Odd Z Axis behavior

Hey all,

Curious if anyone else has been experiencing this; every couple of prints my nozzle gets out of alignment and with starts digging into the bed, or raising way too high. I’ve had my A350 for roughly 2 months and it started pretty good, but after like a month this behavior started. I had to re-calibrate everything and everything was fine for a bit. But now it’s happening pretty much every day, it’s like one of the z axis’s are getting slightly out of sync with the other.

For starters, I pretty much have to re-calibrate daily. Normally, I calibrate my bed using the auto calibration and the stock magnetic bed, but pre-heat the system to a hot end temp of 200 and bed temp of 70 using OctoPrint. This keeps the bed and hot end a lot closer to where the surfaces would be when actually printing due to expansion when heating and blah blah. Anyways, the temp stays consistent throughout the calibration process and at the end I use a feeler gauge of 0.1mm to get the the z offset set. Then in my pre-print config I just have the machine load up the mesh before printing with a M501.

After a calibration the first few prints are just “OK”, meaning they get solid adherence but lots of stringing and blobbing, very poor bridging, and even worse angles (anything over like 50-60 degrees starts looking like a Salvador Dali painting), but overall the prints come out relatively respectable. Nothing near my Ender 3 though… Right now I suspect this is because my bed is too close to the nozzle, so I’m testing that out now… but as I’m writing this out I’m realizing that this might actually be because the z-axis is going out of alignment mid print. It could be one side getting lower than the other, bringing one side of the print closer to the bed, thus dragging up the last layer a bit and causing this stringing/blobbing… I have no idea. I’m not an expert, ha!

What triggered me to finally get online and start asking others was when the auto calibration completely failed today. I mean, it started out fine, but then the sensor seemed to stop working and slowly rammed the nozzle down into the bed after each measurement. I didn’t catch it right away, as normally I go do something else when it’s leveling, but after manually adjusting the z at the end I couldn’t save the final result. I redid the calibration (making sure the temp was still on correctly), and watched it. I watched as after each measurement, the nozzle got closer and closer to the bed, until in the top right corner, the sensor seemed to completely give up and the nozzle went straight down into the bed, moved the bed down with it, and then stopped. Then it brought the nozzle back up, and went to the final position for me to manually set the z-offset. So this explains why I couldn’t save the result - it was missing like 15 points still! So, now I have 5 little holes going up the right side of my bed, with increasing depth, ruining my second magenetic sheet from Snapmaker.

Anyways, this is becoming quite frustrating as I’ve gone through a couple of hot ends and 2 magnetic beds because of this. I’ve tried glass sheets too (borosilicate, single pane, extra stength single pane, mirror, you name it), which I have had varying success with…but honestly now I’m terrified to use them for fear of them breaking when the nozzle slams into the bed. I’ve also tried to disable the motors, bring the z to the absolute top of the axis, to get them lined up, then re-calibrate based on that. Then each print I would disable the motors ahead of time, bring the z all the way to the top again to re-align, and then print. Yeah, that failed miserably…

Right now I manually calibrated the 25 points by hand (I have to do 5x5, 4x4 and 3x3 aren’t enough…the bed isn’t flat enough…probably because of all the nozzle ramming that’s been happening, ha!), and it’s printing seemingly fine again now, though the nozzle seemed to start lower than I calibrated it to…but I didn’t measure it.

Has anyone opened up the axis’s to see if the axis needed to be cleaned or something? I’m not willing to open mine up like I would on my other printers…

Hi @Kaese! Welcome to the salt mines forum!

First off, this might be a bit off topic, but it sounds like you’re doing a calibration with hot hotend? I can’t think of any good that comes from that, although I’m open to discussion on this - I haven’t really thought of it before.

If you take the X axis off the Z towers - do both Z modules move?

Overhangs could be an insufficient cooling issue - to test that, put a desk fan next to the printer while the overhang is printing. At least for me, that helped.

