#linuxcnc | Logs for 2013-10-20

[01:23:45] <MacGalempsy> guys I am having boot issues with my mesa 5i25. a little help?
[01:39:29] <jesseg> Sorry Mac, I have no experience with mesa at all.
[01:39:41] <jesseg> I thought it was a video graphics driver from the dos days :P
[01:39:55] <jesseg> Or was it an FPGA project too? Somehow that has a familiar ring
[01:41:25] <MacGalempsy> maybe a general linux question?
[01:41:35] <jesseg> sure
[01:41:40] <MacGalempsy> when I install the card, the machine will not boot
[01:41:44] <jesseg> linux I do (mostly slackware)
[01:42:01] <jesseg> What distro? What sort of computer?
[01:42:16] <MacGalempsy> it is a ibm m57 sff pc
[01:42:35] <MacGalempsy> the latency test ran great, so I was excited
[01:42:37] <jesseg> Is that a laptop?
[01:42:43] <MacGalempsy> no a desktop
[01:42:52] <jesseg> oh, so it's a regular computer
[01:42:57] <jesseg> ohh - the "card" is the mesa card?
[01:43:04] * jesseg <- not a mesa guy
[01:43:19] <MacGalempsy> yesx
[01:43:24] * jesseg googles mesa cnc
[01:43:33] <MacGalempsy> their support is crap
[01:43:42] <MacGalempsy> online, however, I could call
[01:43:45] <jesseg> Is it a stepper or a servo motor controller?
[01:44:00] <MacGalempsy> servo daughter card
[01:44:13] <jesseg> like a PCI card?
[01:44:18] <MacGalempsy> the daughter card is a 7i77 but it is not hooked up yet
[01:44:31] <jesseg> okay I'm getting the picture - but how does the PC interface to mesa?
[01:44:38] <MacGalempsy> pci card
[01:45:16] <jesseg> http://www.mesanet.com/graphics/parallel/7i76pandgcolo.png
[01:45:19] <jesseg> like that?
[01:46:23] <jesseg> So to clarify, if the PCI card is inserted, then no boot. Otherwise, happy boot (except LinuxCNC is mad about not finding any hardware.)
[01:47:24] <MacGalempsy> yep. you think I need to have the cable hooked up?
[01:47:40] <MacGalempsy> the jumpers are set for no daughter card at the moment
[01:48:00] <jesseg> Not sure if cable makes a difference. Try it if easy. But when you say it won't boot -- how far does it start to boot?
[01:48:06] <jesseg> Do you see the POST messages (ram test, etc.)
[01:48:57] <MacGalempsy> I dont see the post msgs
[01:49:14] <jesseg> And no beeping?
[01:49:20] <jesseg> Does the monitor come out of sleep?
[01:49:47] <MacGalempsy> it comes on, then shuts off, then it powers on but just the fan runs
[01:49:58] <MacGalempsy> the monitor does not come out of sleep
[01:50:31] <jesseg> Have you let it sit for 15 minutes just to make sure it won't time out and throw an error? Can help in debugging
[01:52:49] <jesseg> Have you let it sit for 15 minutes just to make sure it won't time out and throw an error? Can help in debugging
[01:52:58] <jesseg> oh that was a leave not a join.
[01:54:03] <MacGalempsy> havent tried that yet
[01:54:18] <MacGalempsy> there is nothing in the bios that says the port is shut down
[01:55:50] <jesseg> I'm wondering if the mesa card has corrupted code on it that is causing it to assert lines on the PCI buss that is halting the computer
[01:56:03] <jesseg> Try mesa card in another computer, or another slot in same computer
[01:59:15] <MacGalempsy> only one slot. in the computer
[01:59:39] <jesseg> Hmm. well I'd try same mesa card in another computer, and try some other card in questionable slot
[01:59:40] <MacGalempsy> I have a newer alienware that it will fit in
[02:00:51] <jesseg> You might also go into bios on troublesome computer and restore to defaults.. Maybe something's strange set that doesn't work with mesa card
[02:01:00] <jesseg> seems unlikely but it's free to try :P
[02:02:27] <jesseg> Is there a way to reset the mesa card back to factory defaults? (Or is it ram based, needing its code to be loaded on each bootup?)
[02:02:55] <jesseg> well that's all the ideas I have. I'm going to wonder off for a bit, then check back, then go to bed, Lord willing.
[02:09:11] <stillme> How do i get blender 2.62 with bas relief composite node patch?? Does any one know?? or direction on how to compile it??
[02:14:42] <MacGalempsy> jesseg: this machine boots up fine with the card installed
[02:14:57] <MacGalempsy> but it is also windows 7, not linuxcnc
[02:24:55] <jesseg> back
[02:25:05] <jesseg> MacGalempsy, OK... strange..
[02:25:42] <jesseg> However, if the "bad" machine doesn't even do the power on self test (POST) then I highly doubt that it has anything to do with Linux or the operating system or software.
[02:26:13] <MacGalempsy> I think i saw a post setting in the firmware
[02:26:16] <MacGalempsy> or bios
[02:26:28] <MacGalempsy> let me check for it again
[02:26:31] <jesseg> You can get LinuxCNC on a self booting "LIVE" cd. And you can use it that way by putting in a thumb drive - all modifications you make to programs and stuff get saved to the thumb drive
[02:26:37] <jesseg> Oh, yeah, maybe POST is turned off.
[02:26:56] <jesseg> or post display. post always happens, but the display of it can be turned off.
