#linuxcnc | Logs for 2013-04-05

[01:31:09] <Loetmichel> mornin'
[01:57:28] <r00t4rd3d> night
[02:29:42] <DJ9DJ> moin
[02:33:53] <carper64> just deciding what to do today
[05:10:17] <V0idExp> Hello everyone!
[05:11:32] <V0idExp> Is there a way to split core linuxcnc program from GUI, TK, and other graphical stuff? I'm implementing a customized linux distro, based on linuxcnc, which will act as an autonomous CNC device, which is commanded via TCP/IP
[05:12:23] <V0idExp> it would be nice to have a kind of linuxcnc-core, linuxcnc-axis, linuxcnc-utils or something like that...
[05:12:34] <V0idExp> say, it's like a feature request :)
[05:27:13] <archivist> it already splits at a messaging interface http://www.linuxcnc.org/docs/html/code/NML_Messages.html
[05:29:22] <archivist> you should also read the last few months on the developers mailing list for some other ideas being discussed recently
[05:38:56] <mhaberler> V0idExp: I suggest for now you use the NML messaging interface archivist referred to; also, look into the Python linuxcnc module and how you can set it up to work remotely with an emc.var file
[05:39:15] <mhaberler> I really suggest not to use linuxcncrsh or halrmt
[05:40:05] <mhaberler> you will likely run into problems with GUI's that access HAL, like GladeVCP - those cannot work remotely at this point in time; I am working on that.
[05:40:49] <V0idExp> Well... being less experienced, and needing a rapid solution, I've implemented a custom-protocol TCP/IP server, which is a Python socket-based packet listener, which based on commands arrived from socket, calls linuxcnc python functions... :D
[05:41:14] <mhaberler> excellent hack, likely more functional than linuxcncrsh
[05:41:57] <mhaberler> but actually the Python linuxcnc doesnt need that, it should be able to work remotely
[05:42:09] <mhaberler> all it needs is the proper emc.var file I think
[05:42:09] <V0idExp> well, yes, it allows custom behaviour, since commands are processed... also I implemented a custom python HAL component, which interacts with the server, allowing status reporting and polling from client
[05:42:33] <mhaberler> ah ok, sounds like a sensible workaroung for the current situation then
[05:42:42] <mhaberler> HAL would have been the next issue, yes
[05:43:31] <mhaberler> have you managed to tie that into PyVCP or gladevcp? this will be stock later this year, but not before summer
[05:49:09] <V0idExp> well, it's something like this:
[05:49:09] <V0idExp> 1) a python socket-based server, which listens on a given port for custom simple protocol packets, like {type='exec', data='g01 x100 y200'} or {type='home', data='x,y,z'}. In a multithreaded environment, it processes packets and eventually calls the linuxcnc.command or linuxcnc.state functions
[05:49:09] <V0idExp> 2) a python HAL real-time component, which does polling on CNC state, like limit switches, axis position and some pins and implements some "low-level" logic, like some movements based on machine's situation. the component communicates with the server via the same protocol (imagine a packet like {type='state', data='...lot of data...'}, through a UNIX domain socket.
[05:49:09] <V0idExp> 3) a client connects to the "remote cnc machine", and can be written in any language...
[05:49:09] <V0idExp> 4) the next step is to substitute the packet mechanism with some sort of RPC ;)
[05:49:09] <V0idExp> ... this is just for explaining the solution I've adopted... what I need is to reduce to bare minimal necessary the contents and dependencies of the linuxcnc package... since it will be contained on something like an read-only mounted flash card and will not even have a GUI. Imagine a black atom-based box, which just has 50-pin in/outs, a 12V DC jack and an ethernet port. That will eventually a commercial solution...
[05:54:15] <V0idExp1> sorry guys, lost inet connection :(
[05:54:20] <V0idExp1> anyone has replied something?
[05:55:20] <archivist> no
[05:55:43] <ReadError> whats the web interface for linuxcnc?
[05:55:49] <ReadError> i remember seeing something about it
[05:55:52] <ReadError> to run headless
[05:56:37] <V0idExp1> yep, something like that, but instead of a web interface, it will be an RPC service, the interface is a custom closed-source application, which interfaces with machine's user
[05:57:06] <ReadError> #buzzkill
[05:57:09] <ReadError> closed source
[05:58:00] <V0idExp1> well, the "client" will be closed source, the RPC server itself, documented, will be released as LGPL, GPL, BSD or something like that ;)
[05:58:29] <V0idExp1> this is because the client application is based on machine vision computing, which is the cornerstone of our product
[05:59:01] <V0idExp1> but the distro itself, server and everything else which is hosted on "cnc blackbox" will be released for free and will be open
[06:01:15] <mhaberler> V0idExp1: you can always read back in the logs like so:
[06:01:27] <mhaberler> logger[mah]: zappadoing!
[06:01:28] <logger[mah]> mhaberler: Log stored at http://linuxcnc.mah.priv.at/irc/%23linuxcnc/2013-04-05.html
[06:02:02] <V0idExp1> ah, thx! :)
[06:03:01] <mhaberler> Dont expend a lot of effort in inventing your new remote procedure calls which are going to be not needed in a few months - this is needed only temporarily
[06:34:26] <L84Supper> is the hobby and retrofit CNC controller market truly that lucrative?
[06:49:51] <gabewillen> So i am assuming its strictly prohibited to call read() in a hal component on /dev/tty?
[06:59:23] <jthornton> The phaonmneal pweor of the hmuan mnid, aoccdrnig to a rscheearch at Cmabrigde Uinervtisy, it dseno't mtaetr in what oerdr the ltteres in a word are, the olny iproamtnt tihng is that the frsit and last ltteer be in the rghit pclae. The rset can be a taotl mses and you can still raed it whotuit a pboerlm. This is bcuseae the huamn mnid deos not raed ervey lteter by istlef, but the word as a wlohe.
[07:04:01] <DJ9DJ> indeed
[07:04:34] <DJ9DJ> but the research is already some years old ;)
[07:05:46] <jthornton> so am I
[07:05:52] <DJ9DJ> hrhr
[07:06:04] <DJ9DJ> mid-twenty? ;)
[07:06:42] <jthornton> you forgot the multiplier
[07:06:55] <DJ9DJ> :)
[07:26:15] <r00t4rd3d> http://www.leevalley.com/us/wood/page.aspx?p=64300&cat=1,41182,48945
[07:28:20] <archivist> huge... look in the toolbox http://www.collection.archivist.info/archive/DJCPD/PD/2010/2010_07_27_tinware/IMG_0798.JPG
[07:29:37] <Tom_itx> they have numbers that big?
[08:05:54] <pcw_home> gabewillen: You can use read() in a a user space HAL component
[08:21:04] <R2E4_awy> pcw, one step further. I got the vfd installed and running on my R2E4.
[08:21:44] <R2E4_awy> te,mporarily. I am stuck on one of the transformers getting it converted to 220 from 460. There is no markings on it.
[08:22:24] <pcw_home> Not in schematics?
[08:23:18] <R2E4_awy> http://www.youtube.com/watch?v=6q8811s56jw&feature=youtu.be
[08:23:41] <R2E4_awy> IT doesn't show the 220 volt connections in the schematics. IT just shows the 460 volt conenctions.
