Huxley becoming non-responsive mid-print

Hi All,

I have an interesting problem with my Hux. I recently re-built the hot end, and during testing last night got some really good prints in ABS.

This morning however, the last 10 prints I have done ahve all failed mid-print with teh same simptoms;

• The printer becomes totaly un-responsice, the print cannot be paused or cancled in Pronterface
• No temprature data is being sent back to the PC after the printer locks up.
• The extruder driver motor judders back and forth at random a few steps at a time.
• I cannot say 100%, but it appears to happen when moving to the next layer.
• I am getting exactly the same problem when sending g-code from the PC, or printing from a g-file on the SD card.
• The only way to rectify the problem is to power off the printer at the mains, kill pronterface, power the printer back up and then re-starting pronterface to re-connect.

UDATE WITH MORE INFO:

- The lock-up happens in the same place twice when printing the same g-code file.
- The last line sent by pronterface is the same in each instance (line 1704 in [endoftheworldasweknowit.co.uk])


In the file:

M106 S142
G1 F1800.000 E1.99264
G92 E0
G1 Z1.150 F4800.000
G1 X62.290 Y71.064
G1 F1800.000 E1.00000
G1 X83.170 Y71.064 F1800.000 E3.86800
G1 X83.170 Y85.373 E5.83344
G1 X82.241 Y85.373 E5.96103
G1 X82.241 Y76.016 E7.24635
G1 X63.219 Y76.016 E9.85917
G1 X63.219 Y85.373 E11.14449
G1 X62.290 Y85.373 E11.27208
G1 X62.290 Y71.142 E13.22682
G1 X61.771 Y70.545 F4800.000
G1 X64.498 Y70.545 F1080.000 E13.60140
G1 X64.498 Y64.262 E14.46438
G1 X66.025 Y64.262 E14.67417
G1 X66.025 Y70.545 E15.53715 <------------------- The 'Current Line' in pronterface status bar on crash
G1 X79.435 Y70.545 E17.37917
G1 X79.435 Y64.262 E18.24215
G1 X80.962 Y64.262 E18.45194
G1 X80.962 Y70.545 E19.31492
G1 X83.689 Y70.545 E19.68950
G1 X83.689 Y85.893 E21.79769
G1 X81.721 Y85.893 E22.06804
G1 X81.721 Y76.535 E23.35336
G1 X63.739 Y76.535 E25.82343
G1 X63.739 Y85.893 E27.10875
G1 X61.771 Y85.893 E27.37909
G1 X61.771 Y70.623 E29.47658
G1 X62.221 Y70.805 F4800.000
G1 F1800.000 E28.47658
G92 E0
G1 X34.921 Y62.584 F4800.000
G1 F1800.000 E1.00000
G1 X49.229 Y62.584 F1800.000 E2.96544
G1 X49.229 Y83.463 E5.83344
G1 X34.921 Y83.463 E7.79888
G1 X34.921 Y82.535 E7.92648
G1 X44.278 Y82.535 E9.21180
G1 X44.278 Y63.513 E11.82462
G1 X34.921 Y63.513 E13.10995


I had to make a minor change to the pin-out for the hot end (see my last few posts about the blown MOSFET). Howver, H have not touched any of the g-code parsing code, so apart from the pin realocation, the firmware is the latest version of Marlin from the reprappro github [github.com]

If anyone has had a similar problem and can give me some pointers for debugging this, I would be greatfull.

When I get home this evening I will re-upload the Marlin firmware, just in case the upload got corrupted.

Cheers,
Lars

Edited 1 time(s). Last edit at 05/06/2013 07:49AM by LarsIslander.

A Quick Update:

So, I had a Eureka! moment the other day after reading through the gcode file for the above model re-compiled verbosely. I realised that about 10 lines above the lock up, the fan is turned on. I then checked my gcode files everywhere that they have locked up for the same fan turn on (M106 Sxxx) and lo and behold, every single time it crashes is about 10 lines after a fan-turn-on command is sent.

I disabled the fan in Slic3r (by setting 'disable fan for first 9999 layers) and have had successful prints on all parts since.

I'm not going to say that this is really SOLVED, because sometimes a file that has M106 Sxxx commands will print with no problem. My hypothesis is that there is some combination of Fan and other instructions which is causing my minor hack on Marlin (see this thread [forums.reprap.org]) to get confused.

I will keep investigating and update this log with more info if I get anything.