Board Replace vs Component Replace Decision Matrix
Why this matters
Control boards are the most expensive consumable in modern appliances and the most over-prescribed part in the trade. The instinct to "throw the board at it" when a diagnosis is unclear creates expensive returns, dissatisfied customers, and warranty-claim problems with parts distributors who track high return rates per tech. The matrix below codifies the field-tested criteria for condemning a board versus a discrete component. The default assumption should always be that the board is the last suspect, not the first, unless evidence specifically points there.
The four-question gate before any board condemnation
Before ordering a board on any appliance, the tech must answer all four:
- Is there visible board damage? Burned trace, blown SMD component, swollen capacitor, scorching around a triac or relay. Yes = board confirmed. No = continue.
- Has every input the board needs been verified at the board connector? Thermistors ohmed, switches confirmed, supply voltage present at the input pins. If any input is unverified, the board is not condemnable yet.
- Has every output the board issues been verified at the board terminals? Relay commands, triac gate signals, motor drive outputs measured against the OEM service-mode commanded state. If commanded outputs are present and correct, the load is the suspect, not the board.
- Has the OEM diagnostic mode been entered and the fault codes read? Most platforms log fault history accessible via a key-sequence at the keypad. Codes that point at sensor or load faults are definitive even when symptoms look "board-like."
If any of the four answers is "no" or "not yet," the board is not the suspect. Run the diagnostic to completion.
Condemn criteria matrix by appliance class
Refrigerator main board
| Symptom | Board is the suspect when... | Component is the suspect when... |
|---|---|---|
| Compartment overcooling | Thermistor inputs read correct, damper drive output stuck | Thermistor drift, stuck damper, fan stuck on |
| Defrost failure | Heater confirmed open commanded, board not switching the relay | Open heater coil, open bimetal, board logic verified |
| Compressor short-cycle | Inverter board confirmed sending correct command, compressor draws excessive current | Failed compressor, low charge, dirty condenser |
| No display | All supply rails confirmed bad at SMPS output | Bad SMPS, loose UI ribbon, dead battery on memory-backup units |
Range / oven main board
| Symptom | Board is the suspect when... | Component is the suspect when... |
|---|---|---|
| Will not heat (bake) | Thermistor input good, element ohms good, board commanded "heat on" but no voltage out | Open element, drifted thermistor, open thermal fuse |
| Temperature wildly off setpoint | Thermistor verified accurate, board commands wrong duty cycle | Drifted thermistor (most common), door seal blown |
| Self-clean lock failure | Lock motor drive output absent | Open lock motor, broken linkage |
| Display dead | SMPS rails all dead | Ribbon cable unseated, dead memory battery |
Washer main board (front-load and top-load)
| Symptom | Board is the suspect when... | Component is the suspect when... |
|---|---|---|
| Will not start | All inputs (door switch, water level, lid switch) confirmed correct at board | Door switch open, lid lock failed, pressure transducer reading wrong |
| Will not drain | Pump drive output absent during commanded drain | Stuck pump, clogged pump, open pump winding |
| Will not spin | Motor drive output absent or inverter board failed | Failed motor, failed inverter, open hall-effect sensor on rotor |
| Will not fill | Valve drive output absent during commanded fill | Open valve solenoid, closed shutoff, debris in valve screen |
Direct-drive washers (LG, Samsung, many late-model Whirlpool) combine the inverter and the motor control on one PCBA. The inverter section is by far the highest-failure region and the most field-verifiable - check the IGBTs and the bus capacitors for visible damage before condemning the whole board.
Dryer main board
| Symptom | Board is the suspect when... | Component is the suspect when... |
|---|---|---|
| No heat | Element/igniter and thermal fuse confirmed good, board commanded "heat on" but no voltage out | Open element, open igniter, open thermal fuse, drifted thermistor |
| Will not start | Door switch and start switch confirmed correct at board input | Door switch open, start switch open, broken belt (motor will hum) |
| Auto-dry not finishing | Thermistor verified accurate, board commanded wrong cycle length | Drifted thermistor, moisture sensor strip contaminated |
Dryer boards are the lowest-failure boards in the trade by category. Most "control board" calls on dryers are actually thermal fuses or moisture sensors.
Dishwasher main board
| Symptom | Board is the suspect when... | Component is the suspect when... |
|---|---|---|
| Will not start | All door switches and inputs confirmed | Failed door latch, failed touchpad/keypad |
| Drains continuously | Drain pump drive output stuck on | Failed drain pump (rare for this failure mode) |
| Will not heat dry | Element confirmed good, board commanded heat but no voltage out | Open element, open thermal fuse, drifted thermistor |
| No display | All SMPS rails dead | Failed SMPS, water damage to UI board (common on under-counter dishwashers with valve leaks) |
Water damage to dishwasher control boards is real and visible. Open the board, look for white mineral residue or corrosion on the traces - confirmed water exposure justifies board replacement plus the upstream leak fix.
Microwave main board
| Symptom | Board is the suspect when... | Component is the suspect when... |
|---|---|---|
| Will not run | Door interlocks and triac drive verified | Failed door switch (most common), failed triac on board |
| Runs but no heat | HV components all confirmed good (diode, capacitor, magnetron, transformer) | Failed HV diode (most common), failed magnetron |
| Touchpad ghost-touches | Touchpad ribbon confirmed seated and intact | Failed touchpad/keypad (more common than board) |
| Display dark | SMPS rails verified bad | Failed SMPS section of board (board fix), failed display module (display fix) |
The "board is bad" smell test
Boards usually fail in one of three observable ways:
- Dead supply. Line voltage present at the board input, no low-voltage rails out. Nothing on the display, no relay clicks, no response to any input. This is the easiest one to condemn because you can measure it: line in, no 5 V or 12 V out at the rail test points, and the fuse on the board is intact (a blown onboard fuse points downstream, not at the board).
- Dead output stage. Logic is alive and commanding the load, but the relay or triac never switches, or switches and never releases. Confirm by putting the board in test/service mode, commanding the output, and metering for voltage at the board terminal, not at the component. Voltage commanded and none present at the terminal condemns the board. Voltage present and the component does nothing condemns the component.
- Intermittent logic fault. Erratic behavior that tracks with heat, humidity, or vibration: cycles that abort at random points, ghost inputs, a display that drops out and recovers. Usually a cold solder joint, a cracked trace, or corrosion under a connector. Pull the board and inspect under a light before you order - a visible burn mark, a bulged electrolytic cap, or green corrosion around a connector is a confirmed diagnosis; a clean board with an intermittent complaint is not.
Anything that does not fit one of those three deserves another round of component testing before you spend the customer's money on a board.
References
- ANSI/AHAM HRF-1 (refrigerator industry-standard test procedures)
- IEC 60335-1 (general safety requirements for household appliances - electronic control evaluation)
- Whirlpool Service Pointer W11052311 (control board diagnostic procedure and condemnation criteria)
- LG Service Manual general procedures (board-test versus component-test order of operations)
- Samsung Service Bulletin (Twin Cooling Plus diagnostic mode and board fault history retrieval)