Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

RF Engineer from China Give Advice

  1. Oct 17, 2011 #1
    Hello. Yangman here. I come here write an advice. Do not study digital field in EE. Every day I meet some digital circuit chump who call himself engineer because he understand AND NOR NAND NOR and verilog. Just because you can make LED blink on Arduino do not make you real engineer.

    I give an advice that you should study E&M and RF-circuit design. It will put a hair on your chest and make a real man of you, because the math is complex. And you will become a real engineer.

    Yangman
     
  2. jcsd
  3. Oct 17, 2011 #2

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    My chest must be well hairy!
     
  4. Oct 17, 2011 #3
    Do I have a twin in China? I am Yungman and I feel exactly the same way as you can see from all the posts I posted!!!!

    Digital electronics, embedded, FPGA and even firmware are dime a dozen. If I could be promoted to an engineer within two year I started learning electronics, successfully designed a micro processor control system using 8085 and wrote a lot of the application programs using assembly language. This cannot be hard and is dime a dozen!!!

    A lot of the EE student avoid EM at all cost, barely get through it and take the easy way out. There are just so many of those "hardware" engineer around.

    As speed goes up, everything is analog, RF and EM. That's why there is a new category of engineer called "Signal Integrity Engineer" in the last 10 years to HELP those digital engineer to fix their mistake on their layout. I know, I work like a signal integrity engineer for a while and even contracted with KLA to layout their dense CCD board.

    One more thing and that is very important. You think you can get away easy, problem is things change so fast in digital world, you have to constantly study and update your knowledge. Look at Firewire, it was so "in" 10 years ago. You waste time studying all the protocol, it's pretty much gone, USB took over. You spent time study up on one particular IC, then it get replaced. It is a constant thing. It is ok when you are young, wait until you get into 40s, have a family and your brain start lapsing!!! Constantly trying to keep up can be hard. Not to mention there are plenty of young digital engineer full of energy nipping at your behind to replace you!!!

    Analog and RF don't change, it is very hard to get in, but once you are in, things don't change nearly as fast. Particular important is to learn transistor design. From my experience, IC design is mainly transistor design. It is just on a wafer, you just need to learn the limitation of the design on wafer.
     
    Last edited: Oct 17, 2011
  5. Oct 17, 2011 #4

    jim hardy

    User Avatar
    Science Advisor
    Gold Member
    2016 Award

    ""Analog and RF don't change, it is very hard to get in, but once you are in, things don't change nearly as fast. Particular important is to learn transistor design. From my experience, IC design is mainly transistor design. It is just on a wafer, you just need to learn the limitation of the design on wafer. ""

    indeed, if one doesn't know what's just inside the pins of the 'chip' one will make mistakes interfacing to it.


    pay attention to detail. Beware of 'engineers' who want to speak only in terms of "big picture" - in my book that's impressionism which as Bouguereau said is "an excuse to paint not very well".
     
  6. Oct 17, 2011 #5

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    You are so right. No one seems to want to do any sums either. If they can't use an simulation, they don't want to discuss what a circuit will do. You can't burn your fingers on a simulated resistor.
     
  7. Oct 17, 2011 #6
    Until RF where Smith Chart become important, I never even use simulation, never learn PSpice. You really have to go through the circuit enough to get the feel of the circuits...to give you intuition and learn how to think as the circuit. Simulation take away all these.

    Even with RF, when I do simulation with Microwave Office, I always draw out my own Smith Chart first. Then I spend a lot of time varying components and get the feel of how the graph goes how each component value change the graph, spent hours on this.
     
    Last edited: Oct 17, 2011
  8. Oct 17, 2011 #7

    phinds

    User Avatar
    Gold Member
    2016 Award

    So I take it you don't want anyone to be designing computers or any other digital electronics, huh? Sounds like a really bad idea to me.
     
  9. Oct 17, 2011 #8
    I think you can learn designing digital on the fly. It don't not really deserve to be a 4 years degree. Button down and pay the piper, learn the math, EM and analog design so people will not come out ignorant about these. Believe me, I've seen enough engineers that don't know anything about these subjects. You learn digital stuff in like months, not years like RF, EM etc. And particular not that digital speed is getting so high and everything is RF.