[08:25:29] <pcw_home> Is it big? and what does it power?
[08:28:14] <R2E4_awy> http://www.irmtl.com/LinuxCNC/tranny/
[08:28:45] <R2E4_awy> The transfo on the left. It is 90VAC to power the Power Supply for the drives.
[08:31:17] <r00t4rd3d> 5 head
[08:31:19] <r00t4rd3d> http://www.insideindianabusiness.com/newsitem.asp?ID=58474
[08:31:30] <pcw_home> only 2 wires on primary?
[08:31:39] <R2E4_awy> yes
[08:32:45] <pcw_home> Yuck
[08:33:35] <R2E4_awy> Yuck on?
[08:34:40] <pcw_home> I guess the options are find 220 --> 90V xfrmr or use a small 220-220 xfrmr as a booster
[08:35:13] <pcw_home> (or 220 --> 440 xfrmr)
[08:35:38] <R2E4_awy> why 220 -->440 xfo?
[08:36:07] <pcw_home> since thats a standard
[08:36:15] <tjtr33> is the small trafo powered from the big trafo?can you trace small primary to big 2ndry, maybe thru a contactor?
[08:36:53] <R2E4_awy> The small tranny is powered by 460 mains. I am trying to convert it to 220 mains
[08:37:18] <R2E4_awy> the big transfo is no problem as I have documentation on it.
[08:37:55] <tjtr33> then as pcw sez, use a buck-boost to get 2:1 or 1:2 as you need
[08:38:51] <R2E4_awy> I just found a chart in one of the BP manuals.
[08:39:32] <pcw_home> yeah so 220-220 isolation xfrmr in boost mode is probably the easiest
[08:39:34] <pcw_home> I notice thats a split bobbin xfrmr with electrostatic shield so it may be best to retain it
[08:40:26] <R2E4_awy> I found it. jumpers on H1-H4, H3-H6 then line H1 and H6 wooohooo!!
[08:40:30] <gabewillen> Is it possible to read serial data in a real time hal component?
[08:40:38] <pcw_home> Great
[08:41:11] <pcw_home> Sure but you have to do all the low level I/O
[08:41:20] <R2E4_awy> I'll disconnect the secondaries just to be safe....
[08:41:33] <gabewillen> oh so i actually have to flip the bits to read the data i cant use read?
[08:42:03] <pcw_home> no because it not a real time function
[08:42:57] <pcw_home> you can use read in a non-real time HAL comp
[08:43:21] <tjtr33> uses fp?
[08:43:53] <gabewillen> im trying to stream high speed serial data
[08:44:03] <gabewillen> andypugh was working on something similar
[08:44:18] <gabewillen> i guess i could pipe /dev/tty through halstreamer possibly
[08:44:33] <pcw_home> Yes possibly
[08:44:50] <pcw_home> or use userland comp
[08:44:58] <gabewillen> what about speeds though?
[08:45:12] <pcw_home> just depends on how critical the timing is
[08:45:16] <gabewillen> this is an encoder position
[08:45:36] <pcw_home> i dont think speeds are the issue, latency is
[08:46:19] <gabewillen> im sending a floating point position from an encoder at 921600 Baud rate
[08:46:33] <gabewillen> i have a c program that already reads it
[08:46:45] <gabewillen> trying to figure out how to get the data to hal
[08:47:05] <pcw_home> I would write a real time comp that reads the hardware
[08:47:29] <gabewillen> reads straight from the serial port?
[08:47:40] <gabewillen> like the component serport
[08:48:37] <pcw_home> Yes, send request byte each thread and wait for data
[08:50:22] <pcw_home> probably best to use fixed point and sixels or some other framing scheme to keep data length to a minimum
[08:51:15] <pcw_home> and allow on the fly resync on errors
[08:51:35] <gabewillen> wow, this will take a bit of reading up on
[08:52:02] <gabewillen> i was so close to calling you yesterday and ordering a 5i20, had the number typed into my phone and everything. this is my last shot
[08:52:25] <pcw_home> there are some comps that do direct hardware I/O
[08:52:44] <pcw_home> as examples
[08:52:45] <ReadError> pcw_home is the 5i20 creator?
[08:53:00] <tjtr33> argh! serport.comp is in .../hal/drivers and not in .../hal/comp
[08:53:03] <gabewillen> i thought he worked at mesa could be mistaken
[08:53:11] <pcw_home> who wants to know?
[08:53:15] <gabewillen> or owned it lol
[08:53:20] <R2E4_awy> hehe
[08:53:53] <R2E4_awy> pcw is a 5i20, with an irc interface.
[08:54:21] <R2E4_awy> hehe, sorry couldnt resist
[08:55:10] <pcw_home> :-)
[08:55:12] <pcw_home> unfortunately the serport comp sounds promising but just wiggles handshake bits
[08:55:13] <r00t4rd3d> there is gonna be a riot
[08:55:23] <r00t4rd3d> KFC is going totally boneless.
[08:55:32] <ReadError> lies
[08:55:36] <ReadError> its not april first
[08:55:37] <r00t4rd3d> i swear
[08:55:55] <ReadError> why would they do that
[08:55:58] <ReadError> its more expensive
[08:56:05] <r00t4rd3d> http://www.usatoday.com/story/money/business/2013/04/05/kfc-kentucky-fried-chicken-boneless-fast-food-chicken/2011419/
[08:56:35] <pcw_home> Chickenoid
[08:56:46] <r00t4rd3d> la will probably be on fire soon
[09:00:02] <tjtr33> yes serport.comp aint for communications, just digital io, clever name got me
[09:00:33] <gabewillen> i know
[09:00:40] <gabewillen> im going to try streamer
[09:00:54] <gabewillen> just have to figure out how to get the data sent to it correctly
[09:02:12] <pcw_home> if its a continuous stream, you need some kind of framing
[09:04:07] <gabewillen> framing?
[09:04:12] <gabewillen> sorry for my ignorance
[09:04:48] <pcw_home> if you just have a bunch of chars in a FIFO what starts the message?
[09:07:47] <gabewillen> i don't follow... the message streams continuosly
[09:07:53] <pcw_home> a simple low level way would be say 3 byte data message with 7 bits of data per byte, MSB set on first byte for framing
[09:07:55] <pcw_home> (so 3 chars gives you 21 bits of encoder data and about 33 usec per message at 921k)
[09:08:39] <gabewillen> write now im sending the data from the arm processor to the pc using printf("%d", pulses);
[09:09:32] <gabewillen> i figured it would be easier to send an integer value, than a float
[09:10:10] <pcw_home> so its being sent in ASCII
[09:11:39] <pcw_home> that will slow it down but maybe thats OK
[09:11:41] <pcw_home> but even sent in ASCII, it needs a newline or special char for framing
[09:13:26] <gabewillen> oh i see
[09:13:51] <gabewillen> i was using screen to debug the data at first. "%
[09:13:57] <gabewillen> "%d\r\n"
[09:14:21] <tjtr33> you could start reading data in middle of the number unless you use the framing
[09:14:32] <gabewillen> i gotcha
[09:14:34] <pcw_home> yes
[09:14:44] <gabewillen> similar to sending gcode to a cnc via serial
[09:14:50] <gabewillen> the start and end character of each line
[09:14:56] <pcw_home> Yes
[09:14:57] <gabewillen> makes sense now
[09:15:09] <gabewillen> sorry its been a long couple of days, working on this
[09:15:51] <gabewillen> hmmm what would you suggest be the start character?
