#linuxcnc-devel | Logs for 2013-11-12

[00:04:34] <ju-emb> what is "2 extruder folding motors" in your description?
[00:09:03] <pzpz> 2 stepper motor that folde the extruders up when it's not used.
[00:11:12] <pzpz> so what you say is i can get a dro in EMC2?
[00:11:55] <pzpz> 5 axis dro - soo cool
[00:13:18] <ju-emb> For all what you like to include in your machine, I don't think there is another option to linuxcnc ;-)
[00:14:42] <pzpz> my work is linux..
[00:15:43] <ju-emb> what ever that means, emc2 it's name is linuxcnc for quiet a while now
[00:16:24] <pzpz> i work with emc2 in 2009..
[00:17:03] <pzpz> on 3 axis stepper dereml machine. and lpt.
[00:17:17] <ju-emb> so you know how flexible it is.
[00:17:26] <pzpz> no
[00:18:00] <pzpz> for me it was like copy paste.
[00:19:17] <ju-emb> it's really flexible, your limit is your imagination and the money you're willing to spend for I/Os you need
[00:20:49] <pzpz> f*ck i have imagination but no money..
[00:21:28] <ju-emb> Your actual project, I recommend a Mesa card, it blows quiet a lot of headache away
[00:21:28] <pzpz> but it is berrer then money and no imagination...
[00:21:57] <ju-emb> so you can use your imagination to make money
[00:22:42] <pzpz> no.. i make money in IT..
[00:22:50] <pzpz> cnc is a hobby..
[00:23:26] <pzpz> so i make a part list now
[00:23:44] <pzpz> mesa card 5i22
[00:24:32] <ju-emb> what size you plan for your machine?
[00:24:42] <pzpz> G0704
[00:25:23] <pzpz> i live in the 3 floor
[00:25:45] <pzpz> and it is wood floor..
[00:27:09] <pzpz> too big / havy machine will make hole in the floor / i cant take it on the stairs
[00:28:07] <ju-emb> bbl
[00:28:54] <pzpz> what is bbl?
[00:29:03] <pzpz> be back later?
[08:06:49] <skunkworks> logger[mah]_,
[08:06:49] <logger[mah]_> skunkworks: Log stored at http://linuxcnc.mah.priv.at/irc/%23linuxcnc-devel/2013-11-12.html
[09:55:20] <seb_kuzminsky> cradek: i'm ready to push a driver for the Delta VFD-B, a standard modbus VFD: http://www.delta.com.tw/product/em/drive/ac_motor/ac_motor_product.asp?pid=1&cid=1&itid=2
[09:55:28] <seb_kuzminsky> do you want it in 2.5?
[09:56:07] <cradek> is there anything complicated about packaging it?
[09:56:20] <seb_kuzminsky> nope, it's pretty straightforward
[09:56:26] <cradek> then sure
[09:56:31] <seb_kuzminsky> ok :-)
[10:00:49] <seb_kuzminsky> i think this will be the first commit by Yishin Li in our repo, always nice to have new developers joining in
[10:02:06] <cradek> awesome
[10:16:09] <seb_kuzminsky> style question!
[10:16:43] <seb_kuzminsky> the vfdb driver was created by modifying the vfs11 driver (another modbus vfd driver in our tree)
[10:17:11] <seb_kuzminsky> the vfdb driver lacks many features of the vfs11 driver (or maybe those features were incomplete in the vfs11 driver too)
[10:17:29] <seb_kuzminsky> i removed all the dead code (lots of unused hal pins, for example)
[10:18:00] <seb_kuzminsky> would it be clearer to have one easily inspected & verified commit for each thing i remove, or should i squash all 10 or so pin removals into one commit?
[10:40:00] <ju-emb> if I write a component, is there a way to determine what's the actual path where the ini and hal files reside the actual application is started from?
[10:40:02] <ju-emb> I mean is there a variable or so passed to the component, or have I do this by hand?
[10:44:06] <cradek> seb_kuzminsky: I think separate would be slightly easier for review, but I don't think much review is needed if the final result works
[10:48:02] <cradek> ju-emb: what are you trying to do exactly, where you need that information?
[10:49:32] <seb_kuzminsky> ju-emb: the ini file name is in the INIFILE environment variable
[10:50:37] <ju-emb> on init the component, I need to read a table of values from a file, and I like this file to reside in the same location as the actual configuration
[10:51:18] <ju-emb> so I give just the filename as parameter, not the complete path
[10:52:25] <cradek> you are writing a userspace component, not realtime?
[10:52:34] <ju-emb> right
[10:52:47] <cradek> aha
[10:54:15] <seb_kuzminsky> cradek: hrm, i changed the way modbus stuff gets built in master vs 2.5 (2.5 still has the old cloned copy of the libmodbus source, master uses the upstream deb)
[10:54:42] <cradek> ju-emb: I think the ini directory is the cwd, so you don't have to worry about it
[10:54:42] <seb_kuzminsky> the vfd-b driver uses the upstream deb version of libmodbus
[10:54:58] <seb_kuzminsky> so this will take som emore thinking
[10:55:10] <cradek> seb_kuzminsky: darn
[11:53:40] <andypugh> seb_kuzminsky: I just noticed that the "no firmware specified" error is only raised if there is no config string, not if there is no firmware specified. Shouldn't the check be after hm2_parse_config_string ?