[02:27:11] <jesseg> But, it would wake up the monitor and display a logo if it was doing post
[02:27:22] <jesseg> Hmm
[02:27:49] <jesseg> Try unplugging the hard drive cable for the linux computer, and try booting it with the mesa card in, and see if it gives you a disk not found error
[02:27:59] * jesseg wonders off again, will check back in a few
[02:29:12] <MacGalempsy> it is set to quick
[02:29:24] <MacGalempsy> perhaps I need to change that
[02:29:31] <MacGalempsy> to enhanced?
[02:30:10] <MacGalempsy> ok. have to shut this machine down to take out the card.
[02:37:11] <MacGalempsy> crap that did not work either
[02:37:40] <MacGalempsy> letting it run for a bit to see if an error bar shows up. but the green power light doesnt work
[02:38:20] <MacGalempsy> do you think it needs a driver in linux before it will power up?
[02:39:44] <MacGalempsy> I wonder if the 280w psu is too small
[02:42:25] <jesseg> back
[02:42:44] <jesseg> no, it doesn't need a driver to power up.
[02:43:06] <jesseg> The driver would be stored on the hard drive, and the hard drive isn't read until after it powers up, so a driver on the disk can't help it power up
[02:43:17] <jesseg> It may be an incompatibility with Mesa and a cheap mobo?
[02:43:42] <MacGalempsy> andypugh thought this one would work
[02:43:49] <MacGalempsy> so I got it, dang...
[02:44:03] <MacGalempsy> its an IBM m57 sff computer
[02:44:26] <jesseg> I don't know andypugh..
[02:44:55] <jesseg> I don't know what to say. You could try taking all other cards out of the m57 to see if maybe the mesa is conflicting with those.
[02:45:15] <jesseg> You might also try disabling all onboard peripherals in the bios other than video just to see if that's an issue
[02:45:21] * jesseg wanders away
[02:47:54] <Jymmm> MacGalempsy: do you have a larger PS to test with?
[02:51:23] <MacGalempsy> I am trying turning the video to internal instead of auto
[02:52:02] <Jymmm> how many cards are installed?
[02:54:41] <MacGalempsy> jymmm - yes do you think 280w is kind of week?
[02:54:41] <MacGalempsy> *weak?
[02:54:50] <Jymmm> how many cards are installed?
[02:55:42] <MacGalempsy> just one
[02:55:50] <MacGalempsy> no other cards
[02:55:50] <Jymmm> which is?
[02:55:54] <MacGalempsy> 5i25
[02:56:04] <t12> hum
[02:56:10] <t12> see it in lspci?
[02:56:11] <Jymmm> Then why are you switching the video surce?
[02:56:20] <_DJ_> moin
[02:56:25] <MacGalempsy> t12 - the machine will not come on
[02:56:28] <t12> oh sorry
[02:56:30] <t12> just read back
[02:56:37] <_DJ_> morning
[02:56:40] <t12> lenovo/ibm machines were sometimes particular about the pci cards in them
[02:56:43] <t12> on purpose
[02:56:46] <t12> notably on laptops
[02:56:54] <MacGalempsy> crap. anyway around that?
[02:56:55] <t12> like theres a bios whitelist of allowed PCI ids
[02:57:06] <t12> i dont know if that extended to the desktop machines though
[02:57:15] <t12> there were, modifications to the bios
[02:57:18] <t12> it doesnt post at all?
[02:57:23] <t12> like doesnt even start to?
[02:57:25] <Jymmm> MacGalempsy: Stop guessing.
[02:57:48] <Jymmm> t12 isn't helping atm, as he hasn't read back.
[02:58:10] <Jymmm> MacGalempsy: Again, why were you mucking with the video source?
[02:58:37] <Jymmm> MacGalempsy: And do you have a larger PS to test?
[02:58:40] <MacGalempsy> i changed the video from automatic to internal because there is auto. internal. PCI and PCIE. I thought the machine may be trying to load the card as video
[02:58:59] <MacGalempsy> yes I have a few different psu
[02:59:06] <Jymmm> how big?
[02:59:28] <Jymmm> 400W? 600W ?
[02:59:49] <t12> any ebeps
[02:59:50] <t12> beeps
[03:00:19] <Jymmm> t12: Dude, read back before going over the same things
[03:02:57] <MacGalempsy> no bigger psu that I can put in this computer. the t7500 that I it booted up on is 1100w
[03:03:12] <MacGalempsy> the other psu are 100, 150w...
[03:03:40] <Jymmm> MacGalempsy: You booted it form a 1100W PS and still have the issue?
[03:03:50] <Jymmm> MacGalempsy: Not install, JUST test.
[03:03:59] <Jymmm> from*
[03:04:30] <MacGalempsy> I put the card into my workstation with win7 and the machine booted and recognized the card
[03:04:42] <Jymmm> I didn't ask that.
[03:04:49] <MacGalempsy> i put the card back into the linux box and it didnt work
[03:05:02] <MacGalempsy> the win 7 machine is 1100w psu
[03:05:24] <MacGalempsy> i cannot pull the psu from the workstation, it is too big
[03:05:40] <Jymmm> the cable isn't
[03:06:01] <Jymmm> unplug it from the mobo, and plug it into the ibm
[03:06:08] <MacGalempsy> oh. I will try that.
[03:06:26] <MacGalempsy> but I need to shut this computer down...
[03:06:39] <MacGalempsy> be back in a few
[03:09:42] <t12> https://www.dropbox.com/sc/ro2mwx3kn186zku/kFGvmy_cXU
[03:09:44] <t12> getting closer
[03:10:10] <Jymmm> whats under the netbook?
