#emc | Logs for 2009-03-20

[00:20:06] <RayAcayan1> RayAcayan1 is now known as RayAcayan
[00:24:08] <alex_joni> good night all
[01:48:42] <hugomatic1> Hi, do you have any guide for Gecko 540 drive setups? I can't make it work with stepconf, and I can't find a working config either.
[01:51:44] <fenn> god what an awful website
[01:52:01] <hugomatic1> There is a long thread in CNCzone about the charge pump... but I don't have a 'fault' LED when I load my config. Could this be a real problem?
[01:52:19] <hugomatic1> http://cnczone.com/forums/showthread.php?t=72003
[01:54:23] <fenn> it requires EPP mode - does this mean the drive receives more than just step/dir signals?
[01:54:50] <hugomatic1> I have set the EPP mode in the BIOS according to the docs
[01:55:14] <fenn> anyway, you need to have a charge pump in the config somewhere
[01:55:31] <hugomatic1> But in stepconf I set all the pins to unused, except the dir/steps and the charge pump
[01:56:54] <hugomatic1> In stepconf I cannot move the motoers, because the fault LED is on (probably because of the charge pump). In EMC, the Fault goes away when I power the machine.
[01:56:59] <fenn> you have a red LED right?
[01:57:08] <fenn> and it turns green when you run EMC?
[01:57:33] <hugomatic1> exactly, but its always red instepconf
[01:57:50] <fenn> well, i guess that's a bug
[01:58:36] <hugomatic1> I thought I was getting a box that was popular with EMC users :-(
[01:59:01] <SWPadnos> hugomatic1, can you accept dcc?
[01:59:24] <hugomatic1> What's dcc?
[01:59:27] <fenn> i've never heard of G540
[01:59:29] <SWPadnos> most of us don't have a G540. I do, and I made a config for it
[01:59:34] <fenn> dcc is a file sharing protocol that never works
[01:59:42] <toastydeath> ^^^ this
[01:59:42] <SWPadnos> dcc is a file transfer scheme through irc
[01:59:49] <fenn> just use pastebin
[01:59:51] <SWPadnos> this is irc, not dcc :)
[02:00:26] <hugomatic1> I'm using pidgin, and I can give you my email if you want
[02:00:51] <SWPadnos> I'll send a dcc request - we'll see if it works
[02:00:59] <SWPadnos> if not, I'll just stick the file on the web
[02:00:59] <hugomatic1> ok
[02:01:12] <fenn> * fenn mumbles something about sample configs
[02:01:21] <SWPadnos> definitely let me know if it works though, I need to send it to Mariss so he can stick it on their website
[02:01:38] <hugomatic1> I promise :-)
[02:02:00] <SWPadnos> you will need to get accurate latency test times though, accuracy isn't guaranteed unless you put good data in :)
[02:02:13] <SWPadnos> also you'll need to set the scale correctly (steps/inch or cm)
[02:04:26] <hugomatic1> thanks
[02:04:31] <SWPadnos> sure
[02:05:02] <SWPadnos> the latency numbers are probably pretty low in there, I did the testing on a pretty low latency system :)
[02:07:30] <hugomatic1> SWPadnos: amazing! it works like a charm
[02:07:38] <SWPadnos> excellent!
[02:08:01] <hugomatic1> What motors are you using? I have Sherline 2 Amp motors... are yours similar?
[02:09:25] <SWPadnos> I have 3 Keling KL23H284-35-4B
[02:09:55] <SWPadnos> they're the ones that are "made for hte G250" - more or less exactly the same specs as the driver
[02:10:05] <hugomatic1> my pins were not inverted as yours were (didn't get that from the doc). I get a RTAPI ERROR: Unexpected realtime delay each time I load EMC.
[02:10:54] <SWPadnos> did you enter the correct latency times for your PC?
[02:11:15] <SWPadnos> and a realistic maximum velocity
[02:11:25] <hugomatic1> I'll get to the tuning... haven't done any latency tweeking yet
[02:12:26] <SWPadnos> that's not tuning, it's basic stuff
[02:12:34] <hugomatic1> I only have my rotary table here...
[02:12:40] <SWPadnos> like I said, the numbers in there now are probably pretty aggressive
[02:12:57] <SWPadnos> maybe change the latency number to 15000 or so (instead of the 6000 or 7000 it's at now)
[02:13:18] <hugomatic1> top tip
[02:14:05] <SWPadnos> the released version of the config will have low performing but "less likely to die" numbers :)
[02:14:10] <fenn> i thought stepconf wouldnt allow an unrealistic maximum velocity
[02:14:50] <toastydeath> max velocity: 1/9 * c
[02:14:57] <SWPadnos> it won't exceed timing that's figured from the latency numbers, but if the latencies are wrong by a factor of 2, it will allow more than the computer can actually do
[02:15:12] <SWPadnos> (well, if they're wrong at all ...)
[02:15:16] <fenn> toastydeath: but your lathe does warp 6
[02:15:51] <hugomatic1> those numbers are directly in the ini file, right? not in stepconf ...
