Beginners' Curriculum

[jafo]

> Example: a high-pass filter sends high frequencies to ground.

No. A high-pass filter passes high frequencies to OUTput. And since we say "high-pass", presumably low frequencies do not pass to output. They may be diverted to ground, or just not allowed to enter the filter.

Gah, I started writing that the other way around and didn't edit correctly. My bad.

> electrons come from the ground

Electrons are everywhere.

Hence my confusion. I know what scientists say, and I know what (some) electricians have told me, that electrons move from ground to hot. I figured it was another of those cases in which differing explanations only apply in certain situations.

And meaningless. There is a book {u]There Are No Electrons[/u].

Have you ever seen an electron? If you could, would it do you any good?

Now that's an interesting question. I tend to waffle between a realist and an instrumentalist position -- we can measure properties, so there seems to be something there; but all we really know is the results of certain measurements in certain situations, so how much reality can we ascribe to it? But then, it's all metaphor -- poetry, really -- and some metaphors are more useful than others within differing circumstances.

There's "current", a mysterious invisible substance which is only interesting when it is moving.

> If impedance is like water pressure in connected pipes

You are mixing our poor metaphors.

Impedance of a water-pipe is the Length-to-Diameter ratio. A long thin hose is a high-impedance pipe. A short fat storm-drain is a lower-impedance pipe. The Mississippi is a really low impedance. 

> you want to connect the out of a high pressure pipe to the in of a low pressure pipe -- since you want all the water to go from one pipe to another.

Yes, but this is exactly why you must keep our poor analogies straight.

Look at your water supply. There is a 24-inch pipe from the water tower. This splits to 8" pipes in the street. Houses are connected with 1" pipe.

When you turn-on all the water in your house, the high-impedance 1" pipe will not suck enough water to drop the pressure in the 8" street-main. Your neighbors do not get a pressure-drop when your toilet flushes. (Actually the goal is to connect multiple 3" fire-hoses to an 8" main and not have a large drop.)

Other than an unrealistically optimistic view of the plumbing around here, that's quite a lucid explanation.

> Or maybe I simply don't know what "ground" is. (That's where I'm placing my money!)

There is no "ground" except what you pick.

A simple series circuit, such as a flashlight (ponder it!) does not need a ground.

Anything complicated, there are multiple current paths which come together. In a car you have lights and ignition and cigar-lighter. You could run a separate wire from each lamp and gizmo back to the battery. However the car body/frame is always handy, already paid-for, and big steel conducts good enough. We bolt the lamp-socket to the body and call it "ground". Even though the car stands on rubber tires and has no contact with dirt.

Hunh. I think I may have been confusing the ground of a circuit with the ground state of an electron. Combine that with some misinfo from the Net of a Thousand Lies... yeah.

> alternating current -- are electrons reversing directions?

I've looked real close at wires and I do not see any electrons in there, so I can't say they go back and forth. However we have basic experiments (sadly no longer taught in school) which suggests that an electric current will move a compass-needle (magnet) one way or the other depending which way this current flows. And with a low-low frequency AC current I can actually see the needle jiggle back and forth.

It's been a long while since grade school, but I do remember doing that experiment. Plus, it does seem to be congruent with the way that a string or diaphragm vibrates.

> why is it a Good Thing to drop voltage across, say, a drain resistor?

A single JFET (BJT, tube) is like a one-legged bicyclist. Only goes one way.

We can wire two JFETs (etc) in Push-Pull. That costs more.

We can rig a JFET(etc) to "fight" a resistor. This is kinda-semi-like a one-legged bicyclist with a spring or weight on the un-used pedal. We "bias" the rig for a "fair fight", around 50:50. With a 50 pound weight on the odd pedal, the one-legged bicyclist can press zero to 100 pounds. Result is like 50 pounds on each pedal. We can rig the drain resistor to drop 4V, rig the JFET to drop 4V, then swing up 3V and down 3V.

While single-device operation gives less gain and efficiency, two single-ended stages gives more gain than one push-pull stage with generally less complication.

True, but I do prefer asymmetrical waveforms -- if nothing else, the implied extra octaves are the only intervals truly consonant with equal temperament.

Thank you for your reply; once again, you've been very helpful.

[jafo]

no electrons! 
i dig it!

I dig the pun you've invoked -- electrons, ground, dig...

[defaced]

Hence my confusion. I know what scientists say, and I know what (some) electricians have told me, that electrons move from ground to hot. I figured it was another of those cases in which differing explanations only apply in certain situations.

