Top-Down Learning?

Discussion in 'General Electronics Chat' started by GregJ7, Aug 19, 2015.

1. GregJ7 Thread Starter New Member

Jun 7, 2014
18
2
I have been learning some things about hobbyist level electronics for about a year now and have realized that my progress is very slow because I am not learning in the way I learn best. My best understanding is this: All the resources I've run into so far teach in a "bottom-up" fashion. In the case of a capacitor, what might be done is to explain that a capacitor is two metal plates separated by a dialectric, etc. The next step is sometimes to show it being used in a simple diagram and explaining what it is doing. We pull out our oscilloscope and see that the voltage level across the capacitor rises as soon as it receives power, then stops changing. (Technical college classes are usually taught this way. Theory first, application later--and a lot of people are unhappy about it, particularly large technical corporations.)

I learn best with top-down instruction, which one might think of as a divide-and-conquer approach: Codify the problem that you are trying to solve, then ask what would solve it. Then access the tools one has learned about to solve that sort of problem, and then reask the question for each tool that is part of the solution. I want to build a guitar preamp. How shall I do that? Well, I know that a transistor can be used to amplifier a signal and that I'll probably need resistors to change voltages to meet transistor specifications. But resistors sometimes produce significant electrical noise. So I might need a way to filter out some of that noise, etc. The math might be taught at this point as a way to determine component values for a solution. Eventually the student will head toward bottom-up learning on their own when they see that it will be a solution to a problem they have. It happens when they have run into the same situation enough that they want to understand why that pattern exists.

The primary motivation to teach top-down is because it is much easier to hold a student's interest, which is necessary at some level for them to really learn. (An instruction designer might call this being "on task." We might call it being "in the zone.") At present I might not care how a capacitor works, but am interested in how a guitar preamp works. This excites my mind to examine the building blocks needed to design a guitar preamp, but not to learn about capacitors. I'll want to learn about capacitors when a single capacitor is going to solve a problem I have.

Can anyone recommend circuit design learning resources that do this? A couple "practical-oriented" basic resources I have might sound like they would do that, but they don't. There often isn't even an attempt to visually divide a circuit its logical subcircuits. I suspect that people that learned bottom-up have trouble converting what they learned into top-down instruction. "That's not the way I learned it." (That's why instructional design is a degreed field.)

To illustrate my current conundrum, I refer to the Tillman Discrete FET Guitar Preamp near the top of the page at http://www.till.com/articles/GuitarPreamp/. I want to modify this circuit to have a gain control, but have nowhere to start since I have no idea what the purpose of each component was/is, other than the FET. I could tell you what a resistor does (along with some math) and to some degree what a capacitor does, and I can explain to you the physics behind NPN doping, but it's all meaningless in a higher level context ("what is this circuit doing"). The resources I have been learning from never views building up the circuit design from that perspective. Ideally I would learn how to build up a circuit with subcircuits because I have learned what they do as a unit and then also how to alter subcircuit attributes with other components/subcircuits.

By the way, an even simpler example would be how to think of an RC or LRC circuit as a functional unit. The introductory information I've found about these doesn't discuss what an LRC circuit is as a single functional unit other than to say circuit with resistance and a synergy between its inductance and capacitance. That information doesn't help me design a radio circuit.

2. dl324 Distinguished Member

Mar 30, 2015
3,376
651
I think the best way to learn would be to have a project in mind that will require you to learn how it functions.

Taking the "Tillman" preamp and modifying it is a good place to start (though I think his bias over integrated circuits is unfounded; many "audiophiles" suffer from delusions). He already described how to increase gain at certain frequencies, so figuring out how to do it for all frequencies shouldn't be difficult.

To maximize your learning, you should try to understand what each component does. Don't be surprised if you find circuits on the WEB that have extraneous components (that's one way to catch copyright infringement) or the author didn't really know what they were talking about.

Regarding noise. Instead of trying to filter it out, design the circuit to minimize it. There are several types of resistors (carbon composition, carbon film, metal film, wirewound, etc) and each has pros/cons. I know of many audiophiles who claim their circuits sound "better" with carbon comp resistors. The fallacy in that thinking is that carbon comp resistors in vacuum tube circuits are actually causing distortion that some find pleasing. You can't replicate that effect in circuits without vacuum tubes because the distortion is voltage dependent and you need a couple hundred volts across the resistor to get it.

If you're going to work on audio amplifier circuits, you're going to want to learn about how capacitors behave with frequency. Very important for AC coupling for inserting poles/zeros in the frequency response.

Last edited: Aug 19, 2015
3. wayneh Expert

Sep 9, 2010
12,374
3,225
I've spent a lot of time learning in both directions. And you're right, project and goal driven learning is a heckuva lot more fun and satisfying. Nothing makes you pay attention to detail like having a hurdle to jump over in a project you're working on. I'd love to see a lot more project-driven learning used in schools, since motivation to learn is the primary hurdle in young students. And learning how to learn, to go find resources and learn from them, is useful for anyone. Robot competitions are great in this regard but the teaching method could be more broadly applied, in my opinion.