[02:15:54] <toastydeath> to... THE FUTURE
[02:16:16] <SWPadnos> no, stepconf lets you enter them
[02:16:30] <SWPadnos> hmmm. I have a "New540.stepcon" file
[02:16:33] <SWPadnos> I wonder what that is :)
[02:16:37] <fenn> toastydeath: so that would be 392.49805 * c
[02:17:41] <toastydeath> i am both amazed and appaled at your knowledge
[02:18:34] <hugomatic1> 15000 is good
[02:22:50] <SWPadnos> ok, the only change is that the file paths don't have my user name in them in New540.stepconf :)
[02:22:53] <hugomatic1> SWPadnos: thank you very much
[02:22:56] <SWPadnos> sure
[02:29:25] <alexchally> so I am trying to setup emc2 to drive some stepper motors, and for some reason there seems to be no signal coming out of my parallel port
[02:29:55] <alexchally> i put loadrt probe_parport in my .hal file
[02:29:57] <alexchally> no dice
[02:30:06] <alexchally> was wondering if anyone had a clever idea
[02:30:34] <alexchally> note that I got it working once before on a live install, but yhave sense changed computers, so I am fairly sure that all of my wiring, etc etc is working
[02:32:14] <alexchally> bleh, restarting brb
[04:24:26] <Guest240> hello?
[09:20:25] <Guest149> hi alex
[09:21:48] <Guest149> need help guys, I am new to emc and keeps on getting the following message:
[09:22:14] <alex_joni> whoops ;)
[09:27:55] <leonard> need help running 3-axis_turning-machine. 4-axis works fine but 3-axis has errors
[09:38:33] <leonard> please help people
[09:45:40] <archivist> ask the real question
[09:46:54] <leonard> i am new but please look at the following error:
[09:50:29] <archivist> use a pastebin eg http://pastebin.ca
[09:50:52] <archivist> never paste in irc
[09:51:14] <leonard> okay i did not know
[09:55:22] <leonard> i have pasted it? what do i need to do now
[09:57:02] <leonard> 1366061
[09:58:56] <archivist> normally you hand back the full url eg http://pastebin.ca/1366061
[09:59:42] <archivist> near the end core_stepper.hal:83: setp requires 2 arguments, 3 given
[10:02:12] <leonard> how do i fix that?
[10:02:32] <archivist> how did you create the setup
[10:04:20] <alex_joni> leonard: usually that's a problem with the ini file
[10:04:33] <alex_joni> you have something like INPUT_SCALE = 800 0
[10:04:43] <alex_joni> you need to remove the second part of the INPUT_SCALE
[10:04:43] <leonard> loaded ubuntucnc, then i run install file from cooltool. when this did not work i updated emc2.1.5 to emc2.28
[10:05:14] <alex_joni> leonard: the next (2.3.0) emc2 will have cooltool configs in it
[10:07:46] <leonard> How do i remove input_scale? please be patient with me as i am a begginer from south africa
[10:09:58] <archivist> you probably just need to correct the line as it has 1 argument too many
[10:10:07] <archivist> use an editor
[10:20:33] <leonard> which file do i need to edit "3-axis_turning-machine.ini"
[10:24:23] <leonard> is the second part of the INPUT_SCALE THE 0
[10:27:09] <leonard> it is working
[10:30:05] <piasdom> g'mornin all
[10:31:15] <leonard> does emc2 run lathe?
[10:42:37] <leonard> how do i add lathe to emc2?
[10:48:42] <alex_joni> leonard: http://www.linuxcnc.org/docview/devel/html//lathe_lathe-user.html
[10:50:40] <alex_joni> http://www.linuxcnc.org/content/view/28/13/lang,en/
[10:57:13] <leonard> Alex: thank you very much and will let you know how it went
[10:58:13] <piasdom> how do i see what version of emc2 i have ?
[10:59:32] <alex_joni> piasdom: AXIS -> about
[10:59:57] <piasdom> alex_joni; thanks
[11:04:02] <leonard> Alex: i looked at the content and it does not mention which ini file to add the lathe
[11:08:49] <archivist> make a new ini
[11:09:50] <alex_joni> leonard: what machine do you have?
[11:21:58] <leonard> unimat 1 lathe
[11:31:21] <alex_joni> leonard: http://cvs.linuxcnc.org/cvs/emc2/configs/cooltool/
[11:31:40] <alex_joni> those configs run with emc2-2.3.0
[11:31:50] <alex_joni> but with a minor change they can be made to run with 2.2.8 too
[11:42:22] <leonard> is emc2-2.3.0 available for downlaods
[11:44:56] <piasdom> what is cooltool?
[11:52:52] <alex_joni> leonard: only beta versions
[12:16:31] <skunkworks> http://www.cnczone.com/forums/attachment.php?attachmentid=77025&d=1236386181
[12:16:54] <skunkworks> http://www.cnczone.com/forums/showthread.php?t=75035
[12:34:38] <alex_joni> cool
[12:34:45] <SWPadnos> skunkworks, that thread doesn't mention all that it does, which is a lot
[12:36:33] <SWPadnos> Here's what Mariss wrote last night: http://pastebin.ca/1366134
[12:50:47] <archivist> I like "The G380 will not be damaged if the motor is run into a hard-stop (crash into brick wall). This kills the G320 every time."