    You seen digital engineers that are literally blind about real electronics. I sure seen a lot.

    You read digital books? They are really dumb down!!! Any Joe Blow can understand those books. Get a book on one of the major, spend 3 months going through the data sheet and application notes and programming development tools, you'll be the expert of that processor. Try learning PLL etc., you'll find you need to study 4 to 5 semesters of Calculus to get to LaPlace Transform and all to even get started. Ask Jim Hardy, he is very good with Laplace transform and closed loop control. He know RF from amateur radio, he can pickup PLL design in no time. Just has to learn the specifics of PLL like lockin range, pull in etc.

    Bottom line, if you pay the piper and learn the real thing, you can always learn those easy stuff. If you take the easy route, it will be very hard if not impossible to go back into the real electronics. I have been doing both, I don't think the digital, firmware, FPGA can hold candle to the real electronics. It took me 3 weeks to learn programming FPGA and become the chief FPGA designer!!!! Just like learning a computer language and the compiler!!! You learn one software language, you learn the second language in two weeks!!!!

    I have been using the learn as needed attitude all these years and honestly, towards the end, I took out all the experience in the digital design like USB and Firewire from my resume.

    For the young people, don't take the easy way out, do all the math, pay the piper. Once you are over a certain age, you just cannot remember as well and if you find you need it later, it is more than twice as hard to learn.
     
    Last edited: Oct 17, 2011
  10. Oct 18, 2011 #9

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    @yangman
    I can see you have been rattled by this bit it has to be said that many beginners in EE tend gravitate to the quick and easy way in, with elementary logic circuits and simulations. They then seem to think they 'know' all about EE.
     
  11. Oct 18, 2011 #10
    There is something to be said for following your interests rather than whatever path may be perceived as "hardest".

    And let's not forget that the greatest job opportunities are projected to be in software engineering, and you can learn java in 2 days.
     
    Last edited by a moderator: Oct 22, 2011
  12. Oct 18, 2011 #11

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    Learn it or be able to earn good money with it? Companies can employ youngsters on peanuts for simple programming jobs. If you're good then you'll succede at anthing of course but I wonder whether early success will mean good carreer prospects. High ability plus hard work are needed whatever you want to do. Analogue ability is extremely portable and futureproof.
     
    Last edited by a moderator: Oct 22, 2011
  13. Oct 18, 2011 #12

    jim hardy

    User Avatar
    Science Advisor
    Gold Member
    2016 Award

    i've been there.

    there is a valid point in this - get your math and physics under yor belt while you're young.
    Hardly anybody goes back to school for them in mid-life.
    As British mathematician G H Hardy observed, "Math is a young man's game".


     
  14. Oct 18, 2011 #13

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    Quote"When I was a boy of fourteen, my father was so ignorant I could hardly stand to have the old man around. But when I got to be twenty-one, I was astonished at how much the old man had learned in seven years. " Mark Twain
     
  15. Oct 18, 2011 #14

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    Politeness is a rare quality in some cultures.
     
  16. Oct 18, 2011 #15
    I am actually speaking from my life experience. I never have formal electronics education, all the analog electronics that I know were from self studying. In 1979, I attended Heald College for 3 semesters of digital electronics while I was working, I got good in digital and got promoted to become an engineer in 1980. Then I jointed LeCroy and designed a 8085 based control system and wrote a lot of the programs. Thanks to the owner Walter LeCroy taking me under his wing to develop a new project that I got my first taste of high speed analog design. I worked 80 hours week, learning analog electronics and work at the same time. Then I change jobs to Exar to design analog ICs.

    When in 1985, I change jobs to Seimen Medical Div. Designing ultra sound scanner. I was so surprised all the 8085 stuffs were gone, I design the CPU control board using 68000 16 bits processor. Everything was different. That was the first time I realize the digital world is so volatile, if you are out for two years, you are out!!! I since moved myself to the analog group and concentrated on designing the front end low noise circuits.

    All these year, I really don't have much formal knowledge. I only had equivalent of one and half semester of Calculus, no EM. Everything I know was learned on the job. It was not until 2000 when I finally decided to go back and study on my own. I since study math all the way to PDE, EM, RF etc.

