Phase 1 · The machine · L7

First layer and Z-offset

The one measurement where your MK3S and your MK4S are not doing remotely the same thing — one senses metal, the other senses touch.

⏱ ~25 minutes on the machine• 🖨 One calibration print per sheet• Phase 1

◎ Why this lesson

The first layer is the reference every subsequent layer is stacked on. Get it wrong and you get adhesion failures, elephant foot, and dimensional error in Z on every part — none of which any filament setting can rescue. It is also the last thing standing between you and Phase 2 that varies per sheet rather than per machine, so it is the item on the health check most likely to drift without you noticing. And it is the clearest example in this whole course of why the per-machine split exists: your three printers use two genuinely different sensing technologies with two different procedures.

Two independent things have to be right, and conflating them is the usual source of confusion:

Shape — mesh bed levelling The bed is never perfectly flat. The printer probes a grid of points before each print and warps the first layer to follow the surface. This is automatic on all three of your machines and it compensates for shape.
Height — the Z offset How far the nozzle sits above the sheet, everywhere. Mesh levelling cannot fix this, because a surface offset by a constant is still the same shape. This is the number you set, and how you set it depends entirely on your probe.

1Two probes, two philosophies

Why the procedure is not shared.
	PINDA — MK3S	Loadcell — MK4S, Core One
Senses	metal, inductively, from a distance	force — the nozzle physically touching
Knows where the nozzle is?	No. It knows where the probe is	Yes. The nozzle is the sensor
Nozzle-to-sheet distance	set manually, stored, reused	measured automatically before every print
Drifts with temperature	yes — the probe body expands	no
Manual calibration needed	yes, per sheet	no1

That third row is the whole difference. The PINDA tells the MK3S how far the probe is from the steel sheet. Nobody has told it how far the nozzle tip is below the probe — that gap is what Live Adjust Z stores, and it changes when you swap sheets, swap nozzles, or knock the probe.

The loadcell sidesteps the problem entirely by making the nozzle its own sensor. Prusa are explicit: with the loadcell, manual first layer calibration is not necessary, because the value is checked automatically before every print; the only adjustment it needs is during the selftest.1

2Do it — on your machine

Open your machine — only yours.

MK3S · Live Adjust Z, per sheet, stored

Run the calibration print: LCD → Calibration → First layer calibration. It prints a zig-zag pattern and lets you adjust live while it runs.

To adjust: LCD → Live adjust Z while printing, or LCD → Settings → Live adjust Z beforehand. Long-pressing the knob during the first layers is the quick route.2

▲ Which way is which

Counter-clockwise → more negative → nozzle closer to the sheet.
Clockwise → less negative → nozzle further away.
Typical values land between −1.500 and −0.200.2 Adjustments made during a print are saved and persist after power-down — so if you nudge it to rescue one print, you have changed it for every future print on that sheet.2

Do this once per sheet. A textured sheet, a satin sheet and a smooth sheet have different thicknesses and different surfaces; the offset is not shared between them. Log a value per sheet in tuning-log.md, and re-check after any nozzle change.

PINDA temperature drift is real. The probe reads slightly differently cold versus fully heat-soaked. Calibrate on a printer that has been up to temperature for a few minutes, not on one you just switched on.

MK4S · automatic, with a temporary override

The short version: there is nothing to calibrate. Provided the loadcell step of the selftest completed (Lesson 5), first layer height is measured automatically during mesh bed levelling before every print.1

To nudge it during a print: long-press the knob while the first layer is printing. Note the important asymmetry with the MK3S — on a loadcell machine this adjustment is not saved for the next print.2 It rescues the print in front of you and nothing more.

To change it persistently: PrusaSlicer, Printer Settings → General → Z offset.2 Reach for this only when you have a consistent reason — a sheet you always want slightly different, for instance — and log it, because it is now a per-profile value that will silently follow that printer profile around.

If the automatic result is consistently wrong, that is a loadcell problem, not an offset problem. Re-run the loadcell step of the selftest, check the nozzle is clean and cold-wiped, and see Prusa's Loadcell troubleshooting. Don't paper over it with a Z offset.

Core One · as MK4S, plus the enclosure

Identical to the MK4 family — loadcell, automatic per-print measurement, long-press to nudge the current print only, Printer Settings → General → Z offset for a persistent change.1 2

What differs is the thermal environment. A first layer calibrated in a cold chamber and a first layer laid down in a chamber that has been running PETG for an hour are not the same measurement — the bed, the sheet and the frame have all grown. Two consequences:

Calibrate and judge under the door/vent state you actually print that material in. Record which state, as you did in Lesson 5.
If a Core One first layer looks right at the start of a print and changes character by the time it reaches the far corner, suspect thermal expansion in the enclosure before suspecting the offset.