[12:52:07] <SWPadnos> yeah, that's a good one
[12:52:47] <jepler> that list just makes me think the G320 must have been terrible :-P
[12:53:12] <SWPadnos> what I don't like is that the encoder fault detection only detects a broken wire/disconnect - it doesn't detect things like a broken ehncoder wheel, bad LED/bulb, bad (stuck) transmitter, etc.
[12:53:14] <SWPadnos> heh
[13:00:11] <cradek> I wonder if that's the most common encoder failure
[13:00:22] <SWPadnos> I bet it is
[13:00:26] <SWPadnos> at least for new encoders :)
[13:00:32] <cradek> hah
[13:00:48] <SWPadnos> ebay specials it's hard to tell though
[13:01:29] <SWPadnos> Mariss has been using these for testing: http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=102-1308-ND
[13:01:31] <SWPadnos> he loves them
[13:01:49] <SWPadnos> pretty cheap, and programmable resolution
[13:07:20] <skunkworks> it is still going to have atleast 256 count error.
[13:07:30] <skunkworks> +/-
[13:07:33] <SWPadnos> before the drive faults ...
[13:07:40] <skunkworks> right
[13:08:11] <SWPadnos> that may be useful to mach users ;)
[13:54:29] <skunkworks> CPLD design. The G380 can be programmed and customized to behave differently without any hardware changes. Example: Need CW/CCW inputs instead of STP/DIR? No problem.
[13:54:54] <cradek> quadrature! quadrature!
[13:54:56] <SWPadnos> or quadrature, which several people have suggested
[13:54:58] <skunkworks> yah
[13:54:58] <SWPadnos> :)
[13:55:03] <cradek> analog velocity!
[13:55:12] <SWPadnos> actually, that's also in the plan
[13:55:13] <skunkworks> he doesn't seem to want to go there.
[13:55:17] <skunkworks> really?
[13:55:18] <skunkworks> cool
[13:55:31] <SWPadnos> he has floated the idea of providing more or less a FET front end
[13:56:13] <SWPadnos> several of us have also asked for an analog output representing current, for loading measurements
[13:56:38] <SWPadnos> and also outputs for those LEDs, so you could (a) connect them to a visible panel or (b) connect them to the control
[13:59:11] <SWPadnos> I don't know how much of that will ever get released, but those have all been suggested
[14:04:19] <Optic> hi
[14:06:49] <BJT-Work> hi
[14:11:46] <elmood> hi
[14:12:04] <elmood> <- friend of optic... been lurking
[14:12:36] <BJT-Work> hi elmood
[14:12:40] <elmood> hi
[14:23:01] <elmood> moo
[14:29:44] <hypa7ia> i wonder what portion of this channel is hacklabbers
[14:29:49] <jepler> an open-source (gpl hdl + schematic + reference layout) velocity-mode servo amp would be cool
[14:29:58] <Optic> that would be my fault
[14:29:58] <elmood> hyp: you're here too?
[14:30:23] <hypa7ia> no >_>
[14:30:24] <hypa7ia> <_<
[14:30:26] <elmood> haha
[14:30:34] <hypa7ia> * hypa7ia hides
[14:30:37] <elmood> we has laz0r
[14:30:45] <hypa7ia> lazzzzzzzor
[14:31:46] <jepler> (and more relevant to emc than a step-input one)
[14:32:04] <elmood> jepler: what exactly do you want?
[14:32:10] <elmood> i missed the beginning
[14:32:26] <skunkworks> jepler: what feedback? encoder?
[14:32:30] <skunkworks> or tach?
[14:32:41] <SWPadnos> the drive would need velocity feedback
[14:32:56] <SWPadnos> whether that comes from an encoder or a tach is left as an exercise for the designer :)
[14:33:01] <skunkworks> right - could a high count encoder work?
[14:33:04] <skunkworks> ah
[14:33:06] <SWPadnos> (but an encoder won't be as good at low speeds)
[14:33:19] <SWPadnos> unless you compare it to a bad analog tach design :)
[14:33:24] <skunkworks> heh
[14:33:37] <jepler> elmood: I don't think you missed much. I was just thinking about G380 which was talked about above, and imagining what would be more interesting for emc
[14:33:59] <elmood> cool
[14:35:17] <elmood> oooh
[14:35:18] <elmood> PID control
[16:10:22] <alex_joni> jepler: maybe that's the servo drive skunkworks designed
[16:11:11] <jepler> alex_joni: it's not velocity-mode. it would make a good power stage, though (and as we've seen, emc works pretty well with voltage-mode servo amps anyhow)
[16:12:47] <skunkworks> has anyone tried doing velocity and postion from the same encoder? (pid for velocity and position?)
[16:12:49] <alex_joni> yeah
[16:13:04] <skunkworks> would you gain anything?
[16:13:11] <skunkworks> other than complexity ;)
[16:20:27] <jepler> you'd have to talk to someone who uses control theory. :-P if you can run the velocity loop faster (or continuously, as with all-analog amps) it seems like there's going to be some benefit. If you're running them both at the same rate it seems more likely that there's not much benefit (just more tuning parameters).
