Lidar is a form of radar that generates electromagnetic waves across a broad spectrum.

Over the past several years, Time-of-Flight (ToF) distance measurement, a specific type of lidar, has been widely adopted. When a laser is used as the optical source, it enables distance measurement even to small sections at far distances. When used with adjustable optics, it can measure the distance to a specific point and implement a three-dimensional map of the object.

EPC has developed the EPC9126 and EPC9126HC laser drivers to provide the performance benefits of eGaN FETs to lidar systems and enhance the technology level of lidar. Although the basic principle of this laser driver appears simple, parasitic components that many engineers overlook in design give it high-speed, high-voltage, and high-current characteristics.

This technical article was written to provide in-depth information on laser driver design so that users can maximize the benefits of the driver and to answer commonly raised questions.



Basic Design Process
We can now utilize the information covered in the previous article to perform a resonant lidar driver design. Let us start with some laser pulse specifications typically required in system design.

The requirements are as follows:

▲Pulse peak amplitude I_^DLpk
▲FWHM (Full Width Half Maximum) pulse width t_w
▲Pulse repetition frequency PRF
▲Laser diode forward voltage drop V_DLF

Once the basic pulse requirements are selected, the next element needed to complete the design is the power loop inductance L₁. The determination of L₁ was discussed in the previous article, but for now let us assume we have a good estimate value.

Using equations (4) and (6) to determine the value of the resonant capacitor C1, we obtain the following:

Finally, the required bus voltage V_IN is determined through equation (7), and I_DLpk is also used together in selecting an appropriate part number for FET Q₁. Through equations (7), (8), and (9), the remaining values needed to complete the design are determined.


Determining Power Loop Inductance
We have already seen that the required input voltage increases nearly linearly with respect to the laser inductance, and the input voltage determines the ratings of the FET and capacitor.

The laser driver bus voltage also comes from elsewhere, most likely from another boost converter in the circuit. The lower L₁ is, the simpler and more cost-effective the remaining design becomes.

The key principles for minimizing PCB inductance are detailed in reference¹⁾, which introduces several useful techniques for reducing layout inductance.

This material shows that with chip-scale packages of eGaN FETs, the contribution to the power loop inductance of the FET, PCB, bus capacitance, and current sense shunt (if needed) can be maintained below 1nH, and values approaching less than 500pH are achievable.

The parasitic inductance of the EPC9126xx is approximately 1nH depending on the FET and load mounting location. This value is higher than the best achievable value, particularly because it is a compromise to increase design flexibility to accommodate different laser packages.

Let us now examine other inductance sources.

One of the major sources is the laser itself. As noted earlier, through-hole lasers are expected to contribute approximately 5nH in the best case, but often much higher. Surface-mount lasers contribute approximately 1-3nH, meaning the laser becomes the dominant inductance source.

Most of the laser inductance originates from the laser package, including wire bonding. Laser manufacturers are well aware that laser package inductance can overwhelm performance, so advancements in this area are expected in the near future.

Unfortunately, the value of L₁ is difficult to know precisely early in the design phase. Therefore, some iterative work may be necessary. By considering additional voltage margin for the FET in the initial design, some of the unanticipated inductance can be overcome.


EPC9126xx Hardware Driver Design
A typical connection diagram for the EPC9126xx laser driver is shown in [Figure 8].

A complete description of the connections and operation can be found in the Quick Start Guide²⁾³⁾, and here we will briefly review this diagram.

All signal I/O use SMA connectors. This design includes voltage test points obtained from J3, J7, J9, J10 with embedded transmission line probes. The output of the current measurement shunt can be accessed via J6.

Design details, including Gerber layout files and complete schematics, can be found in references⁴⁾⁵⁾. The layout was designed to minimize total inductance according to the principles described in reference¹⁾.

A photograph of the driver and enlarged views of key portions of the design are shown in [Figure 9].

