for the encoder
[21:11:40] <EbiDK> EbiDK is now known as EbiDK|AWAY
[21:12:21] <tarzan> numen: give it a try if big voltage needed use also http://www.c-n-c.cz/download.php?id=6105&sid=1cef88989f604502198d83f14fc6a7c2
[21:12:33] <andypugh> Well, to cut threads you want a spindle encoder. Which leaves you in a bit of a "chicken and egg" situation
[21:12:41] <tarzan> numen: http://www.c-n-c.cz/download.php?id=6109&sid=1cef88989f604502198d83f14fc6a7c2
[21:13:03] <awallin> andypugh: I was thinking I make the shaft on the 1949 Schaublin...
[21:13:28] <tarzan> numen: http://www.2a.pl/~mark/cnc/a3986/
[21:13:40] <andypugh> Cutting threads on a manual lathe isn't particularly hard, and possibly a skill worth learning.
[21:14:15] <tarzan> in fact 4 jaw chuck threading is gonna live forever
[21:14:54] <numen> tarzan with this, only 50v are possible ^^ i want more ^^
[21:15:00] <andypugh> But your proposal would work too. If there is any exposed shaft to use, then I would try to have a register on the outside diameter of the shaft to keep concentricity.
[21:15:43] <numen> for this, i have the externel gate drivers
[21:17:19] <tarzan> numen: build it there is no much more power chips
[21:17:49] <numen> tarzan this is tha standart aplication, where only 50v are possible
[21:18:07] <ries> hey all, when I try 'test this axis' my computer hangs, what can I do?
[21:19:05] <celeron55> try lowering the control thread frequency or whatever it is called
[21:19:15] <tarzan> numen: amperes?
[21:20:06] <numen> about 4 each phasis
[21:20:08] <celeron55> just a guess, but if it's too high, it may take all your cpu cycles and it will hang
[21:20:23] <celeron55> quite unprobable though, i think
[21:20:25] <ries> celeron55: ok.. will do
[21:20:56] <celeron55> if it's at it's default value and you have a decently fast computer, probably not worth even trying
[21:21:18] <celeron55> (but i think i once made it too high and it didn't work)
[21:22:29] <tarzan> numen: it should work try it
[21:22:32] <celeron55> ries: what kind of computer do you have?
[21:23:06] <celeron55> what processor
[21:23:58] <numen> are u shure?
[21:24:29] <tarzan> numen: no but you have no choices
[21:24:39] <ries> celeron55: it's a PIII I think, a Dell laptop runs at 1.2 or 1.8 GHZ..... just testing my HW right now untill I have my real computer for the CNC work
[21:24:57] <numen> ive 10pcs of the a3986 so i will use one for a try^^
[21:25:47] <tarzan> numen: copy links above and try
[21:26:53] <ries> celeron55: currently when I test it I hear my stepper motor giving about a tick per second
[21:38:36] <andypugh> I don't suppose you have a stupidly small maximum speed set?
[21:42:09] <ries> andypugh: yeaaa... but if I set it faster, then the PC hangs
[21:42:15] <ries> and I need to restart it
[21:42:56] <andypugh> I meant maximum motor speed, not base thread period
[21:43:54] <ries> yes, I set it to 10mm/sec
[21:44:39] <frallzor> nice day today, started cutting down the angle iron for the tracks
[21:44:48] <frallzor> oh yeah ries I have a Q for you
[21:45:01] <ries> frallzor: sure
[21:45:21] <frallzor> for the Y-rails, why do they drill 12mm holes for M8 screws?
[21:46:29] <andypugh> ries: Fist things first, what sort of step signal can you see in Halscope?
[21:46:40] <ries> frallzor: what drawing?
[21:46:49] <andypugh> And what is your BASE_THREAD at the moment?
[21:46:49] <frallzor> 12mm on the rails, 12mm on the 50x100 tubing and then the "plate" with M8 threads
[21:46:53] <ries> andypugh: Halscope?
[21:47:07] <andypugh> Are you running the Axis UI?
[21:47:14] <frallzor> 10 20 246
[21:47:30] <ries> andypugh: I am running stepper mill confiuration
[21:47:32] <frallzor> 10 20 440
[21:47:44] <andypugh> Halscope is a virtual oscilloscope that lets you look at all the software signals
[21:48:12] <ries> andypugh: I can start that from the command prompt?