[03:10:23] <t12> programmable power supply
[03:10:40] <t12> really nice for debugging buggy things
[03:10:42] <Jymmm> Nice, amperage?
[03:10:44] <t12> full GPIB control
[03:10:49] <t12> 60V, 2A, 2 channel
[03:10:56] <Jymmm> cool
[03:11:05] <t12> at least GPIB or eth is what i shoot for now on t&m equipment
[03:11:06] <t12> for automation
[03:11:08] <Jymmm> you have a GPIB card in the netbook?
[03:11:15] <t12> nah
[03:11:19] <t12> gpib -> ethernet
[03:11:24] <Jymmm> ah
[03:11:26] <t12> netbook is for talking to the mitsu controller
[03:11:52] <t12> made a ghetto rs485 decoder so i can pull the mitsu encoder data into logic analyzer
[03:11:53] <Jymmm> gotcha
[03:12:01] <t12> so i can get a decent stream of data to try and sort out the protocol
[03:12:13] <t12> that LA cant take differential pods :(
[03:12:23] <Jymmm> does it have an ExpressCard/34 slot?
[03:12:25] <archivist> what LA?
[03:12:29] <t12> logic analyzer
[03:12:40] <archivist> model
[03:12:46] <t12> agilent 16700
[03:13:04] <archivist> I do know what logic analysers are, I have a few
[03:13:08] <t12> oh oh
[03:13:10] <t12> sorry misread
[03:13:15] <t12> not at full mind capacity right now clearly
[03:13:34] <t12> 16700's are suprisingly cheap sometimes
[03:13:42] <archivist> I have a 16500A with 4 scope cards
[03:13:43] <t12> 16500 and 16700's are awesome cause of ethernet programability
[03:13:51] <t12> they're sort of hilarious machines though
[03:13:53] <Jymmm> archivist: you sure? Hard to see any behind that UNIVAC you have in your shop =)
[03:14:03] <t12> how is the scope?
[03:14:14] <t12> is there any way to actually use it in something like realtime? or is it ment for glitch only?
[03:15:01] <Jymmm> t12: does it have an ExpressCard/34 slot?
[03:15:08] <t12> the netbook? i dont think so
[03:15:22] <archivist> t12 not got the manual for mine yet http://www.collection.archivist.info/archive/DJCTM/EQ/2009_01_01_a/IMG_9999.JPG
[03:15:29] <Jymmm> ah ok, mine does so I use a REAL UART serial port
[03:16:01] <t12> remote X11 mode on the 16500's is awesome too
[03:16:07] <t12> though i think the 16500a doesnt have eth?
[03:16:21] <archivist> not that I remember
[03:16:45] <archivist> I need a gpib card one day
[03:16:45] <t12> yeah i'm stuck with all usb things on it
[03:17:06] <t12> that was just a junk pile meachine that was the easiest xp install i could get my hands on
[03:17:30] <Jymmm> archivist: Want the service manual for it?
[03:18:09] <t12> i have a 16500B with a mess of timing/state cards in it
[03:18:13] <t12> that i need to get rid of
[03:18:31] <t12> they're pretty annoying to transport
[03:18:34] <archivist> I think there are some changes A to B
[03:18:39] <Jymmm> archivist: 144 pages
[03:18:48] <t12> C's take ps2 kb/mouse and have stock wth i think
[03:18:54] <t12> s/wth/eth
[03:19:27] <Jymmm> archivist: Confirmed it's from HP 16500A Service Manual
[03:19:33] <t12> i piled up all this LA gear and then rarely use it
[03:19:44] <t12> sort of absurd to use a 16700 to record 1 pin
[03:20:01] <archivist> Jymmm, paper or pdf ?
[03:20:04] <Jymmm> archivist: pdf
[03:20:09] <archivist> :)
[03:20:13] <Jymmm> archivist: http://arcarc.xmission.com/Test%20Equipment/HP/16500-90911.pdf
[03:20:18] <Jymmm> 14MB
[03:20:28] <Jymmm> archivist: Not a scan either, the real thing
[03:21:25] <archivist> what do you mean not a scan, HP went to microfiche and then scanned them for some manuals
[03:21:34] <Jymmm> archivist: Reference Manual http://www.raybel.com/manuals/16500A.pdf
[03:22:04] * archivist starts another download
[03:22:10] <Jymmm> archivist: you want a scrn cap?
[03:22:40] <archivist> for what, same device?
[03:23:24] <jesseg> back
[03:23:33] <archivist> my mirror directory has expanded by two :)
[03:23:57] <Jymmm> archivist: lol, thought you might like those =)
[03:24:28] <t12> amazingly for the 16700's theres some web form
[03:24:33] <t12> where they will mail you the install cds
[03:25:07] <t12> http://software.cos.agilent.com/LogicAnalyzerSW/16700.stm
[03:25:09] <t12> still there!
[03:25:14] <archivist> Jymmm, I dont know what your acronym is reffering to
[03:26:01] <Jymmm> archivist: I know you have limited bandwidth. I was offering a SCRN CAP(screen capture) of the first page of the pdf
[03:26:21] <archivist> HP and TEK dont mind too much about mirrors of their manuals in the wild
[03:26:24] <jesseg> hi MacGalempsy. How's it going?
[03:26:30] <MacGalempsy> well, I didnt to it because it was too much to unplug all the wires
[03:26:39] <jesseg> do what?
