What This Page Explains
This page covers:
- What a self-test actually does
- Startup self-tests
- Power-on self-test
- Sensor tests
- Actuator tests
- Memory checks
- Communication checks
- Calibration checks
- Why self-test names can be misleading
- How to use error codes without over-trusting them
- How to isolate the failed chain
- Common troubleshooting mistakes
- How to verify a repair after the self-test passes
The Simple Version
A self-test usually commands an action, measures the response through a sensor or feedback circuit, and compares that result with an expected range. A “valve test failed” message therefore identifies the check that did not pass; it does not automatically identify the valve as the defective part.
The same failure can come from the valve, driver circuit, supply voltage, blocked tubing, leak, pressure sensor, feedback wiring, configuration, or an invalid setup condition. Record the exact code and test stage, confirm prerequisites and accessories, review logs and service documentation, and isolate the command path from the feedback path. Do not repeatedly bypass or rerun a safety-critical test merely to clear the message.
What Is a Self-Test?
A self-test is an internal diagnostic routine designed to evaluate whether parts of the device appear to be functioning correctly.
Depending on the device, a self-test may check:
- Processor
- Memory
- Power rails
- Sensors
- Motors
- Valves
- Fans
- Batteries
- Communications
- Internal calibration values
Power-On Self-Test
Many devices perform a:
Power-On Self-Test, often abbreviated POST.
This may happen before the normal application starts.
A basic POST can check whether the system can:
- Execute code
- Access memory
- Communicate with required hardware
A Passed POST Does Not Prove Full Device Function
It only proves the tested startup conditions passed.
The device may still have a failure that appears:
- Under load
- During therapy
- After warm-up
- With a specific accessory
Memory Tests
A device may check:
- RAM
- Flash storage
- Configuration memory
If the test fails, possible causes include:
- Memory device failure
- Corrupted data
- Power instability
- Main-board issue
Checksum
Software or calibration data may be protected by a:
Checksum or similar integrity check.
The device calculates a value from stored data and compares it with the expected value.
If they do not match, the data may be:
- Corrupted
- Incomplete
- Incorrect version
Checksum Failure Is Not Always Bad Hardware
It may also follow:
- Interrupted software update
- Incorrect firmware
- Configuration corruption
Power-Rail Tests
Some devices monitor their own internal supply voltages.
For example, software may check whether a rail is within an expected window.
If a 5 V rail measures:
4.2 V
the device may stop startup.
What Could Cause the Low Rail?
Possible causes include:
- Power supply
- Excessive load
- Shorted peripheral
- Bad connector
The self-test may only report:
5 V supply error.
You still need to isolate the cause.
Sensor Tests
A device may check whether a sensor produces a plausible reading.
Example
Pressure sensor at startup should read near atmospheric reference.
If it reports an impossible value, the device may fail self-test.
Possible causes:
- Sensor
- Blocked port
- Damaged tubing
- ADC
- Reference voltage
Plausibility Tests
Many self-tests do not directly prove accuracy.
They ask whether the value is:
Reasonable enough to continue.
A pressure sensor reading:
5 mmHg off
might still pass startup while failing formal calibration later.
Actuator Tests
Devices may command:
- Motor
- Pump
- Valve
- Solenoid
and then look for evidence that it moved or changed the system.
Example: Valve Test
Device commands valve open.
Pressure should fall.
Pressure does not change.
Possible causes:
- Valve stuck
- Driver failed
- Pressure sensor wrong
- Tubing blocked
Again, the failed test covers a chain.
Motor Test
A pump may command the motor and expect:
- Encoder pulses
- Position change
- Current response
If feedback never appears, possibilities include:
- Motor
- Encoder
- Driver
- Mechanical jam
- Wiring
Fan Test
A fan may be commanded to run while the device monitors tachometer feedback.
If RPM remains zero, the device might report:
Fan failure.
Potential causes include:
- Fan motor
- Tach wire
- Connector
- Driver
Communication Tests
Modules often communicate over internal buses.
A host may expect each module to answer during startup.
If one does not respond, the device may report:
- Module missing
- Communication fault
- Initialization failure
Module Error Does Not Always Mean Module Failure
The module may not answer because:
- No power
- Loose connector
- Bus problem
- Host communication issue
Internal Network
Complex medical systems may contain several computers or controllers connected internally.
