Hi all,
Having recently setup Smoothie on my Cartesian Printer, i’m having a hard time getting the 3 Point Bed Compensation working well. I have X & Y Min Endstops but no Z Min, only a Z Probe. I am homing, probing (G32) and then ‘Finding The Bed’ (G30) in the pre-print script.
I can see the Z Axis compensating as i jog the machine and as it prints but the compensation is not great, in fact it is worse than with it turned off.
What are the implications of setting up the Delta Grid Compensation instead? I understand that 0,0 needs to be set to the centre of the bed, but what else would this method need? Will there need to be changes in S3D etc.
Any help would be greatly appreciated.
Thanks
Alex
Imported from wikidot
Hi Alex,
If the Bed Level routine is working, and as you say you can see the Z axis motors compensating during the print, then you must check your Z-limit sensor, which in this case is even your Z-Probe, do the Zprobe test(the one in the screen menu) several times on the same point with cold bed first and hot bed after, check if the result is always the same.
If your sensor is working properly you should see always the same numbers, at least a gap of 0.01/0.03mm, not more, if is more then you have a bad probe(inaccurate), or mechanichal backlash in the Z axis.
Homing is a thing, finding the bed is another thing, as the grid is stored, and till you shut down smoothie or you do again, it never change.
so, in your starting script you should find just the Z-probe at every print, as the grid should be done once and thats that, so you could setup a routine in the Custom Menu and do it once after you heat the bed, then just compensate the gap(if there’s any with the Z-offset and store it)
If you are working on a aluminium bed with the glass on top, the things are even worst , because the distance first, and the difference in temperature on the bed surface are not the same as the aluminium is sensed much less than the iron, so in not the same inductive reading across the bed grid.
Check first if your bed is quite mechanically leveled by testing with the paper in the 3 points of the Zprobe grid first(just to start with the right foot)
We are talking about 0,1mm that make the difference in printing a good 1st layer, so every 100th of mm count.
Personally i got rid of the Z-probe on the extruder, and now i’m using the FSR sensors under my bed, (i have a cartesian printer)no more problems, just turn on, heat the bed, do the first probe and print……
Shame i cannot post pictures of what i did on my printer, it really worth the effort in doing it.
Hope that what i say it make any sense,
Regards
Emanuel
Hi Emanuel,
Thanks for your reply, after having done a lot of physical checks / rebuilding / tweaking i can say that i am pretty confident that the bed is 100% true (straight / flat).
The probe i am using is a BLTouch and its readings seem to be pretty consistent. The bed is constructed as follows:
- 6mm Delrin “Base Bed”.
- 10mm Thick Cork.
- Silicone Heater.
- 4mm Toughened Glass.
I was making one mistake with my firmware setup regarding the probing, I was running the G32 Probing Sequence and then after that i was running G30 at the last probing point and not the first like the Z Probe Section instructs. I had updated the sequence to follow this and the results were slightly better, although i have only been printing in a moderate area in the centre of the bed, so i assume if i were to try and utilise the full 362 x 362mm build area then the problem would progress to be worse towards the outer edges.
I want to try and utilise Delta Grid Levelling but i am unsure on how this should be done as the documentation instructs that the bed centre must be 0,0.
Would other alterations would be required to make this work?? Would the Min Travel be -181 and the Max Travel 181? Do any alterations need to be made in the slicer settings to reflect the changes in the origin?
Any further help would be greatly appreciated.
Thanks
Alex