[11:57:30] <ju-emb> I hate coding in a ssh env!
[11:57:32] <ju-emb> what kernel you guys recommend running on a ubuntu 12.04 machine to build and test code for lcnc?
[11:59:07] <seb_kuzminsky> andypugh: maybe it should use the same check as below? if ((llio->program_fpga != NULL) && (hm2->config.firmware != NULL)) {
[11:59:22] <seb_kuzminsky> err, but with the right number of !'s removed & added etc
[12:00:08] <seb_kuzminsky> ju-emb: i use the stock ubuntu precise kernel for all my sim builds, and the rtai pre-release testing kernel for running real hardware
[12:00:40] <seb_kuzminsky> ju-emb: http://highlab.com/~seb/linuxcnc/rtai-for-3.4-prerelease/
[12:01:03] <ju-emb> thanks seb
[12:11:56] <andypugh> Seb, Yes, that was what I was thinking.
[13:04:20] <pzpz> hi
[13:55:43] <seb_kuzminsky> cradek: we'd have to add a new build-dependency to 2.5 in order to get the vfd-b driver in
[13:55:56] <seb_kuzminsky> and if we did, we'd also get the vfs11 driver
[13:56:30] <seb_kuzminsky> the new dependency would be on libmodbus-dev, which we have in our deb archive on l.o, and which we use in master currently
[14:02:33] <seb_kuzminsky> it seems like adding a new build dependency this late in the 2.5 branch is a bad idea, even if it's a pretty safe one
[14:02:57] <seb_kuzminsky> so my inclination would be to not put vfd-b (and vfs11) in 2.5, and just have them in master instead
[14:05:31] <jepler> seb_kuzminsky: makes sense to me
[16:53:02] <pzpz> i'm starting to make wiring diagram for the mesa 5i22
[16:58:28] <ju-emb> once you are on doing planning wiring, keep in mind that the mesa cards have that 50Pin headers, so you need some sort of PCB soldering connecting them to your machine
[17:09:30] <andypugh> ju-emb: Not necessarily.
[17:12:32] <ju-emb> pzpz mentioned last night, he's short on money, maybe he prefer soldering over buying ;-)
[17:13:40] <andypugh> Bear with me.
[17:14:33] <andypugh> http://uk.rs-online.com/web/p/pcb-connector-housings/6812865/
[17:15:01] <andypugh> Are two-row discrete-wire to header connectors.
[17:15:38] <andypugh> You do need a very small W-crimping too, and the crimps need to be bought separately, but they work well.
[17:17:58] <ju-emb> experience I have is, that those flat cables are the weak point if they are not right connected
[17:18:11] <andypugh> Not flat
[17:19:15] <Tom_itx> C-Grid
[17:19:19] <Tom_itx> that's what those are
[17:19:30] <Tom_itx> if you were to buy them from dk or mouser
[17:19:35] <andypugh> They are listed as Harwin M20
[17:19:56] <andypugh> I am trying to get a picture...
[17:20:28] <Tom_itx> i've used the connectors as singles and put shrinkwrap around them as well
[17:21:18] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/how_to/atmega168/breadboard_basic_side.jpg
[17:21:22] <Tom_itx> somewhat the reverse of that
[17:26:04] <andypugh> Tom_itx: Exactly like this, then? https://picasaweb.google.com/lh/photo/Til4cZnLeEZ6Y2jizdn8-tMTjNZETYmyPJy0liipFm0?feat=directlink
[17:27:18] <andypugh> I made it all rewirable to guarantee I never wanted to :-)
[17:28:06] <Tom_itx> yeah but not all my female pins have shrouds on them, rather just shrinkwrap
[17:28:18] <Tom_itx> i think yours may be as well but hard to see
[17:29:04] <Tom_itx> the ones i wired straight to the 7i43 2nd 50 pin header
[17:29:14] <andypugh> It's not clear, but yes. The box-headers for the ribbon cables all link to a single row of pin headers on the board, and then wired with cromps + heatshrink go to each pin.
[17:29:23] <andypugh> And that is a 7i43.
[17:30:20] <andypugh> But I wanted some isolation for some relays and such, There are a couple of ULN2003 chips on the board.
[17:30:56] <Tom_itx> i _was_ gonna add a 7i84 for stuff like that
[17:31:13] <Tom_itx> wired in to the 7i47
[17:31:38] <Tom_itx> and pendant stuff
[17:33:30] <ju-emb> andypugh: what crimp tool do you use for that? The original Harwin?