[03:26:51] <jesseg> oh the PSU
[03:26:54] <MacGalempsy> the psu on the big box is too much hassle to get out for this test
[03:27:05] <archivist> Jymmm, I am not limited download, I have already got both manuals downloaded thanks
[03:27:32] <Jymmm> archivist: Ah, ok. np. I know you collect all that stuff
[03:27:48] <MacGalempsy> so I need to find a system that is known to work with the mesa 5i25
[03:27:51] <archivist> yes a bit of a manual packrat
[03:28:14] <MacGalempsy> luckily this m57 was only $80
[03:28:14] <Jymmm> archivist: and building your library collection iirc
[03:28:34] <MacGalempsy> well, I will go back to the wiring diagram
[03:29:13] <MacGalempsy> if anyone is interested, I have it up in hangout http://tinyurl.com/o5rk3w9
[03:29:25] <archivist> Jymmm, qties are on http://www.collection.archivist.info/
[03:30:03] <jesseg> MacGalempsy, It's probably not a PSU problem. Unless it's a real muscle machine, 250W ought to be plenty to boot up with one PCI card and one hard drive. I have an AMD 8 core 3.6Ghz system with 16G of ram and TWO 2TB hard drives and extra fans, and it idles at 64 watts INPUT to the power supply.
[03:30:35] <archivist> the startup may need more power
[03:30:38] <MacGalempsy> i figured the psu is ok because its beefier than those in my other pcs
[03:31:01] <jesseg> MacGalempsy, what CPU? 2 core? 32 core?
[03:31:08] <MacGalempsy> duo core
[03:31:13] <jesseg> how fast?
[03:31:25] <MacGalempsy> let me look it up
[03:31:59] <Jymmm> archivist: I just wish you included auto service manuals too =)
[03:32:00] <MacGalempsy> 2.33ghz
[03:32:29] <archivist> Jymmm, I do
[03:32:51] <Jymmm> archivist: 1988 Honda Civic LX ?
[03:32:57] <archivist> nope
[03:33:02] <Jymmm> damn
[03:33:04] <Jymmm> lol
[03:33:28] <jesseg> MacGalempsy, yeah I really don't see how 250W can't be enough. If you have another PSU lying around you could try it, but I'm going to suggest that its an incompatibility with mobo and fpga card
[03:33:36] <archivist> Jymmm, search lucas, a parts maker in my search box
[03:34:34] <MacGalempsy> i agree. there have been other issues with lenovo mobos, so perhaps I have to chalk this one up to trial and error
[03:34:48] <Jymmm> archivist: LOL. Any Honda or Ford?
[03:35:05] <archivist> type the name in the search box
[03:35:39] <archivist> remember I am in the uk so usually not usa models
[03:37:00] <Jymmm> archivist: I understand, but I dont think you care where you get them from either way
[03:37:12] <Jymmm> UK/US/AU/whatever
[03:37:53] <Jymmm> archivist: I'll get you a PDF of FORD you can add to your collection sometime
[03:38:34] <archivist> I generally get real paper manuals, currently stuff I have on PDF is not in the collection proper
[03:39:02] <Jymmm> archivist: I'm not giving up my paper wiring diagram manual just yet =)
[03:39:59] <Jymmm> archivist: I might be buying a civic tomorrow if all works out.
[04:13:05] <t12> now its clock recovery code time
[04:13:08] <t12> this always gets me
[04:14:06] <archivist> choices, make some hardware or do it in software
[04:14:15] <t12> software in this case
[04:14:51] <archivist> is the data rate low enough
[04:15:06] <t12> this is the mitsu encoder data
[04:15:28] <t12> 2.5mbaud
[04:15:39] <archivist> rather fast :)
[04:16:19] <archivist> or are you thinking some avr ?
[04:16:28] <t12> this is just to decode
[04:16:41] <t12> i think i can do this as a hack job though
[04:23:03] <archivist> I want to read a mitutoyo spi interface on my DTI when commanded by gcode so I can check the machine in more detail
[04:32:21] <The_Ball> yay, just made my first OpenServo move
[04:36:56] <awallin> r/c type of servo?