[09:16:13] <gabewillen> white space?
[09:16:30] <pcw_home> if its all real time, you can use a master slave system where linuxCNC sends a request byte and the serial encode device answers
[09:17:03] <gabewillen> thats what i was looking for, i was going to try to do it over ethernet, but serial should be plenty fast enought
[09:17:05] <gabewillen> enough*
[09:17:10] <pcw_home> probably just LF if its ascii on unix
[09:17:38] <gabewillen> 90 rpms, with a 7200cpr encoder
[09:17:45] <gabewillen> is what im shooting for
[09:19:39] <tjtr33> thats 11mS between each read, whats your thread speed?
[09:20:00] <gabewillen> 23000ns
[09:20:07] <tjtr33> woops i ignored the cpr
[09:20:26] <pcw_home> if you can do 921K I think its possible to read one or maybe a couple encoders in real time
[09:20:28] <pcw_home> if the protocol is reduced to simple binary 1 byte request --> 5 or 6 char response
[09:21:16] <gabewillen> how soon could i get the 5i20 combo? its kinda over kill seeing this is an indexing hob sharpener with one axis, and one encoder but it could be extremely simpler
[09:21:24] <pcw_home> if you use halstreamer, you will get glitches occasionally
[09:21:39] <pcw_home> I would use a 5I25
[09:22:03] <gabewillen> thats what i meant
[09:22:08] <pcw_home> (cheaper)
[09:22:46] <gabewillen> yes the cheapest kit... worth it? or should i continue down this path? i already have the arm processor
[09:23:08] <skunkworks> heh - when do you want it to work? ;)
[09:24:09] <pcw_home> I guess it depends on whether you are being paid by the hour
[09:24:25] <gabewillen> on salary
[09:24:29] <gabewillen> lol
[09:25:41] <gabewillen> i already had this working but we need more accuracy hence the 7200 cpr encoder, i had a 2500cpr on there before, but seeing that 28,800 divides evenly into 360 degree's i should be able to eliminate accumulative error
[09:27:23] <tjtr33> pcw_home, "if you can do 921K"... is that about thread speed? serial rate?
[09:27:42] <pcw_home> serial baud rate
[09:29:06] <pcw_home> 921K is like 11 usec/char
[09:30:33] <gabewillen> i figured this would work... but might take some doing
[09:31:07] <gabewillen> could i pipe my c program to streamer?
[09:31:51] <IchGuckLive> hi all B)
[09:32:02] <gabewillen> i tried just editing streamer_usr.c adding the "termios.h" and port settings into the source file
[09:32:31] <gabewillen> its using strtoul to convert to hal_u32
[09:33:00] <gabewillen> the count will never exceed 28,800 as it resets to zero every revolution
[09:33:36] <IchGuckLive> gabewillen: retrofitting a mill
[09:34:00] <gabewillen> nope, just tweaking a hob sharpener i converted a while back
[09:34:23] <pcw_home> probably possible but bear in mind that its not real time so will have occasional timing glitches and since its streamed
[09:34:25] <pcw_home> (not request/reply) the host may not be able to support that fast a data rate continuously
[09:34:32] <IchGuckLive> and the problem is following error on encoder
[09:35:16] <gabewillen> i could drive the entire machine from arm, and communicate over the parallel port
[09:35:48] <IchGuckLive> why not full RT parport
[09:36:20] <tjtr33> 11uSec/char and connected to a hal comp in a thread? and 23uS thread period? i dont see it possible.
[09:36:51] <gabewillen> okay i like that answer
[09:37:08] <gabewillen> it doesn't have to do 90 rpms
[09:37:28] <gabewillen> anything over 50 should suffice
[09:37:54] <gabewillen> its only indexing the gash amount divided into 360
[09:38:10] <tjtr33> pcw_home, is smartserial useful for this problem ( high speed serial data aquisition ) ?
[09:38:10] <gabewillen> so for a 15 gash hob, 24 degrees, and it has about 2 seconds to do that
[09:39:14] <gabewillen> using just lcnc right now i can get about 30rpms
[09:39:47] <gabewillen> repeats with in .0001 after all 15 gashes, thats just through the parallel port
[09:40:46] <tjtr33> using parport, hal, and this serial encoder info?
[09:40:51] <gabewillen> no
[09:41:14] <gabewillen> using hal with the encoder a and b attached to parallel port in pins
[09:41:52] <gabewillen> pid, with max following error of .0125 equal to on line on the encoder
[09:44:58] <tjtr33> ok and it seems that the change to get higher precision and speed shouldnt be a quantum leap
[09:45:14] <tjtr33> the changes being, using the serial data of the encoder
[09:46:24] <tjtr33> and getting > 30 , <= 90 rps
[09:46:39] <tjtr33> > 50
[09:46:58] <chopper79> Hello, I need some pointers on where to start at on making changes to kin modules. I need to modify gantrykins a bit.
[09:49:54] <tjtr33> oh it IS rpm i didnt get rpms sussed till i re-read the gash stuff ( not a gear maker, a mold maker :)
[09:56:51] <tjtr33> if the parport strategy is maxed out, then IchGuckLive's idea about a non motherboard parport sounds like a way to increase throughput (mesa SuperPort or just gpio pins )
[09:57:35] <tjtr33> theres no gain in angular precision yet, but see if those solutions give the desired result for the rpm issue, then work on the serial issue
[09:57:43] <tjtr33> i also have a serial encoder to deal with
[09:57:59] <IchGuckLive> 2 parports or even 3 is good to go
[09:59:44] <r00t4rd3d> drill press lathe chuck that will hold 1.5 - 1.0 - .75 square stock: http://i.imgur.com/qsvkQi3.jpg
[10:00:54] <jdh> drill press lathe chuck?
[10:01:13] <tjtr33> how much faster is a Superport than a motherboard/pci parport? 5:3 ? thats what he needs at minumum
[10:02:27] <pcw_home> on the 5I25 max encoder count rate is 8MHz so maybe200X
[10:03:27] <tjtr33> :)
[10:04:28] <r00t4rd3d> jdh, redneck.
[10:04:44] <pcw_home> same with the stepgen
[10:04:49] <jdh> yes, I am, but I still don't understand.
[10:05:06] <r00t4rd3d> i got the idea from this
[10:05:06] <r00t4rd3d> http://farm3.staticflickr.com/2752/4149578376_141a0ffb41.jpg
[10:05:27] <jdh> ahh... gotcha
[10:05:53] <jdh> really bad 4 jaw chuck
[10:06:27] <jdh> wear safety glasses.
[10:06:35] <r00t4rd3d> lol
[10:06:45] <r00t4rd3d> ive used my drill press as a lathe before
[10:07:01] <r00t4rd3d> center a screw and cut the head off
[10:07:49] <tjtr33> pcw_home, thx, i think gabe can use the AB signals (he did already) and not use the serial ( no idea what hdwr has both), and at least go faster with same precision
[10:07:49] <cpresser> r00t4rd3d: i see you are using vcarve. did you figure how to create hires previews?
[10:08:38] <r00t4rd3d> cpresser, thats aspire but I think its in the menu options
[10:14:45] <r00t4rd3d> why do you know how to set it higher or you want it as high as mine?
