#emc | Logs for 2007-09-22

[00:00:40] <SWPadnos> ok - a tiny bit then stop is OK. but 2.5 seconds is a long period
[00:01:00] <Adam1> if I crank the P over 400 I can induce a crazy oscillation
[00:01:29] <fenn> dont do that with I :)
[00:01:34] <Adam1> I should note that this P value is oscillating when no motion is induced
[00:01:39] <SWPadnos> just to be sure, all other values are 0, right? (I, D, FF0, FF1, FF2)
[00:01:47] <Adam1> it just does when i start it the drives up
[00:01:54] <Adam1> all other values are 0
[00:03:07] <Adam1> So why does the oscialltion period seem so long
[00:03:17] <Adam1> should it be really quick
[00:03:27] <Adam1> or typically really quick
[00:04:09] <SWPadnos> it shouldn't necessarily be really fast, but from the little experience I have, 2.5 seconds seems long
[00:04:26] <SWPadnos> how far is it moving?
[00:04:47] <Adam1> maybe 1-2 degrees
[00:04:57] <SWPadnos> the motor shaft?
[00:05:19] <Adam1> yeah but let me check it is incased and I am basing that off a ratio in my head and viewing the handwheel
[00:05:44] <SWPadnos> is the table moving noticeably?
[00:05:49] <SWPadnos> * SWPadnos thinks the answer is no
[00:06:13] <Adam1> well my scales say .0004 inches in x and y
[00:06:24] <SWPadnos> ok, that's not the oscillation you're looking for :)
[00:06:50] <SWPadnos> also, only do one axis at a time
[00:07:18] <SWPadnos> incidentally, I don't think you can use this method if there's backlash
[00:07:32] <SWPadnos> but I could be wrong about that
[00:08:02] <Adam1> SWPadnos the servo shaft rotates about 7-8 degrees
[00:08:30] <SWPadnos> what's the final ratio motor-> table? 2:1 belt then 5 TPI screw?
[00:09:40] <JymmmEMC> does a 1F cap have an equiv mAh rating?
[00:09:45] <Adam1> 1 servo rev gets me 0.133in on the table
[00:10:59] <SWPadnos> uh - is it some more even number of mm?
[00:11:05] <fenn> JymmmEMC: 0.27mAh/V
[00:11:14] <SWPadnos> heh
[00:11:25] <JymmmEMC> fenn: really?
[00:11:34] <fenn> really. but i dont think that's what you wanted to know
[00:11:54] <fenn> the energy stored in a cap is 0.5CV^2
[00:12:20] <JymmmEMC> fenn: Sorta... I have solar light (LED), but I was wondering if I could replace the dead batteries with a 1F cap instead.
[00:12:30] <Adam1> 21tooth on screw 14tooth on servo, 5 tpi screw
[00:12:52] <fenn> try it an see
[00:13:35] <JymmmEMC> fenn: My faraday flashlight has a super cap in it. so I wasn't sure.
[00:14:17] <SWPadnos> Adam1, I think the motor is bouncing around inside the machine's backlash
[00:14:31] <SWPadnos> 8 degrees at 0.133 inch/rev = ~0.003
[00:14:35] <JymmmEMC> fenn: http://www.neonixie.com/store/product_info.php?products_id=261
[00:14:45] <Adam1> yeah...
[00:15:11] <SWPadnos> if the table only moves 0.0004, then the motor isn't really moving it
[00:15:15] <SWPadnos> much
[00:15:35] <fenn> jymmm for reference, a rechargeable AA has about 10KJ energy storage, whereas a 1F 12V cap has 72J
[00:15:45] <Adam1> I think there is about 0.0036in backlash
[00:16:10] <fenn> 5v -> 12J
[00:16:13] <Adam1> in between 3 and 4 thou
[00:16:25] <SWPadnos> ok. I'm not sure what to sugges.
[00:16:27] <JymmmEMC> fenn: The issue is that the batteries being solar charged reduce their overall life. Where as a cap is used to being charged/discharged.
[00:16:35] <SWPadnos> so I think I'll go buy some donuts - bbiab
[00:16:37] <SWPadnos> :)
[00:16:40] <Adam1> :(
[00:16:44] <Adam1> am I screwed
[00:16:58] <fenn> no, you just have to tune it by hand
[00:17:07] <SWPadnos> I don't think so, you just may not be able to use this method to tune, or someone who has actually tuned a machine can help out :)
[00:17:24] <SWPadnos> I've only been involved in tunning one machine, and I didn't do most of the work
[00:17:34] <SWPadnos> now, it's donut time
[00:17:40] <fenn> Adam1: plot position-fb on halscope
[00:17:44] <JymmmEMC> bu bye homer
[00:17:52] <SWPadnos> D'Oh!
[00:18:59] <Adam1> I havent exactly played with halscope
[00:22:36] <Adam1> What is the position-fb relative too?
[00:22:45] <Adam1> first off I dont have that
[00:23:00] <Adam1> pid.feedback?
[00:27:29] <Adam1> I think I am going to come back on monday and focus on this issue.
[00:27:40] <Adam1> have a great weekend all
[00:28:50] <eric_u> monday is a good day to work on EMC, because your boss is paying you to do it
[00:56:07] <skunkworks> he could always mount encoders on the leadscrew.
[00:56:14] <skunkworks> if all else fails
[00:57:05] <vdKroon> Hi again, maybe any of you can point me to the right direction..
[00:57:19] <skunkworks> * skunkworks points north east
[00:57:26] <vdKroon> I've setup ini file with (I guess) correct values of speed/accel/scale
[00:57:30] <toast> an excellent direction by ay standard
[00:57:37] <vdKroon> but axis are movin slow/erratic so far
[00:58:21] <vdKroon> I have the same config as when I was in mach2, as far as speed, steps/units
[00:59:47] <vdKroon> any ideas on what am i doing wrong
[00:59:49] <vdKroon> ?
[01:00:07] <toast> i don't know anything about emc.
[01:01:28] <SWPadnos> what do you mean by erratic movement?
[01:01:33] <skunkworks> can you pastebin.ca your ini file?
[01:02:11] <vdKroon> uhm, don't know how to tell in english, but, nothing smooth
[01:02:29] <SWPadnos> is this at all speeds, or only at certain speeds?
[01:03:14] <vdKroon> all speeds.. but, the fact is that, when supposing to jog at 2400 mm/min, the travel is much less than that
[01:03:46] <SWPadnos> ok. please put your ini file on http://pastebin.ca
[01:04:23] <vdKroon> I have a scale of 17.777 (800 steps / 45 => resulting of a pulley with 15 teeth, 3 mm pitch)
[01:04:34] <vdKroon> ok, just a moment, i'm on another computer right now
[01:04:54] <SWPadnos> I don't think you should divide by 45
[01:07:10] <skunkworks> so every rotaion moves the slide 45 mm - correct. So yes - I would think 800/45 but I might be wrong :)
[01:07:26] <SWPadnos> no, 3mm pitch means the slide moves 3mm
[01:07:45] <SWPadnos> unless this is a rack/pinion (I may have missed that)
[01:07:46] <vdKroon> sorry
[01:07:57] <vdKroon> I mean 3 mm between tooth
[01:08:09] <vdKroon> anyway, 17.777 is the same value I had with mach, working ok
[01:08:18] <SWPadnos> ok
[01:08:35] <SWPadnos> what acceleration and velocity limits did you have with Mach?
[01:09:09] <skunkworks> what drives are you using?
