Phase 2 · The filament · L15

Max volumetric speed

Your real speed limit — and the reason a fast print can come out matte, weak and under-extruded without a single error message.

⏱ ~50 minutes• 🌀 One spiral-vase cylinder• Phase 2

◎ Why this lesson

Prusa's definition is exact: MVS is "the maximum amount of plastic (mm³/s) your hotend can reliably melt", and it acts as a speed limiter — PrusaSlicer lets your requested speeds through when safe and reduces them only when the hotend would be asked to melt more than it can.1 Set it correctly and you never have to calculate safe speeds for a nozzle-and-layer-height combination again. Leave it at a default and you either lose speed you had, or exceed a limit that has no alarm on it.

▲ MVS never makes anything faster

"MVS never increases your speed, it can only reduce it."1 That is worth holding onto, because it makes the setting safe to raise carefully and dangerous to raise carelessly — nothing will go faster than you asked for, but plenty can go faster than the hotend can supply.

1Four things limit your print speed, and MVS is one

Every extrusion move checks several limits and takes the most restrictive:1

Per-feature speed What you set in Print Settings → Speed for perimeters, infill, bridges and the rest. That's Lesson 16.
Max volumetric speed Reduces the requested speed when flow would exceed the ceiling. Wider and taller extrusions get slowed more, because they carry more plastic per millimetre.1
Overhang and cooling slowdown Overhanging perimeters are slowed based on how unsupported they are.1
Layer time cooldown If a layer is too short in time, all speeds scale down, with Min print speed as the floor.1 That's Lesson 14.

Knowing which of the four is biting is the whole diagnostic skill. Three of them are visible in the preview if you know where to look, and section 5 shows you how.

2Two settings with the same name

MVS exists in two places: Print Settings → Speed → Max volumetric speed and Filament Settings → Advanced → Max volumetric speed. The lower of the two always wins, and setting either to 0 disables that one and defers to the other.1

⌁ This is a feature, not a redundancy

Prusa's framing: the effective MVS is limited by the weaker link, either the hotend or the material — set each in its respective profile and PrusaSlicer uses the lower.1

So the print profile's MVS is a statement about this machine's hotend, and the filament profile's is a statement about this material. Fill in both honestly and the right thing happens automatically on every combination — which for a three-printer, three-material fleet is nine combinations you no longer have to reason about individually.

The failure this prevents: putting the machine's number into the filament profile means your TPU inherits your PLA's ceiling on that printer, and your MK3S's PETG profile carries the Core One's ceiling the moment you copy it across.

⚠ Zero means no limit, not "off"

Setting MVS to zero removes the limit entirely.1 It reads like a disabled setting and behaves like an unlocked one. If you set MVS to 0 while testing and forget, every speed in the profile goes through unchecked — which is exactly the state that produces a fast, matte, weak print with no warning anywhere.

A second trap on the same setting: putting 0 in a per-feature speed field activates AutoSpeed for that feature, where MVS stops being a brake and becomes the basis for calculating speed.1 That's a legitimate mode, but it is not what you want while measuring, and it is easy to trigger by clearing a field.

3The arithmetic

PrusaSlicer's rule is Max speed = MVS / cross-section area, using a stadium-shaped cross-section model for accurate volume calculation.1 Turned around, that is the extrusion equation from Lesson 3 — the one piece of arithmetic the whole course rests on.

Flow, from settings you already have

flow (mm³/s) = layer height × extrusion width × speed (mm/s)

# 0.2 mm layer, 0.45 mm wide, 100 mm/s
0.2 × 0.45 × 100 = 9.0 mm³/s

# the same profile at 250 mm/s
0.2 × 0.45 × 250 = 22.5 mm³/s  ← past an MK3S hotend

# and going taller costs more than going faster feels like
0.3 × 0.45 × 100 = 13.5 mm³/s  ← +50% for one setting

The stadium model means the true cross-section is slightly less than width × height — a rectangle with rounded ends, not a rectangle.1 For tuning purposes the simple product is close enough and errs on the safe side; PrusaSlicer does the precise version at slice time.