[17:33:56] <Tom_itx> i got a cheap one from Radio Shack
[17:34:14] <andypugh> Me too. (Well, from Rapid, actually)
[17:35:01] <Tom_itx> http://www.radioshack.com/product/index.jsp?productId=2103683
[17:35:17] <Tom_itx> they're pretty cheap... i broke one
[17:35:22] <Tom_itx> at the hinge
[17:36:16] <andypugh> I think I use this one: http://www.rapidonline.com/Tools-Equipment/Ratchet-Action-Crimp-Tool-Ht225d-85-0262/?sid=a066007d-87c5-4e0d-bf01-7e4a2ed39760
[17:37:01] <andypugh> But there are chaper ones.
[17:37:09] <andypugh> (And ones about 10x the price, too)
[17:37:17] <Tom_itx> yup
[17:38:35] <ju-emb> 10x the price like the one from Harwin on Mouser
[17:41:25] <Tom_itx> if you use the housing, make sure you get the pins from the same place as they are not all created equal
[17:45:25] <andypugh> Here are some individual wires in a housing connected to a Mesa 7i43 for some testing I was doing. https://picasaweb.google.com/lh/photo/SGrK4YGriunJFEgryZ9votMTjNZETYmyPJy0liipFm0?feat=directlink
[17:50:09] <Tom_itx> not done wireing but some on the right here going to a bundle http://tom-itx.dyndns.org:81/~webpage/cnc/psu/control5.jpg
[17:50:25] <Tom_itx> not a very good pic
[17:51:15] <Tom_itx> you can see my smps for the logic there on the left
[17:51:47] <Tom_itx> the cap in the middle is from a centertap off one of the xfrmrs with a diode
[17:52:18] <Tom_itx> knocks it down to 24v... within the smps input limits
[18:10:17] <KGB-linuxcnc> 03andy 05master 739df95 06linuxcnc 10src/hal/drivers/mesa-hostmot2/hostmot2.c
[18:10:17] <KGB-linuxcnc> Hostmot2: Don't report an error when no firmware is specified for cards that don't need firmware
[18:12:22] <ju-emb> KGB-linuxcnc: are you the Mesa Card specialist here?
[18:12:23] <KGB-linuxcnc> ju-emb: My master told me to not respond.
[18:13:23] <andypugh> The Mesa card specialist is PCW. But I know a fair bit too.
[18:13:46] <andypugh> KGB-linuxcnc: id just the Buildbot reporting on recent updates to the code.
[18:13:47] <KGB-linuxcnc> andypugh: My master told me to not respond.
[18:14:04] <andypugh> And as you can see KGB-linuxcnc is very rude
[18:15:36] <ju-emb> I see, ;-)
[18:15:38] <ju-emb> I'm curious how the parallel interface works on one of that (I think is the 7i43 or so), that connects to a parport driver
[18:21:28] <ju-emb> does that card provide hardware step generation as the 5i23 does?
[18:22:08] <ju-emb> if yes, how works the flow of data from parport to the card?
[18:27:46] <micges> ju-emb: yes 7i43 has hardware step generators
[18:30:37] <ju-emb> anyone here who can tell me about performance of that card on step generation?
[18:31:38] <ju-emb> for me, getting stepp-generation out of the PC to hardware was a great win
[18:33:13] <micges> ju-emb: up to 12 stepgens with more than 2MHz step frequency
[18:33:29] <ju-emb> I use 5i23 , so I'm wondering about how it works with parport vs pci
[18:34:51] <micges> besides that 7i43 is on cable, there is no important difference between them
[18:37:12] <ju-emb> right now, I'm just pregnant with the idea, implementing a parport driver for the beagle and panda, and get stepgeneration out of the cpu into one of those mesa cards
[18:37:47] <ju-emb> actually it's not the cpu, it's pru
[18:46:24] <pzpz> what is the difference between the 5i22-1 and the 5i22-1.5?
[18:47:15] <ju-emb> pzpz: FPGA has more gates on the 1.5
[18:48:08] <pzpz> what do you mean by "gates"
[18:48:56] <ju-emb> logic ports, but for your application the 5i22-1 is just fine
[18:50:21] <pzpz> but both of them is 96 I/O bits
[18:52:13] <ju-emb> that's right.
[18:52:15] <ju-emb> the gates count in a FPGA says something about how much "intelligence" you can implement in one chip
[18:52:52] <pzpz> 1m intelligence for my cnc is ok?
[18:53:08] <ju-emb> that's right
[18:53:55] <pzpz> ok
[18:54:23] <pzpz> i'm looking in the datasheet now for pinout
[18:56:03] <ju-emb> if you're not really into hal config right now, there is a tool to config your machine with mesa cards
[18:57:00] <ju-emb> I think that tool gives you the pins where you can connect stepper signals
[18:57:39] <pzpz> but i like to plan first, make a wire diagram, and see that everything look ok, make a part list and order it.
[18:58:01] <ju-emb> pins that are not dedicated to stepper signals you can use as PGIO
[18:58:38] <pzpz> so what i need to buy next?
[18:58:51] <ju-emb> so, just count the IO pins from the mesa card, but don't fix signals to connectors
[18:59:17] <pzpz> http://www.mesanet.com/pdf/parallel/5i22man.pdf
[18:59:27] <pzpz> page 8
[18:59:38] <pzpz> the pin out start
[19:02:12] <ju-emb> yes, i know that.
