Why Does Plastic Optical Fiber Use 650 nm Red Light Instead of 850 nm?

A wavelength such as 650 nm can look like a simple line in a plastic optical fiber datasheet. In practice, however, it represents a much larger engineering decision.

Conventional PMMA plastic optical fiber (POF) commonly operates with visible red light around 650 nm because the fiber, optical source, transmitter package, receiver, and available optical power budget work effectively together in that region. The wavelength was not selected by the fiber alone, nor simply because red light is visible.

This also explains why standard PMMA POF does not normally use 850 nm, even though 850 nm is widely used in silica multimode fiber systems. The two fiber technologies are based on different materials, different attenuation spectra, and different optoelectronic ecosystems.

The useful question, therefore, is not simply “Why 650 nm?” It is:

Why does a PMMA POF link work well as a complete optical system around 650 nm?

Standard PMMA plastic optical fiber commonly operates around 650 nm because this wavelength falls within a useful low-loss transmission window of PMMA and can be matched with practical red LED transmitters and compatible receivers. The result is an economical, robust short-distance optical link rather than a wavelength selected by the fiber alone.

Several factors have to line up at the same time:

• the fiber must have acceptable attenuation at the operating wavelength;

• a suitable emitter must generate enough optical power in that region;

• that light must couple efficiently into the large POF core;

• the receiver must have adequate sensitivity to the transmitted spectrum;

• the complete link must maintain sufficient optical margin over the required distance and temperature range.

This is why 650 nm should be understood as a system operating region, not merely a material property.

If fiber attenuation were the only consideration, engineers could simply identify the lowest point on an attenuation curve and place the transmitter there.

Real optical links are more complicated.

An extremely low-loss fiber wavelength is of little practical value if efficient, reliable, economical emitters are difficult to obtain there. Similarly, an excellent LED is not useful if the fiber strongly absorbs its output.

A successful link requires overlap between the spectral characteristics of the fiber and the optoelectronic components.

For conventional PMMA POF, the red region around 650 nm became one of the operating regions where those requirements could be satisfied together.

Fiber attenuation is not constant with wavelength.

When optical power travels through a fiber, part of that power is lost through absorption, scattering, imperfections, and other mechanisms. The relative contribution of those mechanisms changes with wavelength.

PMMA therefore does not have one uniform loss value that applies to all colors of light.

PMMA Optical Attenuation Changes with Wavelength

PMMA is an organic polymer whose molecular structure contains carbon-hydrogen bonds. Molecular vibrations associated with these bonds contribute absorption bands and overtones across the optical spectrum.

As a result, the attenuation spectrum contains regions of relatively high and relatively low loss instead of a smooth horizontal line.

Conventional PMMA POF has several relative low-loss regions across the visible spectrum, including an important red transmission window around 650 nm.

For communication engineers, the important consequence is simple:

changing wavelength changes how quickly the fiber consumes optical power.

The red region around 650 nm provides sufficiently low attenuation for the short distances where standard 1 mm PMMA POF is commonly used.

PMMA POF attenuation near this region is normally discussed on a per-meter basis rather than the very low per-kilometer loss levels associated with long-distance silica fiber.

That does not make PMMA a substitute for low-loss long-haul glass fiber. It reflects a different engineering objective.

Industrial PMMA POF links are generally short. Within those distances, 650 nm provides a workable combination of fiber loss, optical source availability, coupling tolerance, and receiver performance.

Not necessarily.

This distinction matters because PMMA has more than one relatively low-loss region in the visible spectrum. Other visible wavelengths can also provide favorable attenuation depending on the fiber composition and measurement conditions.

Therefore, saying that POF uses 650 nm simply because “650 nm is the lowest-loss wavelength of PMMA” is too simplistic.

The more useful explanation is that 650 nm offers a favorable system-level combination of PMMA transmission and mature red optoelectronics.

That combination matters more than identifying the lowest point on the fiber attenuation curve in isolation.

Once the fiber spectrum is considered, the next part of the system is the optical source.

Many industrial POF transmitters use red LEDs whose emission is centered near the PMMA red transmission window. This allows the optical source and fiber to operate as a matched pair.

An LED does not emit all of its optical power at one mathematically exact wavelength.

It has a spectral distribution around a center wavelength. That distribution matters because every part of the emitted spectrum travels through a fiber whose attenuation also varies with wavelength.

Matching a 650 nm Red LED to the PMMA Transmission Window

A POF transmitter must therefore be evaluated by more than a label such as “650 nm.”

Relevant parameters include:

• center wavelength;

• spectral width;

• coupled optical output power;

• coupling efficiency into the fiber;

• wavelength shift with temperature;

• output-power variation with temperature and aging.

If the emitter spectrum moves toward a higher-loss part of the PMMA attenuation curve, the fiber consumes more of the available optical margin.