[04:39:32] <The_Ball> awallin, yeah standard rc
[04:56:57] <t12> S 1111100 0011110 0111111 0011011 0000000 1010111 0110011 0001100 0101011 1111000 0001010
[04:57:00] <t12> S 1111100 0010110 0111111 0000011 1000000 1010011 0110011 0001100 0101011 1111000 1101111
[04:57:02] <t12> S 1011101 0101110 1101111 0011010 0100100 1010111 0110011 0001100 0101011 1111000 1101110
[04:57:06] <t12> S 1111101 0111110 0101111 1001011 0100100 1010111 0110011 0001100 0101011 1111000 0101011
[04:57:09] <t12> S 1111100 0011110 0111111 0011011 1100000 1010111 0110011 0001100 0101011 1111000 1011010
[04:57:12] <t12> getting somewhere
[04:57:14] <t12> or not
[04:58:12] <archivist> seems some consistency
[04:58:18] <t12> yeah
[04:58:24] <t12> nevermind my 7 bit bytes there
[04:59:17] <t12> 1 start bit 11 bytes seems to line up
[04:59:44] <t12> S 10111010 01011100 11011110 00010101 11001001 10111111 01100111 00011001 01010111 11110001 11101111
[04:59:47] <t12> S 10111010 01011100 11011110 00100111 11001001 10111111 01100111 00011001 01010111 11110001 01101100
[04:59:50] <t12> S 10111011 01111100 01011111 00000100 00101001 10111111 01100111 00011001 01010111 11110001 10000110
[04:59:53] <t12> S 11111011 01111100 01011110 00110100 00101001 10111111 01100111 00011001 01010111 11110001 10010101
[05:00:24] <t12> i wonder if the fixed 5 bytes are just for clock recovery/sync
[05:02:18] <archivist> the mitutoyo has odd fixed bit but no mention why
[06:47:08] <exco> Is gimbal lock already a known problem (genserkins, 2.5.3)
[07:00:38] <t12> BIN S 1011101001011100110111101110011101100101111011110110011100011001010101111111000100011011
[07:00:42] <t12> HEX S ba 5c de e7 65 ef 67 19 57 f1 1b
[07:00:45] <t12> DEC S 011100111011001011110 - 1002958
[07:00:47] <t12> BIN S 1011101001011100110111111100010011100101111011110110011100011001010101111111000111101001
[07:00:50] <t12> HEX S ba 5c df c4 e5 ef 67 19 57 f1 e9
[07:00:53] <t12> DEC S 111000100111001011110 - 1003079
[07:00:56] <t12> BIN S 1011101001011100110111101111010011100101111011110110011100011001010101111111000111111010
[07:00:59] <t12> HEX S ba 5c de f4 e5 ef 67 19 57 f1 fa
[07:01:02] <t12> DEC S 011110100111001011110 - 1003102
[08:55:49] <Sven___> Py question, trying to add an event handler to a user space python component but nothing happens when the pin changes state. Is it possible to use events in a user space component or is it just a Glade thing?
[08:56:41] <t12> width=8 poly=0x55 init=0xff refin=false refout=false xorout=0x00 check=0x46 name=(none)
[08:56:45] <t12> is the CRC
[08:57:06] <Sven___> h = hal.component("mytc")
[08:57:15] <Sven___> test = hal_glib.GPin(h.newpin('tooltrigger', hal.HAL_BIT, hal.HAL_OUT))
[08:57:25] <Sven___> test.connect('value-changed', on_trigger_change)
[08:58:05] <Sven___> Is what I am using at the moment, but the method on_trigger_change never kicks in.
[09:20:58] <JT_Shop> Sven___, IIRC you need an event thing to put your part into. have you seen my tutorials
[09:25:09] <kwallace> Isn't on_trigger_change the even thing?
[09:25:19] <kwallace> event
[09:28:53] <kwallace> What is tooltrigger? Is it a button press?
[09:35:02] <kwallace> Does the Glade file have the button tied to a GTK thing in Signals and the thing's value-changed tied to "on_trigger_change".
[09:41:46] <Sven___> kwallace, what was your last question? I lost the connection and the chat history.
[09:42:05] <Sven___> If tooltrigger is a button?
[09:42:55] <Sven___> It is a hal pin going high/low when a tool position is passing by when the tool changer is rotating.
[09:50:39] <kwallace> Sven___: I think I'm way off base. I need to rethink your question, somehow I fixed on Glade.
[09:51:25] <Sven___> Sorry, no glade. It is a user space component loaded with hal.
[09:53:35] <exco_> Is there a branch that already uses Quaternions for genserkins to solve the gimbal problem
[10:13:34] <pcw_home> t12: did you try reveng on a bunch of different data?
[10:28:20] <pcw_home> MacGalempsy: do you have the full modl number of the M57?
[10:28:22] <pcw_home> We have had occasional PCI compatibility problems with some motherboards
[10:28:23] <pcw_home> but unlike the other ones, but this looks like one I can get my hands on easily
[10:29:50] <pcw_home> model number (there are a bunch of differnet M57 models)
[10:39:30] <kwallace2> Sven___: My brain is still mushy, but I would tend to open hal_glib.py and add a print statement to see if value-changed gets pushed.
[10:39:46] <kwallace2> def update(self):
[10:39:46] <kwallace2> tmp = self.get()
[10:39:46] <kwallace2> if tmp != self._prev:
[10:39:46] <kwallace2> self.emit('value-changed')
[10:39:46] <kwallace2> self._prev = tmp
[10:42:22] <pcw_home> t12: looks like plain old async (1 start bit 1 stop bit) 9 chars so 90 bits total (the mesa UART can do this at 2.5 MB)
[11:58:41] <Jymmm> 8ga stranded + 8ga stranded = _____ ga ???
[12:01:58] <JT_Shop> = parallel 8 gauge
[12:03:20] <IchGuckLive> hi all
[12:03:56] <IchGuckLive> what a crap day thunder rain and boring TV
[12:04:47] <Tom_itx> there's always irc
[12:05:55] <IchGuckLive> :-)
[12:06:05] <IchGuckLive> and openttd
[12:06:55] <Tom_itx> Jymmm, http://www.electrician2.com/calculators/wireocpd_ver_1_parallel.htm
[12:08:37] <Jymmm> Tom_itx: doesn't have 8ga
[12:08:47] <Tom_itx> so i see
[12:09:25] <Tom_itx> also would depend on how many strands in the bundle
[12:09:34] <Jymmm> About 5ga http://www.powerstream.com/Wire_Size.htm
[16:15:09] <_DJ_> gn8
[16:38:24] <t12> pca_home: multiple random samples from reveng produce those crc paramaters
[16:38:47] <t12> pca_home: next i'll do the request side, and try to get the other parts of the protocol as well
[16:41:45] <pcw_home> Unlike the Fanuc interface, It appears to be just a standard UART
[16:42:45] <pcw_home> (probably allowed them to use normal UCs for both ends)
[16:43:33] <pcw_home> 9x 8 bit chars each char with a single start and stop bit
[16:45:06] <pcw_home> for the best Mesa interface, I would need to add autosend to our UART (and an option to autosend on timerx)
[16:45:48] <pcw_home> (so the request can be sent automatically)
[16:50:19] <pcw_home> Probably a good thing, thinking about this, I could add autosend with just a FF or two and a FIFO pointer latch
[16:50:22] <pcw_home> (this should also work for SPI) I was considering an alternate xmit buffer but by latching. restoring the FIFO pointers
[16:50:23] <pcw_home> I can re-use the standard xmit buffer for autosend
[16:54:51] <t12> i'm suprised this is actually panning out!
