#linuxcnc-devel Logs
Feb 02 2018
#linuxcnc-devel Calendar
06:56 AM skunkworks_: pcw_home, changed the field voltage to a 12v supply. still get the same error
06:58 AM skunkworks_: every other launch of linuxcnc doesn't see the 7i76. master act similar except I get an extra error - first hm2/hm2_7i92.0: Error after doit clear
07:27 AM pcw_home: weird
07:27 AM skunkworks_: I can try a 5i25 for grins or another 7i92
07:27 AM pcw_home: what bitfile on 7I92?
07:28 AM skunkworks_: 7i92_7i76x2D.bit
07:28 AM pcw_home: (I used 7i92_7i76x1D.bit )
07:28 AM pcw_home: should be the same
07:29 AM pcw_home: Never see such behavior
07:29 AM pcw_home: 5V OK?
07:29 AM skunkworks_: I tried both from fpga and external. but I could try a better external
07:30 AM pcw_home: I used FPGA source
07:30 AM skunkworks_: (the 7i92 is powered by usb)
07:30 AM skunkworks_: I could power the 7i92 by external 5v
07:31 AM skunkworks_: maybe something goofy there
07:32 AM archivist: do some newer usb ports switch power on/off (negotiating charge rates etc)
07:34 AM pcw_home: yeah try a different 5V then a different 7I92
07:34 AM skunkworks_: same
07:35 AM skunkworks_: (2a 5v supply)
07:35 AM skunkworks_: measures 5.04v at the 7i92
07:36 AM skunkworks_: same remote error 8
07:37 AM skunkworks_: the card seems to work..
07:38 AM skunkworks_: and it finds the smart serial board if I plug it in
07:38 AM skunkworks_: (and an the analog inputs work on the smart serial board)
07:39 AM skunkworks_: the stepgens are working
07:40 AM skunkworks_: I could try a different computer
07:40 AM skunkworks_: I always have the odd errors :)
07:43 AM pcw_home: Yeah thats a first AFAIK
07:43 AM skunkworks_: I have tried 2 different interconnect cables too (25 pin) one for sure is IEEE mumble mubble
07:44 AM skunkworks_: the only other thing is the card is set to 10.10.10.10
07:44 AM pcw_home: the errors look higher level than sserial (especially the doit error )
07:45 AM skunkworks_: I will reboot this thing.. It has been in and out of power saving for a long time.
07:45 AM pcw_home: I tried 2.7.12 and master (though the 7i96 branch)
07:46 AM pcw_home: and tested at 10.10.10.10
07:50 AM skunkworks: reboot didn't help
08:01 AM pcw_home: might try the 7I76 on a 5i25, but those are very weird errors
08:04 AM skunkworks: I will
09:27 AM skunkworks: pcw_home it is something in the hal file. I can load and unload the eth driver manually through halcmd with no issue. I deleted everything below the functions and now linuxcnc loads without error mon
09:27 AM skunkworks: (and finds the card every time)
09:39 AM skunkworks: heh
09:39 AM skunkworks: duh
09:40 AM skunkworks_: setp hm2_7i92.0.gpio.004.invert_output true
09:40 AM skunkworks_: setp hm2_7i92.0.gpio.008.invert_output true
09:40 AM skunkworks_: setp hm2_7i92.0.gpio.010.invert_output true
10:10 AM pcw_home: ahh the last one is going to cause pain and suffering on the first sserial port
10:11 AM skunkworks_: :)
10:27 AM skunkworks_: seemed to kinda recover ;)
10:40 AM pcw_home: the drive must probe the sserial devices before the hal settings so the card is discovered but then it goes bad
10:40 AM pcw_home: s/drive/driver/
01:29 PM skunkworks: that works a lot better
01:31 PM pcw_mesa: yeah inverting the serial TX data is going to confuse things a bit
01:31 PM skunkworks: so - channel 0 is 5mv +/-1mv - channel 1 is 1mv +/-1mv. unconnected
01:32 PM skunkworks: sound plausible?
01:34 PM pcw_mesa: possible, I think thats a junkbox 7I87 so the calibration may be off a few MV
01:34 PM skunkworks: sounds good
01:35 PM pcw_mesa: mV I mean, if is MV you better step back
01:35 PM skunkworks: poof!
01:37 PM skunkworks: it really is amazing - just finds everything. (when you don't mess up the hal file)
01:37 PM pcw_mesa: I think the 7I87 has some modes
01:38 PM pcw_mesa: the driver should print them
01:38 PM pcw_mesa: or mesaflash --sserial
01:40 PM skunkworks: I was thinking I could pretty much use this card to fire and sense the arc-ok..
01:40 PM skunkworks: (would just have to use one of the analog inputs as an input - don't know how I feel about that)
01:40 PM skunkworks: ;)
01:41 PM skunkworks: seems like a waste...
