3 ms·

This is my experience too. The best EE I know dropped out of university but knows more about it than anyone I have met. He builds radios, phones and antennas fr

by _wldu 5y ago

This is my experience too. The best EE I know dropped out of university but knows more about it than anyone I have met. He builds radios, phones and antennas from parts as has since he was a kid in the 1950s. I think the same applies to CS. The best ones have been writing games and code since they were kids. It's something that they love to do.

imtringued 5y ago
I'm confused. As someone who dabbled with a basic PCB design. There are a lot of simple circuits that require a lot of theoretical knowledge. Designing the circuit is one thing as you can always copy paste existing designs and modify them. Picking and rating components is another as it requires a complete understanding how your circuit actually works and I was hoping that an EE degree would have helped with that. I guess there is no need for that and I should just "grind" practical circuits until I understand them. This is surprising to me because the nature of imperfect components makes EE far more complicated than software development where you do not have to care about choosing a for loop manufactured by Samsung vs Panasonic that is good up to 1000 iterations or that it is polarized so your for loop can only increment, not decrement. Add a hundred components each with their own characteristics and your head will split.
- hughrr 5y ago
  Well realistically you tend to start seeing patterns after a bit. Then you end up redrawing everything as those patterns in your head. Then you develop an intuitive model of what is sensible and what is not. This is not something they teach you at university. It’s all narrow forward engineering and simple theoretical analysis. There is a better education in having a pile of broken shit with incorrect service manuals and fixing it when it comes to EE. Test gear is fun because you get it for nothing, fix it and then use it to fix the next thing.
- a_conservative 5y ago
  EE drop out here: An EE degree will absolutely help you pick out components for a circuit. Computers and digital logic lend themselves towards a sense of everything being deterministic and understandable. Analog electronics, radio waves, radio wave propagation- I've heard all of those described by EE's (with degrees) as essentially black magic. It isn't black magic obviously, but I think much of those issues end up driving people to experiment with physically building things instead of relying solely on paper. This is just my impression. I think antenna simulation has gotten better recently for example. Another weird thing about EE, that seems to break what many (including me) expect it to be.. it feels like assumptions and approximations layered on top of one another. Take a simple diode. It has a certain current/voltage curve, but what I was taught about designing with them, is that generally you would use an approximation of the curve. Not the actual curve. Oftentimes, in lab, we would treat the voltage drop across a diode as being independent of current, which is absolutely false! But that is the way that EEs think of diodes, because it works! It's a game of learning what assumptions can be made to simplify analysis. You might want to try this for learning about electronics: https://en.wikipedia.org/wiki/The_Art_of_Electronics https://en.wikipedia.org/wiki/The_Art_of_Electronics another good example of how EE's reduce more complicated circuits down to simpler approximations: https://en.wikipedia.org/wiki/Th%C3%A9venin%27s_theorem https://en.wikipedia.org/wiki/Th%C3%A9venin%27s_theorem
  - mousepilot 5y ago
    > But that is the way that EEs think of diodes, because it works! It's a game of learning what assumptions can be made to simplify analysis. I think you are generalizing too much here. There's certainly some cases where a simple model of a diode works, but aren't there other cases where a more exact model is beneficial. I find it hard to believe that EE's graduate believing that one model fits all, unless its a pretty poor curriculum (for whatever reason). I'm not saying it is impossible to find a poor EE program, but we should be blaming those institutions instead of electrical engineering in general.

by _wldu 5y ago

imtringued 5y ago

I'm confused. As someone who dabbled with a basic PCB design. There are a lot of simple circuits that require a lot of theoretical knowledge. Designing the circuit is one thing as you can always copy paste existing designs and modify them. Picking and rating components is another as it requires a complete understanding how your circuit actually works and I was hoping that an EE degree would have helped with that. I guess there is no need for that and I should just "grind" practical circuits until I understand them. This is surprising to me because the nature of imperfect components makes EE far more complicated than software development where you do not have to care about choosing a for loop manufactured by Samsung vs Panasonic that is good up to 1000 iterations or that it is polarized so your for loop can only increment, not decrement. Add a hundred components each with their own characteristics and your head will split.

hughrr 5y ago
Well realistically you tend to start seeing patterns after a bit. Then you end up redrawing everything as those patterns in your head. Then you develop an intuitive model of what is sensible and what is not. This is not something they teach you at university. It’s all narrow forward engineering and simple theoretical analysis. There is a better education in having a pile of broken shit with incorrect service manuals and fixing it when it comes to EE. Test gear is fun because you get it for nothing, fix it and then use it to fix the next thing.
a_conservative 5y ago
EE drop out here: An EE degree will absolutely help you pick out components for a circuit. Computers and digital logic lend themselves towards a sense of everything being deterministic and understandable. Analog electronics, radio waves, radio wave propagation- I've heard all of those described by EE's (with degrees) as essentially black magic. It isn't black magic obviously, but I think much of those issues end up driving people to experiment with physically building things instead of relying solely on paper. This is just my impression. I think antenna simulation has gotten better recently for example. Another weird thing about EE, that seems to break what many (including me) expect it to be.. it feels like assumptions and approximations layered on top of one another. Take a simple diode. It has a certain current/voltage curve, but what I was taught about designing with them, is that generally you would use an approximation of the curve. Not the actual curve. Oftentimes, in lab, we would treat the voltage drop across a diode as being independent of current, which is absolutely false! But that is the way that EEs think of diodes, because it works! It's a game of learning what assumptions can be made to simplify analysis. You might want to try this for learning about electronics: https://en.wikipedia.org/wiki/The_Art_of_Electronics https://en.wikipedia.org/wiki/The_Art_of_Electronics another good example of how EE's reduce more complicated circuits down to simpler approximations: https://en.wikipedia.org/wiki/Th%C3%A9venin%27s_theorem https://en.wikipedia.org/wiki/Th%C3%A9venin%27s_theorem
- mousepilot 5y ago
  > But that is the way that EEs think of diodes, because it works! It's a game of learning what assumptions can be made to simplify analysis. I think you are generalizing too much here. There's certainly some cases where a simple model of a diode works, but aren't there other cases where a more exact model is beneficial. I find it hard to believe that EE's graduate believing that one model fits all, unless its a pretty poor curriculum (for whatever reason). I'm not saying it is impossible to find a poor EE program, but we should be blaming those institutions instead of electrical engineering in general.