To minimize inductance L1, the energy storage capacitor C1 (C11, C12, C13, C14, C15 on the PCB) and current measurement shunt (R12, R13, R14, R15, R16) are composed of five 0402-size surface-mount packages connected in parallel.

The spacing between the top plane and ground plane is 250μm (10mil) to minimize inductance. Blind and buried vias and microvias were not used to minimize cost.

To achieve excellent shunt performance for current measurement, the shunt resistor had to be mounted upside down, reducing the shunt equivalent series inductance from 200pH to 40pH, resulting in a four-fold increase in shunt bandwidth.⁶⁾ Additional discussion on this will be detailed in the current sensing section.


Experimental Results
Both the EPC9126 and EPC9126HC were tested using the Excelitas TPGAD1S09H surface-mount laser, currently the lowest inductance surface-mount high-output pulse laser readily available at the time of writing this article. In each case, circuit testing yielded an input voltage of 75V.


EPC9126
The power loop inductance L₁ was estimated at 2.3nH. With a design of I_DLpk=35A and 3.5ns pulse width, C₁=1.2nF and V_IN=60V. The capacitor value used was C₁=1.1nF, the closest standard part value available. NP0/C0G ceramic capacitors were used due to stable capacitance and low losses.

Test results for V_IN=75V are shown in [Figure 10]. Peak current, I_DLpk=35A, is reached at t_w=3.4ns. This corresponds to the maximum power input for the laser P_DLpk＞300W. The discrepancy between results and calculations occurs because of different capacitance values and inductance estimation errors, additional voltage drops across the shunt, and the fact that the laser diode forward voltage drop is not actually a fixed value.


EPC9126HC
The power loop inductance L₁ was estimated at 2.0nH, reduced due to the larger footprint of the EPC2001C compared to the EPC2016C. With a design of I_DLpk=70A and 5ns pulse width, C₁=2.85nF and V_IN=78V. The capacitor value used was C₁=2.8nF, the closest standard part value available. NP0/C0G ceramic capacitors were used due to stable capacitance and low losses.

Test results for V_IN=75V are shown in [Figure 11]. Peak current, I_DLpk=63A, is reached at t_w=5.0ns. This corresponds to the maximum power input for the laser P_DLpk＞1300W. The discrepancy between results and calculations occurs because of slightly different capacitance values and inductance estimation errors, additional voltage drops across the shunt, and the fact that the laser diode forward voltage drop is not actually a fixed value.


Tips and Useful Information for the 9126
The EPC9126xx is highly flexible and can be used to try new ideas or gain detailed insight into the actual behavior of components for fast, high-current pulses. In this section, we will discuss design details and provide additional recommendations for different directions.

[Figure 12] is the block diagram of the EPC9126xx and provides a useful reference for this section.

In the final installment that follows, we will explore RF techniques for the high-speed performance of laser pulse drivers, specifically the use of controlled impedance and the standard 50Ω impedance for cables and measurements.


References
_{1) D. Reusch and J. Strydom, "Understanding the effect of PCB layout on circuit performance in a high frequency gallium nitride based point of load converter," 2013 Twenty-Eighth Annual IEEE Applied Power Electronics Conference and Exposition (APEC), 2013, pp. 649–655.
2) Efficient Power Conversion Corp., "EPC9126 Lidar Development Board Quick Start Guide, Rev. 2.5," 2016.
3) Efficient Power Conversion Corp., "EPC9126HC Lidar Development Board Quick Start Guide, Rev. 1.0," 2017.
4) EPC9126 Lidar Demo Board (https://epc-co.com/epc/Products/DemoBoards/EPC9126.aspx)
5) EPC9126HC Lidar Demo Board (https://epc-co.com/epc/Products/DemoBoards/EPC9126.aspx)
6) J. Glaser, " High Power Nanosecond Pulse Laser Driver using a GaN FET", PCIM Europe 2018 Proceedings, 2018.}

#Regarding this article, e4ds.com is conducting experimental discussions on GaN technology for lidar. If you have questions about this technology or wish to submit additional contributions, please leave a comment below.