[16:54:56] <t12> i'm used to these things ending in frusteration
[17:13:01] <t12> is there really a stop bit?
[17:20:17] <pcw_home> Yep
[17:20:18] <pcw_home> 1 10111010 01 01110011 01 11101110 01 11011001 01 11101111 01 10011100 01 10010101 01 11111100 01 00011011 0
[17:20:20] <pcw_home> 1 10111010 01 01110011 01 11111100 01 00111001 01 11101111 01 10011100 01 10010101 01 11111100 01 11101001 0
[17:20:21] <pcw_home> 1 10111010 01 01110011 01 11101111 01 00111001 01 11101111 01 10011100 01 10010101 01 11111100 01 11111010 0
[17:20:23] <pcw_home> S DA SS 73 SS COUNT? SS COUNT? SS EF SS 9C SS 95 SS FC SS CRC
[17:20:34] <pcw_home> (paste into a fixed width font)
[17:21:01] <pcw_home> the 01s are stop/starts
[17:21:25] <pcw_home> so just a UART
[17:21:52] <alex_joni> whee.. ascii art ;)
[17:22:25] <t12> ahh i see
[17:22:29] <t12> i decoded right through the stops i guess
[17:22:43] <t12> but i guess that just becomes a fixed shift and offset in the computed number
[17:23:44] <t12> its a 17bit encoder so the position might fall over into that EF byte
[17:23:56] <t12> i think they make 18s too
[17:24:32] <pcw_home> I presume its MSB first
[17:24:42] <t12> i only have 4MB of LA buffer and oersampled alot
[17:24:48] <t12> so i may not have covered a full rotation
[17:25:28] <pcw_home> in any case the Mesa UART should work fine +- some autosend patches for best performance
[17:25:51] <pcw_home> (and low jitter request pulse send)
[17:25:56] <t12> i think LSB first
[17:26:04] <t12> the earlier bits flip more
[17:26:14] <pcw_home> OK
[17:26:21] <t12> http://dodeca-t.com/output.txt
[17:28:02] <pcw_home> in any case, looks pretty simple. I got a Yaskawa drive to try with the 2 wire
[17:28:04] <pcw_home> Yaskawa encoder so I will mess with that next week, it may be similar
[17:28:39] <t12> yeah theres only so many ways to send the data
[17:29:17] <pcw_home> I guess theres Matshushta as well
[17:29:43] <pcw_home> and Okuma
[17:29:53] <t12> any idea of te rank order popularity of these various motors?
[17:30:16] <pcw_home> though Okuma may well be parallel/ multiple bytes
[17:30:35] <pcw_home> no Idea though Fanuc seems the most common
[17:30:39] <t12> theres a lot of pins on their encoder connector
[17:31:12] <pcw_home> Yeah but not enough for full parallel so it may be muxed bytes or some such
[17:32:03] <t12> ok lunch time
[17:32:06] <pcw_home> they tend to be expensive (Okuma absolute encoders) So I haven't snagged one yet
[17:32:46] <pcw_home> do you know what the request pulse is?
[17:49:54] <t12> ill figure the req out today
[17:50:14] <t12> also my crc stuff is likely wrong since i included the stop/start bits
[17:52:17] <pcw_home> Yeah though they are constant so reveng figured out a working CRC for you!
[17:58:40] <s1dev> anyone know of a cheap source of steel in the DFW area?
[18:06:28] <skunkworks_> andypugh: how are your new glade lathe routines?
[18:07:42] <andypugh> They work for me. There are quirks that I am working around rather than solving. I can't keyboard jog if the cursor is in the Glade window, and I typically need to press the "Go" button twice.
[18:08:16] <andypugh> s1dev: eBay. Seriously.
[18:08:33] <s1dev> what about shipping?
[18:08:49] <andypugh> You typically pay that anyway.
[18:09:23] <s1dev> is it significant?
[18:09:56] <andypugh> Depends what you want. Do you want multiple 20' lengths, or are you (like me) typically wanting 6" ? Different markets and different answers.
[18:10:14] <s1dev> 3'ish
[18:12:39] <andypugh> eBay is great for linking folk happy to buy the 3m bars, cut them and post. But that is their business. If you actually want a 3m length, then go to a stockholder (who only sell 3m lengths). If you want more than 10 x 3m lengths, then you need to be talking to the forge direct.
[18:12:55] <andypugh> 3' is awkwardly in the middle.
[18:15:12] <andypugh> For 3' of 25mm bar I would buy the 3m length and keep it. For 3' of 6" bar I would look on eBay. Not that the ratios are any different, but the sums involved are. I don't mind having £50 propped up in the corner of the garage not being used. £500 is a different thing.