[01:09:26] <skunkworks> (maybe there needs to be steplen and such setup)
[01:10:04] <vdKroon> uhm, the drivers are based on PIC, it's a controller named 'C4' by a spanish manufacturer
[01:10:33] <vdKroon> with mach I was at 50 mm/s velocity and accel of 60 mm/s*s
[01:10:48] <SWPadnos> hmmm
[01:11:19] <vdKroon> going to paste ini file
[01:11:19] <SWPadnos> if you put your ini file on pastebin, we can look at it :)
[01:11:21] <SWPadnos> ok
[01:11:37] <toast> * toast waits for the deluge of text
[01:11:48] <jlmjvm> 2*60
[01:12:00] <SWPadnos> =120
[01:12:05] <jlmjvm> sorry
[01:12:14] <toast> irccalc
[01:12:24] <toast> "it may be wrong, but at least it's interactive!"
[01:12:25] <SWPadnos> !EINVAL
[01:12:42] <toast> i like my ti-89
[01:16:32] <jlmjvm> SWPadnos:could my poor latency numbers be because i have 1 onboard parport and 1 pci parport?
[01:16:40] <SWPadnos> not likely
[01:16:53] <SWPadnos> the latency test is run without accessing the parport
[01:17:00] <vdKroon> here it is: http://www.pastebin.ca/706579
[01:17:02] <jlmjvm> k
[01:18:44] <vdKroon> sorry
[01:18:53] <vdKroon> that was not the good ini
[01:18:58] <SWPadnos> oh
[01:19:01] <vdKroon> is this: http://www.pastebin.ca/706581
[01:19:26] <vdKroon> sorry for that
[01:19:40] <SWPadnos> no problem
[01:20:51] <SWPadnos> hmmm. if the hardware can handle 50mm/s vel and 60 mm/s^2 accel, then I don't see any problem with the ini
[01:20:57] <SWPadnos> dod you run a latency test?
[01:21:00] <SWPadnos> did
[01:21:03] <vdKroon> uhm, no
[01:21:19] <vdKroon> how can I do that?
[01:21:46] <SWPadnos> ok. the only thing that I can think of at the moment is to reduce the bASE_PERIOD to something like 25000 or 20000, but you don't want to do that unless your PC can handle it
[01:22:07] <vdKroon> it's a 1800 Athlon
[01:22:20] <vdKroon> (I think)
[01:22:49] <vdKroon> should I try that?
[01:22:49] <SWPadnos> it can probably handle 20000 or lower then - try 25000 and see if the GUI seems to get "sluggish"
[01:23:07] <SWPadnos> sure - the worst that can happen is your PC will hard-lock, requiring a reset :)
[01:23:22] <vdKroon> ok, going to try, brb
[01:32:59] <vdKroon> oooook
[01:33:01] <vdKroon> solved
[01:33:09] <SWPadnos> yay!
[01:33:11] <vdKroon> it's not the PC but the driver
[01:33:13] <SWPadnos> heh
[01:33:17] <SWPadnos> loose wires?
[01:33:22] <vdKroon> nope
[01:33:41] <vdKroon> tried at 20000 -> problem get worse
[01:33:49] <vdKroon> tried at 150000 -> problem solved
[01:34:04] <vdKroon> I think driver can't handle speed
[01:34:07] <skunkworks> I bet it is steplen and such problem..
[01:34:27] <vdKroon> now it runs smooth and max speed
[01:34:27] <skunkworks> his input scale is so low that it will probably run at that.
[01:34:44] <SWPadnos> yep. it's likely that you need to set the step length and step space for your driver
[01:34:58] <skunkworks> Did you say what drivers your using?
[01:35:02] <SWPadnos> skunkworks, it's only ~800 Hz, but the resolution still needs to be better than that
[01:35:20] <SWPadnos> "C4" by a spanish company
[01:35:31] <vdKroon> All that I have is a schematic
[01:35:38] <skunkworks> huh
[01:35:51] <SWPadnos> is there any information online?
[01:36:02] <vdKroon> In spain, electronic manufacturers generally suck
[01:36:09] <vdKroon> yep. let me see
[01:36:34] <SWPadnos> or, does the driver use some chip to do the work (like an L298 or Allegro 3977(?) )
[01:36:58] <vdKroon> pfffffff nop, just PICS
[01:36:59] <vdKroon> http://www.cenece.com/images/ElecC42.jpg
[01:37:14] <SWPadnos> oh
[01:37:25] <vdKroon> http://www.cenece.com/ceneceC4.htm
[01:37:43] <vdKroon> this driver came from a friend
[01:37:53] <vdKroon> I think it's time to trash it
[01:38:11] <SWPadnos> one sec. let me check out the PIC source code. I may be able to tell you what to set steplen and stepspace to
[01:38:19] <vdKroon> wow, thanks
[01:38:28] <SWPadnos> oh. nevermind. it's hex, not source
[01:38:39] <SWPadnos> and I'm not sure I have a disassembler
[01:39:06] <vdKroon> yep, it's hex
[01:39:36] <SWPadnos> hmmm. 4 MHz crystals
[01:40:02] <SWPadnos> that means 1 microsecond internal clock, and the interrupt latency is 4 cycles IIRC
[01:41:25] <SWPadnos> actually, it didn't work well at 50 microsecond BASE_PERIOD. try to find out at what setting it starts to sound good
[01:41:52] <skunkworks> that would be a good test to figure out minimim steplength
[01:42:40] <SWPadnos> or stepspace
[01:42:52] <vdKroon> ok, should I try that?
[01:42:57] <skunkworks> right
[01:43:06] <vdKroon> going down from 100000 until problem occurs?
[01:45:03] <SWPadnos> yes
[01:45:08] <SWPadnos> try it in steps of 10000 or so
[01:45:29] <SWPadnos> (or do a binary search, if you're into that :) )
[01:47:05] <skunkworks> I suppose someone could do a search and see what mach uses for a default stepspace.
[01:47:17] <SWPadnos> or see what he was using...
[01:47:25] <SWPadnos> I think there are settings for that
[01:55:12] <toast_> thundercats.
[01:59:42] <toast_> toast_ is now known as toastydeath
[02:02:19] <vdKroon> ok
[02:02:35] <vdKroon> heh, fact is that nit was setup at 100000 but 200000
[02:02:40] <vdKroon> *not
[02:02:53] <vdKroon> I lowered until 180000
[02:03:00] <vdKroon> at 170000 problem appears again
[02:06:18] <SWPadnos> wow. that's a very very long time
[02:06:28] <vdKroon> :/
[02:06:43] <SWPadnos> did you have to set Mach to hold the step output for longer than default (or to have longer spacing between steps)?
[02:07:02] <vdKroon> yes, I think at maximum..
[02:07:12] <vdKroon> 15 uS, can be?
[02:07:14] <SWPadnos> do you recall what that value is
[02:07:17] <vdKroon> don't remember
[02:07:18] <SWPadnos> no, longer
[02:07:26] <SWPadnos> sounds more like 200 micoseconds
[02:07:28] <SWPadnos> micro
[02:08:17] <SWPadnos> if 170000 sounds good, then you can work with that
[02:08:19] <vdKroon> I think mach stated that 'values over 15 - danger of lockup' or somethign like that
[02:08:23] <SWPadnos> heh
[02:08:30] <SWPadnos> EMC doesn't suffer from that problem :)
[02:08:49] <SWPadnos> I think the longest you can set is about 4 seconds though, so there is a limit
[02:08:53] <vdKroon> Ok, but, sincerely, better change driver?
[02:09:04] <SWPadnos> you don't need to, if it makes the motors move
[02:09:20] <vdKroon> uhm, yeh it does
[02:09:32] <vdKroon> I was thinking about getting one with microstep
[02:09:39] <SWPadnos> that would probably improve things somewhat
[02:09:49] <vdKroon> this is full/half step
[02:09:54] <SWPadnos> you certainly aren't limited by the step rate the PC can generate
[02:11:31] <SWPadnos> so, you have the option of adding some lines to the ini file, or changing one of the hal files. which would you prefer?
