Do you suppose that the older Debian distribution is more stable than 8.6 ? Or the previous kernel ? Specifically, what kind of shielding do you have on your cabling ? I switched from the older version since I would get lockups after very long prints but with this setup, it's only an hour to two hours in and it fails. I was tempted to perfom a mock print, e.g. no filament, setting the extruder to 0 and the HBP to 35 and give it a whirl. Then try the extruder at 36 and the HBP at 0 to see if a crash occurs. This would be to see if a specific ADC channel is giving grief. Is there any possibility of oxidation occurring at the BBB analog in pins ? I haven't tried using a different channel.
On Wednesday, December 28, 2016 at 12:14:10 AM UTC-5, Daren Schwenke wrote: > > Better info: > machinekit@beaglebone:/usr/bin$ dpkg -l machinekit* > ii machinekit 0.1.14581376 armhf PC based motion controller > for re > machinekit@beaglebone:/usr/bin$ cat /etc/dogtag > Machinekit Debian Image 2016-02-07 > > > On Wednesday, December 28, 2016 at 12:05:04 AM UTC-5, Daren Schwenke wrote: >> >> I've been running the following older debian 7 image for many months, >> with many 6+ hour prints, and zero lockups (knock on wood). >> Not sure what version of this image this was, but here are some file >> dates: >> machinekit@beaglebone:/etc$ uname -a >> Linux beaglebone 3.8.13-xenomai-r78 #1 Sat Sep 26 16:54:43 UTC 2015 >> armv7l GNU/Linux >> machinekit@beaglebone:/etc$ ls -al /usr/bin/linuxcnc >> -rwxr-xr-x 1 root root 24129 Mar 16 2016 /usr/bin/linuxcnc >> machinekit@beaglebone:/usr/bin$ ls -al /usr/bin/hal_temp_bbb >> -rwxr-xr-x 1 root root 9567 Mar 16 2016 /usr/bin/hal_temp_bbb >> >> >> I had experienced lockups in other versions, but my hardware has also >> changed since then with better shielding and smoother power to the BBB. >> haltalk (using Machineface) in that build can consume 30% proc at times, >> and this seems to cause random blobbies in the prints. >> The effect is worse at high segment speeds (160mm/sec, all arcs). >> Disconnecting mostly eliminates them and the print continues, so that's >> what I do. It works so I'm not messing with it. >> >> >> On Tuesday, December 20, 2016 at 9:40:39 AM UTC-5, Jonathan Cohen wrote: >>> >>> I took my printer down since the hot end was locking up during a long >>> print. I have posted about this over the past year. I recently purchased >>> a new microsd card, put on a new image of Jessie 8.6. I also replaced the >>> hotend thermistor, redid all of the wiring to decrease the effect of EMI >>> near the cartridge heater leads. I got off one reasonable 4hr test print. >>> However, on an overnight, the ADC locked up at maybe the 5 hr mark on a >>> 6.5 hr print (not heating). So, I recall Charles writing about the ADC >>> driver being a bit temperamental, sensitive to over-voltage and such. I am >>> powering the beaglebone through the AUX in on the CRAMPS board through a 5V >>> LM2596 buck regulator ( >>> https://www.amazon.com/gp/product/B00WEBJRE8/ref=oh_aui_search_detailpage?ie=UTF8&psc=1). >>> >>> I recently went a made sure that it was set to 5V to be sure. Prior to >>> that I was using a 5V DC adapter where I still had this issue. My cable >>> management is reasonable so I do not think wires are moving about. >>> >>> 1. So what I want to know is whether this is a particular problem that >>> is always going to be sporadic during a print ? >>> 2. Is there no straightforward solution ? I tried several microSD >>> cards, thinking it was a quality issue or corruption of the card but that >>> did not fix it. I also though my thermistor may have been bad, or the >>> wiring out both so I swapped that out, to no avail. >>> 3. There was mention of a buggy ADC driver in an older kernel. Has >>> this same driver been brought along to the new Debian release / >>> 4. I did see in the forums about using an external ADC, is this possible >>> to implement on the CRAMPS (I am not sure if the SPI is broken out). >>> 5. Is this an issue that everyone who uses the Beaglebone/CRAMPS combo, >>> sees ? >>> 6. I slice in Cura, is this a G-code problem ? ( I am not using the >>> remapped G-codes0 >>> 7. 5. Does this occur with the Replicape board ? >>> 8. Is it time to throw in the towel ? >>> >>> I really like the using an embedded platform for running a printer. >>> This is a bit frustrating as I can not figure it out. Everyone on this >>> board has been really helpful. I have scoured the board looking for >>> solutions, but am still at a loss (probably since I am not a programmer or >>> engineer !) >>> >>> Thanks, >>> >>> Jonathan >>> >>> -- 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.