As part of the sensor suite for autonomous vehicles, lidar range maps play an important role in object identification in the surrounding environment, alongside car radar and other sensors or imaging systems. Building a functional driver circuit with small form factor and sleek packaging is critical for enabling lidar imaging/rangefinding around a self-driving vehicle.
These same circuits can be adapted to other lidar applications, such as atmospheric monitoring, pollution plume tracking, turbulence measurements in aircraft, and other precise measurements. The primary factors that determine the usefulness of your particular lidar system are the power output, pulse time, and repetition rate. If you design the right driver circuit, or properly adapt your diode to a driving IC, you can ensure that your lidar system will operate with high resolution and range.
Driving a Pulsed Laser Diode - The Transmit Side
Pulsed laser diodes are driven with high voltage, low duty cycle PWM pulses (usually ~1% duty cycle at hundreds of kHz) in order to reach 100 ns or faster pulse widths. Driving a pulsed laser diode with a smaller rise time provides higher resolution images and allows for a faster scanning rate. The short rise times required in driver ICs and in custom circuits requires use of GaAs devices for longer pulses, while GaN is the best choice for shorter pulses.
If you’re designing your own driver circuit, the critical components are the FET driver and transmit amplifier stages. The signal to drive the pulsed laser diode is initially amplified with an FET driver, which then switches on a high current FET transimpedance amplifier with high gain to deliver the required drive current. A block diagram of this circuit is shown below.
Pulsed laser diode driver circuit block diagram
Impeccable impedance matching and controlled impedance routing is required between all components to suppress reflections as these lead to periodic fluctuations in the output power. This applies on the transmit and receive sides of the device. Both sides of the signal chain are normally integrated onto the same board along with the optics to ensure everything fits in a compact package. Note that impedance matching is critically important with a prepacked driver IC, and precise impedance matching with the laser diode is required.
However you choose to drive your pulsed laser diode, you’ll need to ensure the jitter in the output is extremely low. This is critical because, when you’re working with signals that travel at the speed of light, jitter of 1 ns equates to a distance error of 30 cm. You’ll need to bring that jitter down by a factor ~10 to ensure accurate distance measurements.
This requires sufficient decoupling on the power rails, which is typically supplied with a large capacitor across the output from the laser firing circuit. Since you’ll be working with a PWM signal with very fast transition times, you should also take advantage of interplane capacitance in your stackup to help damp any ringing at the power pin of the driver. Note that PDN noise typically induces jitter of ~1 ps/mV in all circuits to which it is connected.
The Receive Side
On the receive side, the reflected/scattered lidar pulse is received with a photodiode array or other detector, and the received signal is used for a time-of-flight measurement, which can easily be performed with time-to-digital converter ICs. Afterwards, the received signal at each emission angle is sent to an ADC and is used to build a depth map from the time-of-flight measurements.
Because jitter adds in quadrature, you’ll need to remove jitter before the amplification stages on the transmit and receive sides. Pulsed laser diode driver ICs typically include a fractional PLL that converts a reference clock to match the scanning rate in the system. This converted clock signal is then used on the receive side for time-of-flight measurements and for serializing output data from the ADC.
Note that we’ve focused on pulsed laser diode driver circuit layout here, but a continuous wave (CW) laser can be operated as a pulsed laser. However, if you intend to drive a CW laser diode as a pulsed laser diode, you should gather an an autocorrelation measurement to determine the pulse length with, which is difficult without sensitive optical equipment and a fine mechanical delay stage.
Finally, the output power from a laser diode and the sensitivity of the detector on the receive side are highly sensitive to temperature. In general, the efficiency of a laser diode and sensitivity of the detector both decrease at higher temperature. A temperature increase of both components is inevitable during operation, requiring a creative thermal management strategy. This might include a small cooling fan, but in my opinion, a better choice is to use a high thermal conductivity substrate and try to dissipate heat to the enclosure as this uses fewer moving parts.
Use Your MCAD Tools
Any optical system carries precise mechanical tolerances, and pulsed laser diodes for lidar systems are no different. Lidar systems for autonomous vehicles will need to rotate around the entirety of a vehicle to provide depth images of the entire surrounding environment. Other systems can remain static, but they still require precise positioning with respect to any other optical components in these systems. During the layout phase, you should use your MCAD tools to verify precise positioning on the board and in your enclosure.
With the powerful PCB design and analysis tools in Altium Designer®, you’ll have a complete electromechanical design solution that is ideal for building optical and optoelectronic systems. You’ll also have access to signal integrity tools that can help you address impedance matching between your pulsed laser diode driver circuit and your diode.
About the AuthorMore Content by Zachariah Peterson