Questions on trying to dial in my first layer on a Mini.
My first layer is going down a bit rough: there are ridges or fins parallel to the direction of the interior printing where one line meets the next. Is this normal? It shows up most in the first layer, by the next layer, it looks better, and I’m happy with the surface quality on finished parts I’ve printed. I think I’ve got the e steps and the filament diameter dialed in well.
To better diagnose, I printed a 75mm test disk at .20mm thick (starting from the Cura PLA standard quality profile, I reset the thickness of all layers to .20mm). Measuring out on the perimeter, on the smooth tracks which are laid down first, I get about between .007" and .008" which equates to about .18 or .19mm. measuring toward the center, where the ridges are, I get .22mm. Which number should I be looking at to determine if my z-offset height is were it should be? Is there a better way to determine the actual printed z dimension being laid down? Maybe print a single trace?
I printed another 75mm disk at .40mm thick. It printed in 2 layers. the ridges fro the first layer were no longer visible, and the ridges from the second layer were much less pronounced - almost as good as the finished surface on some 5mm thick prints I did earlier. Out on the smooth perimeter, I get .32mm. In the center, where the ridges are, I get readings from .39mm to .41mm
Does this sound like something I should be trying to fine-tune?
Bed adhesion is good (in fact, I wouldn’t mind a bit less, but I haven’t tried to adjust for it, since I figure things may change once the PEI surface is no longer brand new)
According to my experience, the best way to control the Z offset is to print a single perimeter line. I’m always using the prime perimeter that is printed around a part before the real part. In this case, ignore the first few printed cm as the nozzle is not fully primed there. Also keep away from edges if you print a square, best is to print a circle. If you do so, you should get a quite constant measurement.
OK. I wasn’t sure if the skirt was actually laid down with much attention to dimension. I’ll check that, but if the edge of a circle is a good measure, then it appears I’m a bit undersized on the z dimension.
I did replace the PEI sheet on the bed, but used one that was a Lulzbot part (thinking it would be the same thickness. Had not done enough printing and tuning with the old PEI to know if it was OK with that or not.
If I do have a z-offset problem, how do I find out the current setting, and how do I correct it (I’m most familiar with Cura, but also have Pronterface).
For me this first layer doesn’t look too bad. What you can see here is harmonic ripple (the nozzle is “jumping” up and down). Maybe it gets better when you fine-tune your z-offset, also a low print speed and acceleration can help. But it might be always there to some degree…
If you find an offset for z, you can enter it in Cura in Machine->Machine settings.
If your top surface also looks like your 1st layer (regarding the strings in the upper area in your picture), you might try to reduce you e-steps a tiny little bit. But it should be near to perfect!
I see the place to enter the Z offset in Cura machine settings, but how do I find out what that setting is now? It just shows 0 in machine settings. The sheet that came with the printer said z offset was 1.32, but I don’t know if that number is what is actually in the printer now. Also, that numbers was with the original PEI sheet (which has since been changed).
I assume I’ve got to add or subtract from the existing z-offset number, but need to figure out what that existing number is. Also, does the machine settings page save this to the EEPROM, or do I have to do an M500 command (or something else) to write it in. It would be nice to have a setting that “sticks” with the printer, since there are several different laptops that are used to control the printer.
Now that I have my reading glasses on, I can see the edge of the disc and read my micrometer better. The curve along the outer rim of the disc measures from .17mm to just under .18mm. The smoothest part of the interior of the disk measures is about .23mm. The 2 layer disk I printed looks noticeably better, most of the strings and bumps have been smoothed out. The smooth edge of the disk measures low: it should be .40mm, but actually measures .354mm. The smoothest part of the center measures .394mm. My 5mm thick parts are coming out exactly 5.00mm.
So I’m wondering if I should just leave the z offset alone, or try to tweak things a bit. I do plan on playing around with the e steps a bit more when I get the chance to snag the printer again, using Triffid Hunter’s e steps fine tuning procedure. For now, it’s working well enough for what the kids need to do.
What is your initial line thickness setting set at? The default is 125% which might account for the extra height in the center of the piece as the extra plastic doesn’t have anywhere to go but up.
hemocyanin is definitly right, the Mini out of the box (and the Taz too for that matter) is set to extrude the first layer at greater than 100% of the necessary plastic. This is to ensure that new users in particular get a very good adhesive bond to the PEI. Once you get really familiar with the printer, and have it dialed in perfectly you can actually drop that down to normal, wich makes it theoretically easier for the part to dislodge from the bed, but also makes it so the resulting part doesn’t have the “elephant foot” edge at the base.
Some of the rest of the variance you are seeing is the X and Y bed rods. The mini is smaller than the Taz, so it doesn’t show the droop / hump as much as a Taz can (which is still not all that much at all) but it does have a variance in the middle and edges, usually around 0.03mm +/- 0.01mm.
