How to Read Device Specifications

How to turn manufacturer numbers into useful troubleshooting and verification limits

Medical device specifications are everywhere.

Published August 16, 2026 · Revised September 5, 2026

Back to Biomed Basics

What This Page Explains

This page covers:

The Simple Version

A specification is not just a number beside a plus-or-minus sign. It is a complete statement about a particular function, at a particular test point, under stated conditions. Before using one, identify what function is being tested, what units apply, what setup the manufacturer requires, and which limit actually determines pass or fail. If any one of those pieces is missing, the comparison may look precise while answering the wrong question.

For example, “flow accuracy ±5%” is incomplete by itself. Is that 5% of the reading or of full scale? Does it apply across the entire flow range? Is the value referenced to a particular gas, temperature, pressure, tubing set, or warm-up period? The footnote beside the specification can matter as much as the bold number.

A Practical Reading Workflow

Start with the exact model and, when the manual distinguishes them, the hardware or software revision. Find the performance specification and then find the service procedure that tells you how the manufacturer expects it to be tested. Read the table heading, column heading, units, symbols, and footnotes before calculating a limit. Write the lower and upper limits on your worksheet before you look at the device result. That small habit makes it harder to unconsciously move the goalposts for a result that is close.

Next, confirm that the analyzer and accessories can support the test. Check the analyzer's range, accuracy, resolution, calibration status, and required zeroing. Match the specified load, tubing, sensor, adapter, gas, environmental condition, and warm-up time. If the setup does not reproduce the stated conditions, correct the setup before adjusting the medical device.

Finally, record enough detail that another biomed could repeat the comparison: the source of the specification, test point, calculated limits, analyzer used, actual result, and final disposition. “Passed output test” is weak documentation. “At a 200 J setting, measured 194 J; acceptable range 170–230 J per service manual section X using analyzer asset Y” shows what was actually verified.

What Is a Specification?

A specification describes how the manufacturer says the device should perform.

Examples:

NIBP pressure accuracy: ±3 mmHg.

Defibrillator energy accuracy: ±15%.

Battery runtime: minimum 60 minutes.

Specifications give you a reference.

Without one, a measurement is just a number.

Specifications Answer “What Should Happen?”

Troubleshooting often compares:

Expected

versus:

Actual.

Example:

Expected power supply output:

24 VDC ±5%.

Measured:

23.8 VDC.

Now you have a meaningful comparison.

Read the Entire Specification

Do not grab only the bold number.

A specification may say:

Accuracy ±2% between 20°C and 30°C using manufacturer-approved sensor after calibration.

Every part matters.

You need to know:

Conditions Can Change the Limit

A device may meet one accuracy specification in normal range and a different one near extremes.

Example:

Flow accuracy:

±3% from 1 to 20 L/min.

Different accuracy below:

1 L/min.

Do not automatically apply one limit everywhere.

Units Matter

Always verify units.

Examples:

A correct number in the wrong unit is still wrong.

Example

Pressure analyzer:

100 cmH2O.

Device specification:

100 mmHg.

Those are not the same pressure.

Make sure both values use the same unit before comparing them.

Range

Range describes the span over which a device or function is designed to operate or measure.

Example:

Temperature measurement range:

25°C to 45°C.

That means the device is intended to measure within that range.

It does not automatically tell you the accuracy.

Range Is Not Accuracy

Suppose:

Measurement range:

0 to 300 mmHg.

Accuracy:

±3 mmHg.

Those are completely different specifications.

Range tells you:

Where it can measure.

Accuracy tells you:

How close the result should be.

Accuracy

Accuracy describes how close a measured value should be to the actual or reference value.

Example:

Specification:

±2 mmHg.

Reference:

100 mmHg.

Acceptable device result may be:

98 to 102 mmHg.

That is a basic tolerance window around the reference.

Percent Accuracy

Some specifications use percentages.

Example:

±5%.

Reference:

200 J.

Five percent of 200 J is:

10 J.

So the acceptable range may be:

190 to 210 J.

Always confirm the manufacturer defines the specification that way.

Fixed Plus Percentage Specifications

Some specifications are more complicated.

Example:

±2% of reading + 1 digit.

Now you have more than one term.

Follow the exact formula in the manual.

