Calibration and first layer problem

I think my LulzBot Mini may have a calibration problem, you can skip the video to 00:45
The nozzle overshoot and bend the print bed a bit. It more often to happen on the right front corner.
Is it normal ?
http://youtu.be/mOah3cBa44E

It does not happen all the time but when it happens, it is very very high chance of fail to print the 1st layer.
Today, I tried to print 3 times, all fail.

In the video, you can see the nozzle just moving on the bed but nothing comes out.
I have cleaned the bed and the nozzle, but it does not help.
I don’t know are these problems related.
http://youtu.be/P1iG4kN0wCs

The print setting:
Material ease of use: Beginner
Material PLA (Village Plastics)
Profile: Standard

Please help.

During that probing sequence the nozzle needs to make an electrical contact with each of the four corner washers. It appears as if your nozzle is still.a bit dirty after the cleaning process. If you haven’t already you should grab some of the blue non-abrasive 3M scrubbing pads and manually clean the nozzle to get any plastic residue off of it. Also change your cleaning pad if you haven’t in a while (you can flip the one in there over).

It also looks like your on an older version of the Mini firmware. The newer firmware has extra code to identify when the probing at one of the corners has resulted in an out of bound value (indicating the possibility of a still dirty nozzle) then sends the head back to be cleaned again.

Looks like in your video, there is some dirt on your front right washer. Use a light abrasive to clean it off or you can loosen the screw and rotate the washer so a cleaner face is on the inside corner where the nozzle will probe. Certainly can’t hurt to try.

Offending dirt shown below:


I don’t think your nozzle is the issue, if the nozzle were dirty you would get bad readings on all the corners. The other 3 corners seem fine.


-Jim

@jim2386 - I think it actually IS a dirty nozzle. Your image got me thinking and I watched the video again. Take a look at these two images. The offending dirt you point out was delivered by the nozzle and has actually stuck to the washer from being pressed on to it so hard.

In this image you can see the head on its way to the front left corner after the cleaning but there is still a lot of filament on the nozzle. Look at the shape of it compared to in your picture (click the picture to make it larger). Also, look at the front right washer in my image compared to yours. The dirt you identify was actually delivered by the nozzle (but didn’t affect the front left probe sequence) and since the payload was dropped off the back two probing sequences went OK as well!


So, it looks like my initial suggestion of changing (or flipping) the cleaning pad is necessary, possibly along with an update of the firmware with the improved probing logic.

That nozzle is clearly not clean - after theoretically being cleaned:
after_clean.jpg
-Doug

Doug,

Good eye! That would explain why the rear two washers probed okay since the dirt was deposited on the second washer. Please don’t think I was trying to negate anything you said. I was just pointing out what I saw :slight_smile: Luckily I think with the initial dirty nozzle and the deposit on the second washer, I think we can’t explain everything here.

That being said, when you do get the nozzle clean and get 4x good probes, make sure you measure the thickness of your bed skirt off your first print. Assuming you didn’t change anything in cura, the skirt should be 0.425mm thick. If it’s not, you’ll need to change your Z offset settings under expert mode in Machine: Machine settings. Don’t mess with this, though, until you clean the washer and the nozzle and get a good print first!

-Jim

That being said…does that mean the metal washer did a better job cleaning the nozzle than the wiper pad did? :slight_smile:


Stainless steel wipers anyone? :smiley:

</THIS IS A JOKE>

P.S. NEVER DO THIS IN REAL LIFE!

I realize you were jokeing, but just in case anyone in the future finds this thread and doesn’t realze that, never ever use a conductive nozzle wiper. You risk shorting the heater core to the auto leveling system and frying your Rambo board. Touching a flat washer is ok because the washer can’t bridge those two systems. somethign like a wire brush or steel wool on the other hand has a very good chance of doing so. And then your printer is unhappy.

Yes, just having some fun. I’ve updated my post so the newbies don’t think it’s some brilliant idea :smiley:

Hi @jim2386 - No worries and not at all. As I said, your post got me thinking which is why I went back to look at the video again. To a degree I had also thought maybe the nozzle was missing the washer all together.

I had just seen that same “corner plunge” thing too many times myself, especially with my first roll of PLA. Even times the nozzle looked clean from the front you’d be surprised how much filament was still stuck on the back side. Hopefully this help the OP out.

When the mini does the bed leveling check it should lightly tap the washer. If you see it pushing down and the corner bending down at all your nozzle is dirty and the calibration is going to be crap. If it does that turn the power of clean the nozzle and start over.

I don’t think you mentioned if filament started coming out a bit later during the first layer. I started messing with the Cura settings and found that if I change the “Initial layer thickness (mm)” from 0.425mm to 0.1mm when using a layer height of 0.18mm that nothing comes out of the nozzle when doing the two rings around the print and flow does not start until about 20 to 30 cm of output nozzle movement into the real print. With the default settings and standard print layer height filament starts coming out half way during the loop around the part and is flowing nicely during the real print.

My theory is that some of the filament oozes out during the preheat so the retraction during cleaning pulls out a little too far or the push back after the cleaning is not enough. If I was more up on G-code I would like to make the start up sequence draw 3 or 4 loops around my print instead of just 2 to get around this issue. I think that would be more of a Cura hack than a G-code hack since knowledge of the bounding geometry of the print is required.

I think the misplaced […] button behind “Platform adhesion type” dropdown does exactly what you want. These are the “skirt” settings and you can choose the number of loops (line count) or minimal skirt length (mm) and start distance, but that’s not your problem.

I would also try updating to the newest firmware for the Mini. If it doesn’t get a signal from one of the washers it goes back to the wipe portion and then tries to level again. To many failures and it will go to a home position so you can do a good cleaning on the nozzle. Since updating to this i haven’t had any issues with bed leveling.

Here is the link to the firmware

https://devel.lulzbot.com/mini/Gladiola/software/firmware/

Thanks. I’ve been away from the forum for a bit. I missed that setting and will give it a try on my next print.