I had considered making the transition but didn't pursue it. From a technical perspective, it's possible. The problem is that if you're paid as a senior digital engineer, you need to start from the bottom as a junior analog engineer -- no one wants to pay you at a senior level. There are some logical transition paths: do mixed signal verification or digital calibration circuits. The bigger question is if you actually want to do this. (I didn't -- while interesting, the semi sector sucks as an employee!)
By the way, having a digital background could be useful. Analog designers tend to have no ability to program or script which could make a designer much more productive. And, they could expand the solution domain to include small calibration circuits which are trivial for digital designers to implement.
@@k@sh: "Other option is start from digital hoping for switch later" - analog engineers, is that possible that you could start with digital and morph into analog later?
It happened that way in my career to some extent. I was never hoping for the transition, but circumstances have moved me into more analog design. I would certainly not put down "analog engineer" on any job application, but I have seen some success on small analog projects and often have this overlap between analog and digital.
Typically I am an analog app-note designer. I start with an app note and move on from there.
Perhaps our friend here is fortunate enough to not have experienced what many others have.
Here in Canada I have worked for BIG internationals and small start ups. The BIG organizations keep you in the "pigeon" hole by force, implied or direct. Any attempt to cross train (your own time or company time) is considered insubordination and a sign of unrest. These are the same companies that pay lip service to the language of, "training", "professional development", and another big one is, "think outside of the box", but reality is very different. What they really mean is, keep doing your work; don't think so far outside of the box that you burst the wall of the box; you can push a little, but only VERY little on that wall. Movement within the organization or suggesting small reshuffling of responsibilities amongst your colleagues is also highly dangerous. Don't think, don't ask questions, just stay put (in your box) and deliver what is needed. NO discussion.
Secondly, if you are over about 35 years of age, you are finished. By that time you have either been over worked, and thus burned out, or, you are now perceived as a threat because you have energy left and they have discovered that you actually can think for yourself (not being a blind company man) and thus you do understand how the company (system) works. In either case you are no longer malleable (as a freshly minted graduate would be) and original thought is deemed dangerous to the company.
The small companies do want you to conduct, "professional development", but not on their time, and not in the portion of the day that they demand as overtime.
In either case, learning new ways of working (developing yourself) must be done quietly, on your own time.
Then comes the issue of applying what you have learned once you have gained some proficiency. How do you show on a resume legitimate work experience if you can't apply it where you presently work? The ONLY thing is to start your own business on the side. Lets face it, how often do you tell yourself that you may not develop professionally and apply it when it benefits your own bottom line?
So, with regard to the the question of too few or too many analog engineers; my experience is, plenty are willing, but only the young malleable few a tolerated for a while.
@kfield: Really speaking even I am finding answer to that question. But better to start career in something instead waiting for exact match for analog "transistor level design". One poosible field to later morph into I see is high-speed interfaces, memory control circuitry, etc.
Sometimes pool may appear smaller when analog enthusiast in pool ignored due to lack of experience.
One way to get experience is PhD, but then research guy finds professorship to be much better option than industry where his design delivery time will always be compared to faster design cycle in digital
Other option is start from digital hoping for switch later. But once in digital domain, they get attached to aura of working on leading edge node & ease of autmated design.
Agreed, re: reuse of IP and no one designing a new RF chip. Look at last week's announcement of Broadcom planning to exit the baseband industry. That means fewer RF and mixed-signal jobs in the industry.
At the last company I was with the analog designers were afraid of a "super" A/D that just samples the antenna and does everything else with DSP. I'm not sure how feasible it is power-wise, but people definitely want it, since digital scales with Moore's law and analog doesn't.
There will be some hotspots where people make lots of money, but nothing to make a 30 year career out of.
I don't see lot of design work in Analog. Most of it now is modifications. Plus the analog IP reduce the need of too many analog designers. No body designs a new RF chip now. Only when a new standard comes. But getting good analog designers is difficult. The real desiign engineers. Otherwise there are many.
My last supervisor is a world class power supply designer (along with lots of basic analog knowledge). 5 years ago (at age 67) he left our place to retire. Since then he has been working two jobs, making far more than he ever made with us and loving every minute. He could have as many projects as he can handle.
@kcarreon You're in an ideal situation, suck up as much knowledge as you can from those experienced analog designers and keep in mind the guys who can make $50 to $250k doing a design in advanced materials. And I think "artisinal circuits" are the next big trend.
A Book For All Reasons Bernard Cole1 Comment Robert Oshana's recent book "Software Engineering for Embedded Systems (Newnes/Elsevier)," written and edited with Mark Kraeling, is a 'book for all reasons.' At almost 1,200 pages, it ...