Fotodiodos de InGaAs frente a los de silicio: Elección del detector adecuado para equipos OTDR

So, you’re in the middle of designing your next-generation OTDR (Optical Time Domain Reflectometer). You’ve got the laser sources figured out, the pulse width modulation is looking good, but now you’re staring at the receiver end. The detector.

It’s always the same debate in the R&D lab, isn’t it? Do we try to cut costs with Silicon (Si), or do we commit to Indium Gallium Arsenide (InGaAs)?

If you look at the datasheets quickly, you might think you have wiggle room. But here is the hard truth: if your equipment is targeting the standard telecom windows (1310nm and 1550nm), getting the wrong photodiode isn’t just a minor spec difference—it’s the difference between a top-tier instrument and a paperweight.

En BeePhoton, we’ve seen plenty of engineers try to “hack” the system by pushing detectors to their absolute limits. Sometimes it works. Most of the time, it results in a noisy mess. Today, I want to walk you through the real, gritty differences between InGaAs vs Silicon photodiodes, specifically for fiber optic testing components. No fluff, just the engineering reality.

The Wavelength Battleground: Where Physics Draws the Line

Before we even talk about pricing or packaging, we have to respect the physics. The fundamental difference between these two materials lies in their bandgap energy. This isn’t just textbook theory; it dictates the upper limit of the wavelength your OTDR detector can actually see.

Silicon: The King of Short Range (850nm)

Silicon is cheap. It’s abundant. The manufacturing processes are mature because the entire semiconductor industry runs on it.

Silicon has a bandgap energy of approximately 1.12 eV at room temperature. What does that mean for us? It means Silicon is fantastic at absorbing photons in the visible range and the Near-Infrared (NIR) up to about 1000nm or 1100nm.

If you are building an OTDR specifically for Multimode Fiber (MMF) testing—mostly LANs and data centers running at 850nm—Silicon is your best friend. It has peak responsivity right around 900nm.

However, and this is a big however, once you push past 1100nm, Silicon becomes transparent. The photons from a 1310nm laser just pass right through the material without generating electron-hole pairs. No current. No signal.

InGaAs: The Telecom Standard (1310nm/1550nm/1625nm)

Aquí es donde Fotodiodos PIN de InGaAs enter the chat. Indium Gallium Arsenide has a smaller bandgap (around 0.75 eV depending on the alloy ratio). This allows it to absorb lower-energy photons—specifically those found in the Single Mode Fiber (SMF) bands.

For an OTDR measuring long-haul networks, FTTx, or PONs, you are working at 1310nm, 1490nm, 1550nm, and 1625nm. Silicon is useless here. You literally have no choice but to go with InGaAs if you want to detect anything.

Engineer’s Note: I once saw a competitor try to market a “universal” probe using a Ge (Germanium) detector to save money over InGaAs. Germanium works in these wavelengths, but the dark current (noise) is significantly higher. If you care about dynamic range (and in OTDRs, dynamic range is everything), InGaAs is superior to Germanium.

Key Technical Specs: Responsivity and Noise

Let’s get into the math. Don’t worry, I know standard WordPress editors hate LaTeX, so I’ll keep these formulas clean and copy-paste friendly.

When you are selecting a OTDR detector, you are balancing two things: How much signal can I get (Responsivity) and how little noise can I tolerate (NEP/Dark Current).

1. Responsivity (R)

Responsivity tells you how much electrical current you get for a given amount of optical power. The formula looks like this:

R = Ip / Popt

Dónde:

• R es la respuesta (amperios/vatio)
• Ip is the generated photocurrent (Amps)
• Popt is the incident optical power (Watts)

Or, relating it to quantum efficiency (QE):

R = (QE * q * lambda) / (h * c)

• lambda = wavelength
• q = electron charge
• h = Planck’s constant
• c = speed of light

Here is the kicker:

• En 850nm, a good Silicon detector might have an R of 0.5 A/W.
• En 1550nm, a high-quality InGaAs detector from BeePhoton will typically have an R of 0.85 to 0.95 A/W.

Because InGaAs is more efficient at these longer wavelengths, your OTDR can detect fainter reflections from further down the fiber. This directly translates to a longer measurement distance.

2. Dynamic Range and Dark Current

In an OTDR, you are blasting a strong pulse and listening for the tiny, tiny backscatter echoes. This requires a detector that recovers quickly (low capacitance) and has a very low noise floor.

Dark current (Id) is the current that flows even when no light is hitting the detector. It’s noise.

Noise Equivalent Power (NEP) is a function of this dark current:

NEP = (sqrt(2 * q * Id)) / R

If you use a sub-par detector with high dark current, your “noise floor” rises. The faint backscatter from 100km away gets buried in the noise. You can average the signal all you want, but you can’t fix bad hardware physics.

Fotodiodos PIN de InGaAs generally offer very low dark currents (often in the nA or even pA range for small active areas), which helps you achieve that 40dB+ dynamic range everyone wants.

Comparison Table: The Cheat Sheet

I put this table together so you can see the differences side-by-side. Useful for when you need to justify the BOM cost to your manager.