[02:11:55] <vdKroon> what's best for the system?
[02:12:14] <SWPadnos> either. it may be easierto upgrade if you only make changes to the ini file
[02:12:26] <SWPadnos> since there will only be one file to fix up
[02:12:40] <vdKroon> Ok
[02:13:19] <vdKroon> ini file then
[02:13:41] <SWPadnos> if you look at the [HAL] section of the ini file, you'll probably see a line something like this:
[02:13:54] <SWPadnos> # HALCMD = save neta
[02:14:18] <SWPadnos> the commands I'm going to tell you need to go in the same section
[02:14:23] <vdKroon> Ok
[02:14:46] <vdKroon> found it
[02:15:51] <SWPadnos> first, I'm going to have you add a variable to the file, so it's easier to change all of the step lengths or spaces later
[02:16:09] <vdKroon> perfect
[02:16:12] <SWPadnos> add this line to the [HAL] section:
[02:16:21] <SWPadnos> STEPLEN = 180000
[02:16:28] <SWPadnos> STEPSPACE = 180000
[02:16:35] <SWPadnos> oops - these lines :)
[02:16:38] <SWPadnos> then:
[02:16:51] <vdKroon> done
[02:17:06] <SWPadnos> HALCMD = setp stepgen.0.steplen [HAL]STEPLEN
[02:17:14] <SWPadnos> HALCMD = setp stepgen.0.stepspace [HAL]STEPSPACE
[02:17:25] <SWPadnos> repeat those last two for stepgen.1 and stepgen.2
[02:17:59] <SWPadnos> change the BASE_PERIOD back to something between 25000 and 50000
[02:18:42] <SWPadnos> so you should have added 8 lines to the HAL section, and changed the BASE_PERIOD
[02:19:23] <vdKroon> correct
[02:19:32] <vdKroon> don
[02:19:33] <vdKroon> one
[02:19:34] <vdKroon> done
[02:19:34] <SWPadnos> ok
[02:19:37] <SWPadnos> heh
[02:19:45] <vdKroon> :)
[02:19:54] <vdKroon> should I try now?
[02:20:03] <SWPadnos> you should be able to see which is the critical timing parameter by changing only one of STEPLEN or STEPSPACE
[02:20:05] <SWPadnos> sure
[02:20:12] <vdKroon> ok
[02:20:22] <vdKroon> ok, going to try that
[02:20:45] <SWPadnos> if you set one of them to 50000, you'll see if it's the problem (or one of the problems, they may both have issues)
[02:21:03] <vdKroon> ok, brb (the machine & pc are in another room)
[02:21:08] <SWPadnos> ok
[02:27:39] <tomp> is that 4 pcs 16c84? and is the input step & direction?
[02:28:11] <vdKroon> wow, kind strange
[02:28:14] <vdKroon> no effect at all
[02:28:25] <vdKroon> same problem if base_period is not raised
[02:28:26] <tomp> must be, and the pic drives the h bridge
[02:28:38] <SWPadnos> tomp, yep - it's a unipolar drive too
[02:28:52] <SWPadnos> hmmm
[02:29:09] <SWPadnos> I wonder if those hal commands worked right
[02:29:15] <vdKroon> tried lenght/space between 50000 and 200000, all combinations
[02:29:32] <vdKroon> if base is not raised to 180000 => same result
[02:30:02] <SWPadnos> can you make sure the step length and space are correct?
[02:30:17] <vdKroon> yes, they are, I doublechecked it
[02:30:20] <SWPadnos> hmmm
[02:30:27] <vdKroon> (or quedchecked) ;)
[02:30:33] <vdKroon> *quad
[02:31:01] <SWPadnos> heh
[02:31:11] <SWPadnos> you checked with halcmd?
[02:31:23] <vdKroon> umpf., nop
[02:31:26] <SWPadnos> ok :)
[02:31:27] <vdKroon> how to do that?
[02:31:32] <vdKroon> [HAL]
[02:31:32] <vdKroon> STEPLEN = 180000
[02:31:32] <vdKroon> STEPSPACE = 180000
[02:31:32] <vdKroon> HALCMD = setp stepgen.0.steplen [HAL]STEPLEN
[02:31:32] <vdKroon> HALCMD = setp stepgen.0.stepspace [HAL]STEPSPACE
[02:31:32] <vdKroon> HALCMD = setp stepgen.1.steplen [HAL]STEPLEN
[02:31:34] <vdKroon> HALCMD = setp stepgen.1.stepspace [HAL]STEPSPACE
[02:31:36] <vdKroon> HALCMD = setp stepgen.2.steplen [HAL]STEPLEN
[02:31:38] <vdKroon> HALCMD = setp stepgen.3.stepspace [HAL]STEPSPACE
[02:31:44] <vdKroon> sorry, las line was corrected
[02:31:58] <vdKroon> STEPLEN = 180000
[02:31:58] <vdKroon> STEPSPACE = 180000
[02:31:58] <vdKroon> HALCMD = setp stepgen.0.steplen [HAL]STEPLEN
[02:31:58] <vdKroon> HALCMD = setp stepgen.0.stepspace [HAL]STEPSPACE
[02:31:58] <vdKroon> HALCMD = setp stepgen.1.steplen [HAL]STEPLEN
[02:31:59] <vdKroon> HALCMD = setp stepgen.1.stepspace [HAL]STEPSPACE
[02:32:01] <vdKroon> HALCMD = setp stepgen.2.steplen [HAL]STEPLEN
[02:32:03] <vdKroon> HALCMD = setp stepgen.2.stepspace [HAL]STEPSPACE
[02:32:08] <SWPadnos> halcmd show param | grep steplen
[02:32:14] <SWPadnos> and
[02:32:19] <SWPadnos> halcmd show param | grep stepspace
[02:32:32] <SWPadnos> from a terminal, while EMC is running
[02:32:39] <vdKroon> ok
[02:32:39] <JymmmEMC> 70 minutes till quiche is done
[02:32:46] <SWPadnos> I had a dnout
[02:32:48] <SWPadnos> donut
[02:32:53] <JymmmEMC> D'Oh
[02:33:01] <SWPadnos> D'Oh-nut
[02:33:05] <JymmmEMC> SWPadnos: Wifey Away?
[02:33:19] <SWPadnos> camping with a friend. I'm supposed to be working
[02:33:37] <JymmmEMC> You're IRCing while camping?!
[02:33:54] <JymmmEMC> A sign of a true geek
[02:34:01] <SWPadnos> I am *NOT* camping!
[02:34:03] <SWPadnos> :)
[02:34:19] <JymmmEMC> oh the eife is camping, gotcha
[02:34:21] <JymmmEMC> wife
[02:34:42] <SWPadnos> right-o
[02:34:51] <JymmmEMC> You dont camp huh?
[02:34:56] <vdKroon> result:
[02:34:59] <SWPadnos> not too often
[02:35:15] <vdKroon> 05 U32 RW 00000001 stepgen.0.steplen
[02:35:24] <vdKroon> 05 U32 RW 00000001 stepgen.1.steplen
[02:35:27] <vdKroon> 05 U32 RW 00000001 stepgen.2.steplen
[02:35:36] <vdKroon> same thing for stepspace
[02:35:45] <SWPadnos> hmmm
[02:35:53] <SWPadnos> I should probably look at the units for those numbers :)
[02:36:06] <vdKroon> lol
[02:36:07] <JymmmEMC> metric tones
[02:36:37] <vdKroon> that should reflect the value?
[02:36:51] <cradek> SWPadnos: nsec I thought
[02:37:10] <SWPadnos> that's what I thought too
[02:37:28] <cradek> HALCMD = ??
