What Are the Main Components of a Phototherapy Device?

Updated: September 24, 2026 | 16-minute read

Illustrative cutaway of the main components of a phototherapy device

What are the main components of a phototherapy device? In a typical red and near-infrared LED device, the main functional groups are the LED emitters, optical assembly, power supply and LED driver, thermal management system, and control electronics. The enclosure, wiring, connectors, and safety features support their operation. Performance depends on how these parts work together—not simply on whether the device emits red light.

A decorative lamp and a purpose-designed device may look similar, but appearance does not establish optical output, safety, or suitability for a treatment protocol. Those characteristics require defined specifications and supporting evidence.

What are the two main types of phototherapy? There is no universal two-category classification covering every use of phototherapy. Dermatology includes narrowband UVB and other ultraviolet treatments; red and near-infrared photobiomodulation, or PBM, is a different application. Narrowband UVB is itself a widely used treatment, so ultraviolet phototherapy should not be described as exclusively broadband. [1]

This article focuses on the components of red and near-infrared LED phototherapy devices. It explains what each subsystem contributes, which measurements matter, and how buyers can assess engineering evidence. It does not treat a component specification as proof of clinical effectiveness.

The optical subsystem: why "number of LEDs" is the wrong metric

Illustrative LED package and optical assembly comparison

Does a higher LED count mean better phototherapy output?

No. LED count describes the construction, but it does not establish irradiance, coverage, uniformity, or therapeutic effectiveness. A useful comparison also needs the emission spectrum, drive conditions, lens design, operating mode, and measurements at the intended working distance.

The distinction between LED package count and chip count is important. A package may contain one chip or several. A dual-chip package can contain one red and one near-infrared chip, but that arrangement is a design choice rather than a universal definition of "dual-chip."

REDDOT' engineering experience includes changing from separate red and near-infrared emitter positions to independently controlled chips within shared packages. Such a change affects PCB routing, driver channels, heat dissipation, and the spacing of active emitters. It may improve the distribution in a single-channel mode, but the result must be checked with an irradiance map.

It does not make near-infrared emission visible. A status indicator or clear mode display is a more appropriate way to communicate that an invisible channel is operating; looking directly at the emitters is not an output test.

What is wavelength binning, and why does it affect photobiological consistency?

Wavelength binning groups LEDs according to specified spectral characteristics under defined test conditions. A nominal wavelength such as 660 nm identifies a design target; it does not mean that every photon has exactly that wavelength.

Keep three specifications separate:

• Peak wavelength: the wavelength at the maximum of the emission spectrum.
• Spectral bandwidth: the width of that spectrum, often reported as full width at half maximum, or FWHM.
• Bin tolerance: the allowed variation in the specified wavelength characteristic among supplied LEDs.

Bin definitions vary by supplier and part number. For example, Lumileds publishes different wavelength ranges and bin tables for its color LED families. A tolerance such as ±2–3 nm should therefore come from the actual component specification, not be presented as an industry-wide requirement. [2]

Tighter purchasing specifications can improve manufacturing consistency. They do not prove that a nearby wavelength is biologically ineffective, and a manufacturing tolerance is not automatically a clinical treatment window.

For independently controlled channels, measure each channel separately and then the combined mode. Overlapping spectra can make individual contributions difficult to identify in a combined measurement. Component or module testing can supplement the finished-device results when further diagnosis is needed.

How lens angle shapes irradiance at treatment distance

Lenses and diffusers shape the distribution of light leaving the device. A narrower beam can concentrate output into a smaller footprint, while a wider beam can increase coverage. The final irradiance pattern also depends on emitter spacing, lens efficiency, alignment, and overlap between neighboring beams.

Changing a lens angle does not guarantee a particular change in area-average irradiance. A center-point reading and an average over a defined treatment area can respond differently. Compare both the distribution and the coverage at the same distance and operating mode.

