Bed Levelling Issues
I've recently completed my MK4S conversion to Core One and am having bed leveling issues.
It appears that probing the rear left and right corners often doesn't apply enough load on the load cell - it tries a number of times and whilst doing so I can see on the loadcell sensor value that the differential when probing is small (10-30) Nearer the front of the bed its 100-200 and higher.
Has anyone else experienced this issue ?
I can usually get a successful leveling at the third attempt so this isn't a major issue - but it is an inconvenience I'd like to resolve.
Originally it looked like I had massive drift on the loadcell - but this turned out to be a constant force from the heater cable - it was pressing on the nozzle thumb screw. I've resolved this by routing it well away and the loadcell sensor value is stable to within +/- 4 when idle.
RE: Bed Levelling Issues
In the rear left, it might be the cover of the heatbed cable which stands out too high and collides with the Nextruder's fan shroud. There is one screw below the cover which needs to be replaced by a flat-head screw (possibly fixed in the build instructions now?). Also, make sure that the cables themselves are routed flat and do not push up against the cover.
Otherwise, some users have observed unreliable Z positioning (sporadic step losses and/or nasty noises of the Z drive). Making sure that the black trapezoidal nuts sit in the correct lateral positions, where they don't fight with the smooth vertical rods regarding lateral position of the heatbed, has often helped: Drive the heatbed to the lowermost position, loosen the mounting screws for the three trapezoidal nuts to let the nuts settle in the right positions, then gradually tighten the screws again.
Lightly lubricating the Z leadscrews with a bit of lubricant from the included tube may also be beneficial. The POM trapezoidal nuts can supposedly run dry, but several users have reported that lubrication was required to get smooth operation of the Z drive.
RE: Bed Levelling Issues
All looks and sounds ok - I suspect it's more related to the rear corners of the heatbed being "flappy" - i.e. the back is supported in the middle only.
The heatbed does appear to deflect easier - which would result in a lower differential on the loadcell. I wonder if its just too low that the software doesn't register it - although I would expect the probing to probe further ?
Re. the lateral positioning - I did have binding issues on the front left when calibrating and the bed was moved to the very bottom. It failed to rise again on the left side. This appeared to be because the distance between the lead screw and smooth rod was too far and caused it to bind. I resolved this by rotating the smooth rod holder at the base - I'm not sure if I hadn't rotated them correctly in the first place and now they are correct or if, when not rotated correctly, they are closer to the lead screw.
I will loosen and retighten the trapezoid nuts when my current print finishes.
Re. lubrication - I thought this was not recommended as the nuts are self lubricating ?
RE: Bed Levelling Issues
Re. lubrication - I thought this was not recommended as the nuts are self lubricating ?
Yes, as mentioned they are "supposed to" run dry. But several users (myself included) had groaning noises in the Z drive which could not be alleviated by adjustment, but went away with lubrication.
I think the concern is that the grease will attract dust over time. But several long-term owners of Mk4 printers reported that they have operated them with lubricated leadscrews for years, with very sporadic cleaning and re-lubrication, if any. And with the enclosure around the Core One mechanics, this should be even less of an issue. - Anyway, if you don't observe any issues with the Z drive , I would also leave it unlubricated for now.
RE:
Loadcell values: it's difficult to tell if you really just get a delta of 30 or if you simply don't see the higher value due to too low refresh rates. I just lightly touched the nozzle with my finger and it produced a delta of 20-40. I don't think there's a chance of the print bed deflecting at such a low pressure. I tried to carefully press at the rear corners and I couldn't spot any deflection.
Does someone know if/how the Core One writes more accurate load cell values into one of the logging possibilities?
POM nuts: they aren't "self lubricating", they are just really hard, smooth and have a very low friction coefficient. So they run decently smooth with almost zero wear, but they can still produce annoying vibrations.
I'm yet to lubricate my leadscrews due to dust accumulation (new machine, want to keep it as clean as possible) and rather trying different Z-axis speeds, but I think I'll have to in the near future. It runs smoothly, but the frequency of the noise matches my room mode. Not loud, but really annoying.
The amount of dust isn't really a mechanical concern and more a visual thing for me. My used and really poorly treated mk3s still has a smooth Z-axis, although you can see wear from metal, carbon and GF particles. It sat in a workshop for a few years before I got it and was lubricated from day 1.
RE: Bed Levelling Issues
The refresh rate /is/ slow - but it does so many probes that I would expect to have seen a higher difference at some point than I have.
When it probes the rest of the bed (mostly one touch required only) I do occasionally get to see a high difference.
So my conclusion that it isn't ever getting as high on the rear corners is sound I think - I would have seen it at some point.
Would be really interested to know just what the probing is looking for - what its window of values is etc..
RE:
Ah okay.. Did you try to disable the hotend heating for the MBL probing? Makes my printer only probe once per point and also got rid of Z-homing in thin air or nozzle cleaning errors.
To test this: Prusa Slicer, expert mode, printer, custom G-code, Start G-code: scroll down to the line with MBL and put M104 S0 in the line above it. You'll get very slightly inaccurate probing due to the nozzle dropping from 170°C to around 115-140°C during it, but it cleaned up my probing and I'm yet to see any impact on 1st layer quality. The change in length only happens for the nozzle between the tip and touching the heatbreak, which I'd say is around 30 mm maximum.
For brass, that's L = 30 mm, delta temp = 50°, => l = 0.03 mm in the worst case.
Not nothing and it can impact the looks for perfect 1st layers, but it has no functional impact.
I hope Prusa will acknowledge and tackle the loadcell vs heater issues and code something like a very short heating pause, when the bed moves against the nozzle and then heating again during the XY-moves.
RE: Bed Levelling Issues
Had the same issue. I seem to have it sorted out. Check your heater wires for hotend and give them some slack. Give your nozzle a good clean also.
RE:
My issues were probably related to the cable in the back underneath the bed. During Z axis calibration, this cable must be routed properly so that it does not obstruct the downward movement of the bed during Z axis calibration. Place be bed 5 cm above the floor, than make sure the cable stays nicely in the back, then lower the bed; make sure the cable stays in the back; then perform the z calibration.