[SOLVED] EMC stops functioning but GUI continue

More
23 Jan 2010 18:27 - 15 Feb 2010 18:28 #1601 by Pentax
So i just built a retrofit for an old CNC. it has servos with nothing but tachs, and i'm using glass scales to cheat as encoders. It works fine, for about a minute. then the servos stop moving, the DRO stops working, and the little tool in the backplot area will move given jog commands and the velocity seems to be active when I jog. I'm not sure what to say about this, or even if there is a log file somehwere since the program doesn't actually crash. Its odd that the GUI virtual tool would move without the motoros moving and no feedback. If i close and reopen EMC the DRO will work instantly, but the servos are completley unprediactable whether they will function or not. My amp is a basic velocity -+10 system that (servo dynamics boards) that only takes single ended input, and they show green status lights through the whole thing.

does anyone have a clue might what be happening? I get following errors a lot, and I'm not sure what that's about. there is very little baclash on the system and with increased FERROR and MINFERROR it seems to function. Basically it works great for about a minute, then it just dies, with no indicated error from my amps (when the servos become unresponsive, they are still braked).

thanks,
A
Last edit: 15 Feb 2010 18:28 by Pentax.

Please Log in or Create an account to join the conversation.

More
23 Jan 2010 20:42 #1603 by PCW
Can you post the dmesg log to somewhere like filebin.ca?

in a terminal window:
dmesg > file_to_upload

Please Log in or Create an account to join the conversation.

More
23 Jan 2010 21:10 #1604 by Pentax
I'll work on getting a full dmesg on here. I also noticed, that i restart EMC with the AMPS on, it will error out and give me a big history (i assume to be the dmesg?) I look at the processes and i can fine a common element, which makes sense. watchdog is biting. i'm going to try to FALSE out the dog, and i'll put the full report on here as well.

Thanks,
A

Please Log in or Create an account to join the conversation.

More
23 Jan 2010 22:27 - 23 Jan 2010 22:28 #1605 by Pentax
Included is the Dmesg and config files i'm running. its my first time tuning PID and i'm having a few issues, but I'm constantly being interrupted by the biting dog.