[19:02:15] <ju-emb> But the FPGA Firmware uses some predefined Pins for Step Dir Enable signals
[19:02:16] <ju-emb> If you plan your wiring, you have to use the dedicated Pins on the connector
[19:03:40] <pzpz> so lats do it like block diagram
[19:04:32] <ju-emb> that's a great idea
[19:04:33] <ju-emb> just fix the pin count, not the pin location on the header
[19:05:30] <pzpz> what pin location?
[19:06:30] <ju-emb> each pin on each header has a number representing it's position on the header
[19:07:25] <ju-emb> so you have header1 pin 27 referred as 1.27 for example
[19:07:52] <pzpz> ohh the P?
[19:07:59] <ju-emb> that assignment you shouldn't do right now
[19:08:17] <ju-emb> yes you have P1-4
[19:08:31] <pzpz> i see P2-P5?
[19:08:41] <pzpz> whare is P1?
[19:09:12] <ju-emb> Sorry, I'm doing that out of the head, not with the datasheet open
[19:09:31] <pzpz> ok..
[19:09:43] <ju-emb> you have four 50 Pin headers on the 5i22
[19:10:21] <ju-emb> ok, now datasheet open
[19:10:33] <ju-emb> the pins are named in sequence
[19:10:35] <pzpz> it is like IDE connector?
[19:10:46] <pzpz> ok i see
[19:10:48] <ju-emb> 0 .. 95
[19:11:20] <pzpz> so it 96 i/o
[19:12:06] <ju-emb> so each pin has it's number, but what I'm talking about is,
[19:12:09] <ju-emb> don't assign a signal at each pin number yet
[19:12:35] <pzpz> almost half of the pins in the connecte are GND..
[19:13:19] <ju-emb> that's for noise immunity, it's an electronic rule
[19:14:03] <pzpz> it is grat!
[19:14:22] <ju-emb> ??
[19:14:31] <pzpz> grate*
[19:15:50] <pzpz> *excellent
[19:15:54] <ju-emb> on the cable flat you have always a signal then gnd signal gnd
[19:16:46] <ju-emb> so you don't get ghost signals made by inductivity
[19:17:10] <pzpz> this is fine.
[19:17:41] <pzpz> i need a special cable?
[19:19:23] <pzpz> i have a-lot of IDE but it is 40 pins..
[19:19:57] <ju-emb> I use the same flat cable as you use inside your PCs for CD or DVD drive, but there you need some counterpart to it, that involves soldering.
[19:19:59] <ju-emb> today we just talked about this and there are another solutions too
[19:21:04] <ju-emb> http://tom-itx.dyndns.org:81/~webpage/how_to/atmega168/breadboard_basic_side.jpg
[19:21:06] <ju-emb> https://picasaweb.google.com/lh/photo/Til4cZnLeEZ6Y2jizdn8-tMTjNZETYmyPJy0liipFm0?feat=directlink
[19:21:29] <ju-emb> have a look at this links
[19:23:44] <pzpz> http://www.ebay.com/itm/23inch-Internal-IDC-50-Pin-SCSI-Male-Female-Extension-Ribbon-Cable-CablesOnline-/280754237369
[19:23:58] <pzpz> 4 cables like that?
[19:25:24] <pzpz> http://www.mesanet.com/graphics/parallel/5i22colo.png
[19:25:26] <ju-emb> yes, but with two female connectors
[19:25:38] <pzpz> ok
[19:25:48] <pzpz> now i see the card
[19:27:09] <pzpz> is the some as scsi connector so no problem..
[19:27:30] <ju-emb> so if you know how to solder, you buy a few of those prototyping boards and 50 Pin headers
[19:27:53] <ju-emb> that's right, the old SCSI use that connectors
[19:28:48] <kwallace1> These are handy: http://www.winfordeng.com/products/cat_pbc.php
[19:29:13] <ju-emb> not really
[19:29:32] <kwallace1> Not really handy?
[19:29:56] <pzpz> what i need to put at the other end of the cables?
[19:30:31] <ju-emb> right
[19:30:33] <ju-emb> if you use the gecko drives, there are screw terminals on it, so you have to transform from Flat cable to Screw Terminals
[19:31:04] <kwallace1> I find them handy to get cables plugged into a breadboard, otherwise they aren't handy.
[19:31:53] <micges> http://www.winfordeng.com/products/brk2x25.php
[19:31:55] <kwallace1> On the other end maybe: http://www.winfordeng.com/products/cat_brk.php
[19:32:28] <pzpz> so i need to convert the 96 i/o form the 4 50 pins flat cables to 96 terminal blocks
[19:32:49] <pzpz> what about the GND pins?
[19:33:01] <pzpz> and noise problems?
[19:33:51] <Tom_itx> every other pin is a GND
[19:34:33] <pzpz> so what to do with tham?
[19:34:58] <Tom_itx> i used one and ignored most of them
[19:35:04] <ju-emb> yes, your drivers need gnd beside of the signals so you can go one gnd, step dir enable to the driver
[19:35:06] <kwallace1> The adapter that micges linked to has solder points for the signals and ground, so just solder jumpers to all the grounds.
