First Layer

Im in need of help to get a better first layer. the attached photo of a pla print was originally face down on my taz5. How can I improve this print as this is the side I need to look good.

Hello gemmark. Did you read my Guides linked in my signature? They will help you with a strategy to get the great prints you are looking for. This looks like a typical case of “first layer blues” - a combination of a too low nozzle at Z=0 and poorly prepared build surface. There may be some extrusion issue mixed in there too.

Was this your first print? How much experience do you have?

On PEI, I print PLA and get absolutely flawless bottom surfaces 100% of the time for the 3D printed fly reels I manufacture. It can be done and you are already a big step in the right direction with a PEI bed and TAZ 5.

I have not had the opportunity to read the entire guide, but did skim through it. I have very little experience with this printer but very eager to learn. The build plate was heated to the suggested temp and it seems to be level. I did print the rocktopuss as a first print and did notice a similar first layer but not as poor as my file. In a quick summary, what would you suggest?


Read my guides and follow them :slight_smile:

First, set the Z height properly. I’d reads it up by .02mm and retry. Keep doing this until the first layer looks good. Best to use my single layer object rather than your print, just to save time and filament. If you can’t get the simple single layer object to print nicely with a Z layer change (by far the most common issue) then we need to be systematic and go through the Guide.

Are you using Slic3r?

If you are, there’s a specific setting you can adjust for the first layer. You can make it bigger or smaller in height than all other layers. It’s completely customizable to your needs.

Also - are you using the material profiles that Lulzbot provides?

I do use Slic3r for some things. However, I assert tha the intent is not to compensate for an incorrect Z setting, it is to legitimately change the layer height of the 1st layer. If you use this feature and print the 1t layer, it should be exactly the height that you set. If you use this feature to compensate for printer adjustment, when you attempt to reprint the gcode file in the future when your printer might have changed, it will likely fail and you’ll need to replace. Whereas, if you adjust the printer to be accurate, you can print the same gcode file anytime in the future perfectly.

It looks like your z-stop is set too and therefore the filament is not sticking where it should on the PEI bed.

To initial z-height:

  • Home XY and Z
  • Slide a business card between the nozzle and bed. The card should slide through with a bit of resistance.
  • If the card just slides through, or doesn’t slide through, adjust the height through the Z-axis end-stop knob. Clockwise to lower the nozzzle, CCW to increase height of nozzle. Quarter turns are a good increment.
  • After each adjustnment, home the Z to register the adjustnment
  • When done, move to each corner of the bed and check for a level bed.
  • If the gap between nozzle and bed are different, adjust the corner screws to raise or lower the bed… DON’T re-adjust the Z-stop knob.

Once you’ve got a good initial height, use the “Z-Offset” setting in your slicer software to tweak the nozzle height for best adhesion. If your off-set gets past .3 or .5, then adjust the Z-stop via the knob as mentioned above.

kcchen_00, I disagree, see this marked up photo:

These are clearly areas where the nozzle is smooshing the filament into the bed - it is too close, not too far away. I’d also say from the even distribution of “smoosh” that at least over the area of this part the bed is reasonably level.

And as I said somewhere in an earlier post today, it is bad practice to compensate for Z height first layer issues in the slicer gcode. Set your Z=0 height correctly and be done with it. You’ll be able to reuse gcode in the future that way. Otherwise, you have to slice and remember what to change every time you print.

Ok… I stand corrected. Maybe there is good smoosh. But how do you explain the areas where the filament width is clearly delineated? If the initial height was too low it would be uniform one way or the other across the bottom. Looks like an uneven bed surface… Almost like air or dust under the painters tape/kapton.

Edit: To check whether the print surface is even, use the edge of a ruler and a flashlight. If you can seek peeks of light under the ruler then the surface may not be even. Best done after leveling the bed to ensure its not deformed from twisting.

I also agree a properly adjusted bed and initial z-height cannot be overlooked. Using the z-offset offers the precision of .1, or .01, increments the initial z-height to vary adhesion to the PEI. That precision can’t be achieved with the z-stop adjustment knob. Partially because there is play introduced by the spring, and the bed deflects when using rigid feelers. I’m actually thinking a rigid standoff would keep the X0Y0 more consistent… versus the springs which deflect.

Ill give it a shot later today and will post my results.

kcchen_00 - I’ve seen very similar patterns with too low a nozzle. What happens is the filament can’t flow out properly, building back pressure. This leads to the visible filament strands. Then, when some pressure threshold is reached, the hotend releases that backed up filament, resulting in the shiny “smooshed” areas you see. This does not completely make sense from what I see in the photo but, but fix one thing at a time starting with the biggest issue will usually obviate the secondary issues that were being masked.