[18:15:41] <DaViruz> thats a horrible mix of measuring systems right there
[18:17:50] <andypugh> :-) I did it on purpose.
[18:18:23] <andypugh> But I really can buy a 3m length of 1" bar
[18:18:33] <s1dev> 0.o
[18:19:17] <andypugh> s1dev: I am in the UK. We are _nearly_ metric.
[18:25:16] <Jymmm> If I press on the brake pedal when I have the drum removed, would it be "normal" for the piston to squirt fluid out?
[18:25:41] <Tom_itx> not unless it went beyond the cylinder end
[18:26:02] <Jymmm> I think it might have. I was bleeding the brakes and forgot to but the drum on
[18:26:13] <Tom_itx> guess you get to start over
[18:26:14] <Jymmm> (Just bought a 88 civic today)
[18:27:54] <andypugh> Jymmm: Yes. You may need to be careful in ensuring that the piston goes back in properly.
[18:28:05] <Jymmm> Tom_itx: Just hope I have enough fluid
[18:28:30] <Tom_itx> your local auto store has more
[18:28:40] <andypugh> In case you are in any doubt, you just messed up.
[18:28:46] <Jymmm> andypugh: It's stuck in the "out" position right now. I'll open the bleeder valve and hope it retracts.
[18:28:48] <Tom_itx> it would be good to get it to run clean anyway
[18:29:25] <Tom_itx> don't screw up the rubber cup when you retract it
[18:29:54] <Jymmm> I'll hold it back
[18:30:27] <Jymmm> Tom_itx: how do you get it to run clean?
[18:30:34] <Jymmm> or "flush" it out?
[18:30:49] <Valen> bleed the brakes
[18:30:54] <Valen> press the pedal down
[18:30:58] <Valen> open the bleed
[18:31:08] <Valen> close the bleed when the pedal hits the floor
[18:31:14] <Valen> slowly raise pedal
[18:31:38] <Tom_itx> stick a tube on the bleeder and put the end of it in a jar
[18:31:47] <Jymmm> I filled a container full of clean fluid; attched a hose to the bleeder and stuck it in the fluid. opened vlave, and the clean fluid was no longer clean
[18:31:52] <Valen> my way works with 2 people
[18:32:16] <Valen> the clean fluid comes from the resivour, the jar is there so when it sucks back it doesn't suck air
[18:32:40] <Valen> I suggest getting a brake bleeder kit, they are like $10 and have a little one way valve in them
[18:32:47] <Jymmm> Valen: Right, but it spit out dirty fluid into the jar
[18:33:03] <Tom_itx> keep spitting until it's clean
[18:33:16] <andypugh> t12: Odd what you can tell about folk from IRC. I think I can guess that you speak French and use an AZERTY keyboard, otherwise "pca_home" is a really wierd typo.
[18:33:21] <Valen> if you do it with 2 people and a clear tube you can watch it change as soon as the fluid is clear
[18:33:53] <Tom_itx> yeah it's easier with 2 ppl
[18:34:27] <Jymmm> the revervoir emptys faster thna the jar =)
[18:34:34] <Tom_itx> keep it full
[18:34:39] <andypugh> Jymmm: http://www.ebay.com/itm/Mity-Vac-Vacuum-Pump-Brake-Bleeder-Kit-MV8000-/130935113182
[18:34:49] <Tom_itx> start with the furthest wheel
[18:34:57] <Jymmm> I am
[18:35:03] <Jymmm> andypugh: what does the pump do?
[18:35:13] <andypugh> It sucks
[18:35:24] <Jymmm> andypugh: instead of using the brake pedal?
[18:35:27] <andypugh> Yes
[18:35:30] <Jymmm> ah
[18:35:48] <Jymmm> but dont you have to "work" the fluid thru the master cyl?
[18:35:56] <andypugh> So you suck at the brake end
[18:36:03] <Jymmm> via the brake pedal?
[18:36:15] <t12> principal component analysis leaking through
[18:36:18] <andypugh> There is a "recuperation" hole.
[18:36:22] <t12> i'm just sloppy at the computer
[18:36:25] <Jymmm> andypugh: But doesn't the master cyl have to move?
[18:36:35] <Jymmm> andypugh: ah ok
[18:36:53] <Jymmm> andypugh: THAT i didnt know =)
[18:37:19] <t12> i dont speak french or have an azerty kb tho
[18:38:00] <andypugh> <hands back the Sherlock Homes certificate>
[18:39:53] <andypugh> Jymmm: You may not be able to see it, but on bike master cylinders you can see that the air bubbles come up through a little hole in the master cylinder, and you can also often see that the piston seal passes that little hole.
[18:40:30] <Tom_itx> yup, don't let the master cyl get too low
[18:40:32] <Jymmm> andypugh: Ah, ok. I'll look after hitting the AP store =)
[18:41:02] <Tom_itx> i generally start with the RR
[18:41:05] <Tom_itx> then the RF
[18:41:23] <Tom_itx> well, after the LR
[18:41:43] <Jymmm> Well, he LR is jacked up atm =)
[18:42:04] <andypugh> My dad had a Landrover where we never managed to bleed the brakes.
[18:42:18] <Tom_itx> some are alot more difficult to bleed for sure
[18:42:19] <andypugh> We replaced everything except the servo.
[18:42:22] <Jymmm> lol, these were TOO soft for my tste
[18:42:28] <Jymmm> taste
[18:42:43] <Tom_itx> auto or manual?