A narrow lens angle alone does not prove deeper tissue delivery. Tissue exposure also depends on wavelength, anatomy, absorption, scattering, illuminated area, and exposure geometry. A surface measurement cannot establish the dose reaching a joint or other internal target. [3]

Every irradiance claim should identify:

• Distance and the surface from which that distance is measured.
• Active channels, intensity setting, and continuous or pulsed mode.
• Measurement area and grid, including whether the result is a point value or an area average.
• Instrument, calibration status, and relevant measurement uncertainty.
• Warm-up condition and ambient conditions.

The inverse-square law should not be used automatically to convert one close-range panel measurement into another. Large LED arrays are extended sources at short distances, and their optics further affect the distribution. Near-field measurements or an appropriate validated optical model are needed. [4]

Wavelength selection and its biological rationale

Red and near-infrared wavelengths are both studied in photobiomodulation. Proposed mechanisms include interactions with cellular photoacceptors and downstream signaling processes. Their contribution can depend on the wavelength, tissue, and experimental conditions. Hamblin' review discusses cytochrome c oxidase, calcium-related pathways, nitric oxide, and reactive oxygen species; it does not establish one mandatory wavelength for all PBM applications. [5]

In some tissues and exposure conditions, near-infrared light can travel farther than visible red light. However, "850 nm reaches bone" or "660 nm works only at the surface" is too absolute. Detectable transmitted light, sufficient target-tissue exposure, and demonstrated clinical benefit are different findings. [3]

Similarly, a 1:1 red-to-near-infrared LED-count ratio does not establish equal optical power in the two channels. Channel output must be measured. It also does not establish an optimal ratio for every intended use.

Is there radiation in phototherapy? Yes. Red and near-infrared devices emit non-ionizing optical radiation. X-rays are also electromagnetic radiation, but they occupy a much higher-energy, ionizing range. "Non-ionizing" does not mean harmless at every exposure.

Eye and skin safety require assessment of the actual emission and exposure conditions. IEC 62471 provides a framework for evaluating photobiological hazards from lamps and lamp systems, including LEDs. It is a safety standard, not evidence of therapeutic effectiveness. [6]

The power and driver subsystem: where irradiance stability is actually controlled

Conceptual power supply and LED driver flow for a phototherapy device

The LED driver board regulates the electrical conditions supplied to the emitters. A well-designed constant-current driver helps maintain the selected current within its specified operating range. This supports repeatable optical output, but it does not eliminate temperature effects or LED aging.

The power supply and driver may be separate assemblies or integrated into one design. Their combined performance should be evaluated under the intended input supply, operating modes, and thermal conditions.

A wide input-voltage rating shows the intended supply range. It does not, by itself, prove identical irradiance throughout that range. Verification should compare LED current and optical output under representative supply conditions and after warm-up.

Useful engineering checks include current regulation, electrical ripple, startup behavior, channel interaction, connector temperature, and protective responses to abnormal loads. Optical measurements should confirm the effect of these electrical characteristics rather than infer it from the circuit description alone.

Electrical safety standards depend on the intended use and product classification. IEC 60335-2-27 addresses relevant household and similar appliances for skin exposure to optical radiation and expressly excludes appliances for medical purposes. Medical electrical equipment requires evaluation against the applicable IEC 60601 framework; IEC 60601-2-57 addresses relevant non-laser light-source equipment. The applicable edition and market requirements must also be checked. [7][8]

REDDOT' engineering records include a power-connection review prompted by multiple devices sharing a supply path. The transferable lesson is to evaluate the complete installation: upstream connectors and cables may carry the combined current of downstream units. A single-device bench test cannot establish the permissible load for every linked configuration.

Connector ratings, fuses, switches, wiring, permitted connection arrangements, and installation instructions must agree. Any design change needs appropriate verification before release.

Pulse frequency and intensity control electronics

The control electronics select channels, adjust intensity, manage timing, and generate pulses when that function is included. Their presence does not guarantee that displayed settings match the actual optical waveform.

A meaningful pulse specification includes frequency, pulse width, duty cycle, waveform, and peak and average output. If a control uses "0" to indicate continuous operation, that meaning should be explicit; it should not be treated as an ordinary pulse frequency.