The LED and fiber are therefore spectrally coupled components of the same link.

For conventional short-reach PMMA links, LEDs offer a practical combination of optical output, simple drive electronics, mechanical robustness, and cost.

The large core of standard POF also makes efficient coupling possible without the extremely tight alignment tolerances associated with many small-core optical systems.

This does not mean an LED is universally superior to a laser. It means that for many conventional industrial POF links, a red LED provides the level of performance needed without adding unnecessary optical complexity.

This is where comparisons between POF and glass fiber often become misleading.

An engineer familiar with data-center optics may see 850 nm used throughout multimode fiber systems and reasonably ask why the same wavelength is not simply used for POF.

The answer is that 850 nm belongs to a different fiber-and-emitter combination.

PMMA and silica are chemically different materials. Their absorption mechanisms and useful wavelength windows are therefore different.

A wavelength that works very well in silica is not automatically a good operating wavelength in PMMA.

For conventional PMMA, attenuation rises substantially as the operating wavelength moves away from its useful visible transmission windows into much of the near-infrared region. That additional fiber loss consumes the available optical power budget more rapidly.

An 850 nm source can still inject light into PMMA. The issue is not whether light physically enters the fiber.

The issue is whether the resulting link loss is acceptable for the intended distance and receiver sensitivity.

Silica multimode fiber followed a different engineering path.

Modern OM3 and OM4 multimode fibers are specifically optimized for laser transmission around 850 nm and are commonly paired with 850 nm VCSEL transmitters.

The wider system is therefore built around a different combination:

silica multimode fiber + 850 nm VCSEL + high-speed receiver + standardized transceiver interface

Conventional industrial POF evolved around another combination:

650 nm PMMA POF vs 850 nm Silica Multimode Fiber

PMMA fiber + approximately 650 nm red emitter + POF transmitter/receiver + large-core optical interface

The difference is therefore not:

650 nm = older technology
850 nm = more advanced technology

The real difference is that each fiber material has been optimized together with its own emitter and receiver ecosystem.

Neither wavelength is inherently better.

Each one makes sense in the optical system for which it was developed.

The easiest way to understand industrial POF is to stop treating wavelength as a standalone specification.

A working link can be represented as:

Fiber + LED + Transmitter + Receiver: The Complete POF Link

Electrical signal → transmitter → LED → POF → receiver → electrical signal

Each stage affects whether enough optical power reaches the receiver with sufficient margin.

The transmitter determines the optical power launched into the fiber and the spectrum over which that power is distributed.

Its real behavior also changes with operating conditions.

A red LED's center wavelength is not completely fixed. It shifts with operating temperature, while its optical output power can also vary across the specified temperature range.

This means a transmitter labeled “650 nm” should not be treated as a perfectly monochromatic source operating at exactly 650 nm under every condition.

The practical concern is whether the transmitter spectrum continues to overlap favorably with the PMMA transmission window across the required operating range.

Fiber attenuation can be approximated as:

Fiber loss = attenuation at the operating wavelength × fiber length

But fiber loss is only one part of a real system.

Connectors, coupling interfaces, component tolerances, temperature effects, transmitter aging, and design reserve can also consume margin.

A simplified optical link relationship is therefore:

Available transmitter-to-receiver power budget − coupling and connection losses − fiber loss − required margin = remaining link margin

In practice, maximum link length is determined by the combined effect of transmitter output, receiver sensitivity, fiber attenuation, connection losses, temperature variation, and the margin reserved for reliable operation.

At the other end of the cable, the receiver has a minimum optical input requirement.

If too much power is lost before the signal reaches that point, reliable operation can no longer be guaranteed.

This is why wavelength affects transmission distance even if the transmitter electronics and cable geometry remain unchanged.

A higher attenuation coefficient means more loss per meter. If every other part of the system remains the same, the available optical margin is exhausted over a shorter distance.

The answer depends on how far the wavelength moves and where it moves on the fiber attenuation spectrum.

Broadcom's HFBR application guidance provides a useful engineering example.

In the referenced POF link, Broadcom reports that fiber attenuation at 650 nm is about 0.05 dB/m lower than at 660 nm. Earlier POF systems also used 660 nm-class GaAsP emitters, while later AlInGaP devices moved the operating region closer to 650 nm.

A difference of 0.05 dB/m may look small.

Across 50 m of fiber, however, that difference corresponds to approximately 2.5 dB of optical loss.

That is enough to matter in a link operating close to its power-budget limit.

The example illustrates an important principle: even within the red region, wavelength selection can influence usable distance.

Moving from 650 to 660 nm still keeps the emitter near the conventional red operating region.

Moving from 650 to 850 nm is a fundamentally different change.

Standard PMMA's attenuation characteristics are much less favorable across much of this near-infrared region. More optical power is therefore lost in the fiber, leaving less margin at the receiver.