That all said, one problem I've seen with task-driven learning is that it can be shallow. The student learns just enough to clear the hurdle but doesn't learn the fundamentals that can give a broader perspective and a deeper learning experience. Indeed the student may see theoretical study as a waste of his precious time as he tries to get his robot to perform the next task. Why do I have to slow down and learn about induction? Or myriad other topics? There needs to be a balance between application work to stimulate and motivate, and a good theoretical coverage of the related topics that are being taught.

#12, elec_mech, MikeML and 2 others like this.
4. crutschow Expert

Mar 14, 2008
13,475
3,361
The reason you can't find resources to do what you want is that it likely doesn't work.
Without knowing how a capacitor or inductor works (or even a resistor), I don't see how you can understand how a circuit works from a "top-down" view.
I think you are trying to put the cart before the horse.

MikeML likes this.
5. BReeves Member

Nov 24, 2012
412
64
I've always done better self taught from studying books after deciding what is was I wanted to accomplish, guess that is top down. Most of what I know was acquired before the internet existed and books were my salvation.

6. ISB123 Well-Known Member

May 21, 2014
1,240
531
You learn by designing circuits.Theory is useless if you don't apply it.Start by making a NE555 based project and as you design the circuit you should learn what is the function of which component and what it does.

7. nsaspook AAC Fanatic!

Aug 27, 2009
3,009
2,350
Top-down, upside down, bottoms up, spin around. All theory with no design or building makes Jack a dull boy but Jack with no basic theory is an electronic Script Kiddie.

#12, kubeek, OBW0549 and 1 other person like this.
8. AnalogKid Distinguished Member

Aug 1, 2013
4,685
1,297
That's because a lot of people and corporations do not understand the difference between information and knowledge, are devoted to instant gratification, and have a bizarrely short-sighted view of return on investment, rate of return, etc.

He who loves practice without theory is like the sailor who boards ship without a rudder and compass and never knows where he may cast.
- Leonardo da Vinci

nsaspook and JohnInTX like this.
9. bwilliams60 Active Member

Nov 18, 2012
734
94
I am fairly new to the electronics world (couple years in) but I know all the tools and all the components but not entirely sure how they all come together. I have found though that by reading a lot of the forum discussions, how certain things come together and how to test them. I find myself now trying to work my way through problems on here before the answer comes and then seeing how close I was to being right. Then I back-peddle and find out where I went wrong. I am a Professor myself but I find that every student has their own ways of learning. Hands on is by far the best teacher but good guidance is always a blessing.

10. AnalogKid Distinguished Member

Aug 1, 2013
4,685
1,297
I respectfully disagree. Electronics isn't shop class. Much of it is not intuitive, some of it is counter-intuitive, and none of it is visible. I clearly remember and completely get the excitement of building something for the first time and having it blink or beep. But deep down that isn't learning, and letting students think that all electronic concepts come with such feel-good packages is misleading and disingenuous. Rule-of-thumb and Ohm's Law appear relatively simple and produce quick results, but the most difficult aspect of both is knowing when *not* to use them. That comes with instruction, not demonstration.

Give a man a fish, teach a man to fish, etc.

ak

atferrari and nsaspook like this.
11. nsaspook AAC Fanatic!

Aug 27, 2009
3,009
2,350
Electronics is as much art as science in the end and great artists didn't get that way from just doodling even with great natural talent. They learned the mathematics of drawing, geometry, perspective, lighting (even if the instructor hid those facts from them) as the foundation for even abstract art.
Circuit Theory is a very important foundation in learning what NOT to do. There are usually a 100 bad ways to make something and usually only a few good ways. Knowing the good ways requires insight and insight is learned by knowing what's crap. Knowing what's crap is learned by knowing the mundane details and methods of how things work so you can sort through the crap to find the nugget buried in it . This requires intuition and intuition is a combination of experience and abstract knowledge (that can look like luck or even magic) that leads you to the correct path even when you don't know all the facts. If you need robotic droids to push buttons then by all means skimp on theory but if you want technicians and engineers to keep technology moving forward then there are no short-cuts.

MikeML likes this.
12. ScottWang Moderator

Aug 23, 2012
4,925
777
1. To buy some kits to solder them and make them to work, study the theories of those kits, and to know the function for each component.
2. To see some more circuits from internet and to study their theories.
3. To learn the ee symbols and the basic functions of ee components as TTL 74LS00 series, CMOS 74HC00 series, resistors, capacitors, diodes, led, bjts, mosfet, relay, buzzer, speaker, etc...
4. Study and use the basic ee formula as Ohm's law V = I*R, R = V/I, I = V/R, W = V*I, etc ...
5. To learn the basic circuit as RC integrating circuit, RC differentiating circuit, LC filter, etc...
6. To buy a oscilloscope, to make a basic power supply, using LM317, LM337, BJT and zener diode, etc ...
7. Doing the experiment and study the ee theories repeat and repeat ...
8. Try to solve the problem by yourself when you meet it before you post the question to the forum.