[02:37:31] <SWPadnos> I don't see why those commands wouldn't work
[02:37:34] <cradek> what's supposed to run those?
[02:37:44] <SWPadnos> they should get run by the run script
[02:37:53] <cradek> oh really
[02:37:53] <SWPadnos> I think
[02:38:09] <SWPadnos> look at all the stepper/ ini files - they have one line commented out
[02:38:17] <SWPadnos> #HALCMD = save neta
[02:38:25] <vdKroon> yes, it's commented
[02:38:33] <cradek> you're assuming a lot methinks
[02:38:36] <SWPadnos> heh - could be
[02:39:02] <cradek> oh ok I see where that's supposed to happen
[02:39:10] <cradek> echo "Running HAL command: $HALCOMMAND" >>$PRINT_FILE
[02:39:16] <cradek> you should see these messages in the debug output
[02:39:30] <SWPadnos> well, the easy way around that problem is to move those lines into the core_stepper.hal file
[02:39:52] <vdKroon> ok, going to do that
[02:39:58] <SWPadnos> but
[02:40:00] <vdKroon> yes?
[02:40:16] <SWPadnos> don't move the STEPLEN= and STEPSPACE= lines
[02:40:21] <vdKroon> ok
[02:40:25] <SWPadnos> and also remove "HALCMD=" from the other lines
[02:40:28] <vdKroon> I left them at ini file?
[02:40:33] <vdKroon> *leave
[02:40:34] <SWPadnos> just the setp command itself
[02:40:46] <SWPadnos> no, I'd take them out of the ini, it'll be confusing later if you leave them in
[02:41:11] <vdKroon> so, this syntax: setp stepgen.0.steplen [HAL]STEPLEN
[02:41:13] <vdKroon> on hal file
[02:41:29] <SWPadnos> yes, one sec and I'll see where it would make sense to put them
[02:41:33] <vdKroon> ok
[02:41:43] <SWPadnos> ah - at the end
[02:41:54] <vdKroon> ok, and where to put values?
[02:42:01] <SWPadnos> there are pretty obvious groups of stepgen setup lines there
[02:42:09] <SWPadnos> the values stay in the ini file
[02:42:15] <vdKroon> ok
[02:42:18] <jepler> In emc 2.1.7, stepgen.X.steplen and other timings are "in make-pulses periods", so 1 means "one BASE_PERIOD" 2 means "2 * BASE_PERIOD", and so on
[02:42:22] <SWPadnos> those are the STEPLEN= and STEPSPACE= lines
[02:42:30] <SWPadnos> oh, that could be a problem
[02:42:37] <vdKroon> i'm on 2.1.6
[02:42:41] <jepler> same in 2.1.6
[02:42:43] <SWPadnos> same on 2.1.6
[02:43:07] <SWPadnos> oh - oops. I probably sourced emc-environemnt when I checked the manpage for that
[02:43:16] <SWPadnos> ^had
[02:43:58] <SWPadnos> so, change the BASE_PERIOD to 25000
[02:44:02] <vdKroon> ok
[02:44:09] <SWPadnos> and change the two numbers from 180000 to 8
[02:44:23] <vdKroon> and move the six lines to hal file
[02:44:26] <SWPadnos> and you can then check to see what happens as you reduce them down to 2
[02:44:37] <SWPadnos> hmmm
[02:44:39] <vdKroon> or not?
[02:44:39] <SWPadnos> one sec
[02:44:48] <jepler> goodnight all
[02:44:53] <vdKroon> maybe just with this *new* units may work
[02:45:00] <vdKroon> goodnigh jepler
[02:45:04] <vdKroon> thanks for help
[02:45:27] <SWPadnos> see you jepler
[02:45:42] <SWPadnos> actually, yes, just move those setp commands
[02:45:53] <SWPadnos> then there's no question that they'll work
[02:46:11] <vdKroon> ok, BASE_PERIOD to 25000
[02:46:18] <vdKroon> two numbers from 180000 to 8
[02:46:26] <vdKroon> six lines to end of HAL file
[02:46:27] <vdKroon> correct?
[02:47:02] <SWPadnos> yep
[02:47:04] <vdKroon> ok, brb
[02:52:28] <tomp> looks like the units are for hot wire foam cutters. thats getting popular, like there's some big need foam airfoils
[02:53:42] <vdKroon> yes, the C4 was intended for foam cuters
[02:54:10] <vdKroon> SWPadnos: That worked
[02:54:34] <vdKroon> Leaving BASE_PERIOD at 25000, I can change LEN/SPACE to 7
[02:57:46] <SWPadnos> great
[02:59:07] <tomp> vdKroon: are your kinematics for xyuv? i that how you program it?
[02:59:49] <vdKroon> xyuv?
[02:59:58] <SWPadnos> xyz
[03:00:01] <tomp> what do you call your axis?
[03:00:10] <vdKroon> uhm, yeah xyz
[03:00:12] <vdKroon> but
[03:00:19] <vdKroon> Just tw days with EMC
[03:00:25] <vdKroon> no idea about kinematics
[03:00:40] <vdKroon> really no much idea about cnc overall :/
[03:00:46] <vdKroon> can you explain a bit?
[03:01:22] <tomp> a wings surface can be traced by rolling a pencil over its surface.
[03:01:38] <tomp> the tip of the pencil traces a different path than the other end
[03:01:47] <tomp> the tip might be called xy
[03:01:57] <tomp> the other end might be called uv
[03:02:06] <tomp> make sense?
[03:02:19] <vdKroon> oh, ok, but I'm not on foam cutting
[03:02:32] <tomp> ah, (never mind :)
[03:02:50] <vdKroon> the C4 was originally build for foam cutters
[03:03:10] <vdKroon> but after some time, they started to sell it as mill driver
[03:03:32] <vdKroon> as time goes by, I more realize that is not a good driver for this application
[03:03:35] <tomp> i see, you use it more general
[03:03:45] <vdKroon> I got it from a friend when just bought the machine
[03:04:25] <vdKroon> I use it on a gantry-style milling machine
[03:04:47] <vdKroon> but I think i'm going to move to another driver
[03:05:43] <tomp> it used irf540's , what voltage do you supply?
[03:07:56] <vdKroon> ahm, the supply also came with controller, 8V 12A + 5V
[03:08:06] <vdKroon> I think it's not good either
[03:13:10] <tomp> bye
[03:13:32] <vdKroon> Ok, have to go sleep
[03:13:35] <vdKroon> 5:13 am here
[03:13:38] <SWPadnos> night
[03:13:44] <tomp> vdKroon: best of luck , gnite
[03:13:46] <SWPadnos> Germany?
[03:13:53] <vdKroon> Spain ;)
[03:13:57] <SWPadnos> ah,ok :)
[03:14:02] <vdKroon> although Dutch in origin
[03:14:08] <SWPadnos> cool
[03:14:18] <vdKroon> Thank you all for your support
[03:14:30] <SWPadnos> you're welcome. good luck with the machine
[03:14:36] <vdKroon> thank you very much
[03:14:42] <vdKroon> cya!
[03:14:47] <SWPadnos> bye
[03:15:25] <tomp> an idea, the foam cutter had a temperature control unit, must have some feedback... if there's a lag in the temperature response, then adaptive feed could slow the rate to compensate
[03:15:36] <tomp> feed rate
[03:15:49] <SWPadnos> sure
[03:19:01] <tomp> we were looking at a book over on #cam, it's listed in a few places on the web > 200$.... but the author sells it as e-book for 25$ :) http://www.amazon.com/Python-Tkinter-Programming-John-Grayson/dp/1884777813 versus http://manning.com/grayson/
[03:21:59] <JymmmEMC> lag in temp response?!
[03:23:09] <JymmmEMC> MORE WATTS!!!
