For kicks, I just moved 843,844 from hal_pru_generic step/dir to toggling with emcmot-0-enable, and they toggle with estop so it's not an issue with the specific pins being burned here either.
On Sunday, November 5, 2017 at 7:37:20 AM UTC-5, Daren Schwenke wrote: > > That's the confusing part. Things work fine with hal_bb_gpio, and not > with hal_pru_generic. > > The estop hardware loop output goes high when I press the virtual estop, > which feeds back into my estop input and it enables, and my LED verifies > that. > Which of course sends my machine-power on p810 high as it's tied to > emcmot-0-enable, and my LED verifies that. > > And the halcmd show says I should have PWM at a given level with > everything enabled and such... and no output. > > > On Sunday, November 5, 2017 at 7:22:30 AM UTC-5, Charles Steinkuehler > wrote: >> >> I'm not sure what's wrong, but since it sounds like absolutely nothing >> is happening, I suggest verifying some of the basics: Make sure you >> can twiddle I/O pins via config-pin and halcmd: >> >> Since you're using a 4.x kernel, I don't think the overlay lines in >> your bebopr_cape.bbio file will do anything. Make sure you have the >> universal cape loaded via u-boot overlays and test twiddling an I/O >> pin using the config-pin command. >> >> If that works, try loading just the HAL part of your configuration >> manually (I think you can just run the arcus_3d_c1.py and bebopr.py >> files). If that doesn't work, you can just make a simple test HAL >> file to load the hal_bb_gpio driver and create a servo thread. >> >> Make sure you start the threads (halcmd start) and see if twiddling >> the I/O pins (ie: bb_gpio.p8.out-07 & friends) via halcmd set has any >> affect. >> >> If either of those fail, the PRU is probably not working for the same >> reason...get the above fixed and things ought to work better. >> >> On 11/4/2017 11:24 PM, Daren Schwenke wrote: >> > I noticed most of my pins were tied to PRU 1, so I moved one pin so all >> the >> > stepgens now sit on PRU 1, and changed to PRU=1. >> > No change. >> > The board is sitting on my desk with nothing other than a jumper >> closing my >> > estop loop from P807 to P809. That toggles appropriately when powering >> the >> > machine. >> > I'm down to testing with just an LED+resistor at <4ma draw and I get >> > nothing from any of the hal_pru_generic output pins, and no errors I >> can >> > see. >> > Here is a pastebin of my /var/log/linuxcnc.log from a single >> startup,estop >> > off, and shutdown. I had to trim some of the shutdown from the end. >> > https://pastebin.com/96ZWP0KD >> > >> > On Saturday, November 4, 2017 at 7:27:51 PM UTC-4, Daren Schwenke >> wrote: >> >> >> >> >> >> >> >> On Saturday, November 4, 2017 at 7:21:21 PM UTC-4, Daren Schwenke >> wrote: >> >>> >> >>> Suppose this would be useful info. Started with this image: >> >>> >> >>> >> https://rcn-ee.com/rootfs/bb.org/testing/2017-10-29/stretch-iot/BBB-blank-debian-9.2-iot-armhf-2017-10-29-4gb.img.xz >> >> >>> Did the stuff for switching to rt kernel the /opt/scripts/tools way: >> >>> root@beaglebone:~# cd /opt/scripts/tools/ >> >>> root@beaglebone:/opt/scripts/tools# git pull >> >>> Already up-to-date. >> >>> root@beaglebone:/opt/scripts/tools# ./update_kernel.sh >> --ti-rt-channel >> >>> --lts-4_4 >> >>> >> >>> Added machinekit repo, installed machinekit-rt-preempt >> >>> hal_pru_generic is of course there, so is the .so file >> >>> >> >> debian@beaglebone:~/Arcus-3D-C1-BeBoPr$ ls -al >> >> /usr/lib/linuxcnc/rt-preempt/pru_generic.bin >> >> lrwxrwxrwx 1 root root 40 Nov 3 06:17 >> >> /usr/lib/linuxcnc/rt-preempt/pru_generic.bin -> >> >> /usr/lib/linuxcnc/prubin/pru_generic.bin >> >> debian@beaglebone:~/Arcus-3D-C1-BeBoPr$ ls -al >> >> /usr/lib/linuxcnc/prubin/pru_generic.bin >> >> -rw-r--r-- 1 root root 848 Nov 2 16:31 >> >> /usr/lib/linuxcnc/prubin/pru_generic.bin >> >> debian@beaglebone:~/Arcus-3D-C1-BeBoPr$ ls -al >> >> /usr/lib/linuxcnc/rt-preempt/hal_pru* >> >> -rw-r--r-- 1 root root 18704 Nov 2 16:31 >> >> /usr/lib/linuxcnc/rt-preempt/hal_prudebug.so >> >> -rw-r--r-- 1 root root 52360 Nov 2 16:31 >> >> /usr/lib/linuxcnc/rt-preempt/hal_pru_generic.so >> >> -rw-r--r-- 1 root root 19008 Nov 2 16:31 >> >> /usr/lib/linuxcnc/rt-preempt/hal_pru.so >> >> debian@beaglebone:~/Arcus-3D-C1-BeBoPr$ >> >> >> >> >> >> >> >>> >> >>> >> >>> On Saturday, November 4, 2017 at 7:00:23 PM UTC-4, Daren Schwenke >> wrote: >> >>>> >> >>>> BBGW with a 'mostly' BeBoPr cape. Hand wired, very basic, with some >> >>>> pins moved so they didn't step on wifi. >> >>>> Booting from eMMC. >> >>>> >> >>>> Config here. Starts, no errors. >> >>>> https://github.com/Arcus-3d/Arcus-3D-C1-BeBoPr >> >>>> >> >>>> GPIO works, triggering estop and machine power correctly. >> >>>> halcmd shows it should be outputting PWM. >> >>>> 576 bit IN TRUE hpg.pwmgen.00.out.02.enable >> >> >>>> --l- <== f0-pwm-enable >> >>>> 576 u32 IN 0x0000034D hpg.pwmgen.00.out.02.pin >> >> >>>> --l- >> >>>> 576 float IN 1 hpg.pwmgen.00.out.02.scale >> >> >>>> 0.000010 --l- >> >>>> 576 float IN 0.6980393 hpg.pwmgen.00.out.02.value >> >> >>>> 0.000010 --l- <== f0-pwm >> >>>> >> >>>> >> >>>> >> >>>> Starting with PWM, cause that was easy to measure with my improvised >> >>>> ghetto soundcard oscilloscope. >> >>>> Verified... pins, not flipping. >> >>>> >> >>>> Relevant lines from /boot/uEnv.txt >> >>>> debian@beaglebone:~$ egrep -v "^#|^\s*$" /boot/uEnv.txt >> >>>> uname_r=4.4.91-ti-rt-r136 >> >>>> enable_uboot_overlays=1 >> >>>> uboot_overlay_pru=/lib/firmware/AM335X-PRU-UIO-00A0.dtbo >> >>>> enable_uboot_cape_universal=1 >> >>>> cmdline=coherent_pool=1M net.ifnames=0 quiet >> >>>> debian@beaglebone:~$ >> >>>> >> >>>> Any ideas? >> >>>> >> >>> >> > >> >> >> -- >> Charles Steinkuehler >> cha...@steinkuehler.net >> > -- website: http://www.machinekit.io blog: http://blog.machinekit.io github: https://github.com/machinekit --- You received this message because you are subscribed to the Google Groups "Machinekit" group. To unsubscribe from this group and stop receiving emails from it, send an email to machinekit+unsubscr...@googlegroups.com. Visit this group at https://groups.google.com/group/machinekit. For more options, visit https://groups.google.com/d/optout.