3Judging it — the only skill that transfers

Whichever machine you're on, the judgement is the same and it is worth more than any of the procedures above. Look at the calibration square with a bright light at a low angle.

Too high — nozzle far

# what you see
lines round in section
visible gaps between them
you can pick at an edge
corners lift as it cools

# what it causes
adhesion failure
weak bottom layers

Too low — nozzle close

# what you see
ridged, ploughed surface
lines squeezed wide, uneven
translucent thin patches
material piling at turns

# what it causes
elephant foot
clogs, Z dimension error

Right is between them: a smooth, continuous, slightly glossy surface with no gaps and no ridges, where individual lines have merged into a single sheet but you can still just make out where they are. It should come off the sheet as one piece, and the underside should be uniform.

⚠ PETG needs a different first layer to PLA

Two things change, and both come from your own guide rather than a Prusa KB article. First, PETG bonds to bare smooth PEI hard enough to tear the coating off the sheet — use a glue-stick separator, or print it on satin/textured. Second, reduce the squish relative to PLA: on the MK3S, Live Z one or two notches further away. Because Live Adjust Z is a per-sheet machine setting rather than a per-filament one, this is a note in your log and a manual step, not something the profile can do for you.

4Read the first layer

Click each observation. Decide first whether it's height, shape, adhesion or something else entirely.

Height, shape, or not the first layer at all?

Round separated lines you can pick apart with a fingernailToo high
Ridged surface with thin translucent patchesToo low
Perfect in the centre, too high at the front-left cornerShape
PETG print won't release, and the sheet coating is markedMaterial
First layer is fine cold, different after 20 minutesThermal

5Check yourself

From memory. Question 2 catches almost everyone once.

Question 1 of 3

Why does the MK3S need a manual Z offset when the MK4S doesn't?

Question 2 of 3

You long-press the knob and nudge Z during a print. On which machines does that stick?

Question 3 of 3

Your first layer is perfect centrally and too high in one corner. What's wrong?

✎ Your task · do this with your tutor

Calibrate and photograph a first layer

On the machine you've been working through Phase 1 with, print a first-layer test — the MK3S's built-in calibration, or a 60 × 60 mm single-layer square on the MK4S and Core One. Then:

Photograph it under a bright light at a low angle. Low angle matters — flat lighting hides everything.
Write down your verdict before you send it: too high, too low, or right — and which specific feature in the photo you're basing that on.
Log the values. MK3S: the Live Adjust Z value and which sheet. MK4S / Core One: confirmation that it was automatic, plus any Z offset you've set in PrusaSlicer and why.
Repeat for each sheet you own. This is the only Phase 1 item that is per-sheet rather than per machine.

→ Paste the photo and your verdict. I'll tell you whether I agree — and more usefully, whether the feature you based it on is the one I'd have used.

What you can now say

Shape and height are separate problems. Mesh levelling fixes shape; the Z offset fixes height; neither substitutes for the other.
PINDA senses the sheet, the loadcell senses the nozzle — which is why one needs a manual offset and the other doesn't.
An in-print nudge persists on the MK3S and evaporates on the MK4S and Core One.
Too high gives round separated lines; too low gives ridges and thin patches. Judge under a low-angle light.
The offset is per sheet, not per machine — and PETG wants less squish than PLA, plus a separator on smooth PEI.

📖

Primary source · read this next

Live adjust Z — Prusa Knowledge Base

Short, with the photographs of good and bad first layers that no written description replaces. Read the PINDA and loadcell sections both, even though only one applies to you — knowing what the other machine does is how you avoid applying its procedure by accident.

💬 I'm your tutor — ask me anything. First layer is the most photograph-able thing in this course, so send pictures. If you're between two verdicts, send it anyway and say which two — being uncertain about the right things is more useful than being confidently wrong. Confused beats stuck.

Linked: Glossary → Live Adjust Z, PINDA, loadcell, mesh bed levelling, build plate, selftest, tuning log. Keep it close with → The Calibration Card.

§References

Prusa Research. Loadcell (MK4/S, MK3.9/S, XL). Manual first layer calibration is not necessary with the loadcell — values are checked automatically before each print, and the only adjustment needed is during the selftest. Link.
Prusa Research. Live adjust Z. The MK3/S/+ and MK4/S menu paths, the −1.500 to −0.200 typical range, the counter-clockwise-is-closer convention, that adjustments made during printing persist after power-down on PINDA machines, that the loadcell equivalent is not saved for the next print, and the PrusaSlicer Printers → General → Z offset route. Link.
This course. Master tuning guide, §4 — PETG on smooth PEI needs a separator, and wants slightly less first-layer squish than PLA. Community-derived; see RESOURCES.md § Gaps. Link.

Lesson L7 of the prusa-profile-tuning course · all lessons · The Calibration Card · glossary