[21:48:13] <frallzor> while the 10 20 235 if threaded with M8
[21:48:57] <andypugh> I believe so, or from Machine->Halscope in the Axis UI. If you are in stepconf then I am not sure
[21:50:06] <andypugh> Yes, halscope at the command line works
[21:50:22] <ries> let me try that in a sec... looking up the answer for frallzor
[21:50:51] <frallzor> Its both odd and logical, odd since its quite a difference, logical; room for adjustment
[21:52:22] <andypugh> frallzor: Adjustment seems very probable. Does the arrangement mean that the area under the bolt heads is 8mm, or 12mm?
[21:52:31] <frallzor> 12
[21:53:05] <andypugh> Hmm, that strikes me as a bit scruffy :-)
[21:53:23] <frallzor> 12 12 and then 8 to fasten the part with the first 12 to the second 12
[21:53:36] <frallzor> 9mm Id buy
[21:53:40] <frallzor> but 12...
[21:54:08] <andypugh> What size counterbore?
[21:54:17] <frallzor> countebore?
[21:54:26] <ries> frallzor: I tell you honestly that I never used that clab strip
[21:54:32] <numen> can anyone here cut me 2 parts with a laser?
[21:54:40] <frallzor> how did you fasten rails then?
[21:54:58] <frallzor> tap the tube?
[21:55:06] <ries> clamp... we use use the L, lined and measured it up... then drilled it with m4... then drilled to Mxx to tap M12 in it
[21:55:30] <ries> Ahh... wait, we put 3 tack welds on it...
[21:55:39] <ries> then drilled it
[21:56:08] <frallzor> well ill drill them separately, no issue but you did tap the 10x50?
[21:56:10] <frallzor> *100x50
[21:57:31] <ries> for teh X rails we drilled it through and used bold/nut
[21:57:37] <ries> the Y rails need to be tapped
[21:57:47] <frallzor> yeal Ill do that too with them
[21:57:53] <ries> make sure you dill the U bar and L rail together
[21:58:56] <frallzor> so bolted X-rails and then threaded the tube for y-rails, correct right?
[22:01:15] <andypugh> Interesting machining challenge from a friend. He has given me a 3mm OD bit of tube with a 0.4mm hole down the middle. He wants e to machine the last 10mm down to 1mm OD..
[22:01:57] <ries> frallzor: I am going to ask it on the forum... I have no idea how the clamp strip works, how it works and what it's good for
[22:02:46] <frallzor> but I just want to know how you did :P afaik its just for comfort and ease, since its 1:1 with the other holes its easy to clamp it with the rest
[22:02:50] <ries> so bolted X-rails and then threaded the tube for y-rails, correct right? <= That's how I did it
[22:03:11] <ries> if you drill them both at the same time, then they are lined up nearly perfect and you have a tiny bot of room to line it out
[22:03:24] <frallzor> yeah
[22:03:38] <frallzor> I need clamps
[22:03:56] <ries> get a lot of them ;)
[22:04:20] <ries> but we tack welded also, our material was quite good.. even though it was bend steel
[22:04:23] <frallzor> only got a few of those woodworkers usually use
[22:04:50] <frallzor> yeah ill put a few tacks on them too
[22:04:52] <frallzor> just in case
[22:06:23] <ries> andypugh: when I try to run halscope, I get 'could not open shared memory (errno=2)
[22:06:38] <ries> Then rtaapi init failed
[22:06:39] <andypugh> Interesting.
[22:06:58] <andypugh> dmesg?
[22:07:25] <celeron55> you have the default live cd install?
[22:07:57] <archivist> `qaz
[22:07:57] <ries> celeron55: yes I have...
[22:08:08] <ries> the system always have been a bit flacky though, because sometimes latency test didn't work
[22:08:13] <ries> and after 2 reboots it does
[22:08:14] <archivist> err 2 is a non existent file
[22:08:32] <celeron55> that looks like something i'd expect to get when i try installing realtime kernel and emc on some random debian box :P
[22:09:03] <ries> let me run it from teh life CD
[22:09:25] <ries> I did some 'fooling' around, but that was more like removing sounds drivers, cupsd etc.... (why needs cupsd... right???)
