Following up on my last post. How close to done can you get. Power cut stopped the print what must be minutes from the end. I may try to restart it with a G code edit but it’s close enough to done to not bother.
Bad something similar Tuesday after a 11h print. I guess I had only 10 layers left and ran out of filemant while I slept. This can be really annoying.
Unfortunaletly I’ve used a rep2x for this. So no reasonable option for gcode editing 
Same problem with me also. Power cut nd my ups dont have enough backup or sometimes pc bumps coz of fluctuations. Is there no other option apart from Gcode editing?
Noway, that’s so bad. Sorry to see you lost hours of printing. I think we all need to get a UPS if you live in South Africa.
I pulled it of so and started assembly it as I could with no power. I also have to commend @BQ111 on one of the nicest supports structures to break out ever. I’m very impressed by the design.
I had a ups on my other printer before doesn’t seem to work all that we’ll. I suspect unless it’s a online one with pure sine wave the hz throws out the arduino.
I’d recommend measuring what you have from the print, take the stl and cut it, print the remaining parts as separate and then superglue them… it looks like the sides are not meant for actual support but rather fitting and look.
I built it and it’s fine all the structure is there. Once I get it working I’ll decide if I’ll reprint all of it or just add the missing bit. With it open I can just fill it with epoxy and turn it solid but it doesn’t need it.
There is an option. If you can figure out about what layer, you can use a stl editor (Slic3r can do it too) just cut off the part that already printed.
Options (some of them):
How to implement this in Marlin (with several possibilities) using one of the 6 end stop pin available.
-
http://forums.reprap.org/read.php?1,297350
It will issue a M600 (change of filament command) upon detecting the end of filament.
M43 should re-home only XY after changing the filament (as the step motor are not energized, they will lose position).
Scavenging an optical mouse should be easy enough (i have several “dead” mouses at home).
Another shoot at it:
Inline images 1
Yet another:
Yet another (sold by Sparklabs, look neat)
and again:
There is also:
High-end (with correction of feed-rate according to diameter measurement in real-time)
* https://www.youmagine.com/designs/filament-sensor-for-3d-printers-and-filament-extruders
* https://github.com/filipmu/Marlin/tree/Filament-Sensor (respective firmware)
but doesn’t provide (yet) protection for end of filament!!!
Not out of filament, was a neighborhood power failure. I saw the filament sensor implementation in the new version if marlin a few days ago. and will at some point implement. I spent the day building the Ciclop 3D scanner.where the print failed is a non issue.
You’re right! But part of the solutions presented are giving some goods tips about what to do when the steppers aren’t energized anymore and also need to keep a register where to start from when getting the energy back.
I played with it a bit last night. Took a G code file. Opened it in notepad++ then search for the height your print stopped eg 103.4mm essentially delete everything before it upto the end of heating instuctuon at the beginning of the file. I ran a test of it it heated and moved to the starting point only trouble I had was cura moves the bed and head towards the staring point so would foul on the already printed part I will need to figure out how to raise the head above the print and lower to it and should work a treat.
Thats was easy found this: G1 Z15.0 F3300 ; move platform down 15mm
So id guess change the 15 to 15mm higher than the print failed so 118.4 for my test.
@Michael_Scholtz I just sliced the top off of my print and printed it then glued it on.
I’ve done that before to fix, its nice to have a flat surface at the bottom to stick to the print. If you see how editing the Gcode worked you’ll understand. I’ll post the tutorial and pics later. It’s 5 pages long so it’s a fair bit to go through but once you get it it’s 5 minutes to get it done. You literally do a search for the height the print failed and delete the gcode before that that’s been printed. It’s worth killing a test print 1/2 done just to see how it repairs. It also got me thinking about embedding objects into prints as you could cut the Gcode print to a point drop in what you need to and then have it print the rest.
in slicer you can bury your piece into the bed and it works that way too…