    I can tell you at my old age, it is so much harder to study. I can study, wrote down in the note book to make sure I understand, derive all the formula beyond the books called for................Then two weeks later when I go back and read my notes, it looks foreign to me...........did I even wrote all these!!!! I had a good memory when I was young, not quite photographic memory, but even in my upper division Chemistry class, I usually read slowly, think through it once or twice and I got A's in my test. Now, it's like going in one way and leak out at the other end!!! You cannot beat the age. I spend a little over 4 years to study all the math and EM now, that is so slow!!! I had to study 4 or 5 times to really sink in!!!! It has be very frustrating that I just cannot remember things. At the beginning, I just use the excuse that I self study so must take more time as I have no instructor to go to. But this excuse got lame fast!! 4 years to study 6 classes, that is slow. I have a lot of passion in electronics already, can you imagine if you have to study under the gun.....so to speak.

    Can you imagine if you are in a field that constantly changing. Yes, it might be easy to get into the field, but you have to spend the rest of your life chasing newer knowledge. Imagine when you are in the 40s and 50s, you have your family, your mind is failing and you have to compete with the young people that are single and have a good mind and have more updated knowledge. If you took the easy way in school and avoided all the difficult stuff, now you find the younger people nipping at your behind wanting to take over your job?

    That is the main reason I keep talking about this, I have seen people loosing their jobs in digital and firmware field. If you are out for a year or two, you are out!!!! Analog don't change as fast particularly as digital speed getting faster and faster, it become analog.
     
    Last edited: Oct 18, 2011
  17. Oct 18, 2011 #16
    That is the reason I was offended as I am a Chinese too and make me look bad. Most Chinese are not like this and I don't want people to have an impression that this is how we are. One bad apple don't make the whole group bad. Chinese people tend to be more humble, I am actually the rare loud one already!!!!
     
  18. Oct 18, 2011 #17

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    I have a feeling that what you have said makes 'unacceptably good sense'.
    Without some depth of analogue expertise and experience in a design team, they will make all the same old mistakes that were made and learned about back in the 50s and 60s.
     
  19. Oct 18, 2011 #18

    Somewhat contradictory statements...no?

    Edit:
    That being said, I don't disagree that RF/Analog knowledge is important to good engineering. While I do a lot of work in the embedded realm, I do well with analog & RF as well. This helps designs meet EMC requirements the first try.
     
  20. Oct 18, 2011 #19
    I am very happy for Yungman that he has found a field he is happy in.
    But engineers who focus on the digital aspects are not all doomed failures, or uneducated slackers. They still are "real engineers" who are worthy of respect!!

    Not everyone wants to learn one topic and do the same thing for 40 years....

    And, to the other guy, the average software engineers salary is higher than EE right now. The job market dictates salary much more than barrier of entry. For example, memorizing the dictionary is very difficult, but how much can you expect to make with that skill?
     
  21. Oct 18, 2011 #20
    Learning one is easy, but if you have to keep learning one after the other constantly is hard. It is not the concept that is hard, it's that each different type has totally different protocode, like Firewire and USB, even though they are serial communication, it is not hard to learn, but if you get down to programming the registers, it is very specific on the bit map, programming sequence, and over all, the layers are different. You have to go through the pain of adapting your mind set to each type.

    I am sure the architecture of different processors have similarities, but they are different enough you have to constantly spent the effort to adapt to the new architectures.

    Usually the life cycle of digital products are very short, so you are constantly under the gun to produce new product and the jobs are usually high pressure. From my experience, analog and RF jobs are never as high pressure.

    If you design embedded system you should know what I mean. Take for example, The last embedded processor I designed was the Analog Devices ADuC831( if I remember correctly) that is 8051 core with dual DAC and muxed ADC and USB control. Even if you use 8051 before, you still have to go through tens of pages learning how to program the ADC and DAC and other functional blocks. Then you have to learn the development system and all the detail stuff. You go to another family, it is all over again. Of cause you have to take a lot of time learning USB. If some other communication link get popular, you have to learn it all over again. Just like if you are expert on Firewire, now it is all but disappeared, replaced by USB, Firewire chipset were dominated by Ti and USB is by Cypress, so the architecture are so different. It is an endless cycle.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: RF Engineer from China Give Advice
  1. Engineering Advice (Replies: 3)

Loading...