For ideal rectangular pulses with negligible output between pulses:

Average irradiance = peak irradiance × duty cycle.

For constant output, or a stable average over repeated pulses:

Radiant exposure (J/cm²) = average irradiance (mW/cm²) × time (seconds) ÷ 1,000.

If the instrument already reports the time-averaged irradiance, do not multiply by duty cycle again. For changing output, radiant exposure is obtained by integrating irradiance over time. These calculations describe exposure at the measurement plane, not the dose absorbed by an internal tissue.

Verification should cover timer accuracy, dimming behavior, channel selection, optical pulse timing, and behavior after interrupted power. Overtemperature responses, output limits, and any interlocks should be assessed according to the design and its risk analysis. Remote controls and apps also need testing of command handling and connection loss where applicable.

Pulse settings are not interchangeable treatment protocols. FDA' January 2023 PBM guidance draft discusses optical parameters and software verification; it remains a nonbinding draft and must not be described as an implemented regulation. [9]

Thermal management: the component most often ignored until the device fails

Illustrative thermal path through an LED board, interface, and heat sink

Why does thermal management matter for sustained irradiance output?

LEDs convert part of their electrical input into heat. If that heat is not adequately transferred away, junction temperature rises and can affect output, emission spectrum, and reliability. Thermal design must therefore consider both light delivery and the temperatures of components and accessible surfaces. [10]

There is no universal rule that an irradiance above 100 mW/cm² requires fans. Irradiance at a measurement plane is not the same quantity as heat generated inside the device. Cooling requirements depend on electrical loading, conversion efficiency, thermal resistance, enclosure design, ambient temperature, and operating duration.

Three elements commonly form the thermal path:

• Heat sink: spreads and transfers heat through an appropriate combination of material, geometry, and airflow.
• Thermal interface material: reduces contact resistance between mating surfaces when correctly specified and applied.
• Optional active cooling: increases heat removal when the thermal load cannot be adequately managed by the selected passive design.

Passive cooling can be suitable when it maintains the required temperatures. Active cooling introduces additional considerations such as fan reliability, airflow obstruction, noise, and maintenance. Neither approach proves good thermal performance without testing. [10]

What happens to LED lifespan when thermal management is inadequate?

Excessive temperature can accelerate degradation, but the amount depends on the component and its operating conditions. A universal claim that poor cooling "halves lifespan" is not justified without supporting data.

A stated lifetime also needs a definition. L70 commonly describes a predicted or measured time to 70% of initial luminous flux under specified conditions; it is not a universal measure of complete device life. For near-infrared emission, radiometric output maintenance is more relevant than a visible-light lumen metric. [11]

Temperature-rise testing should consider the specified ambient range, operating modes, expected session duration, accessible surfaces, and applicable abnormal conditions. Checking the heat sink alone does not establish the temperature of a skin-contact surface or the performance of an overtemperature protection function.

Enclosure sealing and cooling must be considered together. IP65 describes dust protection and protection against specified water jets; it does not establish airtightness, steam resistance, immersion resistance, or suitability for any high-temperature environment. A sealed design can still dissipate heat through a properly engineered conductive path. [12]

Thermal behavior also belongs in the optical safety evaluation. Measurements should account for relevant operating states and changes in emission over time. Whether those changes affect a hazard classification must be determined by assessment, not assumed from the presence of temperature drift alone.

Grounding continuity in metal enclosures

Protective grounding depends on the equipment' electrical protection design. For a Class I design, conductive parts that require protective earthing must have a reliable connection. Class II equipment instead relies on the required double or reinforced insulation; a metal surface alone does not establish which approach applies.

REDDOT' assembly experience includes finding that tightened enclosure screws did not always provide the intended continuity across painted covers. This illustrates why mechanical tightness and electrical continuity are different checks.

Where protective bonding is required, the connection must be designed, assembled, and verified for that purpose. Paint, coatings, washers, joints, conductor terminations, and assembly variation can affect the result. A dedicated bonding conductor may be part of the solution, subject to design verification.