[ 334.760607] RTAI[math]: loaded.
[ 334.842797] hm2: loading Mesa HostMot2 driver version 0.15
[ 334.848610] hm2_7i43: loading HostMot2 Mesa 7i43 driver version 0.3
[ 335.112787] hm2/hm2_7i43.0: 48 I/O Pins used:
[ 335.112799] hm2/hm2_7i43.0: IO Pin 000 (P4-01): Encoder #1, pin B (Input)[ 335.112806] hm2/hm2_7i43.0: IO Pin 001 (P4-03): Encoder #1, pin A (Input)[ 335.112813] hm2/hm2_7i43.0: IO Pin 002 (P4-05): Encoder #0, pin B (Input)[ 335.112819] hm2/hm2_7i43.0: IO Pin 003 (P4-07): Encoder #0, pin A (Input)[ 335.112826] hm2/hm2_7i43.0: IO Pin 004 (P4-09): Encoder #1, pin Index (Input)
[ 335.112841] hm2/hm2_7i43.0: IO Pin 005 (P4-11): Encoder #0, pin Index (Input)
[ 335.112848] hm2/hm2_7i43.0: IO Pin 006 (P4-13): PWMGen #1, pin Out0 (PWM or Up) (Output)
[ 335.112856] hm2/hm2_7i43.0: IO Pin 007 (P4-15): PWMGen #0, pin Out0 (PWM or Up) (Output)
[ 335.112863] hm2/hm2_7i43.0: IO Pin 008 (P4-17): PWMGen #1, pin Out1 (Dir or Down) (Output)
[ 335.112870] hm2/hm2_7i43.0: IO Pin 009 (P4-19): PWMGen #0, pin Out1 (Dir or Down) (Output)
[ 335.112878] hm2/hm2_7i43.0: IO Pin 010 (P4-21): PWMGen #1, pin Not-Enable (Output)
[ 335.112885] hm2/hm2_7i43.0: IO Pin 011 (P4-23): PWMGen #0, pin Not-Enable (Output)
[ 335.112891] hm2/hm2_7i43.0: IO Pin 012 (P4-25): IOPort
[ 335.112896] hm2/hm2_7i43.0: IO Pin 013 (P4-27): IOPort
[ 335.112902] hm2/hm2_7i43.0: IO Pin 014 (P4-29): IOPort
[ 335.112907] hm2/hm2_7i43.0: IO Pin 015 (P4-31): IOPort
[ 335.112912] hm2/hm2_7i43.0: IO Pin 016 (P4-33): IOPort
[ 335.112918] hm2/hm2_7i43.0: IO Pin 017 (P4-35): IOPort
[ 335.112923] hm2/hm2_7i43.0: IO Pin 018 (P4-37): IOPort
[ 335.112928] hm2/hm2_7i43.0: IO Pin 019 (P4-39): IOPort
[ 335.112934] hm2/hm2_7i43.0: IO Pin 020 (P4-41): IOPort
[ 335.112939] hm2/hm2_7i43.0: IO Pin 021 (P4-43): IOPort
[ 335.112944] hm2/hm2_7i43.0: IO Pin 022 (P4-45): IOPort
[ 335.112949] hm2/hm2_7i43.0: IO Pin 023 (P4-47): IOPort
[ 335.112955] hm2/hm2_7i43.0: IO Pin 024 (P3-01): IOPort
[ 335.112960] hm2/hm2_7i43.0: IO Pin 025 (P3-03): IOPort
[ 335.112965] hm2/hm2_7i43.0: IO Pin 026 (P3-05): IOPort
[ 335.112971] hm2/hm2_7i43.0: IO Pin 027 (P3-07): IOPort
[ 335.112976] hm2/hm2_7i43.0: IO Pin 028 (P3-09): IOPort
[ 335.112981] hm2/hm2_7i43.0: IO Pin 029 (P3-11): IOPort
[ 335.112987] hm2/hm2_7i43.0: IO Pin 030 (P3-13): IOPort
[ 335.112992] hm2/hm2_7i43.0: IO Pin 031 (P3-15): IOPort
[ 335.112997] hm2/hm2_7i43.0: IO Pin 032 (P3-17): IOPort
[ 335.113002] hm2/hm2_7i43.0: IO Pin 033 (P3-19): IOPort
[ 335.113008] hm2/hm2_7i43.0: IO Pin 034 (P3-21): IOPort
[ 335.113013] hm2/hm2_7i43.0: IO Pin 035 (P3-23): IOPort
[ 335.113018] hm2/hm2_7i43.0: IO Pin 036 (P3-25): IOPort
[ 335.113024] hm2/hm2_7i43.0: IO Pin 037 (P3-27): IOPort
[ 335.113029] hm2/hm2_7i43.0: IO Pin 038 (P3-29): IOPort
[ 335.113034] hm2/hm2_7i43.0: IO Pin 039 (P3-31): IOPort
[ 335.113040] hm2/hm2_7i43.0: IO Pin 040 (P3-33): IOPort
[ 335.113045] hm2/hm2_7i43.0: IO Pin 041 (P3-35): IOPort
[ 335.113050] hm2/hm2_7i43.0: IO Pin 042 (P3-37): IOPort
[ 335.113055] hm2/hm2_7i43.0: IO Pin 043 (P3-39): IOPort
[ 335.113061] hm2/hm2_7i43.0: IO Pin 044 (P3-41): IOPort
[ 335.113066] hm2/hm2_7i43.0: IO Pin 045 (P3-43): IOPort
[ 335.113071] hm2/hm2_7i43.0: IO Pin 046 (P3-45): IOPort
[ 335.113077] hm2/hm2_7i43.0: IO Pin 047 (P3-47): IOPort
[ 335.113161] hm2/hm2_7i43.0: registered
[ 335.113168] hm2_7i43.0: board at (ioaddr=0x0378, ioaddr_hi=0x0778, epp_wide ON) found
[ 335.150920] hm2/hm2_7i43.0: Watchdog has bit! (set the .has-bit pin to False to resume)
[ 350.043331] hm2_7i43.0: releasing board
[ 350.043341] hm2/hm2_7i43.0: unregistered
[ 350.043391] hm2_7i43: driver unloaded
[ 350.048738] hm2: unloading
[ 350.053038] PnPBIOS: get_dev_node: function not supported on this system
[ 350.053049] pnp: Failed to disable device 00:08.
[ 350.420561] RTAI[math]: unloaded.
[ 350.448540] RTAI[malloc]: vfreed extent f917e000, size 2097152.
[ 350.448753] RTAI[malloc]: unloaded.
[ 350.548417] RTAI[sched_lxrt]: unloaded (forced hard/soft/hard transitions: traps 0, syscalls 0).
[ 350.552717] I-pipe: Domain RTAI unregistered.
[ 350.552728] RTAI[hal]: unmounted.
sharp@sharp-cnc:~$
Last edit: 23 Jan 2010 22:28 by Pentax.

Please Log in or Create an account to join the conversation.

More
23 Jan 2010 22:27 #1606 by PCW
Have you run a latency test? Depending on the watchdog timeout value, a long pause due to
latency problems could make the watchdog bite (probably a good thing)

Please Log in or Create an account to join the conversation.

More
23 Jan 2010 22:35 - 23 Jan 2010 22:37 #1607 by Pentax
allright, the worst on the latency was a servo thread of 10k (ish) and a base thread of 16k (ish). In the hal file the watchdog timout value is set to 1000000 ns. if i increased that would it help my situation or do you think it is some other issue? Thanks.
Last edit: 23 Jan 2010 22:37 by Pentax.

Please Log in or Create an account to join the conversation.

More
23 Jan 2010 22:46 #1608 by PCW
Whats your servo thread rate, if its the default 1 KHz you have 0 margin on the watchdog (your current watchdog is 1 ms)
I'd set it to 2000000 ns or so for a 1 ms servo thread period

Please Log in or Create an account to join the conversation.

More
23 Jan 2010 22:56 - 23 Jan 2010 23:00 #1609 by Pentax
ok, I can't find where to set the has_bit pin to false.... only where to change the timeout. I'll change it and see what happens? do i need to reset watchdog first? thank you for being so helpful on this.
Last edit: 23 Jan 2010 23:00 by Pentax.

Please Log in or Create an account to join the conversation.

More
24 Jan 2010 00:05 #1610 by Pentax
ok, so i set the dog timeout to something insanely large. and the machine runs better, but if something goes wrong, the servos stop and the virtual tool continues..

Please Log in or Create an account to join the conversation.

More
24 Jan 2010 01:27 #1611 by PCW
what "goes wrong" now

Please Log in or Create an account to join the conversation.

Time to create page: 0.141 seconds
Powered by Kunena Forum