[10:15:01] <cpresser> i want it higher as yours :)
[10:16:02] <r00t4rd3d> i think its just Toolpaths, Preview Simulation Quality, Highest
[10:16:36] <cpresser> mine is set to best (other are default & high)
[10:18:35] <cpresser> i was just hopeing you might have a clue how to turn it even higher
[10:19:45] <r00t4rd3d> no but if you use solid colors it seems to look better then wood grain
[10:21:57] <IchGuckLive> jdh: there are so cheep chucks and you might found a solution to get them on the drillpress
[10:22:10] <IchGuckLive> no need for hazardass mounting
[10:22:30] <jdh> Ich: not mine, I value my limbs more that that.
[10:22:41] <r00t4rd3d> psy
[10:22:48] <IchGuckLive> O.O
[10:22:55] <jdh> I have an R8 4" chuck. Was under $100
[10:23:01] <jdh> and two lathes
[10:23:06] <IchGuckLive> gree
[10:23:10] <IchGuckLive> a
[10:23:12] <jdh> hopefully 3 in a few hours
[10:23:15] <cpresser> r00t4rd3d: http://imagebin.org/252898
[10:23:25] <r00t4rd3d> jdh did they email back?
[10:23:26] <cpresser> thats how mine look if you zoom in a little
[10:23:49] <jdh> r00t: yeah, he just asked for phone number and said he would call.
[10:23:58] <r00t4rd3d> vcarve pro 6.0?
[10:24:08] <cpresser> 7.0
[10:24:14] <cpresser> its out just a few days
[10:24:22] <cpresser> i used 5.0 prior to that
[10:25:21] <r00t4rd3d> wow there is an update :/
[10:29:25] <IchGuckLive> r00t4rd3d: did you try sheetcam on this part stuff it is only 2 klicks for tha hole sheet to get g-code on plasma or mil
[10:41:51] <IchGuckLive> by
[10:42:11] <gabewillen> what is full rt parport
[10:46:08] <r00t4rd3d> my guess is real time
[10:46:23] <r00t4rd3d> parport is parallel port
[10:57:40] <jdh> 10% off everything and free ground shipping sitewide at Enco
[11:00:23] <r00t4rd3d> Im gonna make a pipe
[11:00:57] <jdh> out of wood?
[11:01:10] <r00t4rd3d> yra
[11:01:47] <jdh> how about a brass pipe nipple, 90-elbow and a 1/8-1/4 reducer
[11:02:23] <r00t4rd3d> probably shooting more towards the corn cob style
[11:02:57] <jdh> really.
[11:02:58] <r00t4rd3d> i need to cut my chuch first though
[11:03:04] <r00t4rd3d> chuck
[11:04:18] <jdh> 3/8" end mills are $3, 1/4" ones are $7 that's annoying
[11:04:49] <r00t4rd3d> supply and demand
[11:05:00] <r00t4rd3d> no one wants the 3/8
[11:05:56] <jdh> I have lots of 3/8" ones. They generate a lot more chips than 1/4"
[11:06:46] <R2E4_awy> Where are you getting endmills for 3.00 and 7.00?
[11:07:35] <jdh> HSS from enco http://www.use-enco.com/CGI/INPAGE?PMPAGE=/specials/320-0060
[11:08:13] <r00t4rd3d> http://www.ebay.com/itm/Carbide-Router-Bit-3-20mm-125-1-8-Shanks-NEW-Solid-Carbide-CNC-PCB-/160816395593?pt=LH_DefaultDomain_0&hash=item2571677549
[11:08:18] <r00t4rd3d> those are the bits to hae
[11:08:21] <r00t4rd3d> have*
[11:08:49] <jdh> snap too easy
[11:08:57] <r00t4rd3d> ive never broke one
[11:09:31] <jdh> cut faster/deeper/harder
[11:10:16] <R2E4_awy> he says no cause they will break....hehe
[11:10:57] <r00t4rd3d> lead nut would probably snap first
[11:14:10] <Loetmichel> r00t4rd3d: those are only good for fiber plastics
[11:14:23] <Loetmichel> (FR4, carbon and so on)
[11:14:56] <Loetmichel> For wood and aluminium there is to less swarf room in them
[11:21:07] <ReadError> r00t4rd3d,
[11:21:09] <ReadError> i use those
[11:21:17] <ReadError> cept i pay like 18$ for 10
[11:21:27] <R2E4_awy> where?
[11:22:19] <ReadError> ebay
[11:22:23] <ReadError> drillman1
[11:22:30] <ReadError> hes got a bunch of good stuff
[11:22:38] <ReadError> those diamond ones hold up very well
[11:22:47] <ReadError> except these 1/16"
[11:29:30] <jdh> I want them for 6061
[11:42:49] <r00t4rd3d> all i cut is hardwood with them
[11:43:05] <r00t4rd3d> Loetmichel
[11:44:19] <r00t4rd3d> i have 2 and probably cut 8 miles of oak with them
[11:49:59] <r00t4rd3d> weed and boredom = http://i.imgur.com/ggDu9Pv.jpg
[12:14:16] <c60> I buy lots of stuff from http://888sharpco.com/store/ it's a little more expensive, but it's all dimar.
[12:51:49] <Loetmichel> r00t4rd3d oak is a very hard ad short fibered wood
[12:52:41] <Loetmichel> try pine plywood and it will burn its way trhough in no time
[12:52:47] <Loetmichel> because it will clog
[12:54:15] <Loetmichel> for wood and soft metals i use these a lot of the time: http://www.sorotec.de/shop/product_info.php/info/p475_vhm-fraeser-zweischneider--flach---3-0mm.html
[13:37:27] <Tugge> Good evening. I have a problem with my milling machine. I've been trying to engrave text to aluminum, but the result is not good. Here is the picture of testing different RPMs and feed speed: https://skydrive.live.com/?cid=4d02066e8d01f203&id=4D02066E8D01F203%211177&sff=1&authkey=!AFEDXaosw2rYxzA
[13:38:04] <Tugge> I used some sinol to lubricate aluminum but it didn't help.
[13:38:59] <cradek> what's your max rpm? is that anodized?
[13:40:43] <cradek> guessing that's a 1/4 tool and you're using just the very tip - you probably want 10,000 rpm
[13:41:10] <Tugge> RPM was between 10k and 32k and aluminum was anodized.
[13:41:15] <cradek> if you can't do that, consider a drag engraver
[13:41:42] <cradek> what kind of aluminum is it?
[13:41:50] <cradek> looks gummy
[13:43:37] <Loetmichel> if its anodized an a hss engraving bit: forget it, us a tungsten carbide one
[13:43:42] <Loetmichel> use
[13:43:43] <Tugge> The material is from this enclosure: http://www.hammondmfg.com/1455.htm
[13:44:11] <cradek> I agree with Loetmichel
[13:45:03] <Loetmichel> the anodizing is tha same stuff as the grit on "korund" grinding paper
[13:45:03] <cradek> that cast aluminum is probably very gummy material for milling
[13:45:05] <Tugge> So you think that the problem is with the milling bit?
