[sdiy] how to best learn the trade

[Peter Keller]

Hello,

I started doing electronics as a serious hobby maybe 8 months ago
(with the ultimate goal of making hybrid analog/digital synthesizers)
and I was basically in your position. I didn't even know what Ohm's Law
was! After some serious reading and doing some math (it really isn't that
bad--don't be afraid), I have a decent handle on things and have just
started really understanding the lowly transistor. A couple of very good
books are the Art of Electronics and _especially_ the student handbook.

In addition to the wonderful advice you've already gotten (make LOTS
of different things), I can't recommend enough writing a theory of
operations document, which details _exactly_ why something was made the
way it was. Of course, you aren't going to do this for everything, but
for anything which you decide you want a permanent copy of you _really_
should do it. It'll help you remember why/how you did something when
you need a refresher or if you'd like to modify it later.

Here is an example of something I've made:
http://pages.cs.wisc.edu/~psilord/blog/15.html

It was basically the first real thing I've ever pushed all the way to
completion (it in fact uses the radio shack 3-hole board with power busses
that Anthony mentioned). I was quite proud of it when I finished--even
though I'm sure someone who is seasoned (or maybe just salt and peppered)
in electrical engineering would see plenty of mistakes in it. At the
end is the theory document explaining it and you should read it since
it details my thought process (albeit simple and naive--but documented)
in why I made what I made. The device (electrically) is simple. The
codebase behind it is also simple (I'm a systems programmer by trade,
so simple means something different for me :)), but if you don't know
the assembly or how to code, it can be daunting for sure, but read the
comments and you'll get the idea.

I started from _zero_ knowledge on how to make DIY electronics and
analog electronics in general. The analog stuff in that circuit I made
after reading all kinds of things on the web (like how to make an LED
circuit), reading datasheets for things like the 7805 power regulator,
and studying the hell out of the Art of Electronics book, and it still
took me a while to figure that stuff out.  In fact, I joined this list
cause it looked like it'd be a good place to learn this stuff (and music
synthesis), and it is. :)

One _big_ thing which helped me was that I designed the device in pieces.
Each little hand drawn schematic (in the theory document) represents an
isolated circuit that when it is all put together, make the device. In a
way, this sort of functional approach to learning the whys and wherefores
about electronics is really the way to go. Learn small functional building
blocks (voltage divider, low/high pass filters, diode clamps, etc), then
connect them into larger and larger pieces until you end up with something
that behaves in interesting ways. And it goes the other way too. After
learning the basic blocks, you need to then look at larger schematics
and pick out the smaller building blocks (print the schematics out and
circle the blocks).  I can't stress enough this concept of learning
simple things that you wire together into larger things--electronics
seems wedded to the bottom-up method of learning it. Doing the pattern
recognition of the unknown schematics into smaller pieces really helps
you understand what the person was thinking when they made it.

Another thing is at the outset, buy new parts for things. You'll spend
more money, maybe 200-300 dollars initially to build a collection of
useful stuff, but since things will be new, they will just work. Ripping
stuff out of junked electronics (which I admit I've started doing)
only makes sense when you've cached enough mental knowledge about
electronics to be able to meaningfully salvage something in relation
to your time cost for the salvage and subsequent identification, test,
and clean up of it. If a nice power transformer _that you understand_
costs 50 bucks out of a catalog, but you don't know how to use a salvaged
one out of a radio or something, then spend 50 bucks. Don't electrocute
yourself. :) Later, when transfomers are more clear to your inner eye,
you can salvage them to your heart's content.

Use google. Start your search with "how do I read a schematic" and read
the first 20 links that look promising. I must have spent 2 months
simply reading online and building tiny schematics on my breadboard,
doing the math, checking the results with the multimeter to see if I was
right (did the voltage divider work as intended, is the right amount of
current going through the LED, etc), and doing it again and again and
again until the equations and "image" of the schematic sunk in.

Eventually, you'll discover you need things like a frequency sweep
generator, an oscilloscope, etc. in order to see what the circuits you
are building are doing. But, you can probably get a year or two out of
the hobby before you start building circuits complicated enough to need
those to debug/test them--the mind's eye will get you far. And when you
do need them, you only need like three bench devices (maybe around $1500
total for all three) and then you can go *very* far with the hobby.

For now, get a decent collection of breadboards (maybe 1-2 three strip
ones and 1-2 two strip ones), various passive parts like resistor and
capacitor assortments, NPN/PNP transistors, some diodes, LEDS, and use
the hell out of them when learning to build schematics.

Good luck.

Later,
-pete

P.S. Also, read up on electronics safety. I'm sure you already have, but
do it again. :)