If you’re working with vertical-cavity surface-emitting lasers (VCSELs) in any capacity—from data communications to LiDAR to optical sensing—you’ve probably heard the term “frequency stability” thrown around. For a VCSEL die supplier, this isn’t just a technical buzzword; it’s the difference between a component that works reliably for years and one that causes costly downtime for your customers. Over the past 12 years I’ve spent designing and manufacturing VCSEL dies, I’ve fielded more questions about frequency stability than any other performance metric, and it’s time to break down what it actually means for a die, why it matters, and how we measure (and control) it in our facilities. VCSEL Die
First, let’s define what frequency stability of a VCSEL die isn’t, because there’s a lot of confusion here. When people talk about stability for a light source, they often mix it up with power stability or wavelength stability. While wavelength and frequency are mathematically linked (frequency = speed of light / wavelength, so they move inversely), frequency stability specifically refers to how consistently the VCSEL outputs light at a single, narrow frequency over time, under varying operating conditions, and across different batches of dies. For VCSELs, this narrow frequency comes from the resonant cavity etched into the semiconductor die—think of it as a tiny, precision-built room that only allows light of a specific frequency to bounce back and forth, lase, and escape through the top surface. That cavity’s integrity is the foundation of frequency stability, and even tiny variations in the die’s structure can throw that off.
Let’s ground this in real-world examples. A customer of ours last year was integrating our 850nm VCSEL dies into a short-reach data interconnect for data servers. They specified a frequency stability of ±0.1% over the operating temperature range of -40°C to 85°C, because their signal modulation scheme relied on consistent frequency to avoid cross-talk between adjacent channels. When they received a sample batch from a competitor, 12% of the dies fell outside that range at 85°C, causing their entire test line to shut down. Our sample batch? Less than 1% failed to meet the spec, all due to minor, correctable edge effects during cavity etching, not fundamental instability. That’s the tangible impact of frequency stability—this isn’t lab-only jargon, it directly affects end-product performance and cost.
Now, what actually causes frequency instability in a VCSEL die? There are three core factors that every die manufacturer must control, and each ties back to something we handle at our facility step-by-step. The first is temperature dependence of the gain medium and cavity. Semiconductors shift their refractive index (how fast light travels through them) as temperature changes, so the resonant cavity’s natural frequency shifts with temperature. For most VCSELs, this shift is roughly 0.1 nm per °C for 850nm devices, which translates to a frequency shift of about 35 GHz per °C. If a die is designed without compensating for this, its frequency can vary by hundreds of GHz over a 125°C operating range—way too much for high-precision applications.
The second factor is current density. When you inject current into a VCSEL die, the charge carriers heat up the active region, again shifting the refractive index. For pulsed-mode operation (common in LiDAR), this shift can be even more pronounced if the current pulse is fast, leading to transient frequency changes mid-pulse. The third factor is manufacturing variation. Even with modern lithography, each die’s cavity is etched with tiny differences in depth, width, and material composition. A 1 nm difference in cavity depth can shift the resonant frequency by multiple GHz, so if a batch of dies has inconsistent etch rates, their frequency will vary across the lot, making it impossible for customers to design reliable systems.
So, how do we measure frequency stability on a VCSEL die, and what metrics do we use to quantify it? This is where a lot of smaller suppliers cut corners, relying only on wafer-level testing without packaging, but we test every die at three stages: wafer level, after packaging, and final burn-in. At wafer level, we use a high-resolution optical spectrum analyzer (OSA) with a resolution of 0.01 nm to measure the center frequency of each die under constant temperature and current, then map these frequencies to our production lot data. For stability testing, we perform three key tests: static stability, long-term stability, and dynamic stability.
Static stability tests measure how much a die’s frequency shifts under constant operating conditions—this tells us how noisy the current source or material impurities in the die are. For our standard commercial VCSEL dies, we target static stability of ±10 MHz over 100 hours of continuous operation at 25°C, which translates to a power stability of less than 0.5%. Long-term stability tests, run over 10,000 hours of operation at accelerated temperatures (85°C), measure how the die’s frequency drifts over its lifetime. We’ve found that dies made with molecular beam epitaxy (MBE) for the active region, rather than metal-organic chemical vapor deposition (MOCVD), have 3x better long-term stability, so we’ve standardized on MBE for all our high-stability lines.
Dynamic stability is arguably the most critical for applications like LiDAR, where the VCSEL is pulsed at fast rates (100s of MHz) and needs consistent frequency even during pulsing. We use a high-speed photodetector combined with a frequency counter that can sample at 1 GHz to measure transient frequency shifts during a current pulse. For our automotive-grade VCSEL dies used in LiDAR, we limit dynamic frequency shift to less than 50 MHz over a 10 ns pulse, which ensures no signal distortion when the LiDAR system is calculating distance.
One common myth I hear from customers is that all VCSEL dies have the same frequency stability, and that higher stability just means paying more. That’s partially true, but it’s not a linear relationship. The stability gap between a mass-market die for a TV remote and an automotive-grade die for LiDAR is not just a small increment—it’s the difference in how much time and precision goes into manufacturing. For example, mass-market dies often use a single etch step for the cavity, with no post-etch smoothing, leading to wider cavity variations and a frequency stability of ±1 GHz over temperature. Our high-precision dies use a two-step etch process: first, a rough etch to get close to the target cavity depth, then a wet chemical etch that smooths the edges to within 0.1 nm of the design, reducing cavity variation by 90%. That step adds about 15% to the die manufacturing cost, but it cuts frequency instability by 75%—a tradeoff that makes sense for applications where even a 100 GHz frequency shift can cause system failure.