To clarify what you've been told in a hopefully simple to understand explanation:

At one point in time, scientists thought that the positive charge moved.  Thus, "current" is defined as a positive charge flowing from a positive voltage to a negative (typically grounded) voltage.  Lots of books were written and electronic theory was in happy land.  Then someone got an ah-ha moment and realized that the positive charge doesn't move, but the negative charge does (electron), and when it moves it goes from a negative voltage to a positive voltage.  But what about all those books on current flow?  Now they're all wrong?  Nope, they're right, just *backwards*, which is fine because we can make a conventions.  So the convention stuck and is the reason why today we have current flow (or sometimes called conventional flow or conventional current) and electron flow.  If you consider the basic structure of an atom (positive nucleus, negative electron cloud), The Strong Force, and that like charges repel and different charges attract, this whole thing beings to make sense. 

For the most part, it doesn't make a bit of difference which one you use as long as you're a) consistent and b) understand how to get back to reality if you need to.  Current flow and electron flow have the same magnitude, but different directions (this is why PRR said it doesn't make a difference what moves, it's the movement itself we care about - most of the time).  On a circuit level, current flow works just fine and is what 99.9% of people will use.  Most people don't really ever need anything beyond that.  However, on a component level (tubes, semiconductors, polarized caps, etc), electron flow is more or less the only way you can accurately describe what's going on inside the device and have it make any bit of logical sense. 

[BubbaMc]

There's some good stuff in here: http://www.eevblog.com/episodes/

[PRR]

> didn't edit correctly. My bad.

I do that a lot too.

But there's a lesson. In technical fields, words usually have an "exact" meaning. Confusing meanings leads to more confusion.

Example:
> electricians have told me, that electrons move from ground to hot.

No. Electrons move from Negative to Positive.

In your car, the battery Negative is wired to frame, Positive goes through switches to "hot" loads. However in a 1941 Plymouth the battery POSitive goes to frame, the NEGative is "hot". It all works fine. As long as the current loop through battery, switch, and load is complete (no rotted/broken wire) and takes no short-cut (rotted cloth insulation laying on frame).

Ground/Hot is NOT the same as Negative/Positive. Choice of which end of battery is "common" to most loads is arbitrary, although there may be a practical preference. (Circuits with tubes or NPN transistors may be much simpler wired negative-ground; conversely PNP favors positive ground.)

In AC systems, if you call common/ground "zero", the "hot" wires will swing positive AND negative from zero, and the electrons will move one way and then the other way.

And many professional electricians are thin on theory. There are many-many ways to run electric power, but safety and maintainability trumps creativity, so most electric connections are done "by rote".

> an unrealistically optimistic view of the plumbing around here

Totally made-up dreaming. There's no water in the street here except when it rains. I have a half-power pump thru a too-long too-thin (high impedance) pipe. If I run the outside spigot, the water upstairs stops. This can all be explained as a low-voltage source, high impedance feed lines, a low-impedance spigot shunting a high impedance upstairs feed with significant voltage requirement. So if you can intuit poor plumbing, you can intuit what happens in electricity.

> the ground state of an electron

What the heck have you been reading?? Solid-state physics has no business in a pedal-builder's studies.

OK. Accepting that there are electrons, and that metals have lots of "loose" electrons.... put ball-bearings on your car hood. They roll freely. Now put a lot of dents in your hood. The balls tend to sit in the dents, but will roll with slight provocation. A ball in a hood-dent is an electron in its ground state.

But this is really not relevant here.

> I do prefer asymmetrical waveforms

Nothing to do with push-pull versus single-ended.

You say you like asymmetric waves... but are they bigger on top or on bottom? In audio, generally we get waves either way, and potentially equally strong either way. The amplifier must be agnostic about which-way. Therefore the amplifier must be symmetric, to handle asymmetry either-way. A raw single-ended amp won't do that, so we add "bias" to get it set-up half-way between its limits. SE amps tend to have some residual asymmetry; we can design so this is very-very small.

> scientists thought that the positive charge moved.

Not actually true. The idea of "a charge" came later. However there were two types, which obeyed the well-known arithmetic of positive and negative real numbers. Which was which? Without a clue, one direction was called Positive.

> on a component level (tubes, semiconductors, polarized caps, etc), electron flow is more or less the only way you can accurately describe what's going on

Tubes are fair evidence of Electrons. I accept that the only charge-carrier inside a vacuum is electrons boiled from the surface of a negative cathode and collected by a positive plate. However in Circuit thinking, I never think this way. Current comes off the battery, through the load, from plate to cathode to negative.

Semiconductors have BOTH polarities of charge-carrier. A "Hole" is just as real (or not) as an "Electron".