Do not simplify it incorrectly.

Tolerance

Tolerance is the acceptable amount of variation.

Example:

Expected output:

10 VDC.

Tolerance:

±0.5 V.

Acceptable range:

9.5 to 10.5 V.

Tolerance defines the pass/fail boundaries.

Accuracy and Tolerance Are Related but Not Identical

Accuracy describes measurement performance.

Tolerance describes an allowed variation or limit.

In everyday troubleshooting, you may use both terms closely.

But when the manufacturer gives exact wording, use their definition.

Resolution

Resolution describes the smallest change the device can display or distinguish.

Example:

Temperature display:

37.1°C

37.2°C

37.3°C

Resolution may be:

0.1°C.

Resolution Is Not Accuracy

A device that displays:

37.123°C

is not automatically more accurate than one displaying:

37.1°C.

More decimal places do not guarantee a better measurement.

Example

Display resolution:

0.1 mmHg.

Accuracy specification:

±3 mmHg.

The device may display very fine increments while still having a wider allowed error.

Repeatability

Repeatability describes how consistently the device produces the same result under the same conditions.

Example:

Apply the same 100 mmHg pressure five times.

Results:

100.1 100.0 100.2 100.1 100.0

Very repeatable.

But if the actual reference is:

110 mmHg

then the device is consistently wrong.

Repeatability and accuracy are different.

Precision

You may also see the term:

Precision.

Precision generally describes how closely repeated measurements agree with each other.

A system can be:

Precise but inaccurate.

Accurate on average but poorly repeatable.

The exact terminology depends on the manufacturer and measurement field.

Operating Specifications

Not all specifications are measurement accuracy.

Manufacturers may define:

These can explain environmental failures.

Example: Operating Temperature

Specification:

10°C to 40°C.

Device repeatedly fails in a location at:

45°C.

Now environmental conditions matter.

You cannot necessarily judge the device under conditions outside its specified operating range.

Storage Specifications

Storage limits may be wider than operating limits.

Example:

Operating:

10°C to 40°C.

Storage:

-20°C to 60°C.

That does not mean the device can safely operate at -20°C.

Read the correct category.

Power Specifications

A device may specify:

Input:

100–240 VAC

50/60 Hz.

This tells you what AC source it is designed to accept.

It does not tell you the internal DC rails.

Current and Power Ratings

You may see:

2 A maximum.

or:

120 W maximum.

These describe electrical demand or rating.

Do not confuse a maximum rating with normal operating current.

Battery Runtime Specifications

Battery runtime is especially dependent on test conditions.

A manufacturer may specify:

120 minutes under typical operation with a new fully charged battery.

That does not necessarily mean every battery should always deliver exactly two hours under every workload.

Look at:

Minimum vs Typical

This distinction matters.

Specification:

Typical runtime: 4 hours.

That is different from:

Minimum runtime: 4 hours.

Typical values may not be hard pass/fail limits.

Typical Specifications

Words like:

usually deserve attention.

They may indicate expected behavior rather than guaranteed tolerance.

Do not create your own pass/fail criterion from a typical value unless your procedure allows it.

Minimum Specifications

Example:

Battery runtime:

Minimum 60 minutes.

Measured:

58 minutes.

That may be a clear fail if the procedure uses that minimum.

Maximum Specifications

Example:

Leak rate:

Maximum 100 mL/min.

Measured:

125 mL/min.

Fail.

The direction of the specification matters.

Nominal Values

Nominal means a named or approximate design value.

Example:

12 V battery.

Actual fully charged voltage may be higher than:

12.0 V.

Do not use nominal values as exact pass/fail limits unless instructed.

Test Conditions Matter

A specification may depend on:

Read the notes around the table.

Example: Ventilator Accuracy

Tidal volume specification may apply only under:

Using the wrong setup can create a false failure.

Test Point Matters

A specification may apply at one specific point.

Example:

Power supply output:

24 VDC ±5% at connector J3.

You measure:

22 V at a distant board.

That does not automatically prove the supply violates specification.

There may be voltage drop between those points.

Use the specified test location.

Input vs Output Specifications

Do not mix them.

Example:

Infusion pump programmed:

100 mL/hr.

Analyzer measures:

97 mL/hr.

You need the delivery accuracy specification.

Not the display resolution specification.

Alarm Specifications

Alarm behavior may have specifications for:

If testing an alarm, use the relevant alarm requirement.

Do not simply decide:

It sounded close enough.

Alarm Delay Example

Manufacturer states:

Alarm activates within 10 seconds after threshold condition.

Measured:

6 seconds.

Pass.

Measured:

22 seconds.

May fail.

You need the expected timing before making the judgment.

Mechanical Specifications

Mechanical equipment may specify:

Example:

Table longitudinal travel:

230 mm ±10 mm.

Use the actual specification for that movement.

Load Specifications

A device may behave differently at different loads.

Example:

Power supply:

24 VDC ±5% at rated load.

If you measure it completely unloaded, you may not be testing the same condition.

Manufacturer Test Procedures Matter

A specification and a test procedure often belong together.

The specification tells you:

What limit applies.

The procedure tells you:

How to measure it.

Do not separate them casually.

Do Not Invent Your Own Test Setup

If the manufacturer says:

then changing those conditions may invalidate the comparison.

Test Equipment Accuracy Matters

Suppose device tolerance is:

±1%.

Your analyzer uncertainty is:

±5%.

That analyzer may not be appropriate for proving the device meets specification.

Your reference needs to be capable of supporting the measurement.

Specification vs Test Equipment Resolution

Example:

Required measurement:

10.00 V ±0.01 V.

Meter only displays:

10.0 V.

You may not have enough resolution to determine pass/fail reliably.

Compare Like With Like

Before comparing values, confirm:

Then apply the limit.

Example: Simple Fixed Tolerance

Specification:

100 ±5 mmHg.

Acceptable range:

95 to 105 mmHg.

Measured:

103 mmHg.

Pass.

Example: Percentage Tolerance

Specification:

±10%.

Set value:

50 mL.

Ten percent:

5 mL.

Acceptable:

45 to 55 mL.

Measured:

43 mL.

Fail.

Example: Maximum Limit

Leak:

Maximum 50 mL/min.

Measured:

48 mL/min.

Pass.

Measured:

51 mL/min.

Fail.

No plus/minus calculation needed.

Borderline Results

Suppose limit is:

95 to 105.

Measurement:

105.0.

If the specification says inclusive, that may pass.

Measurement:

105.1.

That may fail.

Do not move the limit because the result is close.

“Close Enough” Is Not a Specification

Avoid:

It's only a little outside.

If the required limit is:

±3 mmHg

and the result is:

+5 mmHg,

it is outside the specification.

Then follow the appropriate service procedure.

Do Not Round Too Early

Suppose actual measurement:

105.46.

Limit:

105.4.

Rounding to:

105

could incorrectly make it pass.

Keep enough precision until the comparison is complete.

But Do Not Report Fake Precision

If your analyzer only measures to:

0.1

do not report:

100.037.

Use the precision supported by the test equipment.

Specification Tables

Manufacturer tables may include several columns.

Example:

| Parameter | Range | Accuracy | Resolution | |---|---|---|---| | Pressure | 0–300 mmHg | ±3 mmHg | 1 mmHg |

Read across the correct row.

Do not confuse one column for another.

Footnotes Matter

The table may say:

*Accuracy applies above 50 mmHg.

Below that, another specification may apply.

Always read footnotes.

Software Version Can Affect Specifications

Some performance characteristics or test methods may change with:

Use documentation applicable to the device you are servicing.

Accessories Can Affect Specifications

A measurement specification may require:

An incompatible accessory can make the device appear out of specification.

Example: NIBP

Monitor fails accuracy test.

Before replacing NIBP board, verify:

A leak in the test setup can create a false failure.

Example: Defibrillator Energy

Set:

200 J.

Measured:

196 J.

If specification is:

±15%,

the result is comfortably within tolerance.

Do not adjust equipment because the analyzer did not read exactly:

200.0 J.

Example: Temperature

Reference:

37.0°C.

Device:

37.3°C.

Specification:

±0.2°C.

Fail.

Display resolution:

0.1°C.

The device can display fine increments and still be outside accuracy requirements.

Example: Battery Runtime

Requirement:

Minimum 60 minutes.

Actual:

63 minutes.

Pass under that test.