[13:45:14] <Loetmichel> ... a hss toll is dull in NO time there
[13:45:20] <Loetmichel> yes
[13:45:37] <Loetmichel> use a tungsten carbide engraving bit
[13:45:40] <cradek> you could try carbide with it under oil or water
[13:45:42] <Tugge> Ok. So I need to order better bits.
[13:45:54] <cradek> but the material is still probably a poor match for milling
[13:45:58] <Loetmichel> water with a drop of detergent
[13:46:04] <Loetmichel> or even better: lamp oil
[13:46:22] <cradek> consider adding a plate of appropriate material for your panel instead of engraving the boxes directly
[13:46:40] <Tugge> I believe that lamp oil equals sinol??
[13:46:51] <Loetmichel> cradek: with the right toool the anodized aluminium is engraveable like a charm
[13:47:48] <Loetmichel> i dont know what sinol is
[13:48:14] <Loetmichel> cradek: http://www.cyrom.org/palbum/main.php?g2_itemId=9211
[13:49:01] <Loetmichel> http://www.cyrom.org/hajo/namelampe.avi
[13:49:25] <Loetmichel> Tugge: the video: tahts a carbide bit
[13:49:47] <Loetmichel> ans tha spray is some WD40-copy
[13:51:00] <DJ9DJ> i see no one with my name :(
[13:51:11] <Tugge> Loetmichel: What is your RPM?
[13:51:15] <Loetmichel> 5000
[13:51:21] <Tugge> and feedrate?
[13:51:32] <Loetmichel> DJ9DJ: thats because you dont work for my company
[13:51:55] <Loetmichel> that was a christmas gift from me to my underlings ;-)
[13:51:56] <DJ9DJ> ah, i see
[13:52:14] <Tugge> Could one of the problems to be that my RPM is too high?
[13:52:37] <Loetmichel> it could but is not probable
[13:52:52] <Loetmichel> i had only the 5krpm spindle at that time
[13:53:06] <Loetmichel> it would've been better to use 20k
[13:53:16] <DJ9DJ> but... Loetmichel, you could have made one for me instead one for yourself ;)
[13:53:24] <Loetmichel> the only problem is your tool
[13:53:40] <Loetmichel> this soft aluminium needs a VERY sharp tool
[13:53:57] <Tugge> Okay. I will buy those tungsten carbide v-bits from ebay and try again.
[13:54:00] <DJ9DJ> bubblegum aluminium? ;)
[13:54:11] <Loetmichel> and a hss-tool is dull in anodized aluminium almost at the second you cut the first letter
[13:55:53] <Tugge> Loetmichel: Can you suggest any good brand or models?
[13:56:57] <Loetmichel> no, because i am german
[13:57:26] <Loetmichel> and the shop i buy from would be a bit expensive for you i suppose
[13:58:03] <Loetmichel> i use these a lot: http://www.sorotec.de/shop/product_info.php/info/p179_vhm-fraesstichel-30-.html
[13:58:32] <Tugge> Heh. I was just on that same page checking bits :)
[13:59:12] <DJ9DJ> moi! :)
[13:59:30] <Loetmichel> You have to have a diamond file, thoug
[13:59:36] <Loetmichel> though
[13:59:48] <Loetmichel> the bits come wit a 0,00mm tip
[14:00:03] <Loetmichel> which will break at first contact with the workpiece
[14:00:48] <Loetmichel> you have to file a "flat" on the tip so wide that it cuts a "bottom" in the vgroove to your desired width
[14:00:52] <Tugge> Well that sounds quite expensive.
[14:00:54] <Loetmichel> les see
[14:01:21] <Loetmichel> no, mine came with the leaterman wave, but singe diamond files come at about 5 eur
[14:01:25] <Loetmichel> single
[14:01:48] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_itemId=8961
[14:02:04] <Loetmichel> there i am cutting the "flat head" on a selfground engraving bit
[14:02:27] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_itemId=8958 <- selfground bits made from broken milling bits (TC)
[14:03:27] <Loetmichel> http://www.youtube.com/watch?v=4iaFb6WInuw <- grinding engraving bits...
[14:03:29] <Tecan> (4iaFb6WInuw) "Stichelschleifen" by "Loetmichel" is "People" - Length: 0:04:36
[14:04:58] <Tugge> So that is how you can sharpen those bits.
[14:05:39] <Loetmichel> i dont recommend trying this without a reference
[14:05:51] <Loetmichel> i.e a sharp bit that is working
[14:06:18] <Loetmichel> i needed to grind about 20 or 30 bits to get the first one working
[14:06:20] <Loetmichel> ;-)
[14:06:20] <Tugge> That true. But ofcourse, if you know that your bit is dull, you don't lose anything :)
[14:06:35] <Tugge> What was that blue stuff?
[14:06:46] <Loetmichel> those bits werent dull, they were broken
[14:06:52] <Loetmichel> window cleaner
[14:06:57] <tjtr33> "<gabewillen> what is full rt parport" i think IchGuchLive meant a parport equivalent, like the mesa 5i25/6i25,
[14:06:57] <tjtr33> look at mesa's site. pcw_suggested it could perform much faster than a motherboard parport.
[14:07:16] <Loetmichel> to cool the bit a bit, my fingers tend to smope if the bit get hotter than a few 100 deg. C
[14:07:18] <Loetmichel> ;-)
[14:07:31] <Tugge> Okay. I beleive I have that somewhere (need to ask from my wife...)
[14:07:57] <Loetmichel> it was just standing around, i could have used normal water
[14:07:59] <Loetmichel> ;-)
[14:08:37] <Tugge> So it's just for cooling the bit down?
[14:13:39] <Loetmichel> yes, the diamond disc heats the bit um beond my capability to hold it ;-)
[14:13:44] <Loetmichel> beyond
[14:13:51] <Loetmichel> up
[14:37:29] <Loetmichel> Tugge: but i warn you: it is frustrating to do engraving bits "free hand"
[14:38:01] <Loetmichel> i needed 20-30 tries until the first one had cut right.
[14:38:25] <Loetmichel> and i have plenty of experience with free hand grinding of driss and so on ;-)
[14:41:42] <Tugge> I believe you. But without doing you wont learn, so there is no easy way to be a master :)
[14:42:09] <Tugge> What can you say about these kind of bits? http://www.ebay.com/itm/SET-10-MICRO-ENGRAVING-TOOLS-30-through-120-degrees-Kyocera-Tycom-/380581380499?pt=LH_DefaultDomain_0&hash=item589c6b0193
[14:42:43] <Tugge> These looks a little bit more professional than those Chinese bits.
[14:44:04] <Loetmichel> looks ok to me
[14:46:03] <Loetmichel> s/driss/drills
[14:46:45] <Tugge> Do you prefer more bits with fluids or those V-shape bits? What is the difference between those two?
[14:46:56] <Loetmichel> fluids?
[14:47:08] <Loetmichel> you meant Flutes?
[14:47:26] <Tugge> Those cutting edges of milling bit.
[14:47:42] <Tugge> Or flutes..
[14:47:46] <Tugge> Sorry..
[14:48:37] <Loetmichel> the halved v-shape are good for hard materials like PCB engraving . less good for plastics and easily smearing materials
[14:49:22] <Loetmichel> there are the 1 flute or 2 flute types wit a spiral flute better because the cutting edge angles are better for soft materials there
[14:50:53] <Tugge> So for PCB i should use halved v-shape bit and for aluminum spiral flute type bit.
