Faculty of Engineering, LTH

Denna sida på svenska This page in English

Designing AD-converter circuits in CMOS for base stations


Researcher Siyu Tan observing a CMOS chip in EIT lab using a microscope.

Siyu Tan is observing a CMOS chip in EIT lab. Photo: Siyu Tan.

The emerging 5G network enables reliable connections with significantly reduced latency at an increased transmission rate. The increase in performance requires improvements in all parts of the communication chain. Siyu Tan’s PhD Thesis focuses on converting analog radio signals to digital data on the base station side of the wireless link.

The thesis will be defended on 18 December 2020

The thesis for download [PDF,].

The advancement of wireless communication systems requires high-performance Analog to Digital Converters (ADC), an essential component in a modern high-performance radio transceiver device. The main focus of Siyu Tan’s research is designing high-performance ADCs on integrated circuits. This research analyzes various ADC architectures and their component-level implementations in advanced CMOS technologies and aims to find optimized high-speed analog-to-digital conversion designs.  This research tries to understand the benefits and limitations of specific ADC designs and evaluate the possibilities of integrating them inside the advanced base station devices in the future.

From fixing electronics to designing high-performance AD-converters

When Siyu Tan was a child, he enjoyed collecting non-functioning electronic devices from his friends and trying to fix them. He began to understand what components are used inside a radio, why they are designed like that, and what is the magic behind modulation, demodulation, etc.

“After I grew up, I feel more and more interested in electronic design. Instead of learning the existing circuits, I tried to build my own circuits based on the knowledge and experiences I acquired. Thanks to the growth of the Internet, it became easier for me to understand the challenging circuits by surfing online. It is truly an unforgettable experience,” he says.

“I completed my Master’s study at Lund University, which was a wise choice I made. I’ve learned a lot and gained a huge amount of supports from professors and my supervisors, who have been very knowledgable and patient whenever I’ve faced a challenge. I hope my research outcomes will create a small contribution to the state-of-the-art science world.” 

After you have measured the chip, you will know if your idea works

Siyu explains that the chip circuit design usually takes a very long time from the design phase to the actual measurement. Unlike the circuit-board level design, the internals of a chip is hard to see, and any problems inside the chip become difficult to debug. Therefore, it requires extensive knowledge of circuits and expertise in practices. 

“In my opinion, the whole process from a CMOS circuit design to chip measurement is the most exciting part of this research work, which is also the main topic of my thesis. As a main designer, I am in charge of designing the circuit using computer-aided design (CAD) tools and preparing the chip measurement. I have to carefully think about what to design, why it is beneficial, and its performance compared with the current designs,” he says. 

“Various of small circuit blocks are internally connected in a large circuit and system. Sometimes, a few adjustments here and there make huge differences to the results. The measurements are also challenging. Even though the designer already simulated the circuits extensively before the chip manufacturing, there could still be uncertainties in the actual chip measurements. Will the chip to be tested meet the target specifications? Is it sensitive to the fabrication process, supply voltage, and operating temperature? We will know the answers in the lab.”

Industrial collaborations

“My project is highly related to the industry. I already got the chance to work in Ericsson Research in Lund, together with experienced researchers, to design ADC chips planned to be used in current 5G or even future 6G communication devices,” Siyu says. 

“I learned a great amount of knowledge of the standard circuit design procedures from the industry, and I also got a chance to use a well-equipped lab to measure one of the prototype ADC chips. The experiences I acquired from the industry inspired me during my PhD studies, and also changed my way of thinking for my current and future works.”

“An improved chip I designed is based on an earlier prototype ADC chip designed by experienced researchers in Ericsson Research in Lund, which has been fabricated but not yet measured. This new ADC chip was designed based on industry guidelines, which could be a prototype to be integrated into high-performance communication devices in the future.”

What are your plans?

 “I plan to continue my research career at the EIT-department at Lund University. I will join a joint research work between Ericsson Research in Lund, and the university. The main goal is to improve my current ADC design further and achieve a higher performance than the current design. I will measure the earlier fabricated chip as soon as possible, then learn from the results and improve the design by adding or replacing a few crucial blocks for a satisfying result,” Siyu concludes.