2024-04-15

Christian Elgaard

Title of thesis: Integrated millimeter Wave CMOS Power Amplifiers for 5G Systems.

Link to thesis in Lund University Research Portal.

Defence: Friday April 19th, 09:15, room E:1406.
Zoom link.  Zoom ID: 63926548525.

Describe your research in a popular science way

This dissertation focuses on millimeter-wave power amplifiers built using a common and inexpensive silicon-based technology known as CMOS, for fifth-generation mobile systems and beyond. It comprises four scientific publications based on three measured power amplifiers with increasing complexity, where the third one essentially includes a complete transmitter and parts of a receiver. The third circuit is measured with a signal where the power amplifier sends a whopping 9.6 Gbit/s, equivalent to downloading about an hour's worth of video in just one second. To achieve high-output signal transmission at such high data rates without distorting it to the point of making it difficult or even impossible for the receiver to decode the digital bits while consuming as little power as possible, a Doherty amplifier combined with adaptive bias is used. The Doherty amplifier, invented in 1936, has the special property of consuming very little power when amplifying signals with highly varying amplitudes, a characteristic of 5G (and 6G) signals. To save power, the Doherty amplifier has two amplifiers cooperating: one is on all the time (the main amplifier), while the other (the auxiliary amplifier) is only on at high amplitudes, reducing consumption. However, constructing a Doherty amplifier at such high frequencies as millimeter waves poses a significant challenge, making it interesting from a research perspective. To ensure that the transistors amplify the signal in the desired way, it's necessary to set an appropriate operating point, known as biasing the transistors. Normally, a constant operating point is used, but with adaptive bias, the transistors' operating points are adjusted as the signal's amplitude changes. As demonstrated by the research in the dissertation, using a Doherty amplifier with adaptive bias for the auxiliary amplifier is a good way to mitigate the problems that arise when a power amplifier needs to handle the complex signals of 5G and future mobile systems in a power-efficient manner. The fourth article examines and explains the theoretical aspects of how an adaptive bias signal should be designed to work optimally with the auxiliary amplifier, and the article also contains a detailed description of the construction of a circuit capable of creating such an adaptive bias signal and thus able to change the operating point for the auxiliary amplifier very quickly, i.e., with high bandwidth. The circuit creates the adaptive bias signal by first extracting amplitude information from the Doherty amplifier's input signal. An important theoretical result is that an ideal adaptive bias signal should then be constructed through a nonlinear transfer function from the amplitude information. Measurements and simulations show that the circuit effectively achieves this.

The figure shows a die photo of the power amplifier (PA) described in the first paper included in the dissertation. From right to left showing the pre power amplifier (PPA), PPA output resonator, output stage (partly placed under GND bump), and output combiner. The occupied area is 0.144 mm2.

What made you want to pursue a PhD?

Prior to choosing to become a PhD candidate at EIT I held a research position, within the same field and with tight cooperation with my supervisor Henrik Sjöland, at Ericsson research. When the opportunity came to continue the research that I was already conducting at Ericsson, in the form of an industrial PhD linked to EIT, it was an easy decision to accept.

 
What is the most fascinating or interesting with your thesis subject?

The research and development investments required to launch a global functional cellular network are arguably among the most complex, costly, and far-reaching achievements of mankind. From the complexity and capability of a single transceiver circuit to the complexity of the entire cellular system, one must not overlook the challenges associated with ensuring affordability to enable global widespread access, allowing mobile phones and associated subscriptions to function practically in all corners of the world. At the center of this is the power amplifier and how its limitations are shaping the capacity of the cellular system, and thereby presenting excellent research opportunities.

Do you believe some results from your research will be applied in practice eventually? And if so, how / how?

Yes, I have already seen other published papers in the same field that has used some of the outcome of my research to achieve better performance. I think it is very likely that the proposed idea of using the adaptive bias signal on the auxiliary amplifier of a Doherty amplifier will be used in hardware in both cellular infrastructure and handheld devices. 

What are your plans?

My plan is to continue my current position as Master researcher at Integrated Radio and Systems, RF Frontend and Power Amplifiers at Ericsson Research in Lund and hopefully with some form of cooperation with EIT.