I do not see polarized electrolytic caps as any proof of electrons. On forming, oxygen moves and binds to the aluminum.... but to "understand" you must believe in atoms and digest Oxidation Potential and keep track of signs. This is a deeper mud-hole than just accepting "electrons". I prefer to think I put voltage and current at an electrolyte via metal tabs, and "stuff happens". By observation: the more-positive plate gets a thin skin of non-conductor. Or Chrome or Gold is transferred from bulk to thin film on the other electrode.

[Electron Tornado]

Is it possible to do a better job? The greater community here was mostly non-technical people who have learned by reading and doing.
- what worked for you to learn?
- what stumbling blocks existed and how did you sidestep them?
- what would you advise for a generic beginner curriculum?

I have a relatively technical background, and have done some instructing and little bit of writing. The essence of your question is how to communicate a subject to an audience in a way that will maximize their understanding and retention of the information. In that, there are two major failings of many books and professors.

Mistake one: They forget their audience. Knowledge of a subject is like a layer cake. Knowing your audience will allow you to know which layer to begin a discussion and in what order to proceed to other layers. For example, starting a discussion about transistors with "electrons" and "holes", "N channels" etc is going to make a hobbyist's eyes glaze over. Cover that stuff eventually, but not on page one. Start with some basic terms,  how a transistor is used, and how it works in a simple amplifier circuit. Dissect that, and then go deeper. Start at the "Fischer Price" level, then move to "Tinker Toys and Lego", then "Erector Set" ,etc

How do most hobbyists get into something? They start with a "building block" that they can play with, and learn more about. "Hey, I built a toy! This is really cool! Hmm, I wonder what makes it go...." Start a hobbyist with a basic one transistor amplifier and they can learn an immense amount just from that. Then go deeper. (What is a transistor? Where did that name come from? How does it transist?)

Mistake two: too many books and WAY too many profs forget to show how something relates to and applies the very basic rules we learn, such as Ohm's law, Kirchoff's Laws, etc. I had a prof who would spend an entire lecture doing a derivation that resulted in a long triple integral, but never showed how it related to F=ma.

A great example of how to illustrate such relationships can be found in James Burke's series' Connections. As his discussion progressed during an episode, he was always able to point out the connection with subjects from the beginning of the episode. 


Hope this helps.

[jafo]

But there's a lesson. In technical fields, words usually have an "exact" meaning. Confusing meanings leads to more confusion.

Pretty much my original point -- modulo an awful lot of misconceptions on my part.

Fortunately, the wiki has some very nice links to help with that.

Example:
> electricians have told me, that electrons move from ground to hot.

No. Electrons move from Negative to Positive.

Yep. As you mention later, I was confusing "ground" with "negative" and "hot" with "positive." Is it any wonder I wrote what I posted? Sheesh.

> I do prefer asymmetrical waveforms

Nothing to do with push-pull versus single-ended.

Hmm, I may have jumped a conclusion or two. Push-pulls, when fully symmetrical (class B, right?), produce symmetrical distortions; and symmetrical waveforms contain no even order harmonics. But push-pulls don't have to be identical... neat.

You say you like asymmetric waves... but are they bigger on top or on bottom? In audio, generally we get waves either way, and potentially equally strong either way. The amplifier must be agnostic about which-way. Therefore the amplifier must be symmetric, to handle asymmetry either-way. A raw single-ended amp won't do that, so we add "bias" to get it set-up half-way between its limits. SE amps tend to have some residual asymmetry; we can design so this is very-very small.

Excellent points, but asymmetric distortion -- even just a touch -- is often desirable.

> on a component level (tubes, semiconductors, polarized caps, etc), electron flow is more or less the only way you can accurately describe what's going on

Tubes are fair evidence of Electrons. I accept that the only charge-carrier inside a vacuum is electrons boiled from the surface of a negative cathode and collected by a positive plate. However in Circuit thinking, I never think this way. Current comes off the battery, through the load, from plate to cathode to negative.

Semiconductors have BOTH polarities of charge-carrier. A "Hole" is just as real (or not) as an "Electron".

I do not see polarized electrolytic caps as any proof of electrons. On forming, oxygen moves and binds to the aluminum.... but to "understand" you must believe in atoms and digest Oxidation Potential and keep track of signs. This is a deeper mud-hole than just accepting "electrons". I prefer to think I put voltage and current at an electrolyte via metal tabs, and "stuff happens". By observation: the more-positive plate gets a thin skin of non-conductor. Or Chrome or Gold is transferred from bulk to thin film on the other electrode.