[16:20:32] <jepler> s/uses/knows/
[16:21:53] <skunkworks> hhe
[16:21:56] <skunkworks> heh
[16:22:40] <archivist> I have the book on the theory but it looks horrid
[16:24:18] <archivist> too much maths and not enough coffee
[16:25:23] <skunkworks> with the masa card you could run the velocity loop at close to base period times..
[16:25:54] <skunkworks> * skunkworks still doesn't like analog
[16:25:58] <jepler> on an unrelated topic, what is the kind of switch used in digital camera shutter releases called? It has two contacts, and the first closes before the second one does.
[16:26:23] <skunkworks> btw - dad found he has an old heathkit analog computer trainer..
[16:27:07] <Optic> hihi
[16:28:22] <archivist> skunkworks, some would kill to get that
[16:28:32] <jepler> I thought I'd found them called "double action switches" before, but that doesn't get me any items on mouser
[16:29:52] <archivist> double throw
[16:30:15] <skunkworks> archivist: ebayable item?
[16:30:35] <skunkworks> multi level
[16:30:37] <skunkworks> ?
[16:31:04] <archivist> skunkworks, sure the classic computers crowd will like it
[16:32:40] <jepler> ah, here are some on digikey. they call them "Off-Mom-Mom" switches
[16:34:01] <archivist> mom is momentary(spring does not latch) if thats what you want
[16:34:27] <jepler> right
[16:39:47] <jepler> probably it doesn't make sense for my application..
[16:41:51] <jepler> the specific AVR I've been working with recently is a bit of a pain to put into programming mode: pull HWB low and then strobe RESET low. My first board design uses two separate tactile switches, one for HWB and one for RESET. I was thinking that one of these dual action switches would make that easier, but then you couldn't use the button to reset without going into bootloader mode..
[16:43:39] <jepler> (you can also jump to the bootloader from your firmware, but that assumes you didn't write a buggy firmware..)
[16:43:47] <archivist> I would expect the programmer to be pulling pins to get into the mode not a switch
[16:46:48] <jepler> those modes are available too, I'm just not using them
[16:47:36] <jepler> this is the DFU mode of at90usb-family chips, which allows a firmware upload over USB. It doesn't replace the low-voltage serial programming, it's another alternative
[16:48:03] <jepler> it's very convenient right up to the moment that you write a firmware that is too buggy to automatically reenter the DFU mode
[16:48:14] <jepler> then you have to press the buttons or use a different programming mode
[16:49:29] <archivist> I use pics and just stuff new code in when its buggy :)
[16:50:17] <jepler> I've only got experience with AVRs, so I dunno why I'd switch
[16:51:04] <archivist> dont switch if you dont need to
[17:04:32] <elmood> either is ok, imho
[17:06:03] <archivist> skunkworks, reply to value of heathkit "<LordNLptp> archivist: if it works, probably a fair bit, but i don't know values offhand"
[17:07:29] <JymmmEMC> skunkworks: If Dad finds a Heathkit Atomic clock, let me know
[17:07:42] <archivist> not a chance
[17:11:36] <JymmmEMC> alex_joni: ping
[19:11:56] <MrSunshine> if anyone know their stepper drivers and how they work here, http://mech.vub.ac.be/teaching/info/mechatronica/finished_projects_2005/PICROCK/l298drv.gif <-- the diodes at the output from th eL298, they arent realy needed right ?
[19:13:04] <cradek> sure they are
[19:13:12] <SWPadnos> energy dump, I think
[19:13:26] <archivist> see data sheet for the L298 , required iirc
[19:13:33] <cradek> see the ... yeah, what he says
[19:14:00] <archivist> else things break
[19:15:14] <motioncontrol> good evennig.one question.when i command m3 s1000 and i have spendle at sppet = false because the red line actual line cnc run not is on m3 s100 partprogram line ?
[19:15:42] <MrSunshine> archivist, seen schemas where they arent used also
[19:15:48] <motioncontrol> excuse for error spindle at speed = false
[19:15:56] <archivist> MrSunshine, it does NOT have internal diodes so you do need them
[19:16:31] <motioncontrol> i use the cvs version
[19:16:32] <MrSunshine> ok :)
[19:16:44] <MrSunshine> have to fins myself some that isnt so damn big as the ones i got home thne
[19:16:52] <MrSunshine> ordered the wrong ones, packages were huge :P
[19:17:03] <cradek> they need to be fast. do not try to use rectifier diodes.
[19:17:36] <MrSunshine> do they need to take alot of amparage also ?
[19:17:47] <archivist> rectifier diodes get very hot when abused
[19:18:29] <MrSunshine> aw, its a rectifier diode ive bought :P
[19:18:32] <SWPadnos> motioncontrol, I don't understand your question. can you say what you are doing, what you expect EMC2 to do, and what you see EMC2 do?
[19:18:45] <MrSunshine> what type of diodes do i need then? :/
[19:18:57] <SWPadnos> (that looked a lot like a line from ELIZA, didn't it? :) )
[19:19:21] <archivist> VF ≤ 1.2 V @ I = 2 A D1 to D8 = 2 A Fast diodes trr ≤ 200 ns
[19:19:31] <motioncontrol> yes i repeat and excuse for my english
[19:19:44] <cradek> MrSunshine: according to your gif, you need FR301 diodes
[19:19:50] <SWPadnos> no problem, just trying to get the question so you can get the answer :)
[19:19:53] <cradek> for more information see the L298 data sheet. you can find it online.