[22:10:42] <ries> note:also mach3 on windows XP gives weird results... sometimes it starts... sometimes it doesn't
[22:10:47] <ries> so I think I have HW issues
[22:11:38] <andypugh> Stop! Not so fast
[22:12:00] <andypugh> Hit the "Test axis" button first, then start halscope
[22:12:27] <andypugh> Stepconf by itself doesn't start RTAI
[22:14:06] <ries> will do....
[22:14:14] <ries> give me a minute then :)
[22:27:07] <ries> andypugh: weird... with the life CD it seems to work much better
[22:27:15] <ries> something is broken on the install then
[22:38:09] <ries> frallzor: so I have lift off
[22:38:45] <numen> gn8 every one
[22:38:52] <numen> time to go to bed...
[22:39:00] <ries> night
[22:41:55] <frallzor> lift off? =)
[22:47:15] <ries> frallzor: yeaa I got my computer working with EMC2
[22:48:01] <frallzor> ah
[22:48:26] <frallzor> my table is pretty much done now =)
[22:48:40] <ries> you are going fast :)
[22:48:44] <frallzor> just need to spanke on 4 plates with feets on it
[22:48:47] <frallzor> *spank
[22:48:52] <frallzor> *feet
[22:48:54] <frallzor> bah Im tired
[22:49:29] <frallzor> some welding on gantry and car left
[22:52:02] <frallzor> and the rail-grinding later tomorrow
[22:53:56] <ries> you will have some fun with the rail grinding :)
[22:54:28] <frallzor> cutting is going well and ive adjusted the extra piece of metall to be nice for the angles
[22:54:34] <frallzor> so im all set to cut like crazy
[22:54:35] <ries> is it correct that if the unit is in mm, a leadpitch is in the order 0f 0.0097.... ?
[22:55:07] <ries> frallzor: I only need to do teh z plate, some microswitches and stoppers... cabling on the MM
[22:55:47] <ries> Cabinet is almost done, only I need a transformer currently I use the transformer of my power amp.
[22:56:02] <frallzor> and I also need all the electronics :P
[22:56:09] <frallzor> thatll take a good while =P
[22:58:28] <ries> frallzor: are you planning to use Gecko, PMDX ?
[22:59:05] <frallzor> geckos and pmdx board yes
[22:59:28] <frallzor> If i cant find another nice 7A board for less
[23:03:13] <ries> frallzor: 203V?
[23:03:18] <ries> G203V that is...
[23:05:00] <frallzor> yup
[23:10:15] <ries> good combo I think...
[23:12:41] <frallzor> and then the geared motors from oriental
[23:12:44] <frallzor> seems ok
[23:16:05] <frallzor> now off to bed! GN
[23:21:48] <andypugh> ries: Going back 40 minutes, .0097 might be right in mm per step.
[23:22:20] <ries> andypugh: I am confused... later I was thinking that the setup asks for mm/rev
[23:22:36] <ries> so it might be for me close to 100mm
[23:22:57] <andypugh> The mm/rev and step/rev and microstep and pulley ratio all get multilied together.
[23:25:26] <andypugh> Hmm, but something is wrong, as my SCALE numbers are around 300
[23:26:15] <ries> andypugh: so that must be mm/rev
[23:26:31] <andypugh> Yes, leadscrew pitch in mm.
[23:26:34] <ries> so basicly my pinion... (I have a rack + pinion system, direct driven)
[23:26:52] <ries> I am not a guru... but i's around 96mm I think fo rme..
[23:27:07] <ries> also, when I fill that in my motors almost make a revolution during testing
[23:27:20] <andypugh> Yes, that sounds more promising.
[23:27:39] <ries> I made screenshots (video) and upload them shortly
[23:27:51] <andypugh> At 0.0097 it will be making 1 step every 100mm of movement, which is pretty much what you were seeing
[23:29:26] <andypugh> Though 96mm per rev seems very coarse. The motors can do 1000rpm or so, so you are sacrificing a lot of cutting foce for speed you don't need.
[23:30:09] <ries> andypugh: correct, I am planning to make a beld driven reduction (1:3 or something) in the near feature
[23:30:37] <ries> it's a large machine, 2.5m x 1.5m so I don't need something ultra precise...
[23:30:41] <andypugh> 10:1 would be fine, I think.
[23:30:56] <ries> note taken....
