Integrated analog: what’s missing?

Integrated analog: what’s missing?

Technology News |
By eeNews Europe

Analog functionality has been the stepchild of the electronic world ever since the digital revolution. The digital domain is getting all the attention, the mind share, and the development dollar share ever since Gordon Moore came up with his famous law.

Analog can’t brag about doubling functionality per surface area every 18 months. The smallest features are reserved for the high-performance processors from Intel, AMD, and the ARM licensees. The analog world is still largely working in the micron domain, while digital is at the deep sub-micron level.

Yet, there have been several revolutions in analog functionality since the advent of the digital age.

Here’s one example at work. RF devices, even high-performance devices, are now routinely integrated onto digital CMOS chips. Nobody thinks twice about a WiFi USB dongle, yet all the functions except for the antenna are now integrated on a single chip. Have you opened a DTV dongle? Most have two or even a single chip, handling all of the signal processing from the antenna to the USB interface.

Of course, the amount of actual analog electronics on these chips is reduced to a bare minimum: a low noise amplifier (sometimes), a mixer, a filter or two, and there goes the analog-to-digital converter(s) to put everything into 1s and 0s.

Another example is the replacement of much analog functionality with digital signal processors.

All high-end studio audio equipment is now digital. At the other end, even in extremely price-sensitive applications, digital is also replacing analog. A little while ago I was looking at the specs for a low-power FM transmitter of the type used to send music from your phone to a car radio. Even on these extremely price-sensitive applications, the old-fashioned analog multiplex generator is now digital. No analog filters, just a coupling capacitor at the input, another one at the antenna output, and there you go: The chip contains VCO, synthesizer, modulator, filters, and an I2C or SPI interface.

In my mundane comparison of day-to-day applications, I have also been able to replace much conventional analog circuitry with digital functionality. Many microcontrollers now come with what would have been considered high-performance analog-to-digital converters just a few years ago, and fast enough processors to do much of the previously analog dirty work in the digital domain.

While programmable analog chips from Cypress and Anadigm provide functionality that would be difficult or expensive to emulate completely digitally, the increased performance is not always needed, and the bulk of the market can be served with more conventional A/D then digital then D/A approaches. Of course, there are many potential pitfalls when replacing analog electronics with digital, and anyone thinking that it’s just easier is probably in for some surprises.

However, while there are specific issues, the overall difficulty and, more importantly, the development cost may not be very different, once we factor in the greater ease with which changes can be made further down the development cycle or, should I say it, in production. Then, production costs themselves are likely to be lower for products using digital technology as opposed to analog, because in most digital circuits fewer parts are subject to tight tolerances, reducing your BOM costs.

What has been your experience with programmable analog? We invite your comments and questions. Please join the conversation.

Didier Juges is principal engineer with Crane Aerospace & Electronics and this article first appeared on EE Times’ Planet Analog website.

Related links and articles:

The overt analog at ISSCC

The analog circuitry hidden at ISSCC

Terahertz CMOS debuts at ISSCC

If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :    eeNews on Google News


Linked Articles