Notifications
Clear all

Octoprint / FW 3.10 temp measurement corruption/errors  

  RSS
Rich R
(@rich-r)
Active Member
Octoprint / FW 3.10 temp measurement corruption/errors

Has anyone seen an issue with 3.10 on an MKS3+ where uploading to the SD card from Octoprint results in the printer not being able to measure head temperature reliably?    When I upload gcode directly to the SD card, once it's done the head temp measurement is highly unstable and inaccurate.   The GCODE file is also corrupt and won't print.

 

Resetting the printer and reconnecting results in the measurements settling out.   Here's the measurements before the reset: 

Posted : 26/05/2021 11:22 pm
--
 --
(@)
Illustrious Member
RE: Octoprint / FW 3.10 temp measurement corruption/errors

When is the SD I/O ever reliable? lol.  I would never consider writing to it via Einsy USB.  Heck, even reading the SD directory takes several tries. The entire interface to the LCD panel is - well - hoaky.  And I think I agree that 3.10 probably made it worse. At least I'm seeing more I/O errors now than with 3.9.x ... that is - scrambled directory on the LCD display - 8,3 formatted files, etc.

Posted : 27/05/2021 1:35 am
Rich R
(@rich-r)
Active Member
Topic starter answered:
RE: Octoprint / FW 3.10 temp measurement corruption/errors

@tim-2

I agree, and I've gone to loading the SD card on the PC directly.  Loading the same file remotely as from the PC leads to different filenames for longer names.  I'm pretty sure that the PC is not part of that problem.    

That said, I also do programming on the Arduino class devices and I think probably use the same library cores looking at the 3.10FW sources.  I have problem on Arduino so I'm guessing it's in the hardware or libraries themselves, vs the implementation of the FW itself.

I was more worried if no one else saw issues that I had a bigger problem with the printer itself.   More misery loves company is making me feel better... lol

Posted : 27/05/2021 1:51 am
bobstro
(@bobstro)
Illustrious Member
RE: Octoprint / FW 3.10 temp measurement corruption/errors

Copying to the SD card from OctoPrint is horribly slow. I'm surprised you're able to use it reliably. 

My notes and disclaimers on 3D printing

and miscellaneous other tech projects
He is intelligent, but not experienced. His pattern indicates two dimensional thinking. -- Spock in Star Trek: The Wrath of Khan

Posted : 27/05/2021 2:08 am
--
 --
(@)
Illustrious Member
RE: Octoprint / FW 3.10 temp measurement corruption/errors

@rich-r

No - the LCD board is a kludge. No termination for the two data busses coming from the Einsy. So lines are ringing all over creation. After actually looking at the mess, I've decided it's just simple luck the printer works at all. Change the cable length an inch, stuff the ribbons into the channels a bit too deep, wham.

There are also interrupt handling problems in the filament sensor code which causes other issues. At least the last time I took a deep dive there - I sort of staggered a bit, then told myself to "Run Away!" Much the same recently when I found the LCD board schematics. Sometimes screaming into the void isn't enough.

 

Posted : 27/05/2021 2:10 am
Share: