33

# RANT TIME! Sorry guys, I know this is devRant and probably not a place to post this but am fucking burning with fury and fatigue! I should probably develop elecRant and post it there instead. I FUCKING HATE POWER ELECTRONICS!! I am in my final year of electrical engineering and I can fucking say with confidence that power electronics is the most fucked up unit I have seen in my life. A whole load of useless math from simple RLC circuits just to make students' lives miserable. For those who might not know, power electronics is some unit that involves use of solid state electronics(transistors, diodes etc) for power applications(switching mostly). Basically things like inverters and converters. UPS systems are an example of their applications. Now don't be fooled by how that sounds cool and so smart, this shit is fucked up. These circuits in the attached picture might just seem like simple RLC networks with some BJTs, but they are devils in their own right. They fucking need some advanced unnecessary calculus and Fourier analysis to even calculate the simplest output current!! Worst still, some of these motherfuckers have more than 1 mode of operation,needing one to analyze some fucking 100+ waveforms. I fucking hate this shiit! I hate it! You might say that i am just being lazy and don't want to study. Let me tell you something, FUCK YOU TOO!!

• 9
Fun fact: What you're learning here isn't even very applicable today, because nobody uses BJTs for switching regulators anymore! Unless you're only concerned about the very basic case where the transistor is modeled as an ideal switch, only then the topologies are the same.
• 5
@7400 This education system is already fucked up. All the shit we are learning are totally inapplicable today!
• 13
Confirmed, power electronics and particularly power supply design is absolute hell.. logic circuits are so much easier because there you don't really have to care too much about heat buildup or efficiency. Luckily I'm self-studying so I can choose my own stuff to learn about.. but power supplies, if I can buy them from somewhere, I buy them without a second thought.

And yeah as @7400 already mentioned, BJT's have made place for more efficient MOSFET's these days :)
• 2
@Condor Wow, that's so cool that you are self studying. What areas are you focusing on, light current electronics, digital electronics, high current power systems or others like electromagnetics?
• 8
@CodesNotHot nothing too fancy here. I prefer portable digital electronics that are built with longevity in mind (because less current is less heat and my lithium cells will love me more). Sometimes also power electronics but usually I just build it in the simplest way possible with very little feedback circuitry, if any at all.. I hate control theory. Electromagnetics, eh as far as motors and speakers go I guess. Also, EMC is completely foreign to me :P
• 1
@CodesNotHot Coursera has a specialization on Power electronics from UColorado, it's pretty informative. I skimmed through it and I thought it was pretty nice.

Not that you want yet another course on power electronics, but hey, it's a cool subject.
• 2
@Condor ever tried getting into the deep dark well that is amplifier design? :D
• 7
@RememberMe absolutely! I've been very curious about how to turn my phone into a signal generator that's more useful than just for a few millivolts. The holy grail in my quest for good amplifiers (again power electronics, God save me) would be to be able to use this audio signal and amplify it into not only higher voltage but also higher current.

That said, chances are that I'll be better off with dedicated controls with potentiometers and wrangling the pixies from the wall instead.. but that'd imply generating square waves (probably the easiest), sine waves etc from either 230VAC or ~370VDC (I think, will have to build FULL BRIDGE RECTIFYA to check).

Either way those pixies from the wall should be able to give me up to 40A until the breaker trips, and 16A until my extension cords fry themselves. So yeah, plenty of power available and I can adjust if I need more than those 16A from the extension cords. The issue is really in wrangling them into a different frequency, voltage and things like that.. power electronics are a bitch to work with. Not only do you have to worry about logic and control theory, but heat and efficiency as well. And get all those shits to work perfectly into one smooth setup. Fucking hell!!! I doubt that I currently even have the equipment necessary for it.
• 2
@Condor oh god current amplification is a pain. You could do what I did and use the amplifier from an old speaker set for your signal generator - the problem is that they don't supply much current either, but hey, way better than the amp in your phone.

I'm currently trying to build analog computers to model basic differential equations - it's a really simple and cool idea (not so simple execution but eh). It's all about them amplifiers.
• 7
@RememberMe yeah, currently I've got the amp board from some cheapo USB speakers for that.. currently they've got some small speakers soldered on (on which I've been able to experience firsthand that applying a 3kHz sine makes them run violently, perfect for alarm systems) but eventually I might turn them into low power signal generator drivers yes. Of course reverse engineering would follow afterwards but for the time being.. meh :)

I'm not familiar with analog computers, but I've seen someone put together a 1-bit memory cell out of what I recall to be a flip-flop earlier. Quite interesting!
• 3
@Condor oh, I would call flipflops digital. I mean, yes it's all analog at the low level but you hardly ever think of a flipflop as an analog device, it's always the digital state of the FF that's important.

Analog computers basically exploit the fact that electronic systems can be modelled as differential equations. So if I have a system that I want to simulate via analog computation, I simply design a circuit that expresses the same set of differential equations as the system being modelled. The circuit then becomes a simulation of the model.

Eg. Radioactive decay can be expressed as a differential equation, you can model this using a discharging capacitor because that follows the same kind of differential equation (exponential decay).

They used to use this for all kinds of cool stuff, like industrial process control and ballistics calculations.
• 7
@RememberMe That'd be a very interesting project, but how are you going to deal with the (inevitable) inaccuracy of voltages and their measurements over time? Decimal computers at the time when 5V -5V logic were still a thing didn't become a standard because maintaining voltage levels within half a volt would be too difficult... analog being massively more reliant on exact analog values would be massively more susceptible to small voltage changes.

Analog systems are definitely interesting and my unknowingness about them shouldn't defer you from exploring this - but personally I'd be much likelier to go with a digital system instead, as those are much easier to build and manipulate.
• 1
@Condor oh yes, that's why they were abandoned. Digital is just so much more flexible, and now faster and more reliable too. And much more accurate. But these were used back in the days when digital computing was non-existent or very slow and expensive.

I'm not designing anything mission-critical, lel, so some error is tolerable!
Funnily enough I'm using a bunch of Arduinos to read the outputs of the analog simulation - so it's kind of cheating, but eh.
• 7
@RememberMe Arduino outputs and inputs should be good enough for research I'd say! Sounds like an interesting project.. keep me updated! :D
• 1
Are you allowed to use Pspice at all? It might help you with some of the design tasks so you aren't sitting there iterating on the math all day by hand.
• 3
BJT circuits... Eek.
Sad thing is nearly everything being used for application based teaching is 20 years old and near useless for anything practical nowadays.
• 0
Indian?
• 0
Actually i found these things fairly "easy" and the math behind it isn't to hard,which of course always depends on what you do and learned, but let me tell you one thing: The actual math and physics behind semiconductors is MUCH harder.
• 0
@muma I have done both Semiconductors and power electronics. In fact, i have encountered almost all the math in electrical engineering and i'd rather do semiconductor math. It's fascinating compared to power electronics, trust me