Occasional QGL failures due to power-off gantry tilting?

[Dave32]

V2.4R2, 350, with the mechanical Z endstop, and an Omron probe for QGL/mesh.

Every once in a while, I get an error during the Quad Gantry Level... I believe it was "No trigger on probe after full movement". This happens when it is probing one of the front corners.
What is happening is that at power-off, the heavier rear side of the gantry drops, and sometimes it drops enough that the front/back gantry tilt is beyond what Klipper will command Z to travel in searching for the bed surface. This is logical, and the error is understandable as we wouldn't want the nozzle to collide with one part of the bed while attempting to move overly low to find another.

Has anyone else experienced this, and happen to have a solution? It's a bit frustrating when you've waited 30+ minutes for a warmup, only to find that the print was cancelled (and heaters turned off) because the QGL failed. I've read that some people have resolved it by parking the toolhead front and center after the print ends, but this would make part removal more difficult, so I'm hoping there is a better solution.

Thanks in advance.

 

[Aeronut]

I've run into the same problem on occasion. My solution is to manually perform a QGL just before starting a print. You can also change the number of retries in the CFG file,

 

[Sanity Agathion]

Hi, yes this happens, especially when you park toolhead at the back before powering printer off.

Try adjusting some of these knobs like horizontal Z move or max adjust: https://www.klipper3d.org/Config_Reference.html?h=quad#quad_gantry_level

Or just do a home and QGL before you start heat soaking. It's the easiest and you can put it in your heat soak macro.

 

[Fabreeko]

When i first built my v2.2 i had this happen a lot, i just got in the habit of bringing the rear part of the gantry up manually before turning the printer on.

First time i saw the QGL fail because of this i was super lost , took me an hour to realize its a common occurrence.

 

[Dave32]

Thanks everyone for the suggestions.

My goal is to get this 2.4 to the "click print and walk away" level of reliability. It's something I've been able to achieve with every other printer I've had, just... not this one. It's close, I just get these occasional random failures.

I have lifted the gantry manually before (and of course, it's what I do when QGL fails)... but would like to not make it a habit, as I have no reference (on a print-by-print basis) of whether it was needed, or if I'm actually making things worse. I'm also leery of ending up with the rear of the gantry higher than the front, as with the Z end stop at the rear, this seems like it adds the risk of a bed collision when the nozzle moves to the front for probing?

On the QGL before warmup - That could help, but still means that there is a human in the loop - that I (or someone else) would have to wait around for it to watch for QGL failures at the beginning of every print. We run a lot of prints on this printer, so the time standing around adds up - I'd rather fix this at the machine level instead if possible.

For the Klipper QGL parameters, I already have max_adjust at 10mm (default is 4mm), and horizontal_move_z at 10mm (default is 5mm). 10mm for max_adjust seems like a lot, but perhaps, as a tilt angle over the span of a 350-sized printer, it might not be that much. I'll have to try intentionally skewing it, taking some measurements, and seeing what it takes to induce a failure.

Thanks again, appreciate the help.

 

[Ytrahcodo]

Here's a crazy idea: maybe some mechanical stops could be added to the verticals just below the printable height that stop the gantry from settling below them, then park the printhead beyond the build surface and just a few mm above the stops. If it settles onto the stops they would limit the out of level distance. In this circumstance the nozzle would drop below the level of built surface, so of course it would be necessary that the height would have to be raised again before any horizontal movement.

 

[Sanity Agathion]

Here's a crazy idea: maybe some mechanical stops could be added to the verticals just below the printable height that stop the gantry from settling below them, then park the printhead beyond the build surface and just a few mm above the stops. If it settles onto the stops they would limit the out of level distance. In this circumstance the nozzle would drop below the level of built surface, so of course it would be necessary that the height would have to be raised again before any horizontal movement.

Or have magnetic dock on top of gantry and park there.

 

[Badnoob]

if you tighten your z axis belts a little that will help reduce drop. otherwise, adjust the max adjust setting as Sainty Agathion suggested.