4Measure your ceiling

Print a spiral-vase cylinder — single wall, 0.2 mm layers, continuous extrusion with no travels and no seams to confuse the picture — and ramp the speed through it with M220, the speed factor override.2

⬇ Download · the cylinder

mvs-vase-cylinder.stl — a 40 mm diameter, 40 mm tall cylinder. Slice it in Spiral vase mode at 0.2 mm layers; the M220 ramp below steps every 30 layers, so each speed gets a 6 mm band of wall to go matte in.

Printer Settings → Custom G-code → After layer change G-code

{if layer_num == 30}M220 S150{endif}
{if layer_num == 60}M220 S200{endif}
{if layer_num == 90}M220 S250{endif}
{if layer_num == 120}M220 S300{endif}
; and set MVS to 40 for the test, so it doesn't clamp what you're measuring

Set MVS artificially high — 40 or so — before printing this. The whole point is to find where the hotend fails, and a correctly-set MVS would prevent exactly that. Put it back afterwards.

⌁ M220 ramp generator — and the flow each band demands

Defaults reproduce the ramp above. Add your vase wall speed and the calculator shows the flow sweep, so you know before printing whether the ramp actually crosses your machine's expected ceiling — a ramp that tops out below it finds nothing. Paste the block below whatever is already in the after-layer-change pane; the bottom band runs at 100 %.

First factor % Step % Steps Layers per band Layer height mm Wall speed mm/s Extrusion width mm

Flow at 100 %

Flow at top band

Height used

Past the top of the cylinder. mvs-vase-cylinder.stl is 40 mm — the last bands won't print. Fewer steps or shorter bands.

MVS to 40 before the print, back to the measured value after — and the block comes out of the printer profile with it.

Find where the wall changes Look for the band where the surface goes matte, thin or glassy instead of evenly glossy. That is the hotend running out of melt capacity — it is delivering less plastic than the slicer asked for.
Listen and watch as well Prusa's signs of MVS set too high are extruder clicking, under-extrusion at high speeds, and poor layer adhesion.1 Clicking usually arrives one band before the visual change.
Compute the flow at that band layer height × extrusion width × (profile speed × M220 %). Write the number down.
Take 85 % of it The failure point is not the working point. 85 % leaves room for a colder day, a wetter spool and a slightly different colour.4
Set it in the right profile The material-derived number goes in Filament Settings → Advanced. If you're also establishing the machine's hotend ceiling, that goes in Print Settings → Speed.

Open your machine. The expected answers are far apart.

MK3S · standard all-metal hotend, and the lowest ceiling of the three

Prusa's own hotend table puts a standard all-metal hotend — explicitly including the Prusa MK3 stock hotend — at 8–12 mm³/s.1 This course's guide puts the practical ceiling slightly higher at ~11–13 mm³/s;4 treat 8–12 as the conservative, Prusa-documented figure and anything above it as something you must have measured.

Combined with the material limits1 — PLA 12–20, PETG 8–15, TPU 2–5 — the hotend is the weaker link for PLA on this machine, and the material is the weaker link for TPU. That is exactly the "lower of the two wins" mechanism doing its job.

Working targets to test against4: PLA ~13, PETG ~8, TPU 95A ~1.8–2.5.

The practical consequence: at 0.2 × 0.45, a 13 mm³/s ceiling is about 144 mm/s. You will not reach that on this machine anyway — the Y-axis mass gets you first (Lesson 8). On the MK3S, MVS is mostly a safety net rather than an active constraint. That is fine. Set it correctly anyway, because it is what stops a copied-from-elsewhere speed profile from quietly starving the nozzle.