[19:35:10] <pzpz> just not connect them?
[19:36:40] <micges> kwallace1: thanks for link, I was looking such adapters few times
[19:36:44] <ju-emb> just bridge all the gnds together and solder that qty of gnd terminals as you need
[19:36:54] <pzpz> 45$ wow
[19:37:37] <pzpz> i can make something like that
[19:37:46] <pzpz> no problem at all
[19:37:50] <kwallace1> Flat ribbon is not the best for noise but often is good enough. Twisted pair ribbon is better. Often the far end of the grounds are left unconnected.
[19:38:08] <pzpz> i know how to solder and make pcbs..
[19:38:28] <ju-emb> so no problem with that
[19:38:43] <pzpz> i can get noises??
[19:38:58] <pzpz> it is a common problem with the mesa cards?
[19:39:36] <ju-emb> not with mesa cards, it's a general problem
[19:39:51] <kwallace1> I use the Winford stuff when I'm on the clock, but tend to make what I need when it's my own project. http://www.wallacecompany.com/cnc_lathe/HNC/00004-1a.jpg
[19:40:35] <ju-emb> just keep the ribbon cables short, just to come out of the PC
[19:40:55] <pzpz> no problem.
[19:41:15] <pzpz> half price http://www.ebay.com/itm/IDC50-2x25-Pins-0-1-Male-Header-Breakout-Board-Terminal-Block-Connector-/190903623678?pt=LH_DefaultDomain_0&hash=item2c72be1ffe
[19:42:22] <ju-emb> after that cables go in different directions, and have space between it, space or twist is your friend in noise immunity, but if you twist, you have to know how
[19:42:51] <kwallace1> In my experience, the most important noise item is to have power input line filters on your VFDs and switch mode power supplies. That should prevent most problems right off the bat.
[19:43:45] <pzpz> what do you mean by twist, them?
[19:43:54] <kwallace1> The filter is the shiny bit: http://www.wallacecompany.com/cnc_lathe/HNC/00024-1a.jpg
[19:43:54] <pzpz> is a flat cable...
[19:43:56] <ju-emb> the breakout board is the one you need, but 20 bucks
[19:44:30] <pzpz> i can make it
[19:45:05] <ju-emb> not twisting the flat, from the screw terminal to your drivers and switches
[19:45:06] <pzpz> i have alot of this
[19:46:02] <ju-emb> would say
[19:46:05] <ju-emb> not twisting the flat, from the screw terminal to your drivers and switches
[19:46:06] <pzpz> ohh i see
[19:46:18] <kwallace1> http://en.wikipedia.org/wiki/Balanced_line
[19:46:43] <Tom_itx> like cat 5/6 pairs are inside
[19:48:48] <pzpz> i have roll of it..
[19:49:03] <kwallace1> The twisting of signal and corresponding ground wire cancels noise. CAT5 uses twisted pairs and works well even without a shield.
[19:49:18] <pzpz> fire protected
[19:49:40] <pzpz> it is cat 5E
[19:50:20] <kwallace1> Flat ribbon is great for terminating a lot of wires easily and should work okay for moderate runs.
[19:51:05] <pzpz> you split the ribbon?
[19:51:12] <Tom_itx> if you make a board, use a ribbon to board connector for it
[19:55:13] <Tom_itx> http://www.digikey.com/product-search/en?vendor=0&keywords=69830-006LF
[19:55:17] <Tom_itx> only wider of course
[19:55:49] <Tom_itx> or split the cable off to multiple boards
[19:58:11] <ju-emb> huuh the new rtai kernel is ready, let's give it a try
[19:58:16] <ju-emb> bbl
[20:00:20] <pzpz> http://imgur.com/luwWL1O
[20:00:29] <pzpz> it will be ok?
[20:00:52] <pzpz> kwallace1, Tom_itx ^
[20:01:49] <Tom_itx> do you need all 100 io?
[20:02:03] <Tom_itx> 96 or whatever it is
[20:02:42] <Tom_itx> i would make a board per ribbon personally
[20:03:28] <Tom_itx> and make sure the wires / screw terminals are all accessable
[20:04:11] <pzpz> i can also short all the GND by the pcb and make 97 pins
[20:04:24] <pzpz> 96 i/o + 1 GND
[20:04:46] <Tom_itx> one gnd may or may not carry all the current for the io, that i couldn't say
[20:05:31] <micges> pzpz: what are trying to design?
[20:05:51] <Tom_itx> apparently some sort of breakout board for the ribbon cables
[20:06:28] <pzpz> i think thay short by the board..
[20:07:03] <micges> pzpz: for what will you use 5i22 board?
[20:07:16] <pzpz> micges, the pinout for the 5i22
[20:07:31] <pzpz> my new cnc
[20:07:54] <pzpz> G0704 +5 axis + cool i/o
[20:08:22] <micges> do yo have it?
