Questions: 1. My bed looks like an ocean wave in the mesh. I tried to make it better by adjusting the screws but there isn't much accuracy there. The best I could get was +/- 0.325 which isn't amazing. Any good reason not to mod it to be able to fine tune bed corners? I was thinking swapping out plastic stand offs with springs. 2. The start gcode claims to have mesh bed leveling, but when I go to print it heats everything up and just starts printing. 3. I assume somewhere in the system files the nozzle heights are all specified and can be changed. I took a chance and let it print the first layer from T0 and it was amazingly dead on. I think I just got lucky. Amateur Tip of the Day: When ordering PTFE from E3D for your kit make sure you specify how long you want the PTFE... I just added to cart and got 100mm. Hence the capricorn tubing in my setup that Amazon shipped next day to me.
Ok - I've figured it out two of these. 2. Apparently it doesn't do mesh level before each print, it just uses the built-in stored mesh. I am used to Prusa mesh level at the start of every print. 3. These are in config.g ;tool offsets G10 P0 X-9 Y39 Z-5 ; T0 G10 P1 X-9 Y39 Z-5 ; T1 G10 P2 X-9 Y39 Z-5 ; T2 G10 P3 X-9 Y39 Z-5 ; T3
If you want to mesh level at the start of the print, you could. Just include the GCODE for it in your slicer. It's somewhat unlikely to have changed since the last measurement unless you did change your bed in some way. Thanks for providing some starting points. I think the ordering system recommended a quantity of 70.
I've fixed my bed leveling and went to sub 0.1 height differences across the bed with springs and lock nuts without much effort. I haven't heated it yet to see if things get misaligned though.
Interesting. Can't really understand why a movable spring would give less error than a rigid standoff.
Because the springs are adjustable. I turn the screw to tweak the level of the bed. If you read above you will see that with the rigid standoff the best I could get was 0.325 height difference between corners.
I see. Yeh, you're trying to use mesh bed leveling as a measuring tool to make mesh levelling irrelevant. Any particular reason? OCD perhaps? Right. Mesh leveling is going to take care of all deficiencies.
With +/- 0.32 you end up with a result below (it is exactly +/- 0.32 on a 6mm X 14mm box) across a longer object from side to side if using mesh leveling. This is not the end of the world but means things like snap fit parts or models that require more accuracy could have issues. Prusa printers have mesh leveling but I believe only compensate up to 0.1 which, in this case, leaves 0.22 uncorrected or in a worse case like below as much as 0.54 (difference across positive and negative). Maybe the Duet firmware supports more compensation. I was unable to find that information.
If the mesh bed leveling procedure yields a report that shows +/-0.32mm, it will correct for them as well AFAIK. I mean, you’re right, it’s always better to start off as level as possible, but… You don’t have to go overboard. You could’ve eliminated one axis of skew by adjusting the cantilever angle. The other axis could've been dealt with by shimming the washers on one side, or sandpapering them on the other.
This is in principle a good method of adjustment, however using spring is probably a bad idea and you'll want a solid pillar. As you print and more weight is added to the bed the springs will compress and you'll have distortion in the print. Having a solid adjustment, nuts and screws, would be must better
New plan: Theadlock on top nut, lock nut for adjustment, bottom screw to make sure it doesn't bounce around.
I have 3 complete single material prints, off the TC (pics later): A nearly perfect 20mm cube A benchy that looks like the ship builders were doing great until they took a break for many beers and came back to finish the build Rhino https://www.thingiverse.com/thing:2853276 that I rate a solid average quality with some stringing and pock marks I've been using high quality known working filament for these first prints. I've had minimal experience with bowden printing so I will have to look into tweaking settings. It seems like the TC struggles a lot with retraction.
So this was the way I finally altered mine. You mentioning the lock nut reminded me that @Ephemeris had also designed an easy to adjust mechanism for the bed on the Bigbox. Combining the two ideas, in the end, I simply inserted an M3 nyloc nut into the mix: This now means you can simply screw/unscrew from above in order to make adjustments. OK, so I probably lost 3mm on the max height this way. An alternate standoff could retrieve it. It's a tradeoff I'm OK with for now.
I'm facing a similar problem and therefor also looking for a way to improve the levelling.Your approach lock nut looks interesting, but I don't fully understand it. What exactly do you do to change the height?
You simply screw/unscrew using a hex tool from above. The locknut maintains tightness between the screw/nut and bed. The length of screw sitting in the hex standoff provides sufficient holding to maintain the selected height. The bed itself is not subjected to high levels of vibration so I don't think the set height will change during the course of the print. I run the bed-leveling routine at the start of every print as it's very quick and if there were any movement, this would keep track of that, but i actually don't see any change to the bed leveling graph after a print.
