Coreray: Leading Fiber Optic Switch Manufacturer - MEMS & Mechanical Optical Solutions

    Back in 2010, I watched a $50,000 fiber sensing system become almost useless because someone picked the wrong fiber optic switch. The alarms never stopped. The customer blamed the sensing electronics – but the real culprit was a cheap switch with terrible isolation.

That was my first year at Guangxi Coreray Optical Communication Technology Co., Ltd. (www.coreray.com). Since then, I’ve helped hundreds of engineers avoid the same mistake.

Today, I’ll walk you through the six optical switch parameters that actually matter in real life. No marketing fluff. Just what you need to know before you click “buy”.

1. Insertion Loss (IL) – Don’t Trust the “Typical” Number

    Insertion loss is simply how much light power disappears when it passes through the switch. Lower is better. A good mechanical optical switch should have IL ≤0.5 dB. Our Coreray 1×2 and 1×4 models typically run at 0.2–0.4 dB.

But here’s the trap:
Many suppliers advertise “typical IL 0.4 dB” measured at 25 °C in a clean lab. Put that same switch in an outdoor cabinet at -10 °C, and the IL can jump to 1.2 dB.

Always ask: “Is this IL guaranteed over the full operating temperature range?”

When IL is critical:

·         Long‑haul transmission (>100 km) – every 0.2 dB means another amplifier.

·         High‑sensitivity sensing – extra loss kills your signal‑to‑noise ratio.

When you can be relaxed:

·         Short patch cords inside a lab or data center rack.

·         Applications where you have plenty of optical power budget.

2. Isolation – The Reason Your System May False‑Alarm

Isolation tells you how much light leaks into the “off” port. Higher is better. Standard telecom needs ≥40 dB, but sensing and military often require ≥50 dB or even 60 dB.

A true story: A pipeline company bought cheap 1×8 switches from an online marketplace. Their DAS (distributed acoustic sensing) system triggered false alarms every few minutes. We tested one switch – isolation was only 35 dB, and adjacent channels were cross‑talking like crazy. After switching to Coreray’s MEMS optical switch (55 dB isolation), false alarms dropped by 90 %.

Rule of thumb: If you measure weak signals (vibration, temperature, strain), never settle for less than 50 dB isolation.

3. Switching Time – How Fast Is “Fast Enough”?

Switching time is the delay from command to actual light path change.

·         Mechanical: 5–15 ms

·         MEMS: 1–5 ms

·         Electro‑optic / Acousto‑optic: microseconds to nanoseconds

Most people over‑spec here. Do you really need microsecond switching?

A critical note on “latching” vs “non‑latching”:
For OLP, always choose latching (locking) type. If power fails, a non‑latching switch returns to its default port – which could be the dead fiber. I’ve seen this take down a provincial network for 45 minutes. Don’t let that be you.

4. Repeatability & Reliability – The Real Difference Between Cheap and Good

Repeatability means: after 10,000 or 1 million switches, does the IL stay the same? Reliability covers operating temperature, vibration resistance, and total switching cycles.

Good mechanical switches are rated for 10 million cycles with IL change <0.2 dB. Coreray’s internal testing goes to 20 million cycles with <0.1 dB drift.

Temperature range matters more than you think:

·         Commercial grade: 0 to 70 °C – fine for indoor racks.

·         Industrial grade: -40 to 85 °C – for outdoor cabinets, northern winters.

·         Military grade: -55 to 125 °C – for defense and extreme environments.

We once had a customer in northern China who bought commercial‑grade switches for a roadside cabinet. The first winter at -30 °C, all of them stopped working. Don’t save $20 on a switch that will cost you a $2,000 service call.

5. Port Configuration & Wavelength – Get These Wrong and Nothing Works

Port count is simple: 1×2, 1×8, 2×2, 4×64, etc. Count your inputs and outputs before ordering.

Wavelength is even simpler: your system’s laser wavelength must match the switch’s operating range. Common bands:

·         850 nm – multimode, short distance

·         1310 nm / 1550 nm – single‑mode telecom

·         400–1700 nm – wideband for laser medical, spectroscopy

Pro tip: Some switches claim “wideband” but have very different IL at 850 nm vs 1310 nm. Ask for a wavelength vs. loss curve before buying.

6. Crosstalk & Extinction Ratio – Only Critical for High‑End Applications

Crosstalk is the signal leaking into non‑selected ports. ≥40 dB is standard, ≥50 dB for high‑end. Extinction ratio measures how cleanly the switch turns “off”.

If you’re doing coherent communication, quantum key distribution, or precision interferometry, pay close attention. Otherwise, don’t lose sleep over it.

So, Before You Ask for a Quote…

Ask yourself these six questions:

1.    How far is my signal traveling? (IL requirement)

2.    Are there adjacent channels that could interfere? (Isolation)

3.    How fast must I switch? (Switching time)

4.    Where will this switch be installed? (Temperature, vibration)

5.    How many inputs/outputs and what wavelength? (Ports & λ)

6.    What’s my real budget – including long‑term reliability?

Answer these, and you’ve already eliminated 80 % of the wrong choices.

In the next article, we’ll apply these parameters to real‑world scenarios: telecom, fiber sensing, and lab testing.

Stay tuned on Coreray’s website (www.coreray.com).