[14:50:55] <Loetmichel> and tha aluminium used for "extruding" and anodizing are the type we call "Kaugummi-alu" in german ("bubblegum aluminium") because its tends to melt and smear on teh mill bit
[14:51:02] <Loetmichel> right
[14:53:12] <Loetmichel> and WD40 or cheaper simple lamp oil ("petroleum") is the best cooling fluid for soft aluminium (IMO)
[15:51:17] <chopper79> Hello, I need some help with Gantrykins. Anyone familiar with it?
[16:16:53] <DJ9DJ> gn8
[17:05:41] <gabewillen> alright finally got usbstreamer component working
[17:06:24] <gabewillen> though i have to admit, i used alot of the streamer.c and streamer_usr.c code :)
[17:19:30] <mhaberler> what does it do?
[17:39:22] <gabewillen> alows you to stream data from a arm or microcontroller via usb
[17:39:46] <gabewillen> at 1.5 mb per second though your thread can't read that fast if your using a standard setup
[17:41:15] <gabewillen> i'll find out how well it works in a minute. Going to hook it up to a high resolution encoder and see what happens
[17:41:41] <gabewillen> im thinking about adding an echo back feature to ensure there is still a connection
[17:44:50] <gabewillen> Installing rt-preempt kernel on my 64bit unbuntu 12.04 was the best thing i have done. Latency isn't great but makes for easy testing. it was a hassle to get the bumbleebee nvidia graphics power feature working but. It was well worth it
[18:38:51] <mhaberler> great to hear it works for you
[18:40:10] <mhaberler> I hope the USB delays wont inject too much noise into actual position, that could be an issue
[18:59:04] <tjtr33> gabewillen, this maybe of use, a user said he'd made a fast serial connection with a uart on a mesa board http://tinyurl.com/bofu5w2
[19:01:18] <AR__> so i'm using a 60 deg V-cutter to mill some PCBs
[19:01:23] <AR__> and i'm getting burrs on the traces
[19:01:29] <AR__> no matter what i do
[19:02:08] <tjtr33> gabewillen, andypugh had done it, i'm still reading the thread
[19:04:01] <pfred1> that is because milling PCBs is dumb
[19:04:24] <AR__> yeah i learned that
[19:04:35] <AR__> but it is a good exercise
[19:06:12] <pfred1> try putting down resist with your CNC machine then etching the board
[19:06:32] <pfred1> if you know what you're doing an etch should only take 2 minutes
[19:07:01] <pfred1> toner transfer is kind of dumb too BTW
[19:07:39] <pfred1> trick to etching is have your tank fuming and pump air into it
[19:11:44] <Tom_itx> i've had good luck with it
[19:12:21] <AR__> eh, if i want something done i'd just have it sent out
[19:12:27] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/etching/etch_index.php
[19:12:32] <Tom_itx> my etch tank
[19:12:58] <Tom_itx> pfred1, why do you say it's dumb?
[19:13:41] <AR__> i broke a .023 endmill also
[19:13:45] <AR__> fff
[19:14:19] <pfred1> Tom_itx if it was decent industry would be all over it but they're not
[19:14:39] <Tom_itx> no but it's a simple thing for a hobbyist to do
[19:14:51] <AR__> well it is slower and not feasible for mass production
[19:14:56] <Tom_itx> i agree the industry uses photo etching
[19:15:03] <Tom_itx> i've been thru a board house
[19:15:04] <pfred1> copper doesn't cut very clean
[19:15:07] <Tom_itx> nope
[19:15:11] <Tom_itx> copper is crap to mill
[19:15:22] <Tom_itx> leaves stringy burrs
[19:15:27] <AR__> and it doesnt help that it is on top of fucking fiberglass
[19:15:56] <pfred1> etching kicks ass though
[19:16:26] <Tom_itx> yeah, photo resist costs more than toner transfer though
[19:16:37] <Tom_itx> not counting the printer of course :D
[19:16:45] <pfred1> well not if you're ruining toner cartridges on a regular basis
[19:17:01] <Tom_itx> if i have very many to do, i just use a board house
[19:17:02] <pfred1> I wasted two with PNP Blue
[19:17:10] <Tom_itx> if i have one off i need to test i'll make it myself
[19:17:21] <Tom_itx> the pnp blue sucks
[19:17:27] <pfred1> it sure does
[19:17:33] <Tom_itx> i have some too
[19:17:42] <Tom_itx> but i've managed to get good results with it also
[19:17:48] <pfred1> next time I see mine I have to remember to throw it out
[19:17:52] <Tom_itx> heh
[19:18:03] <pfred1> toner carts for my printer are $100 a pop
[19:18:11] <Tom_itx> i've had good luck with certain photo papers
[19:18:39] <Tom_itx> i was refilling mine for $5 a pop for a while
[19:18:57] <pfred1> well running PNP Blue through mine screwed up the bulb in it or something
[19:18:58] <Tom_itx> i'd collect discarded cartridges from offices
[19:19:05] <pfred1> after that everything printed was all speckly
[19:19:11] <Tom_itx> but then my series ii died
[19:19:32] <pfred1> I tried cleaning it with alcohol and stuff no dice
[19:22:43] <pfred1> this is my etch tank http://www.instructables.com/files/orig/F53/FE12/GSUSNNHA/F53FE12GSUSNNHA.jpg
[19:23:05] <AR__> potatoes?
[19:23:52] <AR__> why are there potatoes in your etch tank
[19:24:26] <pfred1> there is only a test tube in my etch tank
[19:25:19] <pfred1> I don't even need to monitor the temp when it fumes it is hot enough
[19:26:21] <pfred1> that and bubble air through the tank and etches come out fast and sharp
[19:27:15] <pfred1> the best thing for pumping air throug hthe tank I've found is a cheap 12 volt tire air compressor
[19:27:30] <PCW> I think those potatoes are rocks
[19:27:31] <pfred1> it makes a nice pulsating air stream
[19:28:07] <pfred1> I used to use my big air compressor but fell back on to a tire compressor after I'd moved my big compressor it works better
[19:29:08] <pfred1> PCW oh yes the rocks in my water bath that is so the plastic tray doesn't sit right down onto the glass pan
[19:29:25] <pfred1> might melt or something if it did
[19:29:58] <pfred1> river pebbles
[19:32:10] <PCW> potatoes would get funky after a while
[19:32:35] <pfred1> they could get tender and tasty too though
[19:33:13] <PCW> etchant might not help
[19:33:28] <PCW> bbl
[19:33:40] <pfred1> a little seasoning never killed anyone
[19:34:44] <Tom_itx> i use a fishtank bubbler
[19:35:00] <pfred1> I don't have one
[19:35:12] <pfred1> I have caps off old spray cans and rubber hose though
[19:35:42] <pfred1> and enough air compressors too
[19:36:12] <pfred1> it makes a big difference to pump air into the tank
[19:36:19] <pfred1> rocking the tank is for the birds
[19:36:43] <Tom_itx> yeah i agree
[19:37:06] <pfred1> I used to do that and my etches would take like a half of an hour and be all undercut too
[19:37:15] <Tom_itx> i preheated the etchant in hot water in the sink though
[19:37:20] <Tom_itx> don't use a heater
[19:37:35] <Tom_itx> mine come out pretty clean
[19:37:36] <pfred1> yeah I'd turn my heater off and it would stay warm through the etch
[19:37:53] <pfred1> like I said if everything is right a board should only take 2 minutes
[19:37:59] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/temp/tiny/tinyboard1.jpg
[19:38:05] <Tom_itx> there's an attiny10 board
[19:38:14] <Tom_itx> 3 traces under a sot23-6
[19:38:30] <pfred1> you did that with an iron on?
