Voron 350 Magbed and Beacon Probe Compatibility

#voron #v2-4 #beacon #magbed #bed-leveling #3d-printing # Overview Evaluating the Mandala Rose Works 350mm UltraFlat bed for the [[Voron V2.4 Build]], and specifically how the magnet option interacts with a Beacon eddy-current probe. All three MRW variants are the same 9mm 5083 cast aluminum tooling plate at the same ultraflat spec (±0.05mm, ≤0.1mm total) — flatness is *not* the variable. The only difference is the magnets, and the magnets are the entire Beacon question. # Resources - [MRW Voron 350 Bed](https://mandalaroseworks.com/products/voron-350-ultraflat-bed) — product page, all three variants - [MRW Railcore UltraFlat Magbed 350C SMCO](https://mandalaroseworks.com/products/railcore-ultraflat-magbed-350c-smco) — same top-magnet design; this page actually publishes the neodymium spec (N48SH, 150C, ~11 lb pull each) - [Beacon FAQ](https://docs.beacon3d.com/faq/) — authoritative on bed surfaces and magnet interference - [Voron forum — Anybody tried a Beacon probe](https://forum.vorondesign.com/threads/anybody-tried-a-beacon-probe.560/) - [Voron3D Wiki — Beacon3D Probe](https://voron3d.wiki/bedleveling/beacon/) - [max_plastix probe-point calculator for MRW magbeds](https://docs.google.com/spreadsheets/d/19sg3DF3PqzK8-KJ13Eh1PQMJRK1Q9LHXkGnUAmX1ld8/edit#gid=1589762554) — linked from the Railcore page, helps plan mesh points around magnets # Design ## The three MRW variants | Option | Price | What it is | |---|---|---| | UltraFlat — No Magnets | $164 | Bare plate. Bring your own retention (adhesive magnetic sheet, clips, glued PEI) | | UltraFlat — 16× Neo 150C | $254 | High-temp neodymium (N48SH on the Railcore version), recessed just under the surface | | UltraFlat — 16× SMCO 350C | $290 | Samarium-cobalt, same pockets, much higher temp rating and notably weaker field. MRW labels this "Better Beacon Compatibility" | Common to all three: - 355 × 355 × 9mm, ~8 lbs - M5×0.8 threaded fuse and thermistor holes - Two extra M5 holes nearer the plate center so bed wiring can run *between* the mounting extrusions instead of over them - "Top magnet" design on the magbed versions — magnets sit slightly *below* the plate surface for maximum holding power. Requires a good 0.5mm flexplate (Buildtak, KB3D, MRW FireThorn); thinner cheap plates can dimple over the magnets ## How Beacon works, and why magnets break it Beacon is an eddy-current sensor. It energizes a coil at high frequency, induces eddy currents in the conductive steel flexplate, and measures how strongly those currents load the coil. That is a measurement of the **electrical properties of the steel a fraction of a millimeter under the nozzle** — not a measurement of distance directly. Ferromagnetic steel exhibits **anisotropic magnetoresistance (AMR)**: a strong static magnetic field changes the steel's resistivity. So a strong magnet under the plate locally alters the exact property Beacon is measuring, and the probe reports it as a height change that isn't physically there. From the Beacon FAQ: > Beacon is not recommended for use with beds featuring oversized fixed magnets... If the magnets present in the bed are strong enough to reach magnetic saturation in the surface plate, very high magnetic field strengths will be present in the outer surface and Beacon will see sharp peaks. MRW's top-magnet design makes this worse than average, for the same reason it holds so well — the magnets are milled in close to the surface, so field strength at the flexplate is high. MRW's own warning on the product page is about inductive/capacitive/BLTouch probes, but the underlying physics is the same family of problem. ## Why SmCo helps Not the 350°C rating — that's essentially irrelevant here. A V2.4 bed tops out around 110°C and the 150C neodymium is thermally fine at that temperature. The temp rating is a *side effect* of the alloy, not the reason to buy it. What actually matters is that SmCo is a **weaker magnet**: | Property | N48SH (neo) | SmCo 2:17 | |---|---|---| | Remanence Br | ~1.37–1.42 T | ~1.00–1.10 T | | Energy product (BH)max | ~46–49 MGOe | ~26–32 MGOe | | Max operating temp | 150°C | 350°C | Lower flux density at the flexplate means less chance of driving the steel into saturation, so artifacts shrink from "sharp spikes that wreck the mesh" toward "small bumps you can exclude or live with." **The cost:** pull force scales roughly with field squared, so expect on the order of half the holding force per magnet versus the neo option. With 16 magnets on a 350 plate that is still plenty for ABS, but it is a real, noticeable difference. ## Fourth option MRW doesn't list Bare plate ($164) + a **magnetized rubber sheet** on top. From the Beacon FAQ: > Magnetized rubber sheets have a high number of poles, and result in no detectable artifacts. The many alternating poles average out at the sensor's scale. This is the cleanest possible Beacon setup and also the cheapest. Trade-offs: weaker hold, a thermal insulation layer between heater and plate, added stack thickness, and adhesive that ages at ABS chamber temps. ## Mitigations regardless of choice - **Mesh exclusion zones** — Beacon software supports them. 16 magnets on a 350mm bed is a manageable number of dead spots if you know where they are. - **Mesh point placement** — pick probe points that dodge magnet centers (see the max_plastix calculator above). - **Beacon RevH contact mode** — homes Z with the nozzle physically touching the plate, so the **Z offset is immune to magnet interference entirely**. Only the mesh scan is affected. This meaningfully de-risks the magbed options. - Note the same physics applies to Cartographer and Klipper's `probe_eddy_current` — this is not a Beacon-specific flaw. A Klicky/nozzle-contact probe would be fully immune. # Decisions - **Decision**: ==Not yet made — pending magnet coordinates from MRW== - **If magbed convenience is the priority**: SmCo ($290). "Better compatibility" is not "no artifacts," but combined with exclusion zones and RevH contact homing it should be workable. - **If a clean Beacon mesh is the priority**: bare plate ($164) + magnetic sheet. Cleanest sensor behavior, $126 cheaper, weaker hold. - **Rejected**: 16× Neo 150C ($254). Strongest hold but the worst case for Beacon — highest odds of saturating the flexplate and producing sharp mesh peaks. Only sensible if switching to a contact-only probe. - Reported mixed results in the wild: some users run Beacon RevH on stock 150C neo magnets without trouble. It is plate-thickness and magnet-geometry dependent, and Beacon has said quantitative guidance is "coming in the future" but has not published it. # TODO - ==Email MRW ([email protected]) for the 16-magnet coordinate layout on the Voron 350 bed — needed to plan mesh points / exclusion zones before ordering== - ==Confirm which Beacon revision is being ordered; RevH contact mode materially changes the risk calculus== - ==Decide flexplate: needs to be a quality 0.5mm sheet (Buildtak / KB3D / MRW FireThorn) if going magbed== - ==Cross-check bed heater choice — West3D universal edge-to-edge 150C is currently listed in the build note== - ==If going bare plate + magnetic sheet, source a sheet rated for sustained 110°C bed temp== # References - [[Voron V2.4 Build]] - [Beacon FAQ](https://docs.beacon3d.com/faq/) - [MRW Voron 350 Bed](https://mandalaroseworks.com/products/voron-350-ultraflat-bed)