[18:42:49] <Tom_itx> bleed the clutch cyl too
[18:42:57] <Jymmm> auto
[18:43:13] <andypugh> A friend bought it off us. He worked out that the brake servo was a tractor part. But had been fitted by Landrover, the vehicle serial number was 3.
[18:44:29] <andypugh> I think he got it all working by ignoring the bleed sequence in the manual (using a bleed nipple that was on the servo, but shouldn't have been there)
[18:44:41] <Jymmm> I forget, can you use tranny fluid instead of PS fluid?
[18:45:30] <Tom_itx> if i were in the middle of nowhere and that's all i had i would
[18:45:45] <Tom_itx> you're going to the AP store anyway
[18:46:35] <Jymmm> Yeah, was just curious. They say you could on older 1970's cars
[18:48:51] <andypugh> In Cuba apparetly they use Shampoo.
[18:49:11] <Jymmm> good to know =)
[18:49:25] <Jymmm> But all the cars are 1950's American
[18:49:33] <andypugh> (which is largely silicone oil, which I think they have figured out how to purify out)
[18:49:52] <andypugh> Yeah, mainly because you are really rather mean to them.
[18:50:48] <jdh> gotta keep them damned commies in line.
[18:51:22] <jdh> my daughter is trying to convince us to send her to Cuba for a semester to 'study'
[18:52:00] <andypugh> Actually, no, you don't (though I suspect you have your tongue firmly in your cheek there)
[18:52:47] <andypugh> And, havung typed it, what the heck does "tongue in your cheek" mean? what a strange phrase
[18:52:55] <jdh> I'm amazed by the number of my fellow americans who think 'nobody' is allowed to travel to cuba.
[18:53:06] <jdh> vs. just us
[18:53:37] <andypugh> If she _can_ go then I think she would get a lot out of it, to be honest.
[18:54:17] <jdh> the only thing that would prevent her from going is my willingness to pay.
[18:54:46] <andypugh> http://youtu.be/qL0IItt-ff8
[18:57:45] <jdh> not going in my vehicles.
[19:01:30] <andypugh> I suspect it isn't their first choice.
[19:02:58] <andypugh> Brake fluid is either silicone oil (and shampoo is generally silicone oil, dimethicone) or glycol-ester. I don't know what sugar + alchohol makes (other than a good party)
[19:04:19] <jdh> speaking of materials...
[19:04:45] <jdh> is there anything you can use as a cover gas for TIG welding zirconium other than argon and helium?
[19:05:33] <DaViruz> probably most noble gases?
[19:06:02] <DaViruz> i doubt there'd be any advantages over argon though..
[19:06:12] <jdh> certainly not cost.
[19:07:17] <Valen> lol wow he is young there
[19:08:21] <Valen> sugar probably wouldn't do such a good job in a "performance" application
[19:08:43] <Valen> the sugar would caramelise
[19:09:22] <Valen> actually the shampoo is mostly water, so that seems like a bad plan
[19:12:06] <jdh> wiki says DOT2 is castor oil + alcohol
[19:12:29] <Valen> half the power steering stuff still had whale oil in it into the 1970s
[19:15:12] <Tom_itx> you could tell too, it stinks
[19:15:13] <andypugh> Valen: Not that much water in some shampoos.
[19:15:40] <andypugh> I was using whale oil for heat treatment in the 90s
[19:16:36] <andypugh> I think the reasoning was that it worked, they had hundreds of gallons in stock, and _not_ using it all up would be morally worse.
[19:37:11] <jesseg> sounds like a whale of a fun thing
[19:38:46] <jdh> unless you are a whale.
[19:45:55] <andypugh> Whales are ungulates, you know. When they have hooves, they are two-toed.
[19:55:43] <uw> wait wat
[21:31:33] <somenewguy> woo hoo network access in the shop finally!
[21:31:41] <somenewguy> and reliable even
[21:31:54] <Tom_itx> simplifies things ehh?
[21:34:35] <somenewguy> sooooo much
[21:34:50] <somenewguy> don't need to rely on printouts and my horrible memory
[21:35:20] <somenewguy> looking up reference material is a big plus, maybe now I can get my joystick doing the things it should
[21:39:40] <MacGalempsy> evening all
[21:43:42] <somenewguy> that was quick
[22:08:35] <somenewguy> so I see the tool table, but where does the fixture table hide?
[22:12:55] <Tom_itx> in the .var file i think
[22:15:03] <Tom_itx> http://www.linuxcnc.org/docs/2.4/html/gcode_coordinates.html
[22:23:54] <somenewguy> yep, straightforward enough, thanks
[22:24:13] <somenewguy> shame there is nothing in axis to view them directly
[22:24:33] <somenewguy> I have two vices and a goldfishes memory
[22:24:51] <somenewguy> although I bet there is an add on somewhere, now that I have the internet I can find and GET it!
[22:28:30] <NickParker> So is there any sort of guide for getting set up with my mesa 6i25 card? I don't mind if it's absurdly esoteric.
[22:31:13] <NickParker> I've got the card in a PC with reasonably (20k) latency, and I know what sorts of inputs the various systems of my mill need. what do I do from here?
[22:33:54] <NickParker> I've got almost the exact same mill as kwallace btw, and IIRC I need to use type 9 in stepgen
[22:38:46] <Tom_itx> start mapping the io to the machine
[22:39:00] <Tom_itx> pwm, io etc
[22:39:15] <Tom_itx> the bit file will determine how the mesa card is set up
[23:24:56] <NickParker> thanks for the pointer Tom_itx, i'm reading about the bit file now.