You will also get some layer variance anywhere the filliament makes a 180 degree direction turn. It’s basically like a car going around a tight curve, it tends to get forced to the outside of the curve. 3d printers are the same way, so you tend to end up with a slight depression and then a slight overthickness particularily at perimiter edges. There are servo based 3d printer controllers in development that can compensate for that deceleration, but they are expensive and require custom firmware and interfaces.
The plastic will tend to act as a self leveling compound to a certain extent. Subsequint layers are tolerant of small imperfections, espeially bumps. It never hurts to get the bed dialed in further, but there are also some limitations of the machines to work around. You can account for that by lowering speed, or taking your printer apart and doing crazy bearing experments on it, etc. but generally unless you are doing very tiny layer prints you can probably concentrate on final dimensions of finished parts. If those are off, then you need to adjust things.
in the “Advanced” tab of the “use full settings” window, the default (when copying eSun PLA standard profile into the full settings) is for “initial layer thickness (mm)” of .425mm. I set that to 0, which causes the initial layer to print at the same thickness as other layers. So my intitial layer should have been following the .20mm layer height I had set in the basic tab setting for “layer height (mm)”.
There is also an “initial layer width (%)” setting. The default here is 125%. I THINK I left this alone, since I assumed it would compensate for the extra width by spacing the print lines apart a bit more. However, it’s possible I was set at 100% on this, since I have done some prints with the line width set at 100%. Unfortunately, I did not record the setting on that (stupid of me, I know).
When measuring thickness, I measured way out on the edge, where the print was clean from the two outer perimeter circles, or found a spot in the interior that did not have various print artifacts (threads/bumps, etc) - basically, I tried to visually pick the best interior surface to measure.
Hey John, I was seeing the same thing with my prints on my mini. Here’s what Josef from Lulzbot sent me via email. I gotta tell you, those guys are VERY helpful. It has made my purchase of my first ever 3d printer so great that I will buy a Lulzbot when it comes time to upgrade.
I think he means the lower right corner when you type the commands.
From Josef:
It’s possible we are seeing this because the Mini’s auto leveling is leaving the nozzle too close to your heated bed. We can adjust your Printers Z offset through Cura to leave that nozzle higher from the bed and see if that helps. Go ahead and connect to your printer, and bring up the control box. In the lower lefthand corner you will notice a text entry box. Within the box you are going to need to enter some manual commands to update the offset. Make small adjustments when changing, as large ones can cause your printhead to be dug into the bed or print in mid air.
M851 -> Reports current Z offset in mm
M851 ZXXX -> Changes offset to XXX in mm
M500 -> Saves settings
Here is an example:
If your M851 is -1.35 and you want to move the nozzle away from the bed, Then make the number closer to 0.
Make a M851Z-1.25 enter
This will move the the nozzle .1 farther from the bed.
Then do M500 enter to store the change, or it will go back to the previous setting.
When first testing this new offset, keep a careful eye on that first layer. If it appears the nozzle is being dug into the bed, turn off the printer and adjust the offset until it no longer does that.
Oops, in my post above I meant layer width – sorry to conflate the two settings.
Anyway, it is common to see the first layer to be a little wider because of the layer width setting being set at 125% or whatever – basically you are pumping more plastic through the nozzle and if the slicer doesn’t perfectly compensate for the real world, there is likely to be extra plastic looking for a place to go. In the center of the piece, it will likely ooze upward. On the edge, it likely oozes outward.
I saw that, I just didn’t know how to verify what the current setting is, so I knew where to go with the new setting. Also, I wasn’t sure if that box would save it to the EEPROM or not. They may be using one of several laptops to control the printer, and keeping them all in sync may be tough.
That settting in software won’t save to EEPROM. It should stick in software. Its a variable that may need to be adjusted based on the size or how much of the object base is touching the bed.
I hate to say it, but this may be a losing battle… the machine will likely come out of calibration. If its not build-up of material on the nozzle or contacts which prevent proper bed leveling, or a loose belt… Its just going to take some monitoring of the machine unfortunately.
I finally got my hands on the printer again. Tried a few different settings for the z-offset and eventually settled on one that was 0.1mm higher up than it was originally. I used the procedure Rhoderman posted, so the setting is stored and will remain consistent regardless of what computer is driving the printer. I’m really happy with the differences all the adjustments have made. Hopefully, this will also address the difficulty that the students were having removing finished parts from the bed. (It’s hard to tell about adhesion with the smaller parts I was printing today. Hopefully, I’ll get a chance to try it out further tomorrow.)
Thanks for all the tips here and in other threads. Getting the e steps dialed in, and now the z offset has made a big difference. The first layer is much smoother, and the ridges that ran parallel and between each line of print have pretty much disappeared.
The harmonic ripple (different than the ridges I mentioned above) went away early on in the process as soon as I stopped over-extruding so much. It’s got me wondering if what was really happening here was that the nozzle was “surfing” on th efirst layer or two of molten plastic - between the over extrusion and the z-offset being too low.
What you can see here is harmonic ripple (the nozzle is “jumping” up and down). Maybe it gets better when you fine-tune your z-offset, also a low print speed and acceleration can help. But it might be always there to some degree…