Stringing and bridging could be a temp issue? If you have some non conductive heat stable thermal paste put some in the thermistor hole, it’s just sort of floating in there. Staff has doubled down in the past that a floating thermistor is totally fine, but many people had issues with it. I also put some in the heater cartridge hole.

If you’re comfortable tinkering with things, these modules are quite easy to open. Check out here: Rail module exploded diagram?. While you’re in there, definitely make sure there’s a piece of silicone thermal pad on the chip, some are missing it, which is super bad.

You could also pick up some plastic grease for the modules if you want, like this: How to maintain the snapmaker in perfect condition?. A bit above there staff posted a video on how to apply the grease: How to maintain the snapmaker in perfect condition?.

Also, if it’s not obvious (it wasn’t to me and I didn’t see it in any documentation), the weird spring tool in the provided toolbox is for tensioning the dust covers during reassembly. There’s a video they posted, can’t find it at the moment. Also useful if it vibrates loose and just needs a tightening because it’s starting to make noise when the steppers move. I guess they don’t use it all the time, not 100% on when it should be used.

I’ll keep this as short as possible, but there’s lots else to say. One final thing: the z probe is inductive, and is detecting the presence of the spring steel inside the removable print sheet. If you switch to glass or something non-conductive, or even something conductive but further than about 1mm away from the sensor (like an aluminum bed with a 1.5mm print sheet on top like someone else here in the forums), the nozzle will crash into the bed while probing.

EDIT: Also just found this topic, newer than yours by just a bit. Going to follow this one too, sounds like it could be similar? Grinding noise upon saving Z axis calibration

Sooo… the stringing and and blobbing…pretty sure I just figured that out. My nozzle (Not the hot end… the actual nozzle) literally exploded out of the hot end halfway through a print a few minutes ago. The threads on it were hardly there and plastic had oozed down the sides…brand new hot end from Snapmaker direct too, just put it in…last week? Anyways I took pictures of that… never heard of this happening before.

First off, this might be a bit off topic, but it sounds like you’re doing a calibration with hot hotend? I can’t think of any good that comes from that, although I’m open to discussion on this - I haven’t really thought of it before.

For the temperature on during leveling, pretty much everything I’ve learned says you need to have the bed and hot end on when leveling. I’ve gotten significantly better prints this way too on my Ender. The reason for this is thermal expansion, you are not going to have the same print surface cold as you will hot. Same thing with the nozzle… though I have read the nozzle is less important, I keep it at the “standby” material temp (so no oozing), which is like 175.

Overhangs could be an insufficient cooling issue - to test that, put a desk fan next to the printer while the overhang is printing. At least for me, that helped.

I like the thoughts on the overhangs, I definitely considered a cooling issue but I never even though of a small fan! facepalm. I’ll give that a go once I recalibrate with my new nozzle in.

If you take the X axis off the Z towers - do both Z modules move?

If you’re comfortable tinkering with things, these modules are quite easy to open. Check out here:[Rail module exploded diagram?. While you’re in there, definitely make sure there’s a piece of silicone thermal pad on the chip, some are missing it, which is super bad.

You could also pick up some plastic grease for the modules if you want, like this: How to maintain the snapmaker in perfect condition?. A bit above there staff posted a video on how to apply the grease: How to maintain the snapmaker in perfect condition?. [Links removed…snapmaker forums wont let me post these as a “new member”]

As far as taking things apart, I have no problems doing that, however given the cost of the machine I haven’t been willing to. If I end up having to send it back to Snapmaker or something, I didn’t want anything that wasn’t covered in the manual touched. But if Snapmaker says “Go for it” then I will certainly try this. I’ll also try a dry run with the x axis removed, we shall see! Part of me kinda wants one of them to not work :stuck_out_tongue: at least then I’d have an explanation…

I’ll keep this as short as possible, but there’s lots else to say. One final thing: the z probe is inductive, and is detecting the presence of the spring steel inside the removable print sheet. If you switch to glass or something non-conductive, or even something conductive but further than about 1mm away from the sensor (like an aluminum bed with a 1.5mm print sheet on top like someone else here in the forums), the nozzle will crash into the bed while probing.