[19:20:17] <motioncontrol> i write i my partprogram go xo after m3 s1000 after m2
[19:20:29] <archivist> google ST L298 and the datasheet is a click away
[19:21:59] <MrSunshine> i guess i could use the ones ive bought anyways, its a prototype card just to test how it works, wont abuse it to much =)
[19:22:36] <SWPadnos> motioncontrol, do you have spindle speed feedback? (or a signal that tells when the spindle is up to speed)
[19:22:37] <cradek> be careful - wrong diodes will cost you the price of a L298 plus the price of the correct diodes...
[19:22:38] <MrSunshine> schcotsky diodes according to datasheet
[19:22:59] <archivist> what diodes did you get
[19:23:27] <MrSunshine> cradek, well they can take 3A 400V
[19:23:28] <MrSunshine> il check name
[19:23:50] <motioncontrol> i start cnc emc2 and the maschine go at xo , on the display the line go xo is red (the actual line read at cnc), when the x axis at 0 position the emc2 read m3 s1000 and the splindle is run.i have low the pin spindle at speed , but on display the red line for actual line cnc read not is presente.because?
[19:23:53] <MrSunshine> 1N5404
[19:24:20] <cradek> motioncontrol: when motion continues, the line will move to the active motion line
[19:24:34] <archivist> MrSunshine, that diode is a slow rectifier
[19:24:36] <cradek> you have asked this before - the answer is that the line only shows lines that cause motion
[19:24:50] <MrSunshine> archivist, was a page i found that he used them on, so i got them :)
[19:25:03] <motioncontrol> no if you have the spindle at speed = false the motion is stop
[19:25:09] <jmkasunich> MrSunshine: not every electronic circuit on the internet is a good circuit
[19:25:26] <cradek> yes motion will wait for the spindle to be at speed
[19:25:34] <motioncontrol> this fuction only in cvs version
[19:25:35] <jmkasunich> first priority should be the datasheets and application notes from the company that makes the chips
[19:25:36] <pjm_> evening all
[19:25:52] <MrSunshine> jmkasunich, nah i know :)
[19:26:00] <cradek> motioncontrol: maybe I still don't understand your question
[19:26:28] <MrSunshine> UF5407 DO-41 800V 3A ultrafast
[19:26:29] <MrSunshine> that one then :P
[19:26:59] <archivist> jmkasunich, the part on his circuit is ok, he bought a substitute
[19:27:08] <jmkasunich> you probably don't need 800V ;-)
[19:27:54] <cradek> I bet you do want more than 100 though - maybe 200
[19:28:01] <motioncontrol> cradeck excuse, nornaly on another cnc, when cnc read the m3 s1000 the cnc wait the spindle revolution ok(spindle at speed = true) , and display at operation m3 s1000 in line evidence
[19:28:12] <MrSunshine> jmk-st, hehe :)
[19:28:15] <MrSunshine> 60V 5A
[19:28:24] <MrSunshine> 40V 5A
[19:28:27] <MrSunshine> i ment
[19:28:53] <SWPadnos> motioncontrol, are you asking if the M3 S1000 line will be hilighted in the G-code window while the spindle is accelerating?
[19:29:18] <motioncontrol> yes the question is correct
[19:29:19] <cradek> motioncontrol: do you have any problem other than the highlighting of the line is missing?
[19:29:19] <archivist> turn off on the UF5407 is faster than required and ok
[19:30:01] <motioncontrol> no only spindle command and g4 p10 command
[19:30:27] <cradek> yes then the answer is the same as before - the lines that do not cause motion do not get highlighting
[19:30:31] <archivist> MrSunshine, UF5404 would be better it has a lower VF
[19:30:52] <MrSunshine> archivist, thing is that i do not what to buy components from 20 diferent places
[19:31:02] <MrSunshine> as i have to order everything
[19:31:27] <motioncontrol> but the m3 s1000 in acceleration is a motion in velocity mode
[19:31:43] <archivist> VF is above the L298 spec for the UF5407
[19:31:55] <SWPadnos> M3S1000 does not cause any axis motion (on XYZABC...)
[19:32:03] <SWPadnos> so it doesn't get hilighted
[19:32:39] <SWPadnos> the motion controller will (optionally) wait until the spindle is at speed before performing axis motion, but that pause is not motion
[19:33:58] <motioncontrol> yes i undurstand , but i thing is correct for spindle acceleration the m3 s1000 is hightlighting
[19:34:20] <cradek> I agree that would be better - unfortunately it's hard to fix
[19:35:02] <MrSunshine> archivist, where do you find a spec of Vf value for it ?
[19:35:37] <MrSunshine> the L298 that is
[19:35:40] <motioncontrol> because for eample i use the pin spindle at speed for feed enable signal low , when the spindle velocity encoder is equql at spindle command i set true the spindle at speed and the motion is on
[19:36:02] <MrSunshine> < 1.2 V ?