A startup fault can sometimes be an internal networking problem.
Battery Self-Test
A device may evaluate:
- Battery presence
- Voltage
- Temperature
- Communication
It may not perform a full capacity test during startup.
Important Distinction
Battery passes self-test.
That does not necessarily prove:
Battery has adequate runtime.
It may only prove the battery is:
- Present
- Communicating
- Within a plausible voltage range
Calibration Checks
Some systems perform startup calibration or zeroing.
Examples include:
- Pressure zero
- Flow zero
- Optical baseline
If the baseline is outside an expected range, initialization may fail.
Example
Flow sensor calibration fails.
Possible causes:
- Contaminated sensor
- Actual flow during zeroing
- Leak
- Sensor drift
Environment Can Cause Self-Test Failure
Not every failed self-test means hardware broke.
Conditions such as:
- Moisture
- Incorrect accessory setup
- Blocked inlet
- Wrong gas supply
can cause a valid failure.
Example
Ventilator flow-sensor zeroing fails because circuit flow is still present.
The device may be functioning correctly by refusing calibration.
Self-Test Can Protect the User
A failed startup test may intentionally prevent therapy.
Do not bypass a self-test simply to get the device running.
Error Code Is the Starting Point
When a code appears, record it exactly.
Not:
Some startup error.
Instead:
Error 1234: Flow Sensor Zero Failed.
That helps with:
- Service manual lookup
- Technical support
- Trend analysis
Capture the Conditions
Record:
- Time
- Accessories connected
- AC or battery
- Recent repair
- Temperature
Self-test failures may be condition-dependent.
Repeated vs One-Time Failure
A single failed self-test followed by normal operation may be transient.
Repeated failure at the same stage is stronger evidence.
Do Not Just Reboot Until It Passes
Repeatedly power cycling can erase useful clues.
It can also create false confidence.
If a self-test fails once and passes the second time, ask:
Why was it marginal?
Failure at the Same Point
If startup consistently stops at:
42%
or:
Loading respiratory module
that can identify the subsystem being initialized.
Service Manual Flowcharts
OEM service documentation often maps:
- Error code
- Failed test
- Recommended isolation
Use it.
But Think Beyond the Flowchart
A flowchart may say:
Replace module.
Before doing that, verify:
- Module power
- Connector
- Shared bus
when allowed.
Test the Inputs and Outputs
For a failed self-test, ask:
What input does the test depend on?
and:
What output should the device generate?
Example: Pump Motor Test
Input:
Motor command.
Expected output:
Encoder pulses.
If no pulses:
Measure whether:
- Motor receives voltage
- Motor turns
- Encoder signal exists
This breaks the self-test into smaller pieces.
Raw Diagnostic Values
Service mode may show:
- Pressure counts
- Flow counts
- Fan RPM
- Battery voltage
These are useful because they show what the device itself is seeing.
Example
Self-test says:
Pressure sensor failure.
Service mode shows:
Raw sensor value pinned at maximum.
That suggests a specific electrical or sensor fault rather than a random software problem.
Compare With a Known-Good Device
If available, compare:
- Raw values
- Startup sequence
- Sounds
- Timing
Known-good comparison can reveal what is abnormal.
Example
Good ventilator:
Valve clicks at startup.
Bad ventilator:
No click.
Now investigate valve control path.
Listen to the Device
Self-tests often command hardware in a predictable order.
You may hear:
- Fan
- Pump
- Valve
- Relay
Missing sounds can provide clues.
Do Not Rely on Sound Alone
Use it as evidence, not proof.
Sensor Feedback Loops
Many self-tests evaluate a loop.
Example:
Controller → Heater → Temperature Sensor → Controller
If temperature does not rise when heater turns on, the failure could involve:
- Heater
- Driver
- Sensor
Think in Loops, Not Parts
This is one of the strongest ways to interpret diagnostic failures.
Failed Self-Test After Repair
If the device passed before a repair and fails immediately after:
Check what changed.
Possible issues include:
- Connector left loose
- Wrong part
- Calibration missing
- Configuration lost
Board Replacement
After a board replacement, startup failures may come from:
- Firmware mismatch
- Missing option codes
- Serial-number programming
- Calibration data
Do not automatically assume the replacement board is defective.