The practical lesson is straightforward:

replacing a 650 nm POF transmitter with an 850 nm source does not preserve the original link budget simply because both sources can launch light into the fiber.

The complete attenuation spectrum has to be considered.

Industrial equipment does not always operate at room temperature.

Temperature can shift the LED spectrum and change its optical output. Because the fiber loss itself is wavelength-dependent, these changes can affect the available link margin.

A robust POF link therefore has to work not only at a nominal room-temperature condition but across the required operating-temperature range.

This is another reason why the transmitter and fiber specifications must be evaluated together.

The Broadcom/Avago HFBR ecosystem provides a practical example of how an operating wavelength becomes established at the system level.

Broadcom's HFBR-1528Z-class transmitter uses a nominal 650 nm LED and is designed for coupling into 1 mm POF. In the corresponding HFBR link architecture, this combination supports defined short-distance industrial POF links, including configurations specified for distances up to 50 m.

The same 650 nm architecture also appears in faster HFBR families such as the HFBR-0507ETZ, which combines a 650 nm transmitter with 1 mm POF for short-reach 125-Mbaud links.

The 650 nm Industrial HFBR POF Ecosystem

The important point is not the individual part number.

It is the architecture.

The transmitter, receiver, fiber geometry, connector family, optical power levels, and application circuits were designed to work together.

Once such an ecosystem is established, the wavelength becomes embedded in more than the LED itself. It influences:

• transmitter construction;

• receiver spectral response;

• fiber specifications;

• connector and coupling design;

• application circuits;

• link-budget calculations;

• qualification data;

• installed equipment;

• replacement-component compatibility.

Changing from 650 to 850 nm would therefore not be a simple LED substitution. It would require reconsidering the complete optical link.

This is one reason Broadcom/Avago HFBR products and similar industrial POF platforms have remained closely associated with visible red operation.

No.

The term plastic optical fiber describes a broad technology category, not one universal material and wavelength.

The 650 nm discussion applies mainly to conventional PMMA POF widely used in short-distance industrial and consumer communication links.

Other polymer compositions and fiber structures can have very different attenuation characteristics and can support wavelength regions well beyond the conventional PMMA 650 nm window.

This boundary is important.

It would therefore be incorrect to say:

All POF must use 650 nm.

A more accurate statement is:

Around 650 nm is the established operating region for conventional PMMA POF systems because it matches the optical characteristics of PMMA with a mature red-emitter and receiver ecosystem.

The use of 650 nm red light in PMMA POF is not a historical accident and not simply a matter of choosing a visible color.

It comes from a chain of engineering relationships:

Fiber material → attenuation spectrum → practical transmission window → optical source → receiver → power budget → allowable link distance

PMMA provides useful transmission windows in the visible region. Red semiconductor emitters can operate around the approximately 650 nm window with practical optical power, cost, and drive requirements. Receivers and industrial transmitter families were subsequently designed around that combination.

Silica multimode fiber followed another path, with 850 nm VCSELs and laser-optimized OM3/OM4 fiber.

That is why the correct comparison is not simply 650 nm versus 850 nm.

It is:

PMMA + 650 nm optoelectronics versus silica MMF + 850 nm optoelectronics.

In optical-link design, wavelength only makes sense when the complete system is considered.

Conventional PMMA POF commonly uses approximately 650 nm because this region provides a useful low-loss fiber window that can be matched with practical red LED transmitters and compatible receivers. The wavelength is therefore the result of complete link optimization rather than a fiber-only property.

Not necessarily. PMMA has more than one relatively low-loss region in the visible spectrum. The importance of 650 nm comes from the combination of acceptable fiber attenuation and mature red optoelectronic components, not from it being the only or universally lowest attenuation point.

Yes, light at 850 nm can physically propagate through PMMA POF. However, conventional PMMA generally has much less favorable attenuation in that region than around its useful visible transmission windows. The resulting optical loss can therefore reduce the usable transmission distance significantly.

Silica multimode fiber has different material characteristics from PMMA and has been optimized for high-speed transmission around 850 nm. Modern OM3 and OM4 systems combine 850 nm VCSEL transmitters with laser-optimized silica multimode fiber, forming a different optical ecosystem from PMMA POF.

Broadcom/Avago HFBR POF systems combine red transmitters, compatible receivers, 1 mm POF, connectors, and defined optical power budgets as an integrated short-distance optical link. The use of approximately 650 nm therefore reflects the wider PMMA POF system architecture rather than an isolated transmitter choice.

Yes. Fiber attenuation changes with wavelength. If moving to another wavelength increases attenuation, the fiber consumes more optical power per meter. Unless transmitter output, receiver sensitivity, or another part of the system compensates for that additional loss, the maximum reliable transmission distance decreases.