Yes, too close resulting in blobs from the built up backflow is what I’ve seen too… the edges of the print resemble this, but honestly looks more like the filament extruded above the bed lacking detail. I’ve never seen the completely smooth surface and striations of the extruded filament.

Its an interesting one to troubleshoot…

I’ve had my Taz 5 for a year and have printed over a hundred of great prints in PLA, ABS, and HIPS with the original 0.35 mm nozzle. I recently swapped out the Hexagon hot end with the 0.5 mm nozzle. That’s the nozzle they shipped for the 2nd half of Taz5 production. I’ve been having terrible problems ever since. I followed the online instructions for assembling the toolhead with filament in it to align the hot end and I’m using the 0.5 mm Lulzbot Cura profiles, but get lots of bad artifacts on the first layer. All of this has been with PLA.

I also rezeroed the Z-axis, re-leveled the bed, and printed the the “star/plus-shaped” bed calibration gcode.

I never had these sorts of issues over a year of use with the 0.35 mm nozzle. Any ideas? Here are a few jpgs with PLA.



Looks to me like your Z is too low.

Thanks – I went through the “paper thickness” recalibration again and this time calibrate z a bit higher. That made all the difference - thanks!

I do still have another issue that I haven’t been able to find info on. When doing a fill, regardless of whether it’s a bottom layer or top layer, I get bumps that I’ve never had before. I’m using a Taz5, latest Marlin firmware, Cura 19.12, latest Cura profiles and a 0.5 mm nozzle.

I previously used this machine for a year with the 0.3 mm stock (at that time) nozzle and never had these sort of artifacts. In the pictures you can see spots where there are entire lines where there is a bump, almost like it extruded twice along the same ‘track’. It seems to always occur in a place where the slicer moved to somewhere else then came back. Backlash in X or Y positioning?

All the belts on the X and Y axes seem tight and it doesn’t seem like there is any unusual play. Z axis seems firm too. Using Cura PLA fine profiles and same PLA I had no issues with the 0.3 mm nozzle (I always keep in plastic bag with dessicant too between prints).

A bit stumped right now…

Has anyone seen this sort of “rib” feature before? I haven’t been able to narrow down the cause.

Yeah that’s either excess coming out during travel due to overextrusion, or the same symptom due to overtemperature causing the plastic to be overly fluid. The rest of the layer doesn’t look overextruded so I would start by lowering nozzle temperature, upping retraction before travel moves slightly, and check for an air leak in your nozzle body.

I thought it was overextrusion as well at first, but not really anymore. I have previously dialed down from 205 deg C to 190 deg C, to help with that. And I did previously increase the retraction from 1 mm to 5 mm and from 10 mm/sec to 35 mm/sec.

The interesting thing that makes me think its not a retraction issue is that the rib is always for the entire length. Its not like a variable length dribble. And watching the test part print, it is the final final pass it makes. Its like it finishes the layer, then does one more pass, directly on top of plastic already laid for that layer. The pictures below show the time sequence. The 2nd picture shows that the “slit” of not yet laid plastic is the last area to be filled in. It is being filled diagonally from left to right. The final pass is on the right of the slit, and that is where the pronounced bump / rib is deposited. On more complex parts, this happens numerous times when the slicer has decided to go fill a region that was previously left empty.

Could it be a CURA g-code generation issue? Or a positioning issue in the XY axes where it’s not locating the nozzle in XY space exactly correctly and depositing in the wrong location? Or play/slop in the XY axes causing the same? I don’t detect any play in the X or Y axes linearly (along their travel directions), but the Y axis does seem to have some play rotationally, I mean that I can twist the bed and feel some play/give in the y-axis bushings. However I’m not familiar enough to know whether that is the same play that has always been there or if this has developed over the past year of use.

Here are the reprints I did again today to watch the sequence of what’s filled when:

What Tim is talking about is called ‘Z Hop when retracting’, not the retraction distance. It aids the hot nozzle from digging/dragging into the layer between the inner movements as it fills in the remaining sections it creates to spread the heat around during layer printing. The default is set to 0.0 MM(I believe) but a setting of 0.1 MM or more makes a big difference.

I did experiment with Zhop as well. The Lulzbot PLA “fine” profile for Cura had this at 0.1 mm for all of these prints. Testing at 1.0 mm made things no better.

Retract settings have been tried at default (1 mm distance, 10 mm/sec speed) and 5 mm distance with various speeds up to 35 mm/sec based on other posts.

I don’t believe this is a retraction issue since any swath of lines that are high or low are always high or low for their entire length.