Anycubic Kobra Max Heating Element Debug

Anycubic Kobra Max 3D printer product photo — Anycubic Kobra Max

This write-up documents a live debug of an Anycubic Kobra Max that refuses to heat the bed. The machine throws Hotbed heater abnormal, please check it and wiring, then stops. The work so far maps the power path, identifies the hotbed adapter board, and shows that the power MOSFET can still switch when driven manually — which points the fault toward the control / enable path rather than a dead heater pad.

Status (August 2026): in progress. Power MOSFET verified functional with an external ~10 V gate drive. Control-side drive and the source of that gate voltage still under investigation. Board removed for backside inspection.

Printer

Field	Value
Model	Anycubic Kobra Max
Serial	KM2201A0602894
Build volume	400 × 400 × 450 mm
Machine size	665 × 720 × 715 mm
Power	500 W, 110 / 220 V AC
Maker	Shenzhen Anycubic Technology Co., Ltd.

Nameplate on the Anycubic Kobra Max showing model, power rating, and serial KM2201A0602894 — Nameplate used to pin down model and serial before ordering parts or matching wiki guides.

Symptom

With a bed target set (for example 60 °C for PLA), the display stays near ambient and the printer raises the fault dialog. Nozzle control can still look normal while the bed path fails.

Printer touchscreen showing Hotbed heater abnormal please check it and wiring — Fault dialog with bed at 26 / 60 °C and nozzle near target — the bed never ramps.

How the bed is supposed to work

The heated bed is not driven directly by the mainboard high-current rails. Thick wires carry heater power; thin wires are the NTC thermistor (temperature sensor). On this machine, a separate hotbed adapter board sits between the PSU and the heater pad. The mainboard only provides a low-current CONTROL enable. The board MOSFET switches ~24 V on and off (PWM hold) — 24 V is not “25 °C.”

Thick wires — heater pad power path
Thin wires — NTC thermistor (resistance falls as temperature rises; ~100 kΩ at ~25 °C)
CONTROL plug (red / blue) — enable from mainboard into the adapter board
MOSFET — low-side (or similar) switch that actually connects the pad

Bed power voltage guide

State	Across bed + / −
Not heating	~0 V
Heating hard	~24 V
Holding temperature	May pulse / lower average (PWM)

On low-side switch designs one wire can still read ~24 V to ground when idle. The useful reading is the difference across the two thick heater wires. Black is not always chassis 0 V.

Quick fault pattern table

Pattern	Likely cause
0 V across bed while commanded to heat	MOSFET / adapter board / enable path
Voltage OK, pad never warms	Heater pad open or wrong pins
Voltage dies mid-heat	Board overheat, protection, bad crimp
Nonsense temperature on screen	NTC / thin sensor wiring

Board identification

Photos of the electronics bay show two distinct boards: the large Anycubic mainboard and a smaller adapter labeled NF013_Heating-Platform (silkscreen also shows version marks such as V0.0.2 / related revisions). That smaller board is the bed power switch, not the mainboard.

Hotbed adapter board mounted next to the Anycubic mainboard — Adapter on the left (HOTBED / +DC IN / CONTROL), mainboard on the right.

Anycubic mainboard in the electronics bay — Mainboard view — stepper drivers, fan headers, and green power terminal block.

Removed NF013 heating platform board showing HOTBED, DC IN, CONTROL, and 24V OUT connectors — Removed board face: green screw terminals for HOTBED and +DC IN, white CONTROL and +24V OUT headers, toroidal inductor, and MOSFET area. Silkscreen: `NF013_Heating-Platform_V0.0.2`.

Backside of the hotbed adapter board with solder joints and QR label — Backside after removal — mostly through-hole pads and a QR label (`YS220518NF01300101702`). No dense SMD field on this face.

Green DC IN terminal block with black and red power wires — +DC IN terminal block feeding the adapter from the supply rail.

Anycubic publishes a hotbed adapter replacement guide for the related Kobra 2 Max family (same board concept — confirm part number before ordering for the original Kobra Max):

Key components

WSK220N04 power MOSFET

The large package on the adapter is marked WSK220N04 / lot LC05K. This is the high-current switch that connects bed power when the gate is driven hard enough.

Close-up of WSK220N04 MOSFET on the adapter board — WSK220N04 power MOSFET close-up.

MOSFET area with CONTROL connector red and blue wires — MOSFET neighborhood with CONTROL plug (red / blue) seated.

Board area near R8 with red pen marking possible charring — Marked region near R8 / MOSFET — visible discoloration, a suspect for the abnormal-heater fault.

EL817 optocoupler (U1)

U1 is the isolated CONTROL front-end: an EL817-class optocoupler. The mainboard drives the LED side; the phototransistor side (collector on pin 4) participates in turning the MOSFET on. Light carries the binary on/off decision across isolation.

Annotated photo highlighting U1 optocoupler near CONTROL connector — U1 region highlighted next to the CONTROL header.

EL817 optocoupler schematic and pin configuration — EL817 pinout: 1 anode, 2 cathode, 3 emitter, 4 collector.

Debug log

2026-07-26 — setup and first pass

Identified printer from label: Kobra Max, serial KM2201A0602894.
Logged the hotbed heater abnormal fault and walked heater pad / NTC / MOSFET possibilities.
Installed UltiMaker Cura 5.13.0 with the Anycubic Kobra Max profile for later test prints (PLA first).
Bed voltage guide written: idle ~0 V across +/−; heat ~24 V.

2026-08-01 — board ID and FET forced ON

Confirmed the small board is the hotbed adapter (bed power switch), not the mainboard. Thick power path: PSU → adapter → heater pad; mainboard only drives CONTROL.

Field measurements with the unit on, probing the heater negative terminal relative to ground, showed about 22.4 V present — supply side alive. Manually applying ~10 V on the FET gate (V_GS well above a few volts) turned the channel on: the green indicator lit, drain-to-source conducted, and the pad began to take current from the negative 24 V path. Drain node dropped toward roughly 4.3 V under that drive (stronger V_GS would likely pull it lower).

Adapter board with red and green LEDs lit during FET forced-on test — Forced gate drive: green control/status LED on while the MOSFET conducts.

WANPTEK programmable supply set near 24.3 V — Bench supply used around the 24 V class rail during adapter experiments.

V_GS field check (gate to source / GND)

Measured gate (to source / GND)	Expect
~0 V	Bed OFF — no pass
~2–6 V while “heating”	Drive problem / partial ON — danger zone
~8–12 V while heating	Good ON — pass

Takeaway: the FET itself can switch. The open question is where the normal ~10 V-class gate drive is generated (or lost) on the CONTROL / U1 path under mainboard command.

2026-08-10 — board removal

Board pulled to inspect the backside and continue hunting the gate-drive origin. Still open: exact generation path of the ~10 V that successfully forced the FET on during the August 1 test.

Datasheets

Next steps

Trace CONTROL from the mainboard through U1 to the MOSFET gate under a commanded heat cycle.
Measure V_GS while the UI requests 60 °C — compare against the forced-on values.
Check for open resistors, burnt pads, or failed optocoupler around the R8 / U1 / CONTROL island.
If the control chain is dead, replace the hotbed PCBA; if the pad is open, swap the heater element.
After repair: PLA benchy calibration print (Cura Kobra Max profile already installed).

Photo gallery

Select an image to view it full size.