Material choice alone cannot guarantee safety. A conductive steel enclosure can still have an unreliable connection across a coated joint. Likewise, a protective-earth change should not be claimed to resolve display instability or EMC problems unless the cause and result have been demonstrated.

Assembly instructions, operator training, and the applicable electrical safety tests should follow the approved design. Safety-related construction changes must remain traceable.

Quality control systems: why component consistency across a production run matters more than a single sample

Illustrative production inspection with LED and finished-device measurement stations

A single sample provides evidence about that sample; it does not establish the consistency of an entire production run.

LED bins, driver tolerances, thermal interfaces, optical alignment, and assembly processes can vary. A production control plan needs to identify which variations affect the specification and how they will be detected or controlled.

REDDOT LED operates under an ISO 13485:2016 quality management system, with Certificate No. 0220406 issued on July 28, 2025, and holds MDSAP certification, Certificate No. 0220404 issued on the same date. Its documented manufacturing process includes 37 quality inspection steps and traceable inspection records.

ISO 13485 provides a framework for medical-device quality management and applicable regulatory requirements. It does not prescribe a universal 37-step inspection sequence, certify every individual device, or replace design verification. Each manufacturer' inspection plan must reflect its processes, specifications, and risks. [13]

LED binning verification should compare the supplied components with the approved purchasing specification. The test needs the correct wavelength metric: peak wavelength and dominant wavelength are not interchangeable, and dominant wavelength is not an appropriate description of invisible near-infrared emission.

Measurement equipment must also match the quantity being tested. An integrating sphere can support total radiant-flux and spectral measurements when appropriately configured. A calibrated irradiance detector at a defined plane measures incident power per unit area. An integrating-sphere result is not a substitute for an irradiance map at the working distance.

Calibration should cover the required wavelength range and measurement quantity. Instrument settings, detector linearity, saturation, stray light, and dark-signal correction need attention. Dark-signal correction and ambient-light control are distinct tasks. Failure to handle either can bias results, but the direction and significance of the error depend on the setup. [14]

Warm-up should be controlled through a defined procedure. A fixed warm-up period is acceptable when validation shows that it achieves the required stability for that design and test condition. Recording output and spectra during warm-up helps establish that period and reveals drift that a single reading may miss.

For batch comparison, use the same mode, distance reference, measurement area, and calibrated method. Report the average, peak, spatial distribution, and relevant tolerances. Sampling and routine tests should be specified in the control plan; not every qualification test needs to be repeated on every unit.

Traceability should connect the finished device or batch to relevant components, revisions, firmware, inspection results, and approved changes. Regulatory submission and record-retention requirements depend on the applicable market and pathway. It is inaccurate to claim that every submission universally requires the same LED bin records or thermal-interface certificates.

How certification scope maps to component subsystems

Different documents answer different questions. Some evaluate complete equipment, some cover specific hazards, and others concern the manufacturer' quality system. They should not be treated as interchangeable evidence.

CE marking is the manufacturer' declaration of conformity with the applicable EU legislation following the required assessment route. A notified body is involved where that route requires one; CE marking alone is not proof that every subsystem received independent third-party testing. EMC covers immunity as well as emissions. [15][16]

The ETL Listed Mark indicates conformity to the safety standards identified in the listing. It should not be described as verification of treatment efficacy or every advertised irradiance value unless separate evidence explicitly covers those claims. [17]

FDA establishment registration and device listing are separate from product clearance, approval, or authorization. FDA states that registration and listing do not denote those statuses and that it does not issue device registration certificates. [18]

For procurement, request documents that identify the relevant device, configuration, standard or method, issue date, and scope. Review the evidence together with labeling and intended use. The number of certificates is not a reliable substitute for that review.

Key Takeaways

The main components of a phototherapy device work as an integrated system: LED emitters generate light, optics distribute it, the power supply and driver regulate electrical delivery, thermal management controls temperatures, and control electronics manage operation. The enclosure and safety-related construction support all five.

Evaluate the complete device using defined measurements and appropriate safety evidence. LED count, a narrow beam angle, a high center-point irradiance, or a long certificate list cannot independently establish performance or clinical benefit.

