What to Check Before Ordering Parts

How to make sure the part you are about to buy is actually the part that failed

Ordering parts is easy.

Back to Biomed Basics

What This Page Explains

This page covers:

The Simple Version

Before ordering a part, try to answer:

  1. What function failed?
  2. What components are in that path?
  3. What have I ruled out?
  4. Does the failure follow this part?
  5. Is the replacement part actually compatible?
  6. What will need to be done after replacement?

That last question matters more than many people expect.

A board may require:

The repair does not end when the box arrives.

Start With the Failed Function

Do not start with the part number.

Start with:

What stopped working?

Example:

Monitor will not display SpO2.

Possible causes include:

If you immediately order the SpO2 board because the symptom involves SpO2, you may be skipping most of the failure path.

Symptom Is Not Component

A display problem does not automatically mean:

Display assembly.

A charging problem does not automatically mean:

Battery.

A communication error does not automatically mean:

Network card.

The symptom identifies the affected function.

You still have to isolate the failed part.

Follow the Chain

A useful troubleshooting model is:

Input

↓

Cable / Connector

↓

Power

↓

Component

↓

Feedback / Output

Ask which point in that chain stops behaving correctly.

Example: NIBP Will Not Inflate

Possible chain:

Command

↓

Pump driver

↓

Pump

↓

Tubing

↓

Cuff / Pressure system

Before ordering the pump, ask:

If the Pump Receives No Voltage

The pump itself may be perfectly good.

The problem may be:

If Pump Runs but Pressure Never Builds

Now consider:

Again, same symptom.

Different parts.

Known-Good Substitution

One of the strongest reasons to order a part is when the failure follows it.

Example:

Original battery:

Device shuts down.

Known-good battery:

Device runs normally.

Original battery on another compatible device:

Fails again.

Now you have strong evidence.

Failure Follows the Part

This is better than:

Battery is old, so probably battery.

Age can support a hypothesis.

Controlled substitution gives you stronger evidence.

Do Not Substitute Random Parts

Known-good means:

Known good.

Not:

This one was sitting on a shelf.

A questionable replacement can confuse the diagnosis.

Accessories First

Before ordering an expensive internal component, check the easy external pieces.

Examples include:

Example

Device reports:

Temperature probe not detected.

You order temperature input board.

Later you find:

Probe cable open near strain relief.

That was an avoidable parts order.

Inspect the Connector

Connector faults are common because connectors are exposed to:

Look for:

Host Connector vs Accessory Connector

Check both.

A damaged host port can make several good accessories appear bad.

Power Comes Before Logic

Many board replacements happen because a board is:

Not responding.

Before condemning it, verify:

Is the board receiving the power it needs?

Example

Parameter module will not initialize.

Expected 12 V supply:

0 V.

Ordering a replacement module will not fix missing supply power.

Check the Power Rail Under Load

A rail may look correct until the component tries to operate.

Example:

12 V no load.

Drops to 7 V when motor starts.

The load may expose:

Signal Checks

If the component is receiving correct power, check whether it receives the expected command.

Example:

Fan does not run.

Fan supply present.

PWM command absent.

The fan may not be the problem.

Feedback Matters

Some systems need confirmation from the part.

Example:

Motor runs.

Encoder does not report movement.

The device may report a motor fault even though the motor itself turns.

Read the Error Carefully

Do not translate:

Motor Test Failed

into:

Motor bad.

The test may include:

Error Codes Are Evidence, Not Verdicts

Use the error to locate the subsystem.

Then troubleshoot the subsystem.

Service Manual Parts Flow

OEM documentation may tell you to:

  1. Check connector
  2. Measure voltage
  3. Run diagnostic
  4. Replace board

That order exists for a reason.

Do Not Skip Straight to Step 4

Especially if steps 1–3 are easy.

Check Service History

Before ordering a part, ask:

Has this part already been replaced for the same complaint?

Example

Current complaint:

Random reboot.

History:

Main board replaced twice.

Same complaint returned.

Ordering a third board without new evidence is probably not a strong decision.

A Previous Part Replacement Does Not Prove It Was the Cause

It may have:

Look for Repeat Patterns

Repeated failure of the same component may indicate a deeper problem.

Example:

Power supply replaced every year.

Ask:

Part as Victim

The failed component may be the victim of another condition.

Example

Fuse repeatedly blows.

Do not just order more fuses.

Find the overcurrent cause.

Physical Damage

If the part is visibly:

that is useful evidence.

But still ask whether something else caused that damage.

Burned Connector

If one side of a connector is burned, inspect the mating side.

Replacing only half may leave the high-resistance connection intact.

Fluid Intrusion

If a board failed after fluid intrusion, inspect the surrounding area.

A new board installed into a contaminated connector may fail again.

Software and Configuration

Some hardware faults are actually configuration issues.

Example:

Network card replaced.

Still no connection.

Device IP configuration wrong.

The hardware was never the problem.

Firmware Compatibility

A replacement board may require:

Before ordering, verify compatibility.