13. Veracohr Well-Known Member

Jan 3, 2011
559
77
I think that would be a pretty inefficient way of learning. It's basically re-creating the experiments of our predecessors who already figured out this stuff, as a way of teaching. Seems to me it's simply faster to say "this is how it is, because so-and-so figured it out 100 years ago."

I sympathize with your complaint. A lot of times if you're just learning the fundamentals without any context it can seem pretty pointless, and some people don't learn very well in that kind of environment (myself included). I started out trying to learn on my own from books and the internet, and when I kept coming up against walls I decided I needed a formal education. The benefit of real, live teachers, as I figured out, is that they can put those fundamentals into context on the spot when asked. The best way to go about learning is to always think about what you're learning in the context of whatever you're interested in.

Also, I think that learning about a component from only one perspective first can lead to confusion. Say the solution you need to solve is to couple an AC signal to some circuit centered at a different DC voltage. The proper solution to that is a capacitor, but that then leads to discussions on reactance/frequency-dependant impedance, which inevitably leads to discussions on filtering (otherwise why even mention it). Filtering is such a large topic that it would create a huge distraction. By that point you've lost the focus on whatever you started with. If, on the other hand, you learn all the basics of capacitors to begin with, you can later come across applications for them where you don't need to go back to the basics, because you already know them.

So basically I suggest that you just trying thinking differently about what you're learning. It's probably more useful in the long run that searching for a different teaching method.

nsaspook likes this.
14. dl324 Distinguished Member

Mar 30, 2015
3,376
651
With respect to what the OP wants to do. The guitar preamp in question consists of a JFET, 3 resistors, 2 capacitors, and a battery. The URL mentioned also contains a description of the function of each component. The OP wanted to change the gain and how to do that was included (sort of) in the Technical Details section. With all of that information, I believe this endeavor is within the abilities of the OP because he is motivated. Aside from the rant against opamps, not understanding that a JFET also has non-linearities and more noise than a BJT, not knowing how to select a different JFET for the circuit, not mentioning how gain was calculated, and having misconceptions about opamps, BJTs, FETs, and vacuum tubes; the Author was correct enough where it counts.

Being someone who is self taught in most areas, I respectfully disagree with the notion that you need to have formal training to be able to do a job well or correctly. When I was in college, Unix was still Dennis Ritchie's pet project. I taught myself about the Unix Kernel and was more conversant in it's features than 90% of the people I worked with that had MS CS and PhD CE degrees.

People with degrees tend to have strong beliefs that they (degrees) are necessary to do the job. A former Boss would only hire people who had a PhD because he was of the opinion that someone who had a PhD could do "anything". I'm of the opinion that the jobs we did could have been done by an intelligent high school graduate (okay, maybe associates degree) with motivation and some experience. I worked in the semiconductor industry and most of what I worked on won't be in text books for another 10-20 years. Granted most of us had training in the fundamentals, but the order that enables learning isn't fixed.

Individuals who are intelligent and motivated can take short cuts.

15. nsaspook AAC Fanatic!

Aug 27, 2009
3,009
2,350
True that, but I've found in life that short-cuts usually have limitations that can cause small deviations from the known path to be the route to the bottom of the gravel pit.

Last edited: Aug 20, 2015
MikeML likes this.
16. AnalogKid Distinguished Member

Aug 1, 2013
4,685
1,297
Individuals who are intelligent and motivated don't need short cuts.

ak

MikeML likes this.
17. dl324 Distinguished Member

Mar 30, 2015
3,376
651
Maybe short cuts wasn't the correct phrasing. Maybe they don't need to take the traditional path is more appropriate.

When I look at the finalists for the Intel Science Fair in any given year, I'm always amazed at what these high school students have accomplished without the benefit of formal training. Intelligence, motivation, and mentoring have allowed them to make huge strides in learning.

I'm certain the majority of the people frequenting this forum know of Jim Williams. AFAIK, he didn't go to college for more than a term; yet he was a giant in his chosen field.

18. dl324 Distinguished Member

Mar 30, 2015
3,376
651
It's true that most people will take, and benefit from, the traditional path. But there are, and always will be, exceptions; and exceptional people who will accept the challenge of taking a different path and be successful.

19. nsaspook AAC Fanatic!

Aug 27, 2009
3,009
2,350
We agree but you can't base a course curriculum on the exception. One part of the Dunning–Kruger effect is that exceptional people tend to believe that tasks that are easy for them are also easy for others.

One of the hardest steps to learning is getting past stage one. Every day I learn more of my own incompetence.
https://en.wikipedia.org/wiki/Four_stages_of_competence

20. dl324 Distinguished Member

Mar 30, 2015
3,376
651
I don't disagree. The standard curriculum is sufficient for the majority of students. The OP stated that method doesn't work for him...

That may be why I'm a college dropout. I was taking classes to learn more about the circuits I was designing and working on in my job as an R&D Technician. The department chairman at the school I was attending didn't think that was wise and mandated that I follow their prescribed path (he had to approve my course selections). I was there to learn and didn't give a rip about the degree; hence, I'm a dropout because of some academic's opinion and narrow mind...

It didn't matter, I learned what I wanted to learn on my schedule.

nsaspook likes this.