For buyers, the most useful evidence combines optical measurements, temperature-rise results, safety-control verification, electrical and EMC reports, and manufacturing traceability for the intended configuration and use.

FAQ

What are the main components of a phototherapy machine?

A typical red and near-infrared LED phototherapy machine contains LED emitters, an optical assembly, a power supply and LED driver, a thermal management system, and control electronics. Wiring, connectors, the enclosure, and protective features complete the design. Other types of phototherapy machines may use different light sources and architectures.

What does phototherapy consist of?

Phototherapy uses light under defined conditions for a specific therapeutic purpose. For an LED photobiomodulation device, relevant parameters include the emission spectrum, irradiance, exposure duration, illuminated area, distance, and pulse settings where applicable. Irradiance multiplied by time gives radiant exposure at the measurement plane when output is constant; it does not determine internal tissue dose or clinical effectiveness by itself.

What are the three mechanisms of phototherapy?

There is no universally accepted three-mechanism classification that applies to all phototherapy. In red and near-infrared PBM, research examines cellular photoacceptors, mitochondrial responses, and downstream signaling involving molecules such as nitric oxide and reactive oxygen species. These are overlapping processes rather than three independent treatment modes. PBM generally concerns nonthermal biological effects; deliberate tissue heating should not be presented as an essential PBM mechanism. [5][9]

What equipment is used in a phototherapy machine?

Common hardware includes an LED board, driver electronics, a power supply, lenses or diffusers, thermal interfaces, a heat sink, and a timer or microcontroller. Fans, displays, remote controls, and wireless modules are optional. The required safety construction depends on the intended use and electrical design. A spectrometer, integrating sphere, or external irradiance meter is normally test equipment, not necessarily a built-in component.

References

1. American Academy of Dermatology. Psoriasis treatment: Phototherapy. Clinical context for ultraviolet phototherapy, including narrowband UVB.
2. Lumileds. LUXEON Rebel Color LEDs. Manufacturer specifications and linked data sheets for wavelength characteristics and bins; cited for component terminology, not medical effectiveness.
3. Photobiomodulation in the parotid and submandibular glands: a Monte Carlo simulation of photon penetration using 525-nm, 660-nm, and 850-nm wavelengths. Lasers in Medical Science, 2025. DOI: 10.1007/s10103-025-04507-7. Anatomy-specific modeling, not a universal penetration-depth guarantee.
4. Vrije Universiteit Brussel research record. Accurate Determination of the Near Field Illuminance From Narrow Beam LED Arrays Using Ray Files. Near-field measurement and the limits of point-source approximations.
5. Hamblin MR. Mechanisms and applications of the anti-inflammatory effects of photobiomodulation. AIMS Biophysics. 2017;4(3):337–361. DOI: 10.3934/biophy.2017.3.337.
6. IEC. IEC 62471:2006 — Photobiological safety of lamps and lamp systems.
7. IEC. IEC 60335-2-27:2024 — Particular requirements for appliances for skin exposure to optical radiation. Scope excludes appliances for medical purposes.
8. IEC. IEC 60601-2-57:2023 — Particular requirements for the basic safety and essential performance of non-laser light source equipment.
9. FDA. Photobiomodulation Devices — Premarket Notification 510(k) Submissions. January 2023 draft guidance, not for implementation; contains nonbinding recommendations.
10. Cree LED. Thermal Management of XLamp LEDs. Engineering guidance on heat generation, thermal paths, and cooling design.
11. Cree Lighting. Don’t Be Left in the Dark. Explains why L70 lumen maintenance does not define complete system life.
12. IEC. Ingress Protection ratings and IEC 60529 — Degrees of protection provided by enclosures.
13. ISO. ISO 13485:2016 — Medical devices: Quality management systems.
14. Gigahertz-Optik. Parameters of a Spectroradiometer. Instrument characteristics and dark-signal correction.
15. European Union. CE marking: EU requirements.
16. European Commission. Electromagnetic Compatibility Directive.
17. Intertek. ETL Listed Mark.
18. FDA. Important Reminders about Registration and Listing.