01:42 PM pcw_mesa: probably OK though I really should make a plasma oriented sserial remote
01:42 PM skunkworks: that would be cool
01:42 PM pcw_mesa: could even do that minimum peak sensing
01:43 PM pcw_mesa: assuming that's not patented :-(
01:43 PM skunkworks: SOFTWARE MODE AnalogInput_DigitalOutput [index 00]
01:43 PM skunkworks: SOFTWARE MODE AnalogInput16bit*4,8bit*4_DigitalOutput [index 01]
01:43 PM skunkworks: SOFTWARE MODE AnalogInput16bit*6_DigitalOutput [index 02]
01:45 PM pcw_mesa: I think the default is 12 bit data but fast, 16 bit data is slow
01:48 PM pcw_mesa: yeah 3 KHz bandwidth 12 bit 200 Hz bandwidth 16 bits
01:54 PM skunkworks: heh
01:55 PM pcw_mesa: its always averaged/oversampled ( 50 KHz sample rate per channel )
02:42 PM seb_kuzminsky: i wonder if it's time to get rid of our own clone of yapps from 2006 yet
02:58 PM skunkworks: Board hm2_7i92.0.7i87.0.1 Hardware Mode 0 = standard
02:58 PM skunkworks: Board hm2_7i92.0.7i87.0.1 Software Mode 0 = analoginput_digitaloutput
02:58 PM skunkworks: Board hm2_7i92.0.7i87.0.1 Software Mode 1 = analoginput16bit*4,8bit*4_digitaloutput
02:58 PM skunkworks: Board hm2_7i92.0.7i87.0.1 Software Mode 2 = analoginput16bit*6_digitaloutput
03:08 PM linuxcnc-build: build #3522 of 1405.rip-wheezy-armhf is complete: Failure [4failed compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1405.rip-wheezy-armhf/builds/3522 blamelist: Sebastian Kuzminsky <seb@highlab.com>
03:11 PM seb_kuzminsky: Command terminated by signal 11 Command being timed: "make V=1 -j3"
03:11 PM seb_kuzminsky: same old armhf buildslave flakiness
03:34 PM linuxcnc-build: build #5371 of 0000.checkin is complete: Failure [4failed] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/5371 blamelist: Sebastian Kuzminsky <seb@highlab.com>
03:39 PM skunkworks: pcw_mesa, all you would need is poe and you could power everything from the computer! ;)
03:41 PM pcw_mesa: I don't suspect many laptops have POE injectors...
03:41 PM skunkworks: probably
03:42 PM pcw_mesa: I always just steal USB power, (though it wont run a 7I77)
03:46 PM seb_kuzminsky: linuxcnc-build: force build --branch=2.7-packaging-fixes 0000.checkin
03:46 PM linuxcnc-build: build forced [ETA 2h03m34s]
03:46 PM linuxcnc-build: I'll give a shout when the build finishes
03:49 PM skunkworks: I went back to usb power after I figured out the problem'
03:49 PM skunkworks: It is just so convenient.
03:50 PM skunkworks: I can throw it in my back pack and goof around with it anywhere
03:50 PM skunkworks: I should build a little 4 axis disk drive mill that I can carry with me ;)
05:13 PM jepler: mini-dinstall [34430543552] ERROR: Unable to install "/store/debs/mini-dinstall/incoming/transmission_2.84-0.2+deb8u1_kfreebsd-amd64.changes"; adding to screwed list
05:13 PM jepler: funny
05:28 PM linuxcnc-build: Hey! build 0000.checkin #5372 is complete: Success [3build successful]
05:28 PM linuxcnc-build: Build details are at http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/5372
05:44 PM jepler: https://www.youtube.com/watch?v=G_gKIvDgBHs (manual) milling features in the latest update from the Xerox Alto restoration guys
05:48 PM jepler: seb_kuzminsky: I don't honestly know why we have a copy of yapps anymore
05:48 PM jepler: we depend on yapps2-runtime unconditionally, so it must be on all our systems
08:41 PM linuxcnc-build: build #558 of 1540.rip-jessie-armhf is complete: Failure [4failed compile runtests] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1540.rip-jessie-armhf/builds/558 blamelist: Sebastian Kuzminsky <seb@highlab.com>
08:41 PM linuxcnc-build: build #5373 of 0000.checkin is complete: Failure [4failed] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/5373 blamelist: Sebastian Kuzminsky <seb@highlab.com>
09:34 PM jepler: seb_kuzminsky: well I really failed at removing yapps didn't I
09:34 PM jepler: thanks for trying harder
09:35 PM jepler: feel free to squash my commit with yours
10:06 PM seb_kuzminsky: jepler: eh, you did the hard part of figuring out whether to keep it or toss it