[advice] step to take on designing

[chiasyan]

Hi,

Taking an example from eevblog video #95, which explain on designing a linear voltage regulator. The linear voltage requires a control loop in order to output voltage without oscillation.

Yes, I understand that there are whole bunch of theories and parameters need to be considered when designing a stable control loop (pole, zero, bode plot, bla bla bla....).

We do learn the control theory and understand those theories quite well, but when come to practical design, what is the best way of doing this?? Should we do all the complicated maths and then only go for implementation on the design OR we should just use "TRIAL and ERROR" method to archive the goal?

Despite hardware design, software design (algorithms) often requires maths for implementation. For instance, in order to implement digital filter/control on microcontroller, it MAY requires understanding on Z-transform, state-space model, etc. and also performing whole bunch of matrix operations. It might take time to go through theories (argh....) and we might fail/stuck when try to implement those "WORDS" into "CODES". Yeah, often, we might get lucky to "REFER" to some source code from the Internet and often it save us A LOT of time by doing this.

hmm... sometimes, it requires a long revision on the theories that we have learned and we might get stuck in the process while trial and error method shall get all the jobs done in no time Yeap, we might learn something in the process of 'REFERRING" and "TRYING", but we could miss the opportunity to learn even MORE... . Being a student or an engineer, how is your point of view on this issue?

Thank you  
p/s: I need advice so badly....  

We do love electronics, that's why we are here  

[Bored@Work]

Rule number 1: Don't annoy people by

(a) Using stupid subject lines in all caps

(b) Telling people their stuff is "bla bla bla"

(c) Whining because you are to lazy to learn the stuff

(d) Trying to convince us to do your homework by whining and annoying us

[Balaur]

1. Shut-up
2. Observe
3. Learn
4.
5. Profit!!!

Gee, what is with all these young whippersnappers nowadays?

On a more serious note, I could tell you that there is no easy path.

If I could remember and be able to apply all the theory that I've learned in school (and I've passed with flying colors, thank you very much) I would be a much better and well-rounded engineer.

However, life is life and most of the time you will have to work on clearly identified subjects that you should master very well.  As an engineer you are expected to be able to complete a task efficiently and with good results. Part of the effort consists in applying the more or less abstract knowledge you got in school. If you encounter something outside your primary area of competence, your school training should allow you at least to know where and what to look for complete your skills (if you are inclined to do so of course).

In any case, the internet is a very good source of information. While you cannot expect to find something that you can copy and paste to your case, you should be able to get a lot of help.

[djsb]

Just do it.

[Mechatrommer]

if you want to have more insight, this is my advice. take or buy a peanut from a mall whatever. plant it back of your house, and watch closely everyday as you care it, put the water, soil etc etc, see and think how it grows from a smalller thing to become a bigger thing until it grow another peanut. u'll learn something there. Cheers

[Simon]

and the question is ?..... rather a lot of words and not a lot of content, still trying to figure out what is wanted, now thats a bad start to any design including the design of how one will go about designing 

[tyblu]

Just follow the datasheet. Common design path is: 1) find old design; 2) tweak to your specs and prototype; 3) fix prototype; 4) mass produce and pray it works. The math involved is in understanding the initial design and its parameters. You mentioned a regulator: their datasheets will recommend capacitive ranges at their inputs and outputs -- just follow them; experience may tell you to sometimes add 100nF or < to the output because larger electrolytic can't handle high freq.'s well. Theoretical system control and all that jazz is high-risk business, and is restricted to companies who can fund R&D or to universities (or to hobbyists who think it's cool as hell, like myself).