Viewpoint: From BC108 to SiGe BiCMOS – why Analog ICs are great

It all started when I was 13 years old at school. A class mate was an Electronics hobbyist and during break, he talked animatedly about what he was going to build over the weekend. My curiosity piqued, I asked him what magazine he was reading. Within a few days, I purchased the magazine, started reading avidly and discovered the world of Bipolar transistors, namely the BC108!

I saw the following schematic for a two stage amplifier using the BC108:

I didn’t have a clue how it worked or what it did, I just built it using the component values in the magazine. I played around with the amplifier first by connecting a microphone to its input and the output to a crystal ear piece. To my delight, I heard audio and drove my family round the bend walking around with a preamp and a microphone trying to listen to conversation or bird song. Next I connected a tuned circuit to the input of T1 via the ac cap, and I heard MW radio! I was excited with my humble achievements and shortly thereafter, I fell in love with Electronics, which became my day job a decade later.

On a self deprecation note, back at school I was totally rubbish at Maths. I remember my brother saying to me, that Electronics was quite Mathematical, and I’d have to pull up my socks up otherwise I would never realise my dream of becoming an Electronics engineer. Sadly, I failed Maths at O-level and hated the subject at school. I couldn’t see the point of factorisation, or roots of the quadratic or simultaneous equations. However, it was painfully clear that I needed A-level Maths for an entry to BSc course in Electronics. With belt and braces on, I was determined to study A-level Maths and somehow I convinced the college of letting me attend A level Maths classes with no O-level Maths. Looking back I have to say, the teacher I spoke to felt sorry for me and probably thought she’d humour me and say yes.

She did say that if I didn’t pass the first test at A-level Maths, I would have to drop out of A-level for a year and pass my O-level. I agreed to her condition and on the evening before my first day at A-level Maths, I thought I’d study and polish my poor Mathematical skills. I borrowed a Maths book from my brother and opened the first chapter on Pascal’s triangle.

I remember looking at the pattern and could see how the coefficients were derived. But to how to use them? The book I was reading then went on to the following:
(x+1)0=1
(x+1)1=x+1
(x+1)2=x2+2x+1
(x+1)3=x3+3x2+3x+1
So the coefficients from Pascal’s triangle could be used in the order of decreasing powers of x! The next day, amazingly, we studied Pascal’s triangle in class. I was prepared and when the teacher randomly asked me a question, I answered correctly. After a few weeks, the whole class had to sit a test and to my delight I passed, but the most important lesson I learnt was I could actually enjoy Maths and play around with Algebra in the same way I loved making BC108 circuits. Call me a nerd if you like and I suppose I am, I guess.

Fast-forwarding to my first day at Ferranti Microelectronics, Hollinwood, Oldham in 1986, I met my boss Dave Brotton, whom I am still in touch with today. He came to pick me up from reception, took me round to the office where I was going to be based and after the introductions, we got stuck into FAB 3 with a 2um feature size Bipolar transistor process and a transition frequency of 3GHz. Back in those days, an Ft of 3GHz was nearly the state of the art. Dave got me pouring over some layouts of a testchip he had recently completed.

My romance with the BC108 faded for a few years, until I started to really learn circuit design, by talking to designers, picking their brains and reading tonnes of papers. I still remember sitting there with a pencil and paper trying to analyse simple Bipolar circuits, until finally, the penny dropped. Everything started to become clear about how these wonderful circuits worked. As my career progressed and I moved to Plessey Semiconductors, Ft moved up to 15GHz, some 6 years later. At this time, I met the wonderful Nick Cowley, whom I am also in touch with to this day. Nick gave me one of my greatest design challenges….to achieve a highly linear mixer, while keeping the NF below 10dB. Back then just before SiGe, such concepts were the domain of GaAs. Many Bipolar designers thought it impossible to achieve IIP3>+15dBm and IIP2>+40dBm, because they felt the nonlinearity of the resistors would dominate.

The first thing I looked at was the output current of an amplifier with cubic distortion (click to expand):

To expand the above, we have to resort to….PASCAL’s TRIANGLE!!!! My God I thought, I have come a full circle and all the Maths I did at A level and degree level became useful. Another very important lesson in life for me.