Característica	Fotodiodo de silicio (Si)	Fotodiodo InGaAs
Primary Wavelengths	400nm – 1100nm	900nm – 1700nm
Lo mejor para	Visible Light, MMF (850nm)	Telecom, SMF (1310/1550nm), SWIR
Responsivity (Peak)	~0.6 A/W @ 900nm	~0.95 A/W @ 1550nm
Coste	Bajo	Moderate to High
Corriente oscura	Very Low (pA range)	Low (nA range), depends on active area
Speed (Rise Time)	Very Fast	Fast (Suitable for high-speed comms)
OTDR Application	Short-range LAN testers	Metro, Long-haul, PON testers

Real-World Application: Why Active Area Matters

Here is something the datasheets don’t always scream at you: The size of the active area.

In OTDR design, you have a trade-off.

1. Large Active Area (e.g., 300um, 500um): easier to couple the fiber to the detector. You don’t need super precision alignment mechanics.
2. Small Active Area (e.g., 75um, 30um): lower capacitance.

Why does capacitance matter? Bandwidth.

Bandwidth (BW) = 0.35 / tr

Dónde tr is the rise time. High capacitance (large area) slows down the rise time.

For an OTDR, speed determines your Dead Zone. If your detector is slow to recover after the initial high-power reflection of the connector, you might be blind for the first 10 or 20 meters of the fiber. That is unacceptable for FTTH applications where the first splitter is close by.

En BeePhoton, we often recommend specific Fotodiodos PIN de InGaAs with optimized active areas that balance coupling efficiency with low capacitance, ensuring your Event Dead Zone is as short as possible.

A “Secret” Case Study: When “Cheap” Became Expensive

I want to share a story (names changed, obviously) about a client we worked with last year. Let’s call them “OptiTest.”

OptiTest was developing a handheld mini-OTDR for the FTTH market. To compete on price, they decided to source a low-cost InGaAs detector from a generic component supplier, ignoring the shunt resistance specs.

El problema:
Their prototype worked fine in the lab on a 5km spool. But when they took it to the field, the trace became incredibly “fuzzy” after 20km. They were losing about 5dB of dynamic range compared to their theoretical calculations.

The Diagnosis:
They sent the unit to us. We swapped their generic detector with one of our high-shunt-resistance InGaAs photodiodes. The issue wasn’t the responsivity; it was the thermal noise generated by the low shunt resistance of the cheap chip.

La solución:
By switching to a BeePhoton detector, their Noise Equivalent Power (NEP) dropped significantly.

• Resultado: They gained back that 5dB of dynamic range.
• Bonus: The trace was cleaner, requiring less averaging time. This meant the technician could finish the job in 10 seconds instead of 30 seconds. In the field, time is money.

You can check out our range here if you want to avoid that headache: InGaAs PIN Photodiodes Category.

Integrating the Detector: It’s Not Just “Plug and Play”

Choosing the sensor is step one. Step two is the circuit.

When you use InGaAs photodiodes, you are likely pairing it with a Transimpedance Amplifier (TIA).

Vout = -Ip * Rf

• Rf is your feedback resistor.

You might be tempted to make Rf huge to get a massive gain (lots of voltage for a little current). But be careful. A large Rf limits your bandwidth.

f_cutoff = 1 / (2 * pi * Rf * Cf)

• Cf is the feedback capacitance.

Ideally, you want an InGaAs detector that is packaged with a pigtail (fiber attached) to minimize stray light and insertion loss. We handle the alignment in the factory so you don’t have to fiddle with XYZ stages in your production line.

Why Experience Matters in Component Selection

Look, anybody can sell you a chip. There are a million distributors online. But at BeePhoton, we aren’t just moving boxes. We understand the pain of a noisy trace. We know what happens when temperature fluctuations shift your dark current specs.

We have spent years characterizing fiber optic testing components. We don’t just look at the max specs; we look at the min specs, the deviations, and the reliability over time.

When you choose a partner for your supply chain, you need someone who can answer: “What happens to the linearity of this detector if the input power spikes to +10dBm?” (Spoiler: it saturates, and we can help you design a protection circuit).

FAQ: Common Questions about OTDR Detectors

Conclusión

Choosing between InGaAs vs Silicon photodiodes ultimately comes down to your wavelength.

• Building a LAN tester for 850nm? Save money, use Silicon.
• Building a serious OTDR for Telecom/FTTH (1310nm/1550nm)? You need InGaAs.

But beyond the material, pay attention to the active area, capacitance, and shunt resistance. These are the hidden specs that determine if your instrument is “okay” or “market-leading.”

En BeePhoton, we specialize in high-performance detection solutions. We help manufacturers navigate these trade-offs every day.

Ready to upgrade your optical detection?
Don’t let a noisy detector ruin your dynamic range specs.

• Need a quote? Contacte con nosotros
• Have a technical question? Envíenos un correo electrónico a info@photo-detector.com.
• Browse the goods: Consulte nuestro Fotodiodo PIN de InGaAss page.

Let’s build better instruments, together.

---