3D sensing is a depth-sensing technology that augments camera capabilities for facial and object recognition in augmented reality, gaming, autonomous driving, and a wide range of applications. Via a series of questions, Tom Ohlsen, SMU Product Manager for the Keithley Product Line of Tektronix, looks at how this market is growing and some of the common test and measurement challenges faced by companies in this industry.
Mr. Ohlsen’s experience spans over 20 years in the test and measurement industry and includes roles in marketing, R&D and operations. He holds a BS in electrical engineering from the University of Wisconsin and a CORe Credential of Readiness from HBX/Harvard Business School.
Why are you excited about 3D sensing technology?
There is no question that this is a truly exciting space that will have a far-reaching impact across the industry. Test and measurement will play an important role in research as well as product development and manufacturing. 3D sensing technology is about to go full-on mainstream as the likes of Apple, Google and Samsung race to incorporate 3D sensors into their next generation of smartphones. Initially, 3D sensors will be added to the front-facing camera on handsets where they will be used for everything from facial recognition to unlock the phone, 3D selfies, augmented reality gaming, and much more.
Through the use laser diodes 3D sensors can determine the distance and contour of objects based on the time it takes to detect the reflected light while also looking at factors such as the distortion of a reflected grid of light sources. This information is then used to construct 3D images in conjunction with a conventional camera.
What are some of this key trends and challenges facing engineers working to bring 3D sensing solutions to market?
The adoption of 3D sensing for smartphones is driving down size, power and cost—it must fit in a handset and be cognizant of battery power—while driving up performance. This is also enabling expanded use of 3D sensing in other end applications such as automotive and industrial robotics. The need to test these applications is critical and they all have a safety or security component to them. For example, in automotive the application will be safety based driver assist for collision avoidance, blind spot detection, parking assist and driver drowsiness detection—and eventually autonomous driving.
What role does test & measurement play in enabling innovation in this segment?
Both the laser diode that emits the 3D sensing light and the photo diodes that measure the returning light are critical elements of a 3D device and they need to be tested and verified. The “LIV” characteristics of the diode need to be tested at multiple steps in the supply chain process. LIV stands for Light-Current-Voltage. Testing LIV characteristics requires a precise current to be sourced through the diode and synchronized with light intensity and spectral measurements. Reverse bias voltage and current measurements are also tests that must be performed.
What is Tektronix’ role in 3D sensing?
The Keithley division at Tektronix has been working with customers to test various types of light- emitting diodes for more than 20 years. Building on this experience, our engineers have perfected the signal quality and speed requirements for this type of testing through innovations such as TSP and pulsed current SourceMeter SMU instruments. In fact, we’re so entrenched in this industry that customers don’t call the instruments an SMU or Source Meter, they call them the “Keithley.”
What challenges are developers of 3D sensors facing?
One of the bigger challenges is improving manufacturing yield. Commonly used methods for testing laser diodes are slow and can cause good parts to be thrown out or sold as seconds. A better test setup can improve accuracy, yield, throughput and profits.
The main test of a laser diode is the LIV curve, as I mentioned earlier, that simultaneously measures a device’s electrical and optical output power characteristics, ideally early in the process to sort laser diodes or weed out bad devices. In a LIV test, the device under test is hit with a current sweep as the forward voltage drop is recorded along with optical power.
This test is best done in a pulsed fashion early in production before the laser diode is assembled into a module. For diodes still on the wafer or in a bar, pulse testing is essential because, at that point, the devices have no temperature control circuitry. Testing with DC would, at the very least, change their characteristics and at worst, it would destroy them. Later, in production, when they've been assembled into modules with temperature controls, the devices can be DC tested and the results compared to the pulse test. To make matters worse, some devices will pass a DC test and fail a pulsed test. Through close attention to test setup and configuration, test accuracy can be significantly improved at every stage of the process, which ultimately leads to improved yield.
How important is measuring laser diode optical power and how is that accomplished?
This is in fact an important factor. Regardless of the type of laser diode, eventually its optical power does need to be measured in the production environment and, of course, the more accurately the better.
In our experience working with customers around the world, the preferred method for measuring optical power in a production environment is to use an integrating sphere and detector connected to an optical power meter or optical spectrum analyzer and an SMU or DMM. An integrating sphere is essentially a spherical cavity with a diffuse reflective interior surface that’s designed to distribute the optical power from a radiant source uniformly over its interior. A detector on the inside of the sphere therefore receives uniform irradiance allowing for consistent and repeatable measurements
In terms of optical power meters, these consist of either a specialized optical power meter or an electrometer enhanced with special features for photonics applications. In most cases, an electrometer, which is essentially a high refined DC multimeter, can be substituted for an optical power meter and is the more affordable and versatile option.
What about the optical detector inside the integrating sphere? What’s the best type for this application?
There are three common detector materials: silicon (Si), germanium (Ge) and indium gallium arsenide (InGaAs). Each has its advantages and disadvantages. At wavelengths less than about 800nm, silicon is the only choice. But many applications use wavelengths between 1300nm and 1700nm, where InGaAs are best, because the response is uniform and holds up well to about 1700nm. However, InGaAs has a problem with pulse response. An InGaAs detector does not "settle," even within a 10µs pulse. If the pulse width were decreased to 1µs, the problem would be even worse. Ge does not suffer from this effect, so it is preferable for applications that involve short pulses.
What can be done to reduce the cost of testing 3D sensors?
The historical approach has been to use a “home brew” pulsed LIV testing solution. This is now changing by using a pulsed source-measure unit to drive the laser diode, while providing accurate measurement of the source current. The engineer then can use either a precision DMM or picoammeter to measure the photo detector more cost effectively. Tight synchronization is required between the source and detect side of the device for accurate LIV results. For high production environments in which multiple devices are being tested the system can incorporate multi-channel data acquisition for additional cost savings.
With this new solution, the sequencing of the LIV sweep can be automated to provide highly synchronized and accurate results. This leads to higher throughput and yields.