[19:36:10] <archivist> yes
[19:36:32] <MrSunshine> found one with 0.57V Vf at the same place i buy everything else, rated to 40V 3A
[19:36:35] <archivist> thats from the L298 datasheet
[19:36:49] <motioncontrol> in the moment the ffed enable is low the cnc is stop on m3 s1000
[19:36:54] <archivist> but is it fast as well
[19:37:14] <MrSunshine> and says "high frequency operation"
[19:38:41] <MrSunshine> what is the speed of them named in the datasheet?
[19:38:45] <MrSunshine> dv/dt?
[19:39:15] <SWPadnos> MrSunshine, download the data sheet and read the number ...
[19:39:18] <archivist> trr
[19:39:39] <MrSunshine> SWPadnos, cant find any number of speeds :)
[19:39:40] <SWPadnos> and then search somewhere like DigiKey for some diodes that have all the right numbers ...
[19:43:42] <JymmmEMC> alex_joni: ping
[19:44:28] <MrSunshine> archivist, is it named in hz etc or what? .. cant find anything more then dv/dt that seems like a speed incidcator of the diode?
[19:45:40] <MrSunshine> no trr rating
[19:45:52] <archivist> trr is the reverse recovery time quoted in ns in the UF5401 series data sheet
[19:46:38] <MrSunshine> archivist, still i dont want to throw away huge money in chipment cause i need to buy diodes from another place then the one i buy everything else on :P
[19:47:16] <archivist> I understand hence look at others, and check makers data
[19:47:27] <MrSunshine> cant find a rating of trr for the diode anywhere
[19:47:40] <MrSunshine> its just a "-" in all places that sells it
[19:47:49] <MrSunshine> 31DQ04
[19:50:08] <MrSunshine> the one that takes 800V 3A has 75nS trr, but 1.7 Vf
[19:50:08] <MrSunshine> :/
[19:50:56] <MrSunshine> Vf was supposed to be low right? :)
[19:51:11] <MrSunshine> "as low as posible at the worst case of the current load"
[19:53:31] <MrSunshine> i dont get it, why doesnt some datasheets specify stuff like trr?
[19:53:42] <alex_joni> JymmmEMC: pong
[19:53:48] <archivist> the diode does not carry the steppers normal forward current only a pulse
[19:54:11] <JymmmEMC> alex_joni: Did you see the electric car?
[19:54:21] <alex_joni> the one from yesterday?
[19:54:25] <JymmmEMC> yeah
[19:56:56] <roell> hi
[19:58:50] <alex_joni> yup, seen it
[19:58:58] <alex_joni> (like I said yesterday ;)
[20:00:56] <skunkworks> roell: Hi
[20:01:46] <JymmmEMC> alex_joni: Well, if you didn't prepend my nick, I wouldn't have known
[20:05:28] <jmkasunich> MrSunshine: the ones that don't specify the trr are the slow ones that you don't want
[20:05:41] <jmkasunich> for 50/60Hz rectifier duty, trr doesn't matter
[20:06:06] <jmkasunich> so any diode with no trr means it was designed for low frequency rectifier duty
[20:06:23] <jmkasunich> diodes for use in switching power supplies will have trr specified
[20:07:26] <alex_joni> 23:24 < JymmmEMC> alex_joni: did you see it?
[20:07:26] <alex_joni> 23:28 < alex_joni> JymmmEMC: shortly
[20:07:57] <JymmmEMC> weird
[20:08:30] <SWPadnos> shortly makes no sense though
[20:08:30] <alex_joni> that's about 23.5h ago
[20:08:45] <alex_joni> who said I had to make sense anyways?
[20:08:50] <SWPadnos> if you had said "JymmmEMC, briefly", then he would have known ;)
[20:09:17] <alex_joni> excuse my crappy english
[20:09:22] <SWPadnos> heh :)
[20:09:39] <alex_joni> I'll take a crash course :)
[20:10:06] <SWPadnos> well, you're better than 93.22% of this country already, so I wouldn't sweat it
[20:10:06] <skunkworks> it is still better than mine..
[20:10:44] <alex_joni> that still means I would probably flunk a 5th grade test in england :)
[20:10:54] <SWPadnos> yes, of course
[20:11:13] <JymmmEMC> 1.7GB ram lol
[20:11:32] <alex_joni> bbl
[20:13:06] <JymmmEMC> No, wait... down to 1.6GB now =)
[20:14:48] <MrSunshine> jmkasunich, the latest one i looked at did not specify and said high frequency
[20:15:14] <jmkasunich> url for datasheet?
[20:15:16] <MrSunshine> according to another channel they dont specify cause schottky is supposed to be so fast it doesnt mater?
[20:15:52] <MrSunshine> http://parts.digikey.ca/1/1/1037228-diode-schottky-40v-5a-smc-mbrs540t3g.html<-- there it says 0ns
[20:15:54] <jmkasunich> you didn't say it was schottky before (or at least, not clearly)
[20:16:21] <jmkasunich> "they" are right - schottkys technically don't have a reverse recovery time, and most are damned fast
[20:16:34] <MrSunshine> ahh, well it says preferably that in the L298 datasheet :)
[20:16:41] <jmkasunich> however, schottkys tend to be low voltage - like the 40V part you just listed
[20:16:58] <MrSunshine> jmkasunich, ye, but its never going to be used above 30 - 35
[20:17:00] <jmkasunich> the diode voltage rating needs to be somewhat more than the maximum supply voltage you will ever use
[20:17:30] <jmkasunich> oh, this isn't gecko class 48-80V drives?