MK4S · where Prusa's own table has fallen behind

⚠ The KB hotend table predates your machine

Prusa's table lists "Prusa Nextruder (MK4/XL)" at 15–20 mm³/s, high-flow nozzles such as CHT at 20–35, and high-flow hotends at 25–40.1 It does not list the MK4S or the Core One at all, and readers say so in the article's comments — one relaying that Prusa's MK4S launch material quoted 24 mm³/s for Prusament PLA against the MK4's 16, and another reporting a support answer of 24 mm³/s on high-flow nozzles versus 15 on standard.1

Those are comment-thread numbers and marketing copy, not KB specification. They are the best available public figures and they are consistent with this course's ~22–28 mm³/s estimate,4 but they are precisely the kind of number this course exists to replace with one you measured.

Which row applies to you depends on your nozzle. An MK4S with the stock high-flow nozzle sits in the 20–35 band; the same machine with a plain 0.4 nozzle fitted sits much lower. Check what is actually installed before you pick a starting point — this is the one setting where the physical nozzle changes the answer more than the firmware does.

Working targets to test against4: PLA ~24, PETG ~10–13, TPU 95A ~3–4.

And remember the temperature coupling. Prusa note that with larger nozzles you should consider raising nozzle temperature to allow a bigger MVS — PLA 10–20 °C hotter than standard on a 0.6 mm.1 The same logic applies to high flow through a 0.4: if your measured ceiling here is lower than you expected, going back to Lesson 10 and adding 10 °C is a legitimate move. Just re-measure extrusion multiplier afterwards, because temperature is upstream of everything.

Core One · same hotend class, and one enclosure effect worth measuring

Same hotend class as the MK4S, so the same starting bands: PLA ~24, PETG ~13, TPU 95A ~3–5.4 And the same absence from Prusa's published table.1

The enclosure's contribution is ambiguous and worth actually measuring. Warm chamber air means filament entering the extruder is warmer, which helps melting. It also means the heatsink has warmer air to reject heat into, which hurts. Which effect dominates on your machine, in your room, is not something you can look up.

So run the ramp twice for PLA — vented and closed — and note whether the failure band moves. If it moves toward lower flow when closed, you are seeing heat creep territory and that is a real operating constraint on this machine, not a curiosity.

The MK4S and Core One will land close on filament MVS — same melting capacity — while landing far apart on speeds and accelerations in Lesson 16, because one is a bedslinger and the other is CoreXY. Same flow ceiling, different motion ceiling.

5Verify it in the preview

Slice a real part, then switch the preview mode to Volumetric flow rate using the select box at the top left.1 You now have a colour map of exactly how much plastic every move demands, on the model you actually print.

Prusa's own reading guide: on small models and at small layer heights you probably aren't capped at all; on bigger models with tall layers you may see the limit bite, especially on infill.1 And if most of the model is being slowed by MVS, their advice is counter-intuitive and correct — lower the layer height. You get more detail without increasing print time, because the printer can then run faster.1

This is the cheapest habit in the whole course. It costs one dropdown change, it needs no filament, and it turns "I think my speeds are sensible" into "I can see where my speeds are being clamped and by what".

6Find the fault

Click each. All five produce a print that is slower or worse than the profile promised.

Why isn't this printing the way the profile says?

Infill prints far slower than the number in the profileMVS cap
Fast prints come out matte, thin and weakNo ceiling
Only the top of a tapering part prints slowlyWrong limit
One feature runs at a speed you never set anywhereAutoSpeed
TPU strips in the extruder on your fastest printerWrong profile

7Check yourself

From memory. Question 1 has arithmetic in it.

Question 1 of 3

0.2 mm layer, 0.45 mm width, 200 mm/s. What flow does that demand?

Question 2 of 3

You set MVS to 0 during a test and never changed it back. What happens?

Question 3 of 3

Why does MVS deliberately exist in both the print and the filament profile?

