Just installed a 1mm pitch Leadscrew in the latest FB2020. Just finished calibrating the steps/mm. That’ll do, pig, that’ll do. Now to test a really fine print with it. The dual extrusion is on hold until I get the part I need machined, but soon I’ll have superfine dual extrusion printing… hnnng.
Yep, that’ll do. Less than half the thickness of a piece of paper is really good. That’s bed level variability. Nicely done.
Does it stay the same amount of tolerance at double that height or does it become cumulative?
@Jeff_Parish – that was at 0.28 - I need to do some further testing, but I doubt there’ll be much in it with such a fine leadscrew.
Whats ups with that calibrating? Got a 2mm leadscrew. Knowing the pitch, microsteps and motorsteps you just calculate it. Got the same results…
@VolksTrieb the base 3200 steps that was input had me about 1.4 mm over the 10mm it should have been. I would have left it if it’d been close enough.
@Ax_Smith-Laffin thats strange. There must be a fault. Maybe the screw or the motor? This is not acceptable
Tbh, it’s a chinese Leadscrew from Aliexpress, so it’s possible it could be a little out but not by much. I’m not too fussed about having to pin it down as there’s little to no backlash at all in the nut.
Nhung bi mat tren may bay co le ban chua biet
like @VolksTrieb said, you have to calculate the steps of XY&Z. If 3200stp/mm is yours, than dont adjust them to 2800 (like you probably did), instead work on your extrusion. Did you take care about heatshrinking? are all the measurements you did at the same temp? There are so much parameters that can influence the height (especially of small prints) and if you change z-steps, you are doing it wrong :-/
Adjusting steps to fit calibration box is a bad idea.
You should move your z axis 100mm and measure whether it really is 100mm. Then check extrusion.
I use atleast 100mm distances on all my calibration prints. People who use 10 or 20mm crack me up. The accuracy on cheap micrometers is only so good. At 10 or 20mm the inaccuracies and/or any issues with the extrusion width will be exagurated and cause you to over correct your measurements. When doimg thin wall calibrations you want a longer run so the extrusion width has a chance to stabilize as well.
@Nathan_Walkner thats the point. Getting good results like +/-5/100mm. Less is almost impossible due to filament variation and accuracy of the motors and so on. This is not sla.
I agree with all you guys about the limits in accuracy of these printers. That is why a larger sample size is so important. Why I consider IMHO 100mm to be the minimum when calibrating x/y steps/mm. We have all read articles questioning the accuracy of microstepping, etc. Any machine based on grid coordinates also has to round up/down to the next point on the grid if the distance it is being moved falls between two points on the grid.
Lets say your printer is perfectly calibrated for distance, but its extrusion width is off by +/- .05mm. That would mean each side of the extrusion would be over/under 0.025mm. When printing a calibration cube, your cubes length will be off by double that amount since you are measuring 2 walls. Now if you are printing a 10mm cube and compensate for that 0.05mm distance by changing your steps/mm, when you print a 100mm cube the effect on the travel distance will be 0.5mm. I have a printer with a 10" build plate. That’s a potential for it to be off by 1.27mm if I print at the full size of my bed, even though my 10mm “calibration cube” was measured at exactly 10mm!
When you add in the questionable accuracy of cheap digital calipers, extrusion variations, and print artifacts (slight bulges at corners, vertical layer alignment, etc) it means it is very easy to miscalibrate your printer if you are trying to compensate for all these things by only changing the travel distance on your axis. By using a much larger sample size your precision can improve. I’d rather be off 0.04mm on a 100mm or 254mm cube then be off 0.04mm on a 10mm cube.
This is much more applicable to the x/y travel than the z axis, but the same principle applies. if your limit switch triggers your z axis to stop ~+/-25 microns from where you think it is stopping, and you compensate by adjusting the steps/mm, then that error will be amplified when you go higher.
I usually print a 100x100x15mm HOLLOW cube when calibrating my printers. I find thats a good trade off between having a good sample size and print time for my uses.
@Ax_Smith-Laffin Ahhh, I didn’t realize you were changing the micro stepping in Z. The others are correct that the micro stepping should be calculated to the mechanical set up of your printer and then the filament should be calibrated for its physical and thermal properties accordingly. Every filament has different behaviors where the mechanical stepping of your printer should not change.
I have 2mm lead screws on my printers and using my system for calibrated filaments I am either dead on or less than .1mm (.004") off on full print area prints with a proven reliability that my parts will always fit together regardless of which printer they come off of or what filament is used.
With your 1mm lead screws your backlash should be reduced allowing finer layers (0.05mm?) but your calibration still needs to be the filament. If most of your objects will be small then a calibration piece the size of the features you are printing will be just fine. However, with calibrated filament you should be able to scale it up with little variation.
I rarely print much above 30mm, most of what I print is Enclosures and Battery Sleds which range from 13 to 28mm in Z, so this shouldn’t be an issue. Admittedly, I do need to re-check ESteps due to a bit of a firmware issue I’ve had when trying to work out a few kinks. I have done a 100mm vase print and it was about 0.9mm out, but with me not printing much over the above 30mm, it’s a bit moot.