A result of:

59 minutes

is not:

basically an hour.

If 60 is the defined minimum, 59 is below it.

Example: Ventilator Volume

Set:

500 mL.

Measured:

485 mL.

If allowed error is:

±10%,

acceptable range:

450 to 550 mL.

Pass.

Knowing the specification prevents unnecessary adjustment.

Specifications Help Prevent Over-Repair

Without limits, a technician may chase tiny differences.

Example:

Set:

100.

Measured:

99.

If tolerance is:

±5,

the device is already performing correctly.

Do not calibrate equipment simply because the values are not identical.

Specifications Also Prevent Under-Repair

The opposite problem:

It looks close enough.

If it is outside the required limit, it fails.

Specifications keep judgment objective.

Troubleshooting With Specifications

Suppose a power rail should be:

5 VDC ±5%.

Acceptable:

4.75 to 5.25 V.

Measured:

4.1 V.

That gives you strong evidence of a power problem.

Verification With Specifications

After repair:

Measured:

5.02 V.

Within range.

Now you have objective evidence supporting the repair.

Specifications Are Not Diagnoses

A failed specification tells you:

Performance is outside the acceptable limit.

It does not automatically tell you why.

Example:

Pressure accuracy fails.

Possible causes:

The failed test becomes the symptom you troubleshoot.

Do Not Adjust First

If something fails specification, do not immediately calibrate.

First ask:

Why is it out of specification?

If a hose is leaking, calibration is not the right repair.

Real-World Example: Power Supply

Specification:

24 VDC ±5%.

Measured:

22.1 V.

Acceptable minimum:

22.8 V.

Fail.

Now investigate power supply or excessive load.

Real-World Example: Infusion Pump

Set:

100 mL/hr.

Analyzer:

96 mL/hr.

Manufacturer tolerance:

±5%.

Acceptable:

95 to 105 mL/hr.

Pass.

No adjustment needed.

Real-World Example: Pressure

Reference:

200 mmHg.

Device:

194 mmHg.

Specification:

±3 mmHg.

Fail.

You now investigate the measurement path.

Real-World Example: Battery

Runtime specification:

Minimum 90 minutes under defined test conditions.

Result:

45 minutes.

Battery charges to 100%.

The battery still fails the runtime requirement.

Common Mistakes

Confusing Range With Accuracy

They are different.

Confusing Resolution With Accuracy

More digits do not mean more correct.

Ignoring Units

Convert first.

Ignoring Test Conditions

The specification may not apply.

Using a Typical Value as a Hard Limit

Read the wording.

Testing at the Wrong Point

Use the specified location.

Rounding a Failing Result Into a Pass

Keep appropriate precision.

Chasing Perfect Numbers

Within specification is usually the goal.

A Useful Specification Checklist

Before comparing a result, ask:

What parameter am I testing?

What is the specified range?

What is the allowed error?

What units apply?

What test conditions are required?

Is this a minimum, maximum, typical, or tolerance?

Is my test equipment appropriate?

Then compare.

Another Useful Question

Ask:

Am I using the specification the manufacturer actually intended for this exact test?

That can prevent a lot of false failures.

What Did You Actually Prove?

Suppose a ventilator delivers:

500 mL

when set to:

500 mL.

You proved:

The analyzer measured 500 mL under that test condition.

You still need the specification to decide whether that result is acceptable.

Now suppose it delivers:

480 mL

and the allowed tolerance is:

±10%.

You proved:

The measured output is within the manufacturer's allowed performance range at that test point.

That is a meaningful verification.

Final Thoughts for Biomeds

Specifications turn:

Looks good.

into:

Meets the required limit.

They tell you what the manufacturer expects from the equipment.

But the number alone is not enough.

You need to know:

Read the whole specification.

Read the footnotes.

Use the correct test setup.

Compare objectively.

And remember:

A specification tells you whether performance is acceptable.

It does not automatically tell you what part is bad.

That part still requires troubleshooting.

— Jake

Important Note

Device specifications, test methods, tolerances, environmental limits, and acceptance criteria vary by manufacturer, model, hardware revision, and software version. Use current manufacturer documentation and approved test equipment and procedures when making service, calibration, or return-to-service decisions.

Related Biomed Basics

Back to Biomed Basics