[23:30:57] <andypugh> Ah, perhaps 10:1 would take too long to go end-toend then
[23:31:08] <ries> I wonder how that looks like beld driven though....
[23:31:31] <ries> most people with this machine go for 1:3 if my memory serves correctly
[23:31:33] <andypugh> If I was doing a big machine I think I would use direct belt drive.
[23:31:57] <andypugh> (ie toothed belt anchored at each end, and a small pulley meshing with it)
[23:32:06] <ries> that would mean a very long beld on teh X and Y, right?
[23:32:14] <andypugh> Yes.
[23:32:34] <ries> are you not going to get a lot of backslash due to teh belds length?
[23:32:45] <andypugh> But then 2.5m of rack isn't cheap, is it?
[23:34:45] <andypugh> Not much backlash, but certainly some stretch. I understand that with the right preoload it isn't a huge problem, but I have no real experience.
[23:35:14] <andypugh> There is a version of the principle that uses a second belt bonded to the frame to get rid of stretch.
[23:35:55] <ries> andypugh: I got my racks from Japan, transfer by mail was much more expensive then the racks itself
[23:36:39] <ries> the reason they use racks + pinion on this machine is because it's used to do heavy cutting one pass 10mm or more from what I did read
[23:36:53] <ries> so You need a heavy beld to copy with the stresses
[23:36:56] <andypugh> In wood, or steel?
[23:37:04] <ries> people mount 5HP motors on them
[23:37:05] <ries> wood
[23:37:38] <andypugh> You can do 10mm cuts hand-held with a router.
[23:38:37] <andypugh> But anyway, you have racks, they will work well. What I would use on a purely hypothetical machine is irrelevant.
[23:41:16] <ries> andypugh: I just took the designs and build it... I am far from a guru on the subject....
[23:41:23] <andypugh> Fascinating what you can find by accident.
[23:41:25] <andypugh> http://www.sdp-si.com/eStore/CoverPg/Belts.htm
[23:41:33] <ries> I understand teh reason and thoughts behind the design, but couldn't improve it :)
[23:41:56] <andypugh> Don't go imagining that I know a lot myself, I am pretty new to this game.
[23:49:58] <ries> what are you making? Having?
[23:50:20] <andypugh> I am making a clock.
[23:50:44] <andypugh> But about 18 months ago I decided it would be easier with a CNC lathe/mill.
[23:51:01] <andypugh> And that has rather become a project of its own.
[23:52:16] <spasticteapot> How much power would I need if I wanted to drill 1" holes in steel?
[23:52:21] <spasticteapot> And what sort of drillbits should I use?
[23:52:22] <MarkusBec> MarkusBec is now known as MarkusBec_away
[23:52:44] <andypugh> 500w will do it. Use core bits.
[23:53:41] <spasticteapot> Core bits?
[23:53:48] <andypugh> http://www.rotabroach.co.uk/
[23:54:22] <andypugh> It does rather depend on how thick the steel is
[23:54:50] <spasticteapot> 1/4" or less.
[23:54:50] <spasticteapot> I have a mad plan to make money selling project enclosures to audio DIY enthusiasts.
[23:55:24] <andypugh> Hmm. CNC punching would be more usual than drilling then
[23:55:39] <spasticteapot> I don't really know that much about machining, but I do have enough experience making jewelry that I can fudge it.
[23:55:41] <andypugh> Most panel-mount things go in non-round holes
[23:56:32] <spasticteapot> andypugh: I can't afford a CNC machine.
[23:56:42] <spasticteapot> Most of the holes made will be for RCA jacks and the like - very small.
[23:56:48] <andypugh> Q-Max cutters.
[23:57:39] <spasticteapot> I just bought a set of Greenlee metal punches - here in the 'states, they're what the professionals use.
[23:57:51] <spasticteapot> They're very very similar to what you just linked me to.
[23:57:57] <andypugh> Indeed
[23:58:07] <spasticteapot> Could I use them on 3/16" thick metal?
[23:58:44] <spasticteapot> Even steel?
[23:58:44] <spasticteapot> Wow.
[23:59:37] <andypugh> Hmm, perhaps not that thick.
[23:59:49] <spasticteapot> It says up to 10ga. mild steel, so I think I'm out of luck.
[23:59:55] <spasticteapot> Perhaps you could recommend a different material?