After pages of circuit analysis with heavy Algebra, I discovered something I still use today… a means of designing highly linear circuits on Bipolar. Once the cat was out of the bag, there was a buzz at Plessey’s. Everyone including Nick was excited at the prospect of beating GaAs. Sure enough a few years later, after several silicon design iterations, we introduced the SL2030. I have captured a screen shot of the device below (click to expand).

We did it! We could compete head on with GaAs and achieve the high IIP3, and SSB NF numbers. Times moved on and challenges changed and we started working with IBM Microelectronics SiGe BiCMOS, 5HP process. Those were the days of Jim Dunn, who headed up the SiGe processing division at IBM Microelectronics, Burlington, Vermont. I had the pleasure and honour of meeting the great man. We talked about 5HP and 7HP and the use of Carbon doping in the base to improve the fmax of the transistor.

Today some twenty years later, Bipolar transistors are my daily bread and butter, although we do use CMOS as well. The ft of the transistors I am working with today, are in the regions of 250GHz, a factor of x83 greater than FAB3. I collaborate closely with Jack Pekarik, SiGe Process Manager at Global Foundries, Burlington. Jack was hired by Jim Dunn in Burlington and anything Jack doesn’t know about Bipolar transistors, is simply not worth knowing. Jack says, “Recent results show that SiGe has plenty of petrol in the tank to surpass GaAs and compete with InP HBT performance enabling efficient, high data-rate communications systems.”

Thank you kindly Jack! It’s a real honour for me to get such a wonderful quote from a brilliant world leading Silicon processing scientist.

Returning to my beloved home country , the United Kingdom, and looking to the future of British Integrated Circuit engineering, I asked Harriet Green to comment. She came back with the following…

“Why wouldn’t you go into IC design in the U.K.?”

In a world where technology is ubiquitous, businesses are adapting and learning how to integrate technology into their processes. As a result, there has been an unprecedented growth in the global demand for skilled Analog / Mixed Signal and RF IC Design Engineers. Without these engineers, technological progress would stall. From the moment we wake up, to the moment we go to sleep, we use Analog technology in every aspect of our lives.

Here are some reasons why Analog IC is great:

1. Work on cutting edge technology and make an impact on the world we live in today
The Analog IC industry can offer a career working some of the most fascinating and exciting technologies, from 5G, Aerospace tech, Smartphones, Tablet PCs, Portable media devices and many more.

2. Analog IC offers highly competitive salaries and working arrangements
We have seen that with this growth in demand, companies have been offering engineers better basic salaries, better bonus packages and more flexibility with working remotely.

3. Analog is the future
2021 saw governments realise something that everyone in the semis world has known for years: SEMICONDUCTORS RUN THE WORLD. As a result, this has led to governments in the UK and Europe recognising the value of a semiconductor engineer and how fundamental they are to their economies.”

Thank you kindly Harriet! As always Harriet’s insights are not only very meaningful, she is an inspiration for future leaders world over. Again, I feel honoured to receive such a quote from an Industry giant.

My romance with the BC108 has come a full circle. I feel very privileged to continue to contribute in the field of Advanced Analogue Integrated Circuit design and each time I wire up a Bipolar transistor on a CAD tool, I remember it all started with the BC108. I think if there is passion, there is a will to learn.

Author

A veteran of 37 years in the Analog Integrated Circuit Design industry, Ash Madni continues to
contribute to the rapidly changing world of Microelectronics.

Ash has worked for a number of big companies, such as Ferranti, STL, GEC Plessey Semiconductors, Maxim Integrated, Dialog Semiconductors, to name a few. He has also worked in startups, such as Phyworks.

During his tenure at GEC Plessey Semiconductors, Ash raised over 20 patents, including Synthesis Exploiting Algebraic Design, which was used successfully in a number of products. Ash is currently running his consultancy, Madni Technologies (UK) ltd.

For relaxation, Ash enjoys walking, spending time with his wife, daughters, grandchildren and friends. Ash is an active composer in Chamber and Orchestral music. Inspired by the works of Shelley and Stoker, Ash is working on his first SciFi novel Transmutation – the rise of Hemonra. Passionate about the Semiconductor industry, Ash is an active blogger for the Electronics Weekly website.

See also: Viewpoint: The pressures on Integrated Circuit designers