[03:23:47] <JymmmEMC> ~2W/in for EPS
[03:23:55] <JymmmEMC> 1.5 - 2 W
[03:24:06] <tomp> the faster you travel, the more work per minute. work is fueled by temperature, feedrate should reduce temperature. any servo system has lag in response, (it happens after you notice the error ), so varying the rate could limit the temperature drop.
[03:24:53] <tomp> eps = expanded polystyrene?
[03:24:55] <SWPadnos> it's probably doable, but you'd have to be careful of trying to control one thing with two process vars
[03:24:58] <JymmmEMC> tomp: yeah
[03:25:12] <SWPadnos> ie, you can't try to PID the temp at the same time as the feed rate without being very careful
[03:25:14] <JymmmEMC> tomp: But what range of speed change are you talking?
[03:25:33] <tomp> i dont know the process so cant say
[03:25:38] <JymmmEMC> 10ipm to like 80ipm?
[03:25:44] <tomp> still dunno
[03:25:47] <JymmmEMC> k
[03:26:58] <tomp> oh, the variance? i'd guess varying small amounts, like <10% i'd think
[03:27:07] <JymmmEMC> Mind you that 2w/in was for a hand carving EPS - like a hot knife thru butter. No excessive burning/melting with a continous stroke
[03:27:29] <JymmmEMC> You paused, you get a bit extra melting.
[03:27:37] <tomp> 2w/inch of 'knife' ?
[03:27:43] <JymmmEMC> you pause a lot, you get 1/2" hole =)
[03:27:50] <JymmmEMC> no hotwire
[03:27:54] <JymmmEMC> no, hotwire
[03:28:18] <tomp> yeah, wire length
[03:28:34] <JymmmEMC> doens't matter, it's still 2Watts per inch
[03:28:50] <JymmmEMC> 48" == 100 Watts
[03:29:40] <JymmmEMC> But that's for EPS, not necessarily any other formulation. The Blue and PInk stuff has different densities.
[03:30:11] <tomp> interesting, there's 'turns' that a wireedm makes, where the top radius is small and the bottom radius is huge, and that'd mean the top would dwell (overmelt) and the bottom would race ( maybe loose heat to the passing material )
[03:30:45] <JymmmEMC> you taking EDM? Sorry I was talking hot wire foam cutting.
[03:31:00] <SWPadnos> he's talking hot wire
[03:31:00] <toastydeath> tomp: so what happens
[03:31:02] <toastydeath> oh
[03:31:06] <tomp> both, they have similar things to cut but use differnt processes
[03:31:14] <SWPadnos> consider trying to cut out the shape of a lapmshade
[03:31:24] <SWPadnos> at the top, the wire isn't moving as fast as at the bottom
[03:31:34] <SWPadnos> same heat, slower move = burns
[03:31:49] <toastydeath> i didn't know there was heat build up in emd
[03:31:52] <toastydeath> *edm
[03:31:59] <tomp> a cone would be difficult for the heat process, the top is near a poitn ( so near statioanry ) the bottom is a lot of surface travel
[03:32:14] <tomp> no it's not about edm its about form
[03:32:18] <toastydeath> oh
[03:32:27] <tomp> wedm just travels that shape sometimes
[03:33:03] <SWPadnos> any time XY and UV aren't moving identically to each other, there's a slow spot somewhere along the wire
[03:33:14] <tomp> in wedm the cone tip would get a bit 'polished', in hot wire, it might get burned
[03:33:19] <tomp> SWPadnos: right
[03:33:39] <tomp> there's a variance in surface feed
[03:33:47] <SWPadnos> right - somewhere
[03:33:58] <tomp> damn airfoils
[03:34:00] <SWPadnos> heh
[03:34:26] <SWPadnos> actually, nichrome may self-correct that a bit
[03:34:38] <SWPadnos> the hot spots have higher resistance, the cool spots lower
[03:34:52] <SWPadnos> so there may be more energy going to the areas that cool faster
[03:35:24] <SWPadnos> hmmm - I'd have to think too hard to figure that out in my head
[03:36:20] <tomp> wire edm'd donuts hmmmmmmmmmm
[03:36:45] <tomp> and thats enuf from me
[03:36:49] <tomp> gnite
[04:11:08] <toastydeath> blam
[04:19:29] <skinnypuppy1334> Anyone near SouthCaroliona there is a BP Boss for sale Not Mine just found looking on craigslist
[04:19:32] <skinnypuppy1334> http://atlanta.craigslist.org/tls/427508800.html
[04:22:22] <jlmjvm> may have to call about that
[04:22:49] <jlmjvm> skinnypuppy1334:are you in atlanta?
[04:23:26] <skinnypuppy1334> About 100 miles south a little outside macon
[04:23:48] <jlmjvm> huntsville,al here
[04:24:12] <skinnypuppy1334> Cool good to know someone atleast in my timezone
[04:24:24] <skinnypuppy1334> Hows that card working out for you?
[04:24:52] <jlmjvm> bout got it wired up,taking a break for a few mins
[04:25:31] <skinnypuppy1334> Right on , just finished putting a new scale on my Y axis tonight.. long day
[04:25:46] <jlmjvm> glass scale?
[04:26:00] <skinnypuppy1334> Yep I had a broken reader on Y for about the last year
[04:26:19] <skinnypuppy1334> Sargon went out of biz, took a while to find a n.o.s. part
[04:28:11] <jlmjvm> cool
[04:29:51] <jlmjvm> you ever get any motors yet?
[04:58:07] <skinnypuppy1334> I was out of the room , I put some money in the paypal yesterday. Going to watch ebay over the weekend again.
[04:58:47] <skinnypuppy1334> Scale was kind of spur of the moment $200 I wasn't expecting but couldn't turn it down. Had to jump on it
[04:59:30] <skinnypuppy1334> Looks like if I don't see something I'll be ordering some from kelling on monday.
[05:03:17] <skinnypuppy1334> I am curious how I am going to do two Z's. I got 4 geckos and was going to drive the knee with a jog just as a power raise
[05:03:49] <cradek> you have a quill and knee?
[05:04:09] <skinnypuppy1334> BP clone,
[05:04:22] <cradek> I think it's somewhat typical to call the knee 'W' and the quill 'Z'
[05:04:37] <cradek> +Z is quill up, +W is knee down
[05:04:59] <skinnypuppy1334> Ok good to know, gotcha both move the tool from the part
[05:05:24] <cradek> right
[05:05:52] <cradek> emc 2.2 will have full support for XYZW XYZAW etc machines
[05:06:04] <skinnypuppy1334> Drool
[05:06:07] <cradek> (XYZABCUVW)
[05:07:58] <skinnypuppy1334> If it were able to look at the tool length when called and determine whether quill retract is enough or if additional W is needed to load the new tool
[05:08:37] <SWPadnos> just use the knee for tool offsets all the time
[05:08:42] <cradek> yeah it would be interesting to do tool length offset with the knee
[05:08:58] <cradek> you'd always have a full 5" (or whatever) of travel
[05:09:11] <skinnypuppy1334> AAAHHAAA
[05:09:21] <SWPadnos> no, they're out of business
[05:09:32] <skinnypuppy1334> ;o)
[05:09:59] <cradek> there's not explicit support for that but it would be a very easy (source code) change.
[05:10:20] <SWPadnos> actually, you can do it in G-code, with probing
[05:10:28] <cradek> yes
[05:10:31] <SWPadnos> grab tool, probe with W, then move Z only
[05:10:41] <cradek> but with a source code change you could do it with G43
[05:10:50] <SWPadnos> oh, well there you go :)
[05:10:56] <cradek> plenty of neat options to play with
[05:11:06] <skinnypuppy1334> I need to learn some linux related programing. Used to do ok with batchfiles and basic when I was young.