[20:08:49] <pzpz> no
[20:08:55] <pzpz> i want to but it
[20:08:56] <pzpz> http://pastebin.com/zMcBVz1B
[20:09:12] <pzpz> this is what i plan
[20:09:53] <micges> what servos do you want to use?
[20:10:08] <pzpz> i dont know.
[20:10:59] <pzpz> first i was think to use G320X with regular DC brashed motor
[20:11:25] <micges> you must know becouse from that you will know what type of daughter boards you need to connect to 5i22 to make it work
[20:11:27] <skunkworks> http://electronicsam.com/images/KandT/conversion/panel/electricalbox.jpg
[20:11:44] <pzpz> now ju-emb say that with the masa card i can make something like DRO
[20:12:21] <skunkworks> 2 5i20's
[20:14:03] <pzpz> what is the cards at the top?
[20:15:21] <pzpz> alot of wires.....
[20:15:55] <micges> pzpz: 2x mesa 5i20
[20:16:10] <micges> lower you have mesa 7i33
[20:17:34] <pzpz> what machine?
[20:20:18] <pzpz> so you have 144 i/o?
[20:20:35] <pzpz> wow that crazy
[20:23:27] <pzpz> i need to buy with the "Industrial Temperature Range Test"
[20:24:54] <pzpz> micges, skunkworks ?
[20:28:22] <pzpz> the 5i22 is 369$ and the 5i20 is 199$
[20:29:26] <pzpz> what is better 2X5i20 or one 5i22?
[20:43:29] <pzpz> ??
[20:43:45] <pzpz> micges, skunkworks, kwallace1 ?
[20:46:25] <kwallace1> pzpz: This should be close to what you want: http://www.youtube.com/watch?v=UI3nUjWTO5Q
[20:48:35] <kwallace1> http://www.youtube.com/watch?v=oyqx7c_-mrw&feature=plcp
[21:00:55] <pzpz> amazing!
[21:01:29] <pzpz> so i stay with the 5i22
[21:02:29] <kwallace1> I can't seem to find the details on that machine, just this http://www.deweygarrett.com/
[21:02:33] <pzpz> kwallace1, now for the 5 main servo motors, what i need?
[21:03:33] <pzpz> ok...
[21:03:42] <pzpz> lets move on
[21:03:47] <kwallace1> It depends on the amps
[21:04:29] <kwallace1> Dewey uses Pico's amps with Keling brushed motors
[21:04:30] <pzpz> i dont know what controlers / encoder / motors to chose
[21:04:55] <pzpz> Keling?
[21:05:11] <ju-emb> that's why I asked yesterday for the size of your machine
[21:05:13] <pzpz> it is a good brand?
[21:05:25] <pzpz> G0704
[21:05:39] <pzpz> http://www.grizzly.com/products/Drill-Mill-with-Stand/G0704
[21:05:39] <kwallace1> http://www.kelinginc.net/ServoMotor.html
[21:06:33] <kwallace1> You will need ball screws for that machine.
[21:07:42] <kwallace1> They need to be preloaded some how, preferably double nutted.
[21:07:58] <pzpz> i cant work with the lead screws?
[21:08:46] <pzpz> double nut with spring?
[21:08:49] <kwallace1> Not if you are serious about having a decent CNC.
[21:09:09] <ju-emb> it's a question what precision you need, and how long your precision stays that way
[21:10:22] <ju-emb> normal lead screws have a lot of friction so you'll need stronger motors
[21:10:33] <pzpz> +- 0.1 mm..
[21:11:35] <ju-emb> that's not so much
[21:11:43] <pzpz> i think to start with lead screws.. and see how it going..
[21:12:09] <kwallace1> With manual machines the machinist uses a technique to compensate for backlash and is a big part of becoming a machinist. CNC machines don't have a machinist's skill so can not tolerate any backlash.
[21:13:01] <kwallace1> When you take the backlash out of a lead screw the force to drive the screw gets too high.
[21:13:20] <pzpz> i know.
[21:14:11] <ju-emb> in general, ball screws are cheaper than stronger motors and drives
[21:15:13] <kwallace1> So you really need to use preloaded ball screws as well as double angular contact end bearings.
[21:16:54] <kwallace1> Ball screws and bearings can cost as much as a used mill, so it is cheaper to find a used CNC instead of converting a manual machine.
[21:17:08] <pzpz> where i can get a ball screws for the G0704 in reasnabel price?
[21:19:08] <pzpz> but i cant find a used cnc that 2 people can lift in stairs
[21:19:27] <pzpz> and will not make a hole in my floor.....
[21:19:55] <pzpz> 150kg is the limit that i can have..
[21:20:05] <kwallace1> Microkentics was one vendor I was considering http://www.microkinetics.com/conv_kits/
[21:20:55] <kwallace1> The screws are double nutted, reasonably accurate and have oiling ports.
[21:21:46] <pzpz> but 2000$
[21:22:07] <kwallace1> I bought some HiWin screws but they were not preloaded so didn't work.
[21:22:44] <kwallace1> Before that I bought some screws from England but they never showed up.
[21:22:44] <pzpz> what is "preloaded"
[21:24:29] <kwallace1> For a double nut screw each nut is shimmed or other wise adjusted to take out the backlash plus a little bit more.