[20:17:41] <MrSunshine> :P
[20:18:03] <jmkasunich> for a 30V supply, a 40V schottky will be OK, IF (and only if) you have a decent layout and bypass capacitors
[20:18:33] <jmkasunich> at high currents, you can get voltage spikes across stray inductance
[20:18:38] <skunkworks> http://www.cnczone.com/forums/showthread.php?t=76233
[20:18:40] <jmkasunich> but the L298 is only 2A, isn't it?
[20:18:45] <skunkworks> does this look right?
[20:18:46] <skunkworks> http://pastebin.ca/1366473
[20:18:49] <MrSunshine> a bypass in parallel with the diodes ?
[20:18:56] <archivist> no
[20:19:04] <jmkasunich> a bypass across the power supply
[20:19:12] <jmkasunich> (near the L298's and the diodes)
[20:19:18] <MrSunshine> jmk-st, rated to 2A per h-bridge
[20:19:29] <MrSunshine> jmkasunich, ye thats what i mean
[20:19:37] <MrSunshine> well this layout i have now doesnt :/
[20:19:44] <MrSunshine> but it was built to find the bugs =)
[20:19:46] <MrSunshine> and i find more and more
[20:20:06] <jmkasunich> lack of bypass capacitors IS a bug!
[20:20:22] <MrSunshine> =)
[20:21:07] <MrSunshine> found a 60V 5A shcottky now, i guess ill go with that one then :)
[20:21:19] <jmkasunich> sounds better
[20:22:22] <skunkworks> jmkasunich: how are the jobs coming?
[20:22:24] <MrSunshine> the bypass capacitor, any minimum uF on that? :)
[20:22:50] <archivist> type is important
[20:23:12] <jmkasunich> skunkworks: hope to finish this weekend
[20:23:23] <skunkworks> jmkasunich: cool :)
[20:23:49] <archivist> ST recommend a 100nf
[20:23:51] <jmkasunich> MrSunshine: you should have a bulk cap - electrolytic, several hundred to a few thousand uF, and high-speed caps, 0.1uF ceramic
[20:24:09] <MrSunshine> hmm ok :)
[20:24:13] <jmkasunich> they bulk cap may already exist as part of your power supply
[20:24:33] <MrSunshine> got a huge one in the power supply im going to use
[20:24:33] <archivist> ST has 470uf on the appnote
[20:24:52] <jmkasunich> I must be channeling ST today
[20:25:28] <MrSunshine> the one in the powersupply is 22000uF
[20:25:45] <skunkworks> too far away
[20:25:55] <MrSunshine> but i should put one by the L298 also then :)
[20:26:14] <jmkasunich> if you have more than a few inches of wire between the supply and the drive, yes
[20:26:30] <MrSunshine> ye :)
[20:27:33] <MrSunshine> the ceramic caps, should they go betwene the outputs (output0 - 1, then output 2 - 3) or should it be output1 -> GND, output2 -> GND etc? :) Seen some datasheets having ceramic caps straight over the supply lines like that (0 - 1, 2 - 3)
[20:27:45] <jmkasunich> no no no
[20:27:53] <jmkasunich> you have no idea what a bypass cap is, do you?
[20:28:10] <jmkasunich> bypass caps are applied between power supply and ground
[20:28:39] <MrSunshine> bypass caps are from VCC to GND to level out the dips in the voltage
[20:28:54] <MrSunshine> or, power or whatever its named :)
[20:35:04] <jmkasunich> yes
[20:35:23] <jmkasunich> as close as possible to the chips and diodes
[20:35:38] <jmkasunich> (sorry for delay, phone)
[20:36:25] <MrSunshine> ok :)
[20:36:40] <MrSunshine> have to see what i can do =)
[20:38:38] <MrSunshine> jmkasunich, humm, one more more ceramic caps ? ... more and it just adda sto capacitance right ?
[20:41:43] <SWPadnos> different value (and different material) caps are used for different purposes
[20:42:05] <SWPadnos> for bulk caps (the big electrolytics in the power supply), you just add them together
[20:42:45] <SWPadnos> for filtering higher frequencies, you use a small value cap, ideally with low ESR (like a tantalum or electrolytic)
[20:42:58] <SWPadnos> (these are usually in the 0.1 uF range)
[20:43:11] <jmkasunich> multiple caps in parallel add
[20:43:21] <jmkasunich> if you use 4 caps of 0.1uF each, you will get 0.4uF
[20:43:37] <jmkasunich> but the four caps (if properly located) will have less lead inductance than one larger cap
[20:43:52] <SWPadnos> right
[20:44:03] <jmkasunich> that is why you use the small ceramic caps in the first place - the large electrolytic cap has too much inductance to be usefull for very high frequencies
[20:44:24] <SWPadnos> for filtering, you can't just look at the totals and be happy, you have to look at layout and other things that are often not considered
[20:44:27] <jmkasunich> and the small ceramic caps don't have enough capacitance to be usefull for low frequencies
[20:47:08] <MrSunshine> reboot ... computer acting up )
[20:47:09] <MrSunshine> :)
[21:55:57] <MrSunshine> there, 1 big elyt, new diodes, 2 ceramic 0.1uF capacitors added .. anything more? :)
[22:01:34] <alex_joni> fume extraction
[22:01:41] <alex_joni> you need to catch the magic smoke
[22:02:04] <alex_joni> OT: I was wondering what google was going to come up with next: http://www.theregister.co.uk/2009/03/20/gmail_undo_email_option/
[22:02:12] <toastatwork> they sell glass devices you can fill with water to catch that
[22:04:32] <MrSunshine> alex_joni, hehe if there comes magic smoke out of it i guess ive done something wrong ;P
[22:08:29] <ds3> generally speaking, tiny endmills are relatively strong for plunge operations, right?