✎ Your task · do this with your tutor

Find the ceiling, then take 85 % of it

Report which nozzle is physically fitted to the machine you're tuning — stock, high-flow, or something you replaced. This changes which of Prusa's bands applies.
Set MVS to 40, add the M220 ramp, and print the spiral-vase cylinder. Tell me the profile speed you started from so the arithmetic works.
Report the band where the wall changed, and whether the extruder clicked before or after that point. If clicking came first, the click is your real ceiling.
Compute the flow at that band, show the arithmetic, and give me 85 % of it.
Set it in Filament Settings → Advanced. Then slice a real part, switch the preview to Volumetric flow rate, and tell me whether anything is being clamped and where.
Confirm the M220 block is deleted.

→ Point 5 is the one that changes how you work. Once you've seen your own model coloured by flow demand, "is this profile fast" stops being a feeling and starts being something you can look at.

What you can now say

MVS is what the hotend can melt, and it only ever reduces speed — never raises it.
Two MVS settings, and the lower wins: hotend in the print profile, material in the filament profile.
Zero means no limit, and a zero in a speed field means AutoSpeed. Same value, two different meanings.
Measure the failure point, then take 85 %. Clicking often arrives a band before the surface changes.
The Volumetric flow rate preview settles arguments — and if most of a model is clamped, lowering layer height gains detail for free.

📖

Primary source · read this next

Max volumetric speed — Prusa Knowledge Base

The best-written article in Prusa's slicer documentation, and the source of nearly every fact in this lesson. Read the comments as well — they are where you learn that the hotend table hasn't been updated for the MK4S or Core One, which is exactly the kind of gap worth knowing about before you trust a number.

💬 I'm your tutor — ask me anything. Send me your ramp arithmetic and I'll check it before you commit a ceiling to a profile — an MVS that's 15 % too high is invisible until a big part comes out weak. Screenshots of the volumetric flow preview are very easy to read together. Confused beats stuck.

Linked: Glossary → max volumetric speed, volumetric flow rate, extrusion equation, AutoSpeed, extrusion width, layer height, hotend, bedslinger, CoreXY, calibration order. Keep it close with → The Calibration Card.

§References

Prusa Research. Max volumetric speed. MVS as "the maximum amount of plastic (mm³/s) your hotend can reliably melt"; that it "never increases your speed, it can only reduce it"; the four limits checked per extrusion move (per-feature speed, MVS, overhang & cooling, layer time cooldown); Max speed = MVS / cross-section area with a stadium-shaped cross-section model; that MVS is settable in both Print settings → Speed and Filament settings → Advanced with the lowest value winning and zero meaning no limit; AutoSpeed activation when a per-feature speed is set to 0; "Start conservative and increase gradually" with extruder clicking, under-extrusion at high speeds and poor layer adhesion as signs of MVS being too high; the effective MVS being limited by the weaker of hotend and material; the filament table (PLA 12–20, PETG 8–15, ABS/ASA 10–16, TPU 2–5, PC 6–10, PA 8–12 mm³/s); the hotend table (standard all-metal including Prusa MK3 stock 8–12, Prusa Nextruder MK4/XL 15–20, high-flow nozzles 20–35, high-flow hotends 25–40 mm³/s); raising nozzle temperature with larger nozzles; and the Volumetric flow rate preview mode with the advice to lower layer height when MVS is the binding constraint. The MK4S and Core One figures quoted in section 4 come from the article's reader comments, not from Prusa's tables, and are flagged as such in the lesson. Link.
Prusa Research. Buddy firmware specific G-code commands. M220 — "Set speed factor override percentage", parameter S<percent>. Link.
Prusa Research. Cooling. Min print speed and the layer-time brake — the limit most often mistaken for an MVS cap. Link.
This course. Master tuning guide, §2.6, §3 and §4 — the M220 ramp method, the 85 % rule, the practical per-machine flow ceilings (MK3S ~11–13, MK4S and Core One ~22–28 mm³/s), and the per-material MVS starting values. Link.

Lesson L15 of the prusa-profile-tuning course · all lessons · The Calibration Card · glossary