[21:24:48] <pzpz> http://www.automationtechnologiesinc.com/products-page/g0704-kit/g0704-cnc-update-machine-ballscrew-kit
[21:25:46] <kwallace1> For single nut screws, over sized balls can be used but they need to be replaced as the screw wears.
[21:27:39] <pzpz> i dont need that acurate machine
[21:27:39] <pzpz> +-0.1 mm is fine for me
[21:27:39] <pzpz> now i make thing by hand...
[21:28:12] <kwallace1> It's the backlash that's the problem.
[21:30:38] <pzpz> i see a guy in cnc zone that put a spring betweem 2 ball nut.
[21:30:55] <KGB-linuxcnc> 03ysli 05vfd-b-3 47f4fdc 06linuxcnc 10src/Makefile * Update Makefile for vfdb_vfd
[21:30:56] <KGB-linuxcnc> 03ysli 05vfd-b-3 b03906b 06linuxcnc 10src/hal/user_comps/vfdb_vfd/ 03Submakefile 03vfdb_vfd.c * Initial commit for DELTA VFD-B Spindle Driver
[21:30:57] <KGB-linuxcnc> 03jofey 05vfd-b-3 6d4ce97 06linuxcnc 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * Update implementation for DELTA-VFD-B Inverter, vfdb_vfd.c
[21:31:04] <KGB-linuxcnc> 03ysli 05vfd-b-3 2371b66 06linuxcnc 10src/hal/user_comps/vfdb_vfd/Submakefile * vfdb_vfd: add '-m' linking parameter to Submakefile
[21:31:11] <KGB-linuxcnc> 03jofey 05vfd-b-3 60592e5 06linuxcnc 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * Add RPS hal signal to vfdb_vfd.c
[21:31:17] <KGB-linuxcnc> 03jofey 05vfd-b-3 e456422 06linuxcnc 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * vfdb_vfd.c: update VFD-B CMD register definition
[21:31:24] <KGB-linuxcnc> 03jofey 05vfd-b-3 2fe1ea6 06linuxcnc 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * vfdb_vfd.c: can do FORWARD and REVERSE spindle rotation
[21:31:31] <KGB-linuxcnc> 03seb 05vfd-b-3 57d3613 06linuxcnc 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * vfd-b: finish rename from "vfs11" to "vfdb"
[21:31:37] <KGB-linuxcnc> 03seb 05vfd-b-3 6ebaa1b 06linuxcnc 03docs/man/man1/vfdb_vfd.1 * vfd-b: add a manpage
[21:31:43] <KGB-linuxcnc> 03seb 05vfd-b-3 9d7e5a7 06linuxcnc 10docs/man/man1/vfdb_vfd.1 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * vfd-b: make motor frequency & speed into ini parameters
[21:31:50] <KGB-linuxcnc> 03seb 05vfd-b-3 1fe7588 06linuxcnc 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * vfd-b: report at-speed correctly
[21:31:57] <KGB-linuxcnc> 03seb 05vfd-b-3 909d185 06linuxcnc 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * vfd-b: fixup copyright
[21:32:03] <pzpz> ?????
[21:32:03] <KGB-linuxcnc> 03seb 05vfd-b-3 57b4662 06linuxcnc 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * vfd-b: whitespace cleanup
[21:32:09] <KGB-linuxcnc> 03seb 05vfd-b-3 16ed3d8 06linuxcnc 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * vfd-b: remove a redundant assignment
[21:32:16] <KGB-linuxcnc> 03seb 05vfd-b-3 bb67c54 06linuxcnc 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * vfd-b: comment out some nearly dead code...
[21:32:20] <pzpz> KGB-linuxcnc, stop!!!
[21:32:22] <KGB-linuxcnc> 03seb 05vfd-b-3 b6363c1 06linuxcnc 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * vfd-b: remove dead code 1/15
[21:32:29] <KGB-linuxcnc> 03seb 05vfd-b-3 b6deeb1 06linuxcnc 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * vfd-b: remove dead code 2/15
[21:32:34] <KGB-linuxcnc> 03seb 05vfd-b-3 f41fe80 06linuxcnc 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * vfd-b: remove dead code 3/15
[21:32:36] <pzpz> WTF?!
[21:32:41] <KGB-linuxcnc> 03seb 05vfd-b-3 3473d2e 06linuxcnc 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * vfd-b: remove dead code 4/15
[21:32:47] <KGB-linuxcnc> 03seb 05vfd-b-3 f3d019e 06linuxcnc 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * vfd-b: remove dead code 5/15
[21:32:53] <KGB-linuxcnc> 03seb 05vfd-b-3 7a2099a 06linuxcnc 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * vfd-b: remove dead code 6/15
[21:32:55] <kwallace1> We are on the developer channel so these messages are common.