[05:11:10] <SWPadnos> damn. now I've got to get another motor ;)
[05:11:12] <cradek> tool length probing is my favorite new hack
[05:11:29] <toastydeath> probing with what?
[05:11:30] <cradek> you should get three working first...
[05:11:41] <cradek> just poke a switch with the tool
[05:11:41] <SWPadnos> maybe I can make this crappy Ametek work on the knee with 5:2 gearing or something
[05:11:52] <toastydeath> oh
[05:11:59] <SWPadnos> err - 5:1, that was
[05:12:05] <toastydeath> what's the accuracy of the switch
[05:12:06] <toastydeath> usually
[05:12:08] <skinnypuppy1334> Would it be advisable to begin learning to program with C or Python?
[05:12:17] <SWPadnos> you can get very accurate switches
[05:12:18] <cradek> skinnypuppy1334: yes
[05:12:24] <SWPadnos> or use touch sensors
[05:12:26] <toastydeath> cool
[05:12:45] <skinnypuppy1334> If you had a small amount of time to set aside for one or the other which would you choose?
[05:12:46] <cradek> toastydeath: I was getting better than .0005 repeatability, I think it was actually 1 step
[05:13:01] <cradek> one or two maybe
[05:13:19] <cradek> skinnypuppy1334: what do you want to program?
[05:13:32] <cradek> lots of people seem to recommend python for beginners
[05:13:44] <cradek> (it's probably easier to get started with)
[05:14:15] <skinnypuppy1334> Just something to learn , feel in the dark been a long time since I had to think program logic other than g code
[05:16:39] <cradek> hmm, after midnight. goodnight folks
[05:16:48] <SWPadnos> night
[05:17:11] <skinnypuppy1334> gnight cradek thanks
[05:27:18] <toastydeath> eff
[12:18:49] <jimlas53> Can anyone answer a question regarding MSG use? Running EMC2/Axis. Thx
[12:57:59] <jepler> DATAJA
[13:01:04] <jimlas53> jepler, is there a way to prompt the operator without requiring the "OK" acknowledgement?
[13:03:09] <jepler> no.
[13:05:37] <jimlas53> OK, thanks. is there any way to use the "OK" click as an implied "S" to resume the program?
[13:06:25] <jepler> no.
[13:09:01] <jepler> if you're prompting the user to switch tools, you should use hal_manualtoolchange instead of (MSG) and M0
[13:10:48] <jepler> (copy /etc/emc2/sample-configs/sim/axis_manualtoolchange.hal to your inifile directory and list it on a new [HAL]HALFILE line in your .inifile, after the other(s) already there
[13:12:18] <jepler> )
[13:20:09] <jimlas53> OK, done. What a beautiful thing! I have 2 machines running and this will help the process. Thanks again!
[15:57:15] <jmkasunich_> jmkasunich_ is now known as jmkasunich
[16:33:45] <fenn> just what i always wanted, gold-plated hex keys http://www.use-enco.com/CGI/INSRIT?DCMP=EMC-1170059&PMAKA=NX308-0643
[16:45:06] <Roguish> gotta have 'em, cant' win without 'em
[16:52:13] <Roguish> fenn: hey, what editor do you use? gedit is too simple, and i hate vim.
[16:55:37] <fenn> kate is a good gui editor
[16:56:02] <fenn> i use nano and vim on the console
[16:57:59] <fenn> kate is sorta modeled after emacs i think
[16:58:29] <fenn> but windows-equivalent keyboard shortcuts
[17:00:41] <Roguish> does kate do context highlighting? ya know, different colors for differnt things?
[17:00:46] <fenn> yes
[17:01:05] <fenn> sometimes you have to tell it what the language is, it's not quite as smart as vim in that regard
[17:01:44] <Roguish> thanks, i'll try it. is it multidocument?
[17:01:50] <fenn> very much so
[17:02:00] <Roguish> coool. thanks.
[17:02:30] <Roguish> i'm working on some classicladder stuff.
[17:12:33] <Skullworks-PGAB> P2P gnutillia type network for Cell phones? http://news.bbc.co.uk/1/hi/technology/6987784.stm
[18:15:01] <JymmmEMC> Skullworks-PGAB: I like that Sony donated 3M
[18:15:40] <JymmmEMC> s/donated/invested/
[18:24:35] <Skullworks-PGAB> IIRC it was Erickson
[18:25:33] <Skullworks-PGAB> but I think your right - I think Sony own a chunk of Erickson
[18:28:33] <JymmmEMC> I can use my cordless phones like intercom/walk talkies like that already.
[18:29:24] <JymmmEMC> But you can only put them in phone mode -or- P2P mode, not both at the same time.
[18:39:56] <JymmmEMC> heh heh heh http://store1.yimg.com/I/hobased_1966_41665618
[19:07:28] <jlmjvm> i have a 12v 500mA wall power supply that i was gonna use for a usc board,when i cut the plug off it and checked it with a meter its showing 16v,and i have another 1 thats showing 18v
[19:07:52] <JymmmEMC> that's not under load
[19:08:03] <jlmjvm> is it the load
[19:08:45] <JymmmEMC> you connected it to a load and measured it?
[19:09:08] <jlmjvm> no,justchecked the wire ends
[19:09:24] <JymmmEMC> Like I said, it's not under load.
[19:09:39] <JymmmEMC> how much does the USC draw?
[19:09:58] <jlmjvm> k,makes sense,have never used or checked 1 before
[19:10:10] <jlmjvm> 12v 500 mA
[19:10:29] <JymmmEMC> It draws 500ma? If so you should get a bigger one
[19:10:53] <jlmjvm> thats what Jon told me to get
[19:10:58] <JymmmEMC> dont push the PS to it's limits like that - they're cheaply built enough as it is.
[19:11:19] <jlmjvm> he said 9-12v half amp
[19:12:24] <jlmjvm> i have a really nice 12v@1.5 amp regulated supply that i would like to use,waiting for him to email me back
[19:12:57] <JymmmEMC> You can ALWAYS use a power supply that has higher mA rating
[19:13:41] <SWPadnos> you can run it off 9V
[19:13:50] <SWPadnos> it's got a regulator onboard
[19:14:11] <SWPadnos> I think he sent me a wall wart with mine, and it's 9V 300mA
[19:14:39] <jlmjvm> i picked up a 12v 500mA
[19:15:16] <SWPadnos> that'll work, though the regulator may get a little warm if you have all the SSRs on for a while
[19:16:01] <jlmjvm> what about the 12v@1.5 amp regulated supply
[19:16:14] <JymmmEMC> jlmjvm: Thar'll work
[19:16:14] <SWPadnos> don't bother, there's a regulator onboard
[19:16:27] <JymmmEMC> SWPadnos: he already has it
[19:16:52] <jlmjvm> was left over from a previous job
[19:17:12] <SWPadnos> the 12V supplies will work, you just need to keep an eye on the regulator heatsink for a while
[19:17:30] <JymmmEMC> SWPadnos: USC uses what, 5V?
[19:17:32] <SWPadnos> I wouldn't waste a nice bench power supply on the USC :)
[19:17:37] <SWPadnos> yes, it's 5V
[19:17:42] <jlmjvm> k,will do,gonna try to move a motor later today
[19:17:51] <JymmmEMC> jlmjvm: Got a 5V 1A wall wart?
[19:18:10] <SWPadnos> that won't work, since the regulator needs a volt or two of headroom
[19:18:11] <jlmjvm> have a 4.5 700mA
[19:18:28] <SWPadnos> and you do want to leave the regulator in circuit, not bypass it
[19:18:42] <JymmmEMC> jlmjvm: Ok, SWPadnos says just above 5v
[19:19:00] <SWPadnos> SWPadnos says 7V minimum :)
[19:19:18] <JymmmEMC> SWPadnos: Fine 5V + 1.2V
[19:19:39] <JymmmEMC> wait... .707 * 2
[19:19:58] <JymmmEMC> SWPadnos: happy?!