Software Update
Self-tests may change after firmware updates.
New software may:
- Add checks
- Tighten limits
- Change calibration
If a Fleet Fails After the Same Update
That pattern matters.
Environmental Self-Test Failure
Some devices require:
- Correct temperature
- Gas pressure
- Fluid level
before startup completes.
A device can correctly fail when the environment is wrong.
Example
Anesthesia machine fails leak test.
Possible causes:
- Circuit not assembled
- Absorber not seated
- Leak in breathing system
- Internal valve
The machine may be functioning exactly as designed by identifying the leak.
Self-Test vs Full Performance Test
A self-test may tell you:
Internal checks passed.
It does not replace independent performance verification.
Example
Defibrillator passes startup.
That does not independently prove:
200 J output is accurate.
Use a defibrillator analyzer when required.
Example
Pressure sensor has a consistent offset but remains within startup plausibility range.
Self-test passes.
Independent pressure analyzer reveals calibration error.
Independent Test Equipment Matters
Internal diagnostics and external verification complement each other.
Self-Test Failure That Clears After Cleaning
Example:
Optical sensor path contaminated.
Self-test fails.
Approved cleaning restores expected signal.
Now you have evidence that contamination was part of the failure.
Self-Test Failure That Clears After Reseating Connector
That can indicate:
- Loose connector
- Oxidized contact
But ensure the repair is stable.
Do not simply reseat and declare victory without trying to reproduce the original condition.
Real-World Example: Fan Self-Test
Device fails startup:
Fan error.
Fan spins.
Service mode shows:
0 RPM.
The fan motor is not necessarily the problem.
Tach feedback line may be open.
Real-World Example: Pressure Test
Ventilator fails pressure-sensor startup check.
Known-good pressure sensor also fails.
Reference voltage to sensor is missing.
Problem is host electronics, not sensor.
Real-World Example: Module Communication
Monitor reports SpO2 module missing.
Module receives no power because connector pin is damaged.
Replacing module would not fix it.
Real-World Example: Battery Test
Device says:
Battery passed.
Clinical complaint:
Runtime only 10 minutes.
Startup test confirms battery presence, but capacity test shows severe degradation.
The self-test did not evaluate the same requirement.
Common Mistakes
Replacing the Part Named in the Error
Understand the full test chain first.
Rebooting Until the Error Goes Away
Intermittent pass is not proof of repair.
Assuming Self-Test Pass Means Entire Device Is Good
Independent verification may still be required.
Ignoring Accessories
External setup can cause valid test failures.
Ignoring Configuration After Board Replacement
Hardware may be fine.
Clearing Logs Too Early
Preserve evidence first.
Treating Every Error as Hardware
Software and environment can trigger failures too.
A Useful Self-Test Framework
When a self-test fails, ask:
What did the device command?
Then:
What response did it expect?
Then:
What sensor or feedback determines pass/fail?
Then:
What components sit between the command and the feedback?
Then:
Can I independently verify any part of that chain?
This turns:
Self-test failed
into a specific troubleshooting path.
Another Useful Question
Ask:
Is the reported error naming the failed component, or merely the test that could not complete?
That question can save expensive parts.
What Did You Actually Prove?
If a self-test passes after you replace a valve:
You proved:
The self-test now sees the expected response with the replacement valve installed.
If you also test the original valve and confirm it is stuck:
You have stronger evidence.
If the device passes every internal diagnostic:
You proved:
The built-in checks passed under current conditions.
You did not prove every performance specification with an independent reference.
Final Thoughts for Biomeds
Self-tests are valuable because they tell you what the device itself sees.
But they should not do your thinking for you.
The message:
Pressure Sensor Test Failed
does not necessarily mean:
Order a pressure sensor.
Instead think:
Command → Hardware → Physical Effect → Sensor → Feedback → Software Decision
Find where that chain breaks.
That is the difference between using diagnostics and merely reading error messages.
And after the error clears, ask the same question you should ask after any repair:
What did you actually prove?
— Jake
Important Note
Self-test sequences, diagnostic modes, error codes, service limits, calibration requirements, and permitted troubleshooting procedures vary by manufacturer and model. Follow current OEM service documentation and complete all required independent functional, calibration, and safety verification before returning equipment to service.