[21:33:00] <KGB-linuxcnc> 03seb 05vfd-b-3 8db8ff0 06linuxcnc 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * vfd-b: remove dead code 7/15
[21:33:01] <cradek> pzpz: it's a feature. you might prefer the #linuxcnc channel if you're not interested in development
[21:33:06] <KGB-linuxcnc> 03seb 05vfd-b-3 b3602e1 06linuxcnc 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * vfd-b: remove dead code 8/15
[21:33:12] <KGB-linuxcnc> 03seb 05vfd-b-3 863c8df 06linuxcnc 10docs/man/man1/vfdb_vfd.1 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * vfd-b: remove dead code 9/15
[21:33:18] <KGB-linuxcnc> 03seb 05vfd-b-3 13108c8 06linuxcnc 10docs/man/man1/vfdb_vfd.1 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * vfd-b: remove dead code 10/15
[21:33:26] <KGB-linuxcnc> 03seb 05vfd-b-3 4c2bfd2 06linuxcnc 10docs/man/man1/vfdb_vfd.1 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * vfd-b: remove dead code 11/15
[21:33:33] <KGB-linuxcnc> 03seb 05vfd-b-3 8704c6d 06linuxcnc 10docs/man/man1/vfdb_vfd.1 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * vfd-b: remove dead code 12/15
[21:33:39] <KGB-linuxcnc> 03seb 05vfd-b-3 4ad89f3 06linuxcnc 10docs/man/man1/vfdb_vfd.1 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * vfd-b: remove dead code 13/15
[21:33:46] <KGB-linuxcnc> 03seb 05vfd-b-3 de93e0a 06linuxcnc 10docs/man/man1/vfdb_vfd.1 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * vfd-b: remove dead code 14/15
[21:33:53] <KGB-linuxcnc> 03seb 05vfd-b-3 59c4195 06linuxcnc 10docs/man/man1/vfdb_vfd.1 10src/hal/user_comps/vfdb_vfd/vfdb_vfd.c * vfd-b: remove dead code 15/15
[21:33:59] <KGB-linuxcnc> pzpz: My master told me to not respond.
[21:34:21] <pzpz> KGB-linuxcnc, ok
[21:34:22] <KGB-linuxcnc> pzpz: My master told me to not respond.
[21:34:25] <pzpz> KGB-linuxcnc, ok
[21:34:25] <KGB-linuxcnc> pzpz: My master told me to not respond.
[21:34:41] <pzpz> but you respond..
[21:34:52] <kwallace1> LOL
[21:35:11] <kwallace1> KGB is a computer
[21:35:20] <pzpz> i know..
[21:35:32] <pzpz> i make it for the joke...
[21:36:00] <kwallace1> Indeed.
[21:37:12] <pzpz> http://www.automationtechnologiesinc.com/products-page/g0704-kit/g0704-cnc-update-machine-ballscrew-kit
[21:37:21] <pzpz> and i see a guy in cnc zone that put a spring betweem 2 ball nut.
[21:37:36] <pzpz> it will work for me?
[21:38:37] <kwallace1> You could try it, nothing to lose.
[21:39:07] <cradek> pzpz: to be more overt: you should take your questions and this conversation to #linuxcnc, because it is on topic there, and you'll have a larger and more varied group to discuss these things with.
[21:39:59] <pzpz> ok.. i move to linuxcnc
[21:40:11] <cradek> great, thanks
[21:41:11] <pzpz> fine.. but nobody there..
[21:48:57] <cradek> weird
[21:49:55] <cradek> does anyone have (ntp correct) seconds timestamps that say when he quit?
[21:52:16] <micges> 04:32:32
[21:52:35] <micges> (checking time zone..)
[21:53:14] <cradek> thanks I know which minute it was :-)
[21:53:42] <micges> ah sorry didn't read carefully
[21:56:22] <cradek> I found an error message but have no clue what it means
[21:56:33] <cradek> and kgb's web site and vcs both seem gone
[21:56:56] <cradek> yet another vc bot casualty
[21:59:08] <cradek> [root@git /usr/local/src/kgb.trunk]# svn log
[21:59:09] <cradek> svn: E000061: Unable to connect to a repository at URL 'svn://anonscm.debian.org/kgb/kgb/trunk'
[21:59:12] <cradek> svn: E000061: Can't connect to host 'anonscm.debian.org': Connection refused
[21:59:15] <cradek> oh how I hate svn
[22:00:31] <cradek> ah there is evidence of a 1.31; I probably have 1.25
[22:00:34] <cradek> ls
[22:00:36] <cradek> oops
[22:12:33] <cradek> perl dependencies everywhere
[22:12:43] <cradek> in addition to how I hate svn, I also hate everything
[22:13:59] <ju-emb> maybe it's time to grab a cup of coffee
[22:14:31] <cradek> ---> Installing 'p5-common-sense-3.6' from a port (devel/p5-common-sense)
[22:14:39] <cradek> oh come on this is ridiculous
[22:18:15] <cradek> after trying many sites, it finally found common sense in .nz
[23:39:46] <huosen> I'm looking for some information on hardware to be used with linuxcnc, any suggestions?
[23:42:11] <CaptHindsight> http://wiki.linuxcnc.org/cgi-bin/wiki.pl?EMC2_Supported_Hardware