Z height repeatability

[Daid 306]

There has been quite a bit of information and possible mis-information about Z platform adjustments and re-calibrating the Z level.

A while back, at Ultimaker we took a bunch of readings with a very accurate gauge (micron accuracy)

First off. The limit switch. Some people think that the metal lever adds in accuracy. But we could not find any evidence of this. As long as we didn't touch the machine, the homing was within a few microns every single time.

We tested this in 2 different ways.

1) First by just keeping homing repeatedly. Just pressed home on the UC and measured the difference between the previous home position.

2) Turning the lead-screw by hand and home again. This resulted in the same accuracy as using the repeated home.

So I see no reason to remove the metal lever. As the current repeatability of the Z switch is very high.

Next the platform stability itself. This is where the problems happen. Just giving a good tug on the platform could deform the platform arms itself causing a 0.5mm difference in position.

Now, this does seem to depend on the machine, as my home machine hardly needs any calibration after moving. But that's rare.

So if you want to improve repeatability of the platform, you need to stiffen the platform somehow. We have some ideas at Ultimaker, but I would love to hear other ideas on this. Especially on how to fix this for yourself without re-designing the whole platform.

 

[owen 19]

Hi Daid

I have screwed and glued the piece of wood between between 2 arms from underneath. The screws are self tapping screws screwed in from a 45 degree angle at the corners where the arm meets the wood.

Also in my pre Gcode I move the nozzle to X100 Y0 and the push the platform into it 1mm and then back away. This is to overcome historesis. That is if you push the platform down and release it will come back to a different spot then if you push it up and release.

So this makes it so you are always pushing it down slightly and releasing before the start of a print.

Owen

 

[gadgetfreak 7]

Hi Daid

I have screwed and glued the piece of wood between between 2 arms from underneath. The screws are self tapping screws screwed in from a 45 degree angle at the corners where the arm meets the wood.

Also in my pre Gcode I move the nozzle to X100 Y0 and the push the platform into it 1mm and then back away. This is to overcome historesis. That is if you push the platform down and release it will come back to a different spot then if you push it up and release.

So this makes it so you are always pushing it down slightly and releasing before the start of a print.

Owen

 

Owen, this sounds nice, do you have any pictures?

Did you also upgrade to stiffer/heavier springs for the bed and if so - did it make a difference?

/Daniel

 

[joergen 2]

Owen, this sounds nice, do you have any pictures?

Did you also upgrade to stiffer/heavier springs for the bed and if so - did it make a difference?

/Daniel

 

yes, stiffer springs make a HUGE difference on the bed.

in regards to Z axis repeatability, I had chopped off the metal tongue on the z switch long time ago, and installed 2 z axis adjusters, where one turned into a static piece of plastic with one purpose: push the button (http://www.thingiverse.com/thing:19912 or similar). It is spot on, for first layers with 0.2mm height, even when switching between different glass plates on my HP. the only thing I see over time is a drift of the Z=0, where it may start that a print needs 1 click adjust, then 2, and once it hits 4-8 clicks, it's time to adjust the level with a full turn of the M3 screw... not sure what is drifting over time (over 2 weeks or so), but it doesn't bother me as much.

[gadgetfreak 7]

Thanks for input Joergen.

Would the spring make any improvement to X or Y (it seems the bed can shift/vibrate in X and Y sometimes I've seen)?

o you have part numbers or details of the springs you are using?

Daniel

 

[billh 0]

Daid,

I believe print bed zero height / parallelism repeatability is the most common cause of print failures for beginning users and is the biggest waste of time for most users.

The platform rigidity and nut attachment to the platform needs significant improvement for the Ultimaker to reach its potential.

Best, Bill

 

[chopmeister 0]

I never cared much for Z repeatability. I switch layer heights between practically every print. In that scenario it would take ages to setup the Z adjuster (or whatever you have) correctly for a single print and then do it all over again every time. I use a double or triple skirt on every print, which gives me enough time to manually adjust the bed to the appropriate height. I got better at this with time and now I miss maybe one print in twenty five or more. Even when I do, the effort required to restart is much less than fiddling with the Z switch constantly. I just keep it in a position where the nozzle homes safely above the bed, and that's it.

Keeping the platform level was always a much bigger issue for me.

 

[jelleatprotospace 0]

Daid, did you test the endstop at different temperatures?

 

[illuminarti 18]

I am always slightly bemused by the angst over bed height and level. I can see why it might be especially critical for very thin layers, but when printing in the 0.1 to 0.2mm layer height range, it's just not a big deal for me at all. I never have to adjust the z end stop, and I occasionally quickly adjust the bed corner spring screws using a sheet of paper as a feeler gauge. And then it 'just works'.

I'm printing PLA at 220-230ºC onto an unheated bed covered in isopropyl-alcohol-wetted blue tape. I add no extra layer height on the first layer, and normally print my first layer at 75 to 100 mm/s. What am I doing wrong??