Same Part Number Can Have Revisions

Do not assume:

Same-looking board = same board.

Check:

OEM Part Number

Use the actual manufacturer part number when available.

Do not identify parts only by appearance.

Option Codes

Some boards may support multiple device configurations.

Replacement may require enabling:

Serial Number Programming

Some systems store serial identity on a board.

Replacement may require reprogramming.

Calibration Data

Calibration constants may be stored on:

Replacing the part can erase the old calibration.

Ask Before Ordering

What setup will be required after I install this?

That can affect whether local replacement is even practical.

Board-Level vs Module-Level Repair

Some shops repair at component level.

Others replace modules.

Follow:

Do Not Turn Every Board Into a Soldering Project

If board-level repair is not supported or cannot be validated safely, module replacement may be the correct repair.

Cost Matters, But Evidence Comes First

A cheap part is easier to justify probabilistically.

A $25 cable may be worth substituting early.

A $6,000 main board should usually have stronger evidence behind it.

Probability Has a Place

Sometimes perfect proof is impossible.

Intermittent faults may require a judgment call.

The question becomes:

Is this part sufficiently likely to be the cause that replacement is reasonable?

Example

Intermittent reboot.

Logs consistently show internal power-supply fault.

Voltage rail drops during failure.

Power supply is a reasonable replacement even if it does not fail every time.

Document Your Reasoning

A good note might say:

Unit intermittently rebooted during NIBP operation. 5 V rail dropped below specification during pump startup. Input voltage remained stable. Recommending replacement of power-supply module.

That is much stronger than:

Ordering power supply.

Ordering Multiple Parts

Sometimes OEM troubleshooting requires replacing assemblies together.

That can be valid.

But avoid:

Shotgunning parts

simply because you are unsure.

Why Shotgunning Is Bad

If you replace:

all at once and the problem disappears, you may never know what actually fixed it.

That makes the next failure harder.

Change One Thing at a Time When Practical

This preserves diagnostic value.

Emergency Repairs

Sometimes downtime matters enough that you cannot wait for perfect isolation.

You may order probable parts in parallel.

That is a business/operational decision.

Still document what is known and unknown.

Verify Parts Availability Before Disassembly

Some parts may be:

Know the repair path before taking apart critical equipment unnecessarily.

Loaners and Alternatives

If the required part has long lead time, coordinate:

instead of rushing an uncertain repair.

Real-World Example: Dead Touchscreen

Complaint:

Touch does not respond.

Image is normal.

Hard keys work.

Touch diagnostics show no input.

Touch controller power present.

Known-good display/touch assembly restores touch.

Now ordering display assembly is defensible.

Real-World Example: Battery Not Charging

Battery will not charge.

Known-good battery also will not charge.

Charge voltage absent at battery connector.

Ordering battery is unlikely to help.

Real-World Example: NIBP Pump

Pump runs.

Pressure does not rise.

External cuff and hose known good.

Internal leak found at valve manifold.

Pump does not need replacement.

Real-World Example: Network Board

No network link.

Known-good cable and switch port.

Correct power present.

NIC link LEDs remain dark.

Same board fails in second compatible unit.

Now board replacement is well supported.

Common Mistakes

Ordering the Part Named in the Error

Error name may describe the failed test, not component.

Ignoring Accessories

Check cheap, exposed items first.

Ignoring Power

No board works without correct supply.

Replacing Parts That Were Already Replaced for the Same Complaint

Review history.

Ordering Before Checking Compatibility

Revision matters.

Forgetting Post-Replacement Configuration

Install is not the end of repair.

Replacing Several Parts at Once

You lose diagnostic information.

A Useful Pre-Parts Checklist

Before ordering, ask:

What exactly failed?

What is in the failure path?

What have I ruled out?

Does the failure follow this component?

Does the part have proper power and commands?

Is there any upstream cause?

Has this part been replaced before?

Is the replacement revision correct?

What calibration or configuration comes afterward?

What Did You Actually Prove?

If you replace a board and the device works:

You proved:

The device functions with the replacement board installed.

You did not automatically prove:

The original board was the root cause.

Your evidence is stronger if:

Final Thoughts for Biomeds

Parts replacement is part of biomed work.

Parts guessing should not be.

You do not need courtroom-level proof before every repair.

But you should know why you are ordering what you are ordering.

Start with the failed function.

Work the chain.

Rule out the easy external causes.

Check power and communication.

Use known-good substitution where appropriate.

Review the history.

Then order the part that the evidence points toward.

The goal is not to make parts ordering difficult.

The goal is to stop spending time and money replacing components that were never actually broken.

Before you click:

Submit Order

ask one last question:

What did I actually prove?

— Jake

Important Note

Parts identification, hardware revisions, authorized repair level, calibration requirements, firmware compatibility, and post-replacement configuration vary by medical-device manufacturer and model. Follow current service documentation and facility purchasing and repair procedures before installing replacement components.

Related Biomed Basics

Back to Biomed Basics