{"id":3625,"date":"2026-10-09T10:02:05","date_gmt":"2026-10-09T02:02:05","guid":{"rendered":"http:\/\/www.whereonvacation.com\/blog\/?p=3625"},"modified":"2026-10-09T10:02:05","modified_gmt":"2026-10-09T02:02:05","slug":"what-is-the-drive-current-range-for-940nm-vcsel-4513-de814e","status":"publish","type":"post","link":"http:\/\/www.whereonvacation.com\/blog\/2026\/10\/09\/what-is-the-drive-current-range-for-940nm-vcsel-4513-de814e\/","title":{"rendered":"What is the drive current range for 940nm VCSEL?"},"content":{"rendered":"

Hey there, 940nm VCSEL<\/a><\/p>\n

<\/p>\n

If you\u2019re here, you\u2019re probably deep in the weeds with 940nm VCSELs\u2014whether you\u2019re designing a 3D sensing module for a new smartphone, building a LiDAR unit for a robot, or testing proximity sensors for smart home gear. One question I get asked every single day, at least 10 times a week, is: \u201cWhat\u2019s the right drive current range for a 940nm VCSEL?\u201d And I\u2019ll be real with you, it\u2019s not a one-size-fits-all answer. But as someone who\u2019s been selling these things for over 8 years\u2014seen the hype of early 3D sensing, the chaos of supply chain crunches, and all the random use cases no one talks about publicly\u2014I can break this down like we\u2019re chatting over a coffee, no fancy jargon dumps.<\/p>\n

First, let\u2019s get one basic thing straight: 940nm is the sweet spot for most consumer and industrial VCSELs right now. It\u2019s eye-safe for most applications (we don\u2019t want to blind anyone, that\u2019s a big no-no), it plays nice with silicon photodiodes, and the manufacturing is mature enough that we don\u2019t have to charge a small fortune for them. But drive current is where it gets tricky. Too low, and your VCSEL doesn\u2019t turn on (or it\u2019s so dim it\u2019s useless). Too high, and you\u2019re frying the thing\u2014early failure, weird wavelength shifts, maybe even popping it before your prototype hits the testing stage.<\/p>\n

Let\u2019s start with the bare minimum drive current, or what we call the \u201cthreshold current.\u201d That\u2019s the point where the VCSEL starts lasing, not just glowing like an LED. For most standard 940nm VCSEL arrays that are 1mm x 1mm (the most common size we supply, since it works for everything from phones to small LiDARs), that threshold is usually between 0.5 mA and 2 mA per emitter. Wait, hold on\u2014emitter, not the whole array. A standard 1mm x 1mm array has like 10 to 20 individual emitters, right? So if you\u2019re testing a single emitter (super common for R&D), you\u2019re looking at that 0.5mA to 2mA sweet spot to get it lasing. If you\u2019re powering a full array, you add those up\u2014so 10 emitters would be 5mA to 20mA total for threshold. That\u2019s the rock bottom, if you go below that, it\u2019s just an LED, not a VCSEL, and you\u2019re wasting your time.<\/p>\n

Now, the upper limit\u2014this is where most people mess up. What\u2019s the max drive current you can push to a 940nm VCSEL before it dies or stops performing? This depends entirely on two things: how long you want it to last, and what you\u2019re using it for. If you need pulsed operation (which is 99% of the time\u2014LiDAR and 3D sensing use pulses, not continuous wave), you can crank way more current than if you\u2019re running it 24\/7 in continuous wave (CW) mode. Let\u2019s break that down.<\/p>\n

For pulsed operation\u2014this is the big one for our customers. 90% of our VCSELs go into pulsed applications. The max drive current here? For standard 1mm x 1mm arrays, we typically recommend a peak current of 500 mA to 1.5 A. Wait, that\u2019s a huge range, I know. Why? Because pulse width and duty cycle matter so much. If your pulse is super short\u2014like 10 ns or less, which is common for time-of-flight (ToF) LiDAR where you need ultra-fast pulses\u2014you can push up to 1.5 A without any issues. The heat doesn\u2019t have time to build up, so you don\u2019t get thermal damage. But if your pulse is longer, say 100 ns or 1 \u00b5s, you need to back off a bit\u2014maybe 500 mA to 800 mA, because more time means more heat even if the total energy is the same. And duty cycle? If you\u2019re firing pulses 10% of the time vs 0.1% of the time, that changes everything. Higher duty cycle = more average heat = lower peak current you can use without shortening lifespan.<\/p>\n