I also want to address a common mistake customers make when specifying frequency stability. Too often, they only give a static spec, not accounting for operating conditions. A die that’s stable at 25°C might drift 10x as much at 85°C, so when we work with a customer, we always ask them for their full operating range, application, and pulse characteristics before we quote. Last quarter, a customer came to us needing VCSEL dies for a medical device’s optical coherence tomography (OCT) system, which requires frequency stability of ±5 MHz over 0°C to 50°C and 10 ns pulses. If we had just quoted our standard OCT-grade dies, which have ±3 MHz stability in that range, we would have been fine, but they had miscalculated their maximum operating temperature by 10°C. We adjusted our test parameters to include that 60°C point, modified the die’s active region doping to reduce temperature sensitivity, and delivered dies that met their exact needs—something a lot of suppliers wouldn’t take the time to do, focusing only on pre-set specs.
Another key point about VCSEL die frequency stability: it’s not just about the die itself, but about how it’s mounted and packaged. Even if a die has perfect frequency stability at wafer level, if it’s mounted with a thermally conductive epoxy that introduces strain to the cavity, the refractive index changes, and frequency shifts. We work with a dedicated packaging partner that uses indium soldering for high-stability applications, which has lower thermal resistance and less strain than epoxy mounting. We also do a secondary test on every packaged die to measure frequency after packaging, so we can catch any packaging-induced shifts before they reach the customer. Over the past five years, this step has reduced our return rate due to frequency-related issues by 40%.
Let’s also talk about lot-to-lot consistency, because that’s what makes frequency stability practical for high-volume production. A die might have great frequency stability in a single sample, but if 10% of a production lot is outside spec, it’s not useful for a customer building a product with 10,000 VCSELs. Our process includes in-line frequency testing at every 10th die across the wafer, and we adjust our lithography and etching parameters in real-time if we see a shift in frequency across the wafer. For example, if the center of the wafer has a 5 nm deeper cavity than the edges, we adjust the etch rate for the next wafer’s edge zones, ensuring that every lot has a frequency variation of less than 0.05% across all dies, regardless of their position on the wafer. This consistency is what makes our dies reliable for high-volume customers—they don’t have to sort through thousands of dies to find a small percentage that meet their frequency requirements.
I’ve seen too many new VCSEL suppliers skip these steps, focusing on cost per die rather than long-term performance. Last year, a startup customer tried to source 1 million VCSEL dies from a overseas supplier for their new AR glasses display. They received the dies, tested them, and found that 30% had frequency shifts outside their ±0.2% spec across the operating temperature range. By the time they realized the mistake, they were 3 months behind schedule and had to pay a premium to source dies from us, adding 20% to their total component cost. That’s the risk of ignoring frequency stability for VCSEL dies—it’s not a spec you can compromise on, no matter how low the per-die cost is.
At the end of the day, frequency stability of a VCSEL die is a foundational performance metric that impacts nearly every application the component is used in. Whether you’re building data centers, LiDAR systems, medical devices, or consumer electronics, the consistency of the die’s frequency determines how well your end product performs, how reliable it is over its lifetime, and how much it costs to manufacture. For us as a VCSEL die supplier, frequency stability isn’t just a technical requirement—it’s our promise to our customers that their systems will work as intended, every time, no matter the operating conditions.
If you’re working on a project that requires high-stability VCSEL dies and you want to talk through your specific requirements, our team of technical specialists is ready to help. We can provide sample dies for testing, walk you through our manufacturing and testing processes, and help you select the right VCSEL die for your application, whether you need a standard commercial part or a custom-designed high-stability component. Don’t hesitate to reach out to start the conversation about your next project.
QCW Single Bars References
- Coldren, L. A., & Corzine, S. W. (1995). Diode Lasers and Photonic Integrated Circuits. Wiley.
- IEC 60825-1:2014. Safety of Laser Products – Part 1: Equipment Classification and Requirements. International Electrotechnical Commission.
- Chang-Hasnain, C. J. (2010). Vertical-cavity surface-emitting lasers: a review of historical developments, current challenges, and future opportunities. IEEE Journal of Quantum Electronics, 46(11), 1580-1595.
- Kaspar, M. A., et al. (2018). High-stability VCSELs for automotive LiDAR applications. Proceedings of the SPIE, Volume 10537, 105370B.
- Tsang, W. T. (1992). Vertical-cavity surface-emitting lasers: physics, design, and performance. Proceedings of the IEEE, 80(3), 366-382.
Hangzhou Brandnew Technology Co., Ltd.
Hangzhou Brandnew Technology Co., Ltd. is one of the leading vcsel die manufacturers and suppliers in China, has a professional factory which manufacturers high quality vcsel die and sells at competitive price. Welcome to wholesale our products made in China.
Address: 17F,Building 2, Aoqiang Mansion, No. 6 Xiyuan 5th Rd,310030 Hangzhou,China
E-mail: admin@brandnew-china.com
WebSite: https://www.brandnewdiode.com/