[19:20:07] <SWPadnos> 5V + 2
[19:20:22] <SWPadnos> the 7805 isn't an LDO
[19:20:30] <JymmmEMC> junction drop is .707v isn't it?
[19:20:38] <SWPadnos> it's not a diode, it's a regulator
[19:20:53] <JymmmEMC> SWPadnos: it's got diodes in it dad burnit!
[19:21:04] <SWPadnos> yes, diodes * .7 = 2 :)
[19:22:04] <SWPadnos> heh - I guess I estimated well:
[19:22:14] <SWPadnos> note the dropout voltage rating
[19:24:08] <JymmmEMC> jlmjvm: Anyhow, due to the regulator on board, you'll have a little bit of a voltage drop, thus you need 5v + 2v or more to compensate for the voltage drop. If you go too much over the regulator will heat up and your heat sicks will get hot. Jon probably said 9V because it's commonly available enough
[19:24:39] <SWPadnos> actually, I'm the one who said 9V, because that's what I have connected to mine (and I think Jon supplied it)
[19:24:49] <jlmjvm> he told me 9-12v at a half amp
[19:24:59] <SWPadnos> oh, then don't worry about it
[19:25:18] <SWPadnos> I'd be surprised if it used 500mA anyway
[19:25:18] <jlmjvm> was just concerned when it checked 16v
[19:25:22] <JymmmEMC> jlmjvm: If it needs 300mA, you can always go higher rating on the mA though. As it should only draw what it needs.
[19:25:27] <SWPadnos> right
[19:25:39] <SWPadnos> never go over voltage or under current on a power supply
[19:25:58] <JymmmEMC> Unless it shorts out of course.
[19:26:18] <SWPadnos> well, then you're hosed anyway, so you might as well enjoy the fireworks
[19:26:27] <JymmmEMC> jlmjvm: Got Fuse? ;)
[19:26:40] <JymmmEMC> jlmjvm: do you have any resistors?
[19:26:42] <SWPadnos> actually, you shouldn't go too far over either, power supplies rarely regulate well when they're under 10-20% load
[19:26:46] <jlmjvm> prolly need some,lol
[19:26:58] <jlmjvm> yes i have some resistors
[19:27:14] <SWPadnos> don't use a resistor - there's a regulator ...
[19:27:26] <JymmmEMC> use ohms law and calculate 400mA load at 12V
[19:27:42] <JymmmEMC> 12 * .4 = I
[19:27:43] <skinnypuppy1334> 12/.4
[19:27:47] <skinnypuppy1334> ?
[19:27:51] <SWPadnos> 30
[19:28:01] <JymmmEMC> 30 ohm resistor
[19:28:11] <SWPadnos> but, don't do that
[19:28:14] <SWPadnos> it's not necessary
[19:28:23] <jlmjvm> k
[19:28:27] <JymmmEMC> SWPadnos: No no, he is reading 16V from the walmart
[19:28:32] <SWPadnos> it also screws up regulation
[19:28:33] <JymmmEMC> walwart
[19:28:43] <JymmmEMC> SWPadnos: No t connected to the SUC
[19:28:50] <JymmmEMC> no load voltage
[19:28:53] <SWPadnos> I'm betting the wall wary has about 30 gauge wires, so that'll go down to 12-14V as asoon as it's connected
[19:28:59] <SWPadnos> wart
[19:29:26] <JymmmEMC> SWPadnos: Right, that's why I was suggesting that he measure the walwart under a 30ohm load NOT connected to the USC
[19:29:28] <SWPadnos> (kidding, but the coil is likely very small gauge, so the internal resistance will be a factor once there's any current)
[19:29:33] <SWPadnos> ah - ok
[19:29:44] <SWPadnos> I thought you were suggesting a series resistor to drop the voltage to the regulator
[19:29:55] <JymmmEMC> SWPadnos: oh hell no, I'm not THAT bad
[19:30:03] <SWPadnos> well, I didn't know that
[19:30:04] <SWPadnos> L
[19:30:06] <SWPadnos> :)
[19:30:15] <JymmmEMC> jlmjvm: so what is the smallest resistor you have?
[19:30:48] <jlmjvm> i actually went to wal mart looking for a ps,called Jon the next day and told him I couldnt find any of those wal mart type supplys
[19:30:57] <SWPadnos> wall wart, not wal mart!
[19:31:05] <SWPadnos> go to Radio Shack
[19:31:12] <SWPadnos> they still sell wall warts
[19:31:26] <JymmmEMC> SWPadnos: He went to walmart to look for walwarts by the seashore
[19:31:35] <jlmjvm> thats what he told me.
[19:31:51] <jlmjvm> my girlfriend laughed for bout an hour
[19:32:12] <JymmmEMC> jlmjvm: so what is the smallest resistor you have?
[19:32:16] <jlmjvm> we did actually find a 9v 300mA there
[19:32:29] <jlmjvm> lemmee see,i think a 470k
[19:32:45] <JymmmEMC> jlmjvm: any 120, 120K ?
[19:33:28] <SWPadnos> just hook it to the USC and see what the voltage ends up at. it won't hurt it. that regulator can take 30V input
[19:33:57] <SWPadnos> the only problem is under sustained usage, the power dissipated in the regulator is (supply-5V)*current
[19:33:58] <JymmmEMC> DO NOT PLACE THE USC on a static bag while testing
[19:34:08] <jlmjvm> 162k is my smallest
[19:34:16] <SWPadnos> so a 15V supply means 10V drop * 500mA max = 5W
[19:34:35] <jlmjvm> its mounted to a plate
[19:34:47] <SWPadnos> that heatsink is probably capable of handling 5W anyeay, and that's well within the regulator spec of 30V, 15W max dissipation
[19:34:51] <JymmmEMC> jlmjvm: Is the plate electrically conductive?
[19:35:07] <SWPadnos> JymmmEMC, it's mounted - like with standoffs, not sitting on it
[19:35:22] <JymmmEMC> SWPadnos: (isolation)
[19:35:24] <jlmjvm> its the plate that came out of the control cabinet
[19:35:36] <jlmjvm> yes,on standoffs
[19:35:43] <JymmmEMC> jlmjvm: metal or plastic?
[19:35:46] <JymmmEMC> standoffs
[19:35:49] <jlmjvm> metal
[19:35:55] <SWPadnos> there's no need for isolation, the wall wart should be isolated
[19:36:25] <JymmmEMC> SWPadnos: how do you isolate the PC signals from the walmart of it's on metal standoffs?
[19:36:44] <SWPadnos> wrr
[19:36:45] <SWPadnos> err
[19:37:17] <SWPadnos> a walmart wallwart wants watts wasted with wegulators
[19:37:40] <SWPadnos> whoopw - wegwuwators
[19:38:00] <JymmmEMC> SWPadnos: Ok, isn't the USC an opto isolated device? Meaning it has two halfs to it (so to speak)
[19:38:32] <SWPadnos> the parallel port is is not isolated from
[19:38:39] <SWPadnos> the parallel port is not isolated from the PC
[19:38:48] <SWPadnos> so the USC has to share ground with the PC
[19:39:07] <JymmmEMC> SWPadnos: In one-half of it, but on the other side of the optos?
[19:39:21] <SWPadnos> but the mounting holes aren't electrically connected
[19:39:30] <JymmmEMC> SWPadnos: Are you sure?
