Hello all, a little background on me. I'm in my 30's and have worked in the field for about 12 years now. I eventually hit that "wall" and realized that unless I went to school I'd always be a technician no matter where I went. So, a few years ago I started my bachelors in EE part-time. The more work experience I gain the more I hear "analog is dead, everything will be digital eventually. There isn't really a need to spend a lot of time learning about op-amps, FETs, etc"

Now I've done some embedded software, and I've also done some "lightweight" analog circuits before. I agree that digital is orders of magnitude easier in a lot of cases, but at the end of the day won't you always need some sort of transducer to generate a signal for processing? Wouldn't this device have to rely on an analog circuit of some sort? This is just one example, but I think I've made my point.

To sum it all up, I like programming but I'm REALLY interested in mastering the analog side of things. At the end of the day...is it worth it? Thoughts?

 

If it's what you are interested in, then pursue it.

As you noted, digital circuits are of limited use unless they can connect with the analog world -- and that requires analog electronics in most cases.

The "analog is dead" mantra is just that. It is a sound-byte that makes a point regarding a trend and is NOT to be taken literally.

 

The input interface with the physical world is almost always analog, and the outputs usually have to do some analog functions to deliver their information. A box full of ones and zeros can do very important things, including displaying the results, but there will always be a need for analog circuits to collect real world information like Ph, sunlight, and weight. Sometimes an analog method can be much simpler than a digital method. For instance, think about how would you write the code for detecting sunset when car headlights are passing by at random times. In analog, it's just an RC time constant being integrated across several minutes. Speaking metaphorically, that's about 2 lines of code in analog.

 

Don't forget that digital is analog, perhaps only using about 10% of the analog range.
When you bump into mysterious problems in the digital domain, your knowledge of analog is what is going save your bacon.

Sorry, analog is not dead. That is just a catchy sound bite that has no merit as @WBahn says.

 

The real world is analog. But be warned - if you get good at analog you'll be the guy who has to do all the demanding and difficult stuff that is hidden away while the digital types get to do all the showy bells and whistles stuff.

There are unquestionably truly amazing things done with melding of digital circuitry with analog - just look at the development of modems, for example. Inexpensive computing power has put digital circuitry into things that used to be entirely in the analog realm. It can reduce parts counts and cost. It can often make it easier to cope with "Yes, that's what I asked for but that isn't what I need" problems. It can allow hacks who don't really know what they are doing to produce things that are at least reasonably functional.

One thing I recommend to anyone interested in analog is to read everything that the late Jim Williams, most recently with Linear Technology, ever wrote. The ap notes he produced were not just terse "here's how you use this part" things but real discussions of evaluation of problems, approaches to solutions and how to get stuff actually done in the lab.

 

No. Analog ain't dead -- yet.

But for many applications, there is little one can't do digitally that used to require substantial analog front-ends. Even radio to the GHz range.

And I am considerably impressed with modern class D amps that can, with high fidelity and power, drive a set of speakers without an output filter.

What I want is a set of digital speakers that require neither a voice coil or a cone. Think: efficiently drive the air (with fidelity) directly with electrons.

Now that would be neat.