[19:38:40] <Tom_itx> yessir
[19:38:51] <pfred1> I see some raggies
[19:39:07] <Tom_itx> well it ain't perfect but it did work
[19:39:11] <pfred1> in a high frequency circuit they could matter too
[19:39:26] <Tom_itx> like i said, i generally only make them for one off tests nowdays
[19:39:40] <Tom_itx> if i were making a high frequency circuit i'd have them made for me
[19:39:55] <pfred1> be nice to have the capability
[19:40:26] <Tom_itx> i suppose. but my volume isn't worth the effort
[19:40:29] <pfred1> I do electronics to do it not have it done for me
[19:40:44] <pfred1> I do it because i enjoy it
[19:41:03] <Tom_itx> i enjoy it as well
[19:41:13] <pfred1> if I can't do it then I'm not interested
[19:41:29] <Tom_itx> but i'm not gonna hand etch 1k boards to pass on to the public
[19:41:37] <pfred1> me either
[19:41:43] <Tom_itx> i developed the first off myself
[19:41:49] <pfred1> people can go etch their own boards
[19:42:04] <Tom_itx> i got my own toaster oven as well
[19:42:15] <Tom_itx> works great
[19:42:24] <pfred1> me too I have a cuisinart
[19:42:33] <pfred1> I got it at a garage sale for $10
[19:42:40] <pfred1> brand new
[19:42:44] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/toaster_oven/toaster_oven_index.php
[19:43:19] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/toaster_oven/triac_top2.jpg
[19:43:31] <Tom_itx> i tried 'iron on' silkscreen there
[19:43:48] <pfred1> you are ambitious
[19:44:03] <pfred1> I never label projects enough
[19:44:04] <Tom_itx> just toner transfer
[19:44:08] <Tom_itx> with no copper
[19:44:37] <pfred1> I just did a neat mini project
[19:44:42] <pfred1> a real breadboard
[19:45:08] <Tom_itx> anyway.. i'm out for a bit
[19:45:30] <pfred1> I hammered brass escutcheon pins into a scrap piece of pine wood and soldered stuff to it
[20:07:19] <jdh> http://tinyurl.com/crsu8gu my new lathe!
[20:09:28] <pfred1> 4 jaw chuck huh?
[20:10:51] <ProxDem> nice lathe jdh
[20:11:52] <ProxDem> that's pretty kewl wish I had a small one like that
[20:20:18] <r00t4rd3d> man
[20:20:20] <r00t4rd3d> jdh, copy all the parts and make me one :D
[20:20:29] <jdh> I need some tiny steppers for it
[20:20:31] <r00t4rd3d> i got a 17
[20:20:57] <andypugh> http://www.zappautomation.co.uk/en/nema-8-stepper-motors/352-sy20sth42-0804b-high-torque-hybrid-stepper-motors.html
[20:20:58] <andypugh> (NEMA 8, 20mm)
[20:20:59] <andypugh> Very cute :-)
[20:21:06] <jdh> that's a tiny motor
[20:21:08] <jdh> need to scrounge some junk printers
[20:21:09] <r00t4rd3d> see this?
[20:21:09] <r00t4rd3d> http://www.precisiontools.com.sg/CLISBYMINILATHES.htm
[20:22:32] <jdh> this one says made in .au and has an au URL
[20:22:50] <r00t4rd3d> you can get the motor at that site i linked too
[20:23:03] <pfred1> I got piles of motors
[20:23:04] <r00t4rd3d> gear belt everything
[20:23:05] <jdh> spindle motor?
[20:24:47] <pfred1> a sewing machine motor would run that lathe
[20:25:41] <jdh> I"m going to try to turn some delrin
[20:26:02] <pfred1> what are you going to try to turn it into?
[20:27:01] <AR__> his ass
[20:31:19] <pfred1> that would be an expensive sex toy
[20:32:47] <pfred1> I knew a guy who was a total nut for making paraphanelia and he got a Unimat first thing he made on it was a pot pipe
[20:33:32] <pfred1> he was all like there's this design i always wanted to try and see if it'd work
[20:34:09] <AR__> rofl
[20:34:34] <pfred1> oh it was funnier because he was stoned on time release speed and he was at it for like 12 hours straight
[20:36:28] <pfred1> there is a lot of setup time involved with those all in one machines
[20:45:12] <R2E4_> pcw: you here?
[20:47:25] <pfred1> R2E4_ 20:23 < PCW> bbl
[20:52:59] <andypugh> I have had some success with a flatbed plotter and an OHP pen for resist.
[20:53:01] <andypugh> But I can generally wait 10 days for Hong Kong to send me much better PCBs than I could ever make
[20:54:18] <pfred1> i used to love that cartoon
[20:54:28] <pfred1> maybe just the theme song
[20:54:41] <pfred1> Honk Kong Phooie #1 super guy!
[20:55:01] <pfred1> it was a very catchy tune
[20:55:43] <pfred1> I don't think they could get away with making cartoons like that today
[20:55:48] <jdh> eyeballing a 4 jaw chuck doesn't lead to goodness
[20:55:58] <ProxDem> lol
[20:56:07] <andypugh> Put a tool very close to the work to judge the high side. You can get pretty close that way.
[20:56:11] <andypugh> exit
[20:56:11] <jdh> yeah, it's just a lot of back and forth
[20:56:16] <jdh> a cnc toy
[20:56:17] <jdh> it has a spindle motor that works fine
[20:56:27] <pfred1> jdh just keep telling yourself that theoretically 4 jaw chucks are more accurate than scroll chucks
[20:56:56] <pfred1> it seems to help somewhat while fiddling around with them
[20:58:15] <jdh> they have a little higher capacity too, which helps with this
[21:00:06] <jdh> I need an MT0 live center. or dead
[21:04:25] <Tom_itx> andypugh, how do you go about centering work in your rotary axis?
[21:04:37] <Tom_itx> you can't spin it like a regular lathe head to find the high spots
[21:05:51] <Tom_itx> or does yours have collets...
[21:06:06] <Tom_itx> i'm thinking of a 4 jaw rotary head
[21:06:59] <andypugh> Jog it.
[21:07:02] <Tom_itx> or even a workpiece you may purposely want off center.. what's the best way to find the reference point
[21:07:42] <andypugh> The the trick is actually not to "spin" a 4-jaw anyway.
[21:08:01] <Tom_itx> yeah i move it in 90 deg increments when i'm centering one
[21:08:08] <Tom_itx> or 180
[21:08:33] <Tom_itx> with a last word indicator if needed
[21:08:40] <andypugh> You find which adjuster is lowest, loosen it, find the highest, and then tighten it to half-way between the two readings, then back to the first one, and tighten it up, then do the same thing for the other two adjusters.
[21:08:42] <Tom_itx> generally i'm turning the piece anyway
[21:09:07] <andypugh> The trick is to look at the actual numbers on the gauge.