I really wish that Snapmaker hadn’t chosen to use this sensor… I figured this out almost right away. Wish they said something about that as glass sheets are pretty common. I prefer glass, but since it requires me to manually level still… :frowning: I was going to look at what all was involved there to see if it could be replaced with something like a BL Touch, but haven’t looked at anything yet.I have the feeling that much like the rest of the machine…its not gonna be be very easy. Also refer to not wanting to take thing apart :slight_smile:

Also, if it’s not obvious (it wasn’t to me and I didn’t see it in any documentation), the weird spring tool in the provided toolbox is for tensioning the dust covers during reassembly. There’s a video they posted, can’t find it at the moment. Also useful if it vibrates loose and just needs a tightening because it’s starting to make noise when the steppers move. I guess they don’t use it all the time, not 100% on when it should be used.
I’m gonna have to find that video. I definitely had some noise a little while back that I was convinced was a screw coming loose, so I went through and put a tiny bit of threadlock on all of the screws to keep them from vibrating out again. I actually wonder if somehow this is what caused my nozzle issue!

Agreed on the hot bed. If you’re interested, I posted my devations at ambient, and at 60C, in a 10x10 grid, with a dial indicator. The deviation is small, but significant. On achieving a perfect level.

The hot nozzle makes sense if you’re using a feeler gauge. With a separate probe, it shouldn’t matter except right at the end, and if you measure it once you should be able to leave the nozzle cold and adjust it at the end. At least for my bed, I have had good results with the nozzle cold, and it reduces some risk of melting through the print sheet.

I have a glass sheet in the mail, and am doing a magnetic base over it so i can use the same removable print surface. I don’t think there’s anything wrong with the removable print surface, and using it is not at odds with a glass plate, entirely. If you really prefer printing on glass, then yea, that’s an issue.

Also, I have some other topics on Z probes. I’m working with a few others to improve Z sensing in CNC, and there’s definitely some ways to improve the 3D print also. BLTouch won’t be an option until we can safely upload firmware to it, which at the moment requires a chip flash memory programmer; not very accessible to the normal user. But there are other options, including replacing the inductive probe with something else, that would require manually flipping down, or something - like the allen key levelling method. In the past I have also coated the glass bed of a different printer with aluminum foil and glue stick and used the nozzle as an electrical contact sensor - that worked pretty well, and also should be inductive compatible, though initial results seem like getting the aluminum flat and even enough for the inductive sensor is difficult.

I will say I did check the inductive probe accuracy and repeatability, and it’s damn accurate. +/-0.01mm. I think that’s really good, considering we took about 40 samples, and all samples were in that tolerance.

Totally off topic, but me, being relatively new to 3d printing, this is incredibly interesting. Truth be told, I probably have the stuff necessary for flashing… microcontrollers have become a small hobby of mine and I intend to build my own 3d printers in the near future. Clearly I need to be more active on this forum cause it sounds like you guys are doing some really cool work, the inductive nozzle is brilliant

Anyways, thanks for the tip on the hotend, I’ll probably turn that off going forward. I will say I was doing a magenetic sheet over the glass and then removing the sheet to get the z offset during auto-leveling. That actually worked pretty well - I just switched back to the sheet because…well this post in it’s entirety about the Z-Axis.

1 Like

I have read the whole topic, did you set your leveling sensor to the right position?
It seems your sensor is a bit too high to detect some points before tipping into the bed.
There is a topic about in the forum, how to.

1 Like

I have, I’ve used a feeler gauge to set it at the 1mm recommended height from snapmaker.

May you set it a little lower, credit cards (recommended) have about 0.8mm, give it a try with 0.5mm.
Hope this helps!