Now, CW mode\u2014this is less common, but we still get asks for it, like for some continuous proximity sensors or low-power face unlock that doesn\u2019t pulse. For CW, the max drive current is way lower, usually 10 mA to 50 mA for a 1mm x 1mm array. Why so much lower? Because you\u2019re running it non-stop, so heat buildup is constant. Push 100 mA CW and that VCSEL will be dead in a week, maybe less. We always warn customers against CW unless they absolutely have to, because pulsed is way more efficient and kinder to the part.<\/p>\n

Wait, but let\u2019s not forget about custom cases. We make a lot of non-standard 940nm VCSELs too\u2014like smaller arrays, bigger arrays, single emitters for medical devices, even ones packaged for extreme temperatures. For a single emitter (not a full array), max pulsed current is usually 50 mA to 100 mA, and CW is 1 mA to 10 mA. For a bigger array, like 2mm x 2mm for long-range LiDAR, we go up to 3 A peak current in pulses, as long as the pulse width is short enough. And temperature? Oh right, temperature affects this too. If you\u2019re operating at 85\u00b0C (hot as hell, common for automotive LiDAR), you have to lower the drive current by about 20-30% compared to room temp. The VCSEL\u2019s efficiency drops when it\u2019s hot, so too much current will overheat it immediately. I once had a customer test a 2mm array at 105\u00b0C with 3A, and it smoked within 2 minutes\u2014oops, that was a fun support call.<\/p>\n

Another thing: don\u2019t forget about series resistance and your driver circuit. I can give you the current range, but if your driver is garbage, you\u2019re not going to hit that range reliably. We always tell customers to pair our VCSELs with a VCSEL-specific driver, not an LED driver. LED drivers are built for lower frequencies and higher voltage, but VCSELs need fast rise times and precise current control. A bad driver can make your current jump way higher than you set it, killing the VCSEL, or not deliver enough current so your signal is weak.<\/p>\n

Wait, let\u2019s talk about why this matters beyond just \u201cnot breaking the part.\u201d The drive current directly impacts the output power of the VCSEL, right? More current = brighter output, which means longer range for LiDAR, faster response for 3D sensing. But it\u2019s a trade-off. You don\u2019t want to run at max current every time, because it shortens lifespan. Our data shows that if you run a 1mm array at 1 A peak current (pulsed, 10 ns pulse, 0.1% duty cycle), the lifespan is over 100,000 hours\u2014way more than any product will last. But if you run it at 1.5 A, same pulse, lifespan drops to around 10,000 hours, which is enough for a consumer product\u2019s 5-year lifecycle, but not for industrial equipment that lasts 10+ years. So it\u2019s all about balancing performance and longevity.<\/p>\n

I\u2019ve seen so many customers get this wrong. Last year, a startup came to us with a prototype for a new AR glasses depth sensor. They were using a 1mm array, running it at 2 A peak current in pulses, and wondering why their units were dying after a month. Turns out they didn\u2019t account for heat buildup in the glasses (small, enclosed space, no ventilation) and they cranked the current too high. We switched them to a 1.2 A max, adjusted their driver, and their lifespan went up to over a year, which is perfect for their product. Another example: a big automotive LiDAR company was using our 2mm arrays, originally set to 2.5 A peak, but we told them at their operating temp (60-80\u00b0C) they needed to drop to 2 A. They did, and their warranty claims went down by 15%\u2014saving them a ton of money.<\/p>\n

Also, let\u2019s clarify: these ranges are for standard 940nm VCSELs we supply, right? If you\u2019re buying cheap no-name VCSELs from Amazon or AliExpress, the ranges are totally different. Those are usually lower quality, so their max current is way lower, and threshold current is higher. We only supply industrial-grade VCSELs that are tested to MIL-STD standards, so their specs are consistent and reliable. No random batch-to-batch variation, no hidden defects that make one part die at 100mA and another at 200mA.<\/p>\n

Wait, let\u2019s recap so it\u2019s easy to remember, because I know this is a lot of numbers. For our standard 940nm VCSELs (1mm x 1mm array, 10-20 emitters):<\/p>\n