[22:13:38] <pjm_> ds3 they are ok but for 'nice' holes your machine has to be very rigid or they have a tendancy to wobble a little
[22:14:59] <Guest374> they need to be senter cutting
[22:15:05] <Guest374> Guest374 is now known as skunkkworks
[22:15:12] <skunkkworks> center
[22:15:36] <Acidshock> hello everyone! anyone have time to answer what I hope will be a quick question?
[22:15:37] <alex_joni> skunkkworks: we know you can't spell ;)
[22:15:44] <alex_joni> Acidshock: don't ask to ask.. just ask
[22:18:34] <ds3> Hmmm
[22:18:45] <Acidshock> ok... well I am trying to mill out a pcb board and the program I use exports to ISO gcode. I believe its trying to quest a tool change(T4 M06) EMC2 seems to not like this. Is there any way to make it prompt for a tool change?
[22:18:50] <ds3> in that case, I wonder how quick will I blow through 1/8" EM's in ABS
[22:19:06] <alex_joni> emc2 surely likes T4 M6
[22:19:34] <Acidshock> wells its telling me invalid gcode at line 5, which is that
[22:19:36] <alex_joni> depending how you set up your config, it's possible to get a prompt for a tool change
[22:19:47] <alex_joni> anything else on that line?
[22:20:05] <alex_joni> maybe it doesn't like M06 .. try changing it to M6
[22:20:15] <Acidshock> nope... however it is being save on a windows machine.. Maybe trailing carriage return?
[22:20:22] <alex_joni> no, that should be fine
[22:20:29] <Acidshock> ok I can try that
[22:20:35] <alex_joni> and even if it wasn't.. it shouldn't go to line 5 ;)
[22:20:55] <Acidshock> Am I right in thinking that is a tool change? I am not familiar with Gcode but it seems logical
[22:21:14] <Acidshock> tool change/specification
[22:22:00] <alex_joni> T4 specifies the tool to be loaded next
[22:22:05] <alex_joni> and M6 is the actual toolchange
[22:22:12] <alex_joni> Txx stands for tool prepare
[22:22:13] <Acidshock> coolk
[22:22:15] <Acidshock> err cool
[22:22:25] <Acidshock> ok im gonna test it out really quick.
[22:22:28] <Acidshock> Thanks alex!
[22:24:15] <alex_joni> Acidshock: no problem, hope it works
[22:24:27] <alex_joni> if it doesn't put the first 10 lines or so online somewhere
[22:24:36] <alex_joni> maybe on pastebin.ca, so we can take a looksy
[22:52:44] <Acidshock> hey alex... no luck :(
[22:52:48] <Acidshock> I did post the code
[22:52:49] <Acidshock> http://www.pastebin.ca/1366614
[22:52:59] <Acidshock> well some of it
[22:53:02] <Acidshock> lol first few lines
[22:55:46] <Acidshock> emc says unknown g code at line 5
[22:55:57] <Acidshock> with it being m06 or m6
[22:59:31] <jepler> Acidshock: the unknown code is G71 on line 4
[23:00:21] <jepler> if I remove it and add M2 as the last line, it loads in my emc2.2.8 using configs/sim -- it looks like two circles that overlap partially
[23:00:39] <jepler> if the distance units are millimeters, add G21 before any motions in your file
[23:01:26] <jepler> the internet suggests that some control treats g71 and g21 the same, but in emc g71 is an undefined code. use g21 for mm.
[23:03:38] <Acidshock> thanks jepler! now I am not very familiar with gcode. I am relatively new to the processes :) Where would I add this G21 again? Is it just in that 4th line or before each line instead of the g01 and g00?
[23:04:45] <jepler> Acidshock: I think you want to put G21 right where G71 is now
[23:05:01] <jepler> just once will do
[23:06:23] <jepler> and add M2 after the last line -- emc requires M2 at the end of all files (*slight oversimplification, but treat it as a rule anyway)
[23:07:33] <Acidshock> will do! thanks so much again. I also found an article on wiki talking about the change, etc. I will read up on this to try to prevent further questions like this
[23:08:47] <jepler> best of luck
[23:09:31] <jepler> this is the most helpful single page to refer to when trying to figure out what is and is not in emc2's gcode: http://linuxcnc.org/docs/html/gcode.html
[23:09:58] <jepler> bbl
[23:26:22] <alex_joni> good night all
[23:27:21] <roell> night alex_joni