Philosophy is probably outside the scope of this forum, but may be even more interesting. Well, to me, at least. For example, how do you reconcile electron agnosticism with the large amount of measurements have? Sure, local conditions (paging Dr. Higgs) and the quantum nature of our nervous systems come into play, so there is a real ambiguity, but these seem to do more modification than creation. Expound on! I need a break from allaboutcircuits anyway.

[PRR]

> Push-pulls, when fully symmetrical (class B, right?)

Class B and Symmetry are not connected.

A class A audio amplifier must be "biased" to "halfway ON". We set a DC current/voltage which it sits at, and swing audio both ways from there.

An audio push-pull amp can be two class A sections in anti-phase. However it can also be two sections which are biased "OFF" at idle. This gives good improvement in full-signal efficiency and great reduction of power consumed for small and zero signals Since most speech/music is small most of the time, this is very interesting.

The drawback of class B is that devices do not go from OFF to ON cleanly.

> push-pulls don't have to be identical... neat.

No, but now you are juggling too many balls. While there are exceptions, usually the designer wants push-pull to be symetrical, and any strong asymmetry is found (or made) in the single-end stage before it.

> but asymmetric distortion -- even just a touch -- is often desirable.

Well, there is more than one stage between mike/gitar and speaker. And most amplifier stages invert. If 1st stage rounds the top, and 2nd stage rounds (what is now) the bottom, the result tends to symmetry. But never exactly so. Further the mismatch between the two stages working with different signal level and probably different bias levels tends to subtract the low-order distortion products but multiply the high-order distortion products.

"MOST" of the stages in your amplification chain "must be clean". Therefore you should have a solid understanding of "pretty clean amplifiers".

Yes, for the dry sound of naked steel string, one or a reasonable number of "flavor" stages is musically useful. However "musically useful flavor" is most-often clean when playing soft and snarly/screamy when loud. Therefore a "flavor" stage starts with a clean stage, taking a too-large signal when loud, and the overload "flaws" adjusted to musical taste.

Note that most of the pre-1960s amplifiers, still cherished by guitarists, were concieved as "clean" amplifiers, which happen to have pleasing overload. (And some which are not widely cherished today can be shown to have unhappy overload.)

To be a general amplifier designer you should deeply study CLEAN amplifier stages, then extend to overloaded stages.

To be a fuzz-pedal designer, you should know enough to steal decent clean amplifier designs, then study their overload action.

[PRR]

> how do you reconcile electron agnosticism with the large amount of measurements have?

I have almost seen "an electron". There is an experiment with a cloud chamber and an electric field. If done very carefully, you can find bits of vapor which rise or fall or balance according to the electric field you apply. The magic values fall very close to integers (times a constant for chamber-size): 0, 1, 2, 3.... The interpretation is that electric charge comes in indivisible lumps.

"Electron" is a useful mental picture. But the electron is too big for Quantum Mechanics (we find a wave) and much too small for everyday electronics. The only place you might want to think electrons is the thermionic vacuum-tube, where electrons are a reasonable explanation of what happens at the cathode. Even there, it is equally useful to think "hot cathode makes elecric steam" and to observe that tube cathodes make "negative-charge steam". But if you stay outside the tube and look only at the pins, it is equally valid (and a lot less tedious-math) to pretend that positive steam flows down from plate to cathode.

There is a theory that water is molecules of H and O. This does explain some things wonderfully well. However when I am de-leaking my plumbing, or de-clogging my flooded ditches, I don't go thinking "H and O, two H and an O stuck together". I see "water", drips and ponds. Sometimes it moves as a "current".

One thing about water current: we can usually see "why it moves". Either it moves downhill (the low side of the cellar) or we have gone to some trouble to make it move (pumps). In electricity, this "force" is invisible and rarely felt. We have a mysterious current which flows for mysterious reason. Get used to it.

Instead of electrons, it may be better to think "electric current makes a field which deflects a compass needle". While this idea has flaws at a very deep level, it is absolutely reliable in everyday understanding. It isn't necessary to think "How" this mysterious current deflects a magnet.

[fpaul]

Regarding current flow, I remember many years ago sitting in my apartment watching my roomates black and white tv.  Both antenae were broken off with just nubs remaining.  There was a sliding glass door about 15 feet to the right opening up to a small creek.

Suddenly a lime green jagged line came through the glass door, about 4 feet high, moving perpendicular to the floor.  It seemed to me to be moving slowly, although it was probably only a fraction of a second.  My roomate said he blinked and wasn't sure if he saw it or not. It stopped about 6 inches over the tv, split into an inverted "v", and hit the two nubs of the antenae. Simultanously, a rather loud thunderclap was generated in my apartment about 6 feet in front of me.

People can debate this all they want.  For me, I'm voting that the flow in a circuit should be in the direction that green line was taking.