I bought a new Qidi X-Max 3 on their BF sale for 500EUR, which seemed like a great deal and I’ve been wanting an enclosed klipper core-xy printer for some time for more demanding prints in engineering filaments, so I pulled the trigger.
I’ve now had the printer for two days, and it’s somewhat of a disappointment unfortunately.
The printing speed and quality of the prints are good, nothing to complain about here.
My issue is with their bed meshing and first layer. The bed is crooked as hell with a massive 0.45mm deviance from lowest to highest point, but it should not be a big issue with bed meshing because it should compensate. But it just seems like it’s not compensating at all, the parts of prints that are closest to flat plane have great first layer, all other places are either way too squished or not squished at all. This means I can only get a proper first layer on about 1/3 of the bed, either front, middle or back part.
This is in stark contrast to my semi-custom anycubic bedslinger with klipper, which lays down near perfect first layer every time, without adjust Z-offset across the entire plate. And I was honestly expecting the same from the Qidi.
I have run their calibration routine multiple times. I’m fairly well versed in the 3D hobby. Is there some setting in the Qidi that I’m missing? It’s running their own dirty klipper version, who knows what they’ve changed/missed from newest mainline. I am already considering returning as I was expecting more of a turn-key printer that my experience so far.
If your bed is physically tilted and you’ve ensured it’s flat (it seems that you have), you will probably want to manually adjust its straightness relative to the X/Y plane. Or perpendicularity. Is that a word? You know what I mean.
There are four locknuts on posts on the underside of the bed and if you remove the steel build sheet on top you’ll see the heads of the four screws on the other ends of these. There are probably myriad ways to measure its straightness, but Qidi recommend just manually moving the head around and using the textured nozzle offset sheet the printer came with to ensure that the gap is more or less consistent with the tip of the nozzle in the various extremities of the bed. You could also use a feeler gauge for this purpose if you were feeling frisky or wanted something more durable. You’ll want to do this with the bed at room temperature, so that rising or falling temperatures won’t be muddling your results via thermal expansion.
With the locknuts loosened, you can screw the four corners of the bed up and down slightly using the screw heads on top. These are the four screws closest to the corners. Be sure to hold the screw heads in place when you retighten the locknuts beneath, otherwise the act of tightening them will probably also turn the screws slightly and mess up all your hard work. Turning the screw heads clockwise lowers the bed, and turning them counterclockwise raises the bed in that corner.
Do not attempt to auto-home the print head or run a mesh level job without the steel sheet attached to the bed. The probe relies on the presence of the steel sheet and you will drill the nozzle into the magnetic surface of your print bed if you do. Just grab the print head to move it around in the X/Y plane and leave its jog controls alone. Only jog the bed itself.
Qidi have a video detailing this here. Yes, it’s just an MP4 plonked on a Google drive and no, I don’t know why they didn’t just post it you Youtube or something. They seem to distribute most things by just sticking them in a Google drive folder. You get used to it, dealing with Qidi.
I had to go through this rigmarole when I replaced the heated bed a while ago, which Qidi are not keen to tell you in advance. When I just slapped the new part on there as advertised I wound up with one corner of the bed tilted near as makes no difference to a full millimeter below the presumed plane of the Z axis and the other corner maybe 0.5mm above it. Somehow with mesh leveling this more or less still worked, but it’s much improved now that I’ve actually done it right… ish.
What I have now is this:

Look, it’s not exactly an ideal Euclidean plane or whatever the hell. But 0.2228mm from one corner to the other? S’okay? S’alright. That’s little enough that the mesh can compensate for it.
In case anyone is wondering, the mesh leveling appears to use a 9x9 grid. I thought it would be 10x10. I was wrong. That’s only 81 points of measurement which means that vagaries could theoretically fall in between the probed points. It’s not likely these will be Earth-shakingly severe, because the steel surface plate isn’t exactly tinfoil and it’s only so flexible to begin with. And here’s another dumb tip for your travels while we’re at it: Make sure there’s no crap stuck to the backside of your steel plate, or trapped between it and the magnetic base. Scraps of black filament are what get me, because they’re hard to spot. But they’ll cause you no end of grief.
Ah they did make the bed easily adjustable, would have been nice to include that in the manual…
Anyway, i was able to trim it to a 0.17mm deviance across the plate (both heated and cold) and that really helped with first layer consistency. The 0.45mm I had before must just have been way too much for it to compensate.
I don’t own a Qidi, but I do like them.
Some cheap and easy checks to try.
If you have a decent straight edge, (wood and plastic rulers need not apply), check to be sure your bed isn’t warped. The heat bed is only 6mm aluminum and it might be warped. It does happen with Qidi machines. That’s a warranty issue.
It can also be that one of the nuts on one corner of the bed isn’t clocked in unison with the other 3. Multi start screws can have this timing issue. And the only way to fix it is to figure out which is the offending, disassemble the machine just enough to remove the screw, and turn the nut to try a different start. Then test the machine to see if that fixed it. Again, it’s not common, but it does happen.
Yeah warped bed was already checked (it’s reasonably straight at 60°C after 10min of heat soaking in 40°C chamber temp) and it’s straight enough, just slightly tilted along the X-axis so the front is higher than the back.
I suspect the stepper screws of not being properly timed then.
I’m not sure just how the bed homes top or bottom, but I would home the bed, shut the printer off so the steppers are off and not locked, then turn all the screws by hand to deadhead against a hard stop to even them out. The bed should be as level as it can be then.
Pro way: Use a test indicator solidly attached to the print head and move it around the bare heat bed by hand. Turn the screws until they are all zeroed out. I did that to my Prusa Mk3s when I first built it years ago.
Is that over kill? Yes, yes it is. But as an old toolmaker, I have the tools to do that kind of super fussy inspection. I also used a granite surface plate and precision squares to build the frame on to ensure prefect alignment of the frame. Ahhh the boredom of being quarantined.
Are you sure the bed mesh gets loaded? Maybe when you switched profiles in your slicer you lost the start g-code that does that
Yes it performs KAMP when starting a print and I can see that the adaptive mesh it probes is loaded from the web-view.


