I have the latest version of Luban and the 1.10 machine firmware.
I’m following the guide in the support section. I moved the head to the 2 lower screws. and adjusted the sensor height to 1mm. Opened Luban and inputted the 3 codes M502, M500, and G28. At this point the head went to the home position.
The next step is to run the calibration from the touchscreen. The head lowered itself hard into the left rear corner of the build plate. And I mean HARD. Makes me wonder if it bent something.
It’s obvious that the sensor position is behind the back edge of the build plate so that i cannot “see” the plate at all.
The instructions don’t have the head being screwed to it proper location until this step is completed. Is that correct?
What did I do wrong?
You have to mount the tool head to the lower location, adjust the sensor and then mount it back to the original location.
Haven’t had a look at the support article, but this was the description in a forum thread.
The problem of your first leveling was like follows (I run into this issue every time because of my taller build plate)
The z axis travels till position X or triggered sensor. Then the x and y Beginns moving to the first calibration point, which ruins the bed.
The directions tell you to run the calibrate from the touch screen, then put the head back to it’s proper position, then run the calibration again. I imagine a lot of people have F’ed their machine by following these instructions.
As it turns out, the bed plate is warped. It’s high in the center by about .7mm all around and also drops about 1.5mm to the left rear.
They should have machined this part after casting it.
Excuse my, maybe stupid, question.
Is your frame mounted with the nuts down?
Normally people have a difference from the highest to the lowest point of about 0,6mm.
If you have 1,5mm difference, there is something wrong with assembly or manufacturing.
Not a stupid question. Yes the frame is mounted in the correct orientation with the nuts down. I brought the frame to work today and put the machinists straight edge that I use on it. It is definitely warped. I’d like to get a new one. If that is not possible I will machine this one to make it flat and true. I have to talk to my machinist (he’s off today) about a jig to hold it in place.
My plan is to get a .5" thick piece of a36 steel plate. Drill countersunk holes to attach the frame to it. Then have the steel plate machined flat on both sides. then mound the frame to it and machine both sides true.
I would much rather just buy a new frame as long as they have corrected the warping issue.
You should open a support ticket or make it like you posted 
I’ve thought about exactly this same problem. One of the foolish design defects of the linear modules is that they require mounting from the bottom because they didn’t put flanges on the extrusion. What I’ve been thinking about doing is making some flange adapter plates that screw onto the extrusion from the back but afterwards can be screwed down from the top. This seems the only sane way of dealing with any base that’s heavy and stiff enough to be worth making a modification for.
Exactly. I had planned on using the plate first as a jig to machine the base plate. Then as a plate to mount the entire unit to.
For the time being I have put those little washer shaped stickers the are used on 3 ring binder paper holes on each of the outside edge top mounting points of the bed frame. I doubled them up in the corners. After a test print I found that I need to triple them up in 2 of the corners. I then did a manual calibration and a full bed test print I made with F360. The print was great except those 2 corners I am going to add the third layer to. I have considered using ceramic washers and nylon screw to keep the build plate heat from reaching the frame.
That’s a great idea! I was going to glue some shim stock down, but that’s much easier. As long as it doesn’t melt.
To reduce the direct heat transmission to the frame (which results in different amounts of bed warping at different bed temps, due to the different thermal coefficient of expansion of the aluminium frame and the PCB heatbed), I have some M4x12 titanium alloy screws on order to replace the standard M4x10 stainless steel screws. Nylon screws would be much better from a heat transmission perspective, but I don’t think that they would have the necessary tensile strength (and the only ones that I can find have slotted heads, rather than hex socket or phillips head). Titanium alloy conducts a lot less heat (about 7 W/m.K) than stainless steel (about 16 W/m.K).
I’ll be experimenting with a mixture of fibre washers, nylon washers and o-rings (used as washers) between the heatbed and the frame for each of the screws (which is why I’m using M4x12 screws), so that the heatbed is not sitting directly on the frame. Having an o-ring in there will allow some “squish” for tuning the bed level at each screw point, but they may not be necessary if I can use some 3 ring binder paper holes to get the top of the frame level.
I’ll probably also attach some foam rubber to the underside of the heatbed to reduce the indirect heat transfer.
I’ll do some bed level measurements before and after and report on the results.
PS. I’ve been using a 310x310 Creality glass bed on top of my heatbed for the last week - it’s been an amazing improvement, but thermally isolating the heatbed from the frame can only improve things even more. One thing to note about the Creality glass beds is DO NOT USE Isopropyl Alcohol (IPA) to clean them - it wipes off the coating and reduces the stickyness of the bed. 