[19:39:33] <SWPadnos> at least not on the version I have (2.1, from 7/2004)
[19:39:38] <JymmmEMC> ah, ok
[19:39:39] <SWPadnos> yes, I'm looking at one
[19:39:49] <JymmmEMC> SWPadnos: Heh, I never assume on this crap.
[19:40:09] <SWPadnos> neither do I, that's why I'm looking at mine
[19:40:15] <SWPadnos> I had to dig for it you know
[19:41:03] <JymmmEMC> SWPadnos: I've been looking at this for a while http://www.cnc4pc.com/Store/osc/product_info.php?cPath=33&products_id=48
[19:41:38] <jlmjvm> ive used 1 before
[19:41:42] <JymmmEMC> SWPadnos: But it requires 5v from USB port for one-half, and another Ps for the other half of it.
[19:41:49] <fenn> while on the topic of wall warts, i'd just like to mention that those new lightweight switching wall warts like to fry stuff when there's lightning nearby
[19:42:16] <JymmmEMC> fenn: even when properly earth grounded?
[19:42:29] <fenn> they dont have grounds, and if they did, i assume they'd still blow up
[19:42:57] <JymmmEMC> fenn: I've found that PC SW PS have neutral tied to earth ground
[19:43:03] <fenn> that's a PC
[19:43:13] <fenn> i'm talking about cellphone charger thingies
[19:43:14] <JymmmEMC> if you use a open frame SW PS, then no.
[19:43:20] <jlmjvm> it had some problems and we had to send it back,but Arturo refunded the money.Im sure it was a fluke
[19:43:39] <JymmmEMC> jlmjvm: Oh, what was the problem?
[19:44:04] <jlmjvm> charge pump wasnt working right
[19:44:34] <jlmjvm> the customer ended up going with the bob campbell combo board
[19:44:42] <JymmmEMC> oh great <sarcasim>, good to hear/know at least. Was it the breakout board, or the individual charge pump?
[19:45:08] <fenn> a charge pump is just a capacitor and a transistor or two
[19:45:12] <jlmjvm> it was a combo board
[19:45:23] <JymmmEMC> damn
[19:45:54] <JymmmEMC> fenn: ?
[19:45:58] <SWPadnos> even easier - a capacitor and a diode or two
[19:46:11] <jlmjvm> but the guy there gave us no problem on the return,will not hessitate to do business with again
[19:46:56] <fenn> well you need an AC supply for a cockroft-walton multiplier
[19:47:01] <JymmmEMC> I'm reluctant to buy when the issue was the safety device.
[19:47:09] <jlmjvm> easy for yall prolly,im not up to repairing boards quite yet
[19:47:34] <SWPadnos> a pulse trainis an AC supply
[19:47:38] <SWPadnos> train is
[19:48:25] <SWPadnos> OK. time for a Radio Shack run. bbiab
[19:48:29] <fenn> ew
[19:48:30] <JymmmEMC> ew
[19:48:38] <fenn> good luck
[19:48:37] <jlmjvm> later
[19:48:50] <SWPadnos> all I need is alligator clip leads - geez
[19:49:08] <fenn> bet they dont have em
[19:50:32] <fenn> i hope this terranet thing takes off
[19:50:38] <JymmmEMC> ?
[19:50:40] <fenn> peer to peer cellphones
[19:50:43] <JymmmEMC> oh
[19:50:46] <Skullworks-PGAB> yeach the Radio Shacks that are left only seem to sell crap chi - import toys and cell phones.
[19:51:03] <JymmmEMC> Skullworks-PGAB: That's ALWAYS been the case.
[19:51:15] <fenn> nah you're just too young to remember
[19:51:41] <JymmmEMC> fenn: Shit.... we aint tlaking when they had/sold HF rigs now
[19:52:08] <fenn> The company was started as Radio Shack in 1921 in Boston, Massachusetts, by two brothers, Theodore and Milton Deutschmann, who wanted to provide equipment for the cutting-edge field of amateur, or ham, radio.
[19:52:29] <fenn> did they even import stuff from china in 1921?
[19:52:42] <JymmmEMC> Yeah, Chinese
[19:52:56] <JymmmEMC> who you think made the railroads
[19:53:05] <fenn> thought that was way earlier
[19:53:11] <Skullworks-PGAB> it was
[19:53:16] <JymmmEMC> gold rush days late 1800's
[19:53:34] <Skullworks-PGAB> and yeah workers - another chi - export
[19:54:21] <JymmmEMC> Oh, can you connect a SSR directly to a paraport pin?
[19:54:38] <Skullworks-PGAB> but on the east side most the rail workers were irish - potatoe famine refugees
[19:55:51] <Skullworks-PGAB> Jym - not sure the paraport can provide enough current on its own to trip an SSR
[19:56:23] <JymmmEMC> Skullworks-PGAB: http://cgi.ebay.com/Solid-State-Relay-SSR-24-480V-AC-25A-Heat-Sink_W0QQitemZ330166550669QQihZ014QQcategoryZ78207QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
[19:56:35] <JymmmEMC> coltrol current: 3-25mA
[19:56:49] <Skullworks-PGAB> that might work
[19:57:19] <Skullworks-PGAB> I been looking at old stuff - 20-50ma
[19:57:26] <JymmmEMC> ah, gotcha.
[19:58:10] <JymmmEMC> I REALLY want the cnc4pc breakoutboard, but been reluctant about it, then jlmjvm mentioned the charge pump issue.
[19:59:20] <JymmmEMC> Artuto (maker of cnc4pc bb) has it setup to draw the 5v from USB port, and I don't want to use USB to supply power like that. Just wanted a single cable between PC and controller box.
[20:00:14] <JymmmEMC> I dont want to use a wallwart inside the controller, as then I have to make sure it doesn't come unplugged.
[20:00:16] <fenn> so use a usb hub
[20:00:37] <fenn> hub has its own wall wart
[20:00:56] <JymmmEMC> fenn: it's just screw terminals.
[20:01:16] <fenn> the power connection?
[20:01:34] <JymmmEMC> it requires TWO power supplies (for isolation)
[20:01:57] <fenn> oh, then the usb is probably only powering the opto's
[20:02:04] <JymmmEMC> yeah
[20:02:04] <fenn> 30mA or so i bet
[20:02:32] <JymmmEMC> But if I could find a cheap open frame linear PS that would be fine.
[20:02:41] <fenn> for 30mA?
[20:02:44] <fenn> are you nuts?
[20:02:49] <JymmmEMC> Ever take apart a wallwart?
[20:02:53] <fenn> sure
[20:03:05] <JymmmEMC> it's barely held together internally
[20:03:48] <fenn> you are one of those annoying customers i bet
[20:04:03] <JymmmEMC> Full Disclosure baby!
[20:04:43] <JymmmEMC> fenn: What, you going to install a 120V outlet inside the controller for a wallwart?
[20:04:57] <fenn> depends what else is in the controller
[20:05:28] <fenn> i dont see the problem powering the opto's from the PC though
[20:05:40] <fenn> if the PC isnt on, there isnt going to be any data going through them
[20:05:51] <JymmmEMC> fenn: I dont want a second cable going from the PC to the controller box.
[20:06:01] <JymmmEMC> Just the paraport cable
[20:06:10] <fenn> well, too bad
[20:06:17] <JymmmEMC> fenn: fsck you =)
[20:07:40] <fenn> jymmm put some hot glue or epoxy on the wall wart internals if you're worried about it falling apart
[20:07:58] <JymmmEMC> fenn: Yeah, I was thinking hot glue
[20:08:11] <fenn> transformers do heat up though..
[20:08:20] <JymmmEMC> maybe silicon
[20:09:29] <DanielFalck> http://imagebin.org/10554
[20:09:38] <DanielFalck> http://imagebin.org/10553
[20:09:47] <DanielFalck> apt to g code in action