[21:09:13] <Tom_itx> yeah
[21:09:23] <Tom_itx> my lathe has a bit of slop in it though
[21:09:27] <pfred1> Tom_itx last word? but, but but ...
[21:09:47] <Tom_itx> never heard of that?
[21:10:02] <pfred1> I've used them
[21:10:31] <pfred1> I rebuilt a wheel dresser once and the needle didn't move when I indicated it so I thought I broke something
[21:54:22] <r00t4rd3d> http://i.imgur.com/6NYeL9b.jpg
[21:54:55] <r00t4rd3d> that does stuff to my brains
[21:58:46] <jdh> so does that shit you've been smoking.
[22:01:11] <pfred1> you're going to lose your mind eventually
[22:01:17] <pfred1> so you might as well have fun doing it
[22:24:33] <R2E4_> my R2E4 is up and running on the existing control.
[23:24:12] <noob123> can anyone suggest the best value (effiency/cost) CNC that works with linuxcnc ?? Could be on ebay or other site ??
[23:24:57] <noob123> 3 axis, mostly for PCB
[23:27:49] <pfred1> you need a very accurate machine to cut PCB and a high speec spindle helps too
[23:27:55] <pfred1> high speed even
[23:28:20] <pfred1> you're going to either have to go pneumatic or water cooled
[23:29:03] <pfred1> forget the dremel nonsense
[23:30:38] <pfred1> a turn key machine capable of cutting PCB I'm going to guess is going to start at $4,500 ?
[23:31:03] <pfred1> another option is to retrofit a milling machine like an X3
[23:31:13] <pfred1> but you're still going to have to upgrade the spindle somehow
[23:32:00] <noob123> by cut you mean doing the traces or really cuting the board ??
[23:32:23] <pfred1> noob123 if you see pipes for linear guides and partleboard as a cutting table be very wary
[23:32:55] <pfred1> those machines likely cannot hold tolerances to cut PCB
[23:33:57] <pfred1> I know what you want to do and in order to do it successfully it isn't that easy with the wrong machine
[23:34:07] <pfred1> it isn't that easy with the right machine
[23:34:58] <pfred1> copper cladding is what 2 mils thick?
[23:35:39] <noob123> about this
[23:35:43] <pfred1> you're just not going to be that accurate with pieces of pipe and particleboard
[23:36:19] <pfred1> like a fireball 90
[23:36:44] <noob123> checking..
[23:37:05] <pfred1> well a fireball can't cut PCBs
[23:37:10] <pfred1> although they claim that it can
[23:37:59] <noob123> they don look much precise with those pipes....
[23:39:07] <pfred1> noob123 if you really want to mill PCBs the minimum you might get away with is retrofitting one of these and upgrading the spindle on it too http://www.mini-lathe.com/X3_mill/X3rvw/X3.htm
[23:39:15] <Connor> The pipes are just the frame work.. I think they use hardened rods do they not?
[23:39:34] <pfred1> Connor it doesn't matter the table is particleboard
[23:40:00] <Connor> pfred1: No it's not. It's MDF.
[23:40:04] <Connor> Not the same.
[23:40:07] <pfred1> same difference
[23:40:21] <pfred1> it is hardly ground steel
[23:40:53] <pfred1> someone was in here about a year ago trying to do it on a fireball
[23:40:55] <noob123> in this video it seems precise http://www.youtube.com/watch?v=7W9rVt730q8
[23:41:01] <Connor> I just made a fixture for my PCB's.. and milled it out flagt.
[23:41:17] <Connor> flat.
[23:41:27] <pfred1> Connor so you're milling PCBs on a fireball?
[23:41:50] <Connor> No. On my DIY Router.. which, is made of MDF, Supported rails and ball screws..
[23:43:00] <pfred1> THK rails?
[23:43:01] <Connor> http://www.ivdc.com/cnc/cnc13.JPG
[23:43:06] <Connor> No.
[23:43:18] <pfred1> oh those ground rods?
[23:43:19] <Connor> Thats my setup.. with the PCB jig.
[23:43:27] <Connor> Yes. Ground, supported..
[23:43:31] <Connor> as is the fireball.
[23:43:48] <Connor> It uses ACME screws vs ball screws..
[23:44:13] <Connor> the Firball that is.. Mine using 1610's for the X & Y, and a 1605 for the Z
[23:44:27] <pfred1> well I know people do it but then again people bungee jump naked too and yo uwon't catch me doing that either
[23:45:16] <pfred1> to really make an X3 work you do have to retro ball screws into it too
[23:45:30] <Connor> The KEY is making sure the gantry and the Z axis is square..
[23:45:35] <pfred1> I mean you can run the leads it comes with but I don't know with the backlash
[23:46:45] <pfred1> is that a Bosch Pony?
[23:47:32] <Connor> Colt
[23:47:52] <pfred1> I wouldn't think the bearings in it would be accurate enough or the RPM high enough to cut copper cladding too cleanly
[23:48:09] <Connor> 30K RPM ?
[23:48:26] <pfred1> yeah that is pretty slow in the world of high speed spindles
[23:48:49] <pfred1> 80-100K is more in the ballpark
[23:49:49] <noob123> Connor, did you make it by the scratch, or you followed some already done project ?
[23:49:54] <Connor> .0004" Run out
[23:50:01] <Connor> scratch.
[23:51:25] <Connor> http://www.precisebits.com/products/equipment/bosch_colt_kits.asp?tsPT=!!!Bosch-Colt!!!Kit!!!PG!!!#Tabs
[23:51:30] <Connor> Collects for my router..
[23:52:01] <noob123> have you made it available ?? or do you guys recommend some projects that I could Study ??
[23:52:07] <pfred1> all moving gantry machines seem pretty shaky to me
[23:52:34] <Connor> noob123: what you see is what you get.
[23:53:05] <Connor> pfred1: I didn't design it for pcb's only.. but for a PCB specific only one.. yea.. moving table design would probably be better.
[23:53:15] <Connor> one with a much smaller cutting area.
[23:53:15] <pfred1> this looks like it might be able to do it http://www.aliexpress.com/item/mini-CNC-3040CH80-4-axis-Engraving-milling-and-drilling-machine-with-800W-water-cooling-spindle-motor/598487931.html
[23:53:52] <Connor> I am planning on converting the Z axis and table to aluminum at some point..
[23:53:56] <pfred1> catch as catch can with that imported stuff
[23:54:12] <pfred1> Connor get a sink cut out
[23:54:21] <pfred1> granite!
[23:55:06] <Connor> I wanted to put granite under the whole machine because the desk is a bit warped.. caused issues till I shimmed it some.
[23:56:35] <pfred1> noob123 turn key vs. build it yourself the economics can vary widely
[23:57:14] <Connor> I probably have over $1200 in the machine... not count CPU.. but, was spent over time.. and $290 bobo on the balls crews.. had to re-order them machined.
[23:57:18] <pfred1> I've heard of people shelling out serious cash and not getting what they expected
[23:58:01] <pfred1> then again if you build it yourself there is no guarantee it'll be everything you want it to be right off the bat either
[23:58:07] <Connor> I will say, it was a learning experiance. I had fun building it.. I would change a few things.. but, I like my design. 18" x 12" cutting area in a 36" x 25" foot print.