Try your public library for an older ARRL Manual (say from the 1970-1980 era).  These books are a gold mine of information on circuits that will still work today.

Ameritron ALS-500M with the new ALS-500RC ($49.00).

Let me see if I remember the grid naming convention used when I was a youth, and the rationale for each:

Closest to the cathode is the CONTROL grid.  It operates with a relatively low voltage and is biased negative with respect to the cathode.  Since it's the closest grid to the cathode, it CONTROLS the tube's conduction.  Regardless of the much higher positive voltage on the next grid, the CONTROL grid's closeness to the cathode rules.

Next grid out, past the control grid, is the SCREEN grid.  It has a high positive voltage on it, which serves to accelerate the electrons towards the plate.  Because of this, it's designed with a very small cross-sectional area so it doesn't hog the entire tube current flow and most of the electrons will get to the plate.

Finally, at least for a pentode, is the SUPPRESSOR grid.  An electron that gets accelerated by the screen grid is travelling very fast when it hits the plate.  This can cause a condition called "secondary emission", if I remember correctly.  The purpose of a SUPPRESSOR grid, which has a negative bias, is first to slow the electron down a bit, and second to SUPPRESS energetic electrons from hitting the plate and bouncing back, or hitting the plate and knocking other electrons off the plate.

Different tubes, different circuits, different requirements.  These descriptions sort of detail a "vanilla" tube in a "vanilla" circuit, with an explanation of what the various grids do.  Now, if your circuit is cappuccino almond fudge, your mileage may vary.

Yes.  Any Extra Class licensee should.

M0CUQ - I can empathize (or empathise) with you about the 'SO....LID COPY' -- I heard that several times in the late 1950s when I was a Novice.  The more I think about it, the more I convince myself that it was my own mind inserting the long space, anticipating something like 'SO HW CPY?'

That's right - you're a Nederlander!  Sorry; I forgot.

Anyway - try here: http://arj.nvg.org/ham/paddle1.html
This even has a link to a guy with a PAØ call sign!

Tom - that wooden be a bad idea!

"...get the bugs out of it."  Good one!!

You might check back issues of QST.  There were some good articles on the subject in past months.  73 Magazine also had some similar articles, but you'd have to dig a little since 73 is no longer a viable periodical.  WorldRadio might also have some paddle plans.

My High School electronics teacher (and Elmer) Chris Delzer, ex-K0OEI, ex-K7LBQ, taught me that the letter "Q" sounded like " Pay Day Today".  Unfortunately, there are no Q's in this phrase!

Measures' web site has moved to:

http://www.somis.org/

There are a lot of links to articles and other documents (probably including the discussion referred to above).  

There should be absolutely NO problem using RG-58 for this application.

The first time I ever heard of learning code that way was in the old book (later made into a couple of movies) "Cheaper by the Dozen".  In the book (the true story of a family back in the 1920s), the father is an "efficiency expert" who teaches his twelve children Morse code at the dinner table.  The only phrase I remember was that the letter "C" was "CAREless CHILdren".  

This is a self-limiting way to learn code: you hear "DAH di DAH dit", you then think "CAREless CHILdren", and then write down "C".  The idea should be to hear "DAH di DAH dit" and immediately write the "C" down without the extra step.

If the CW signal with which you're driving the amplifier has ripple on it, then that ripple will be present in the output, appropriately amplified.  However, if you're feeding a clean drive signal to the amplifier, the output will be free of hum unless, as he said, you overdrive the amplifier.

I was a radioman in the Navy, and I could "hear" the code with no visual "interface".  However, after I got off watch, I would often go up to the Signal Bridge, especially at night.  When my ship's signalmen would be using flashing light to communicate with the other ships in company, I could copy these visual signals just fine, but I "heard" the code!  

A signal searchlight has mechanically-operated shutters which make a clattering sound, and of course I could hear that sound when my ship's SMs were using the lights -- but each flash of light from the other ships - maybe five miles away -was a 700 Hz "beep" in my head!

I still work for the Navy, and even though much of the visual signalling has become obsolete (there are no "Signalmen" any more; that job has been taken over by the Quartermasters).  Heck - there aren't any RADIOMEN any more!  The old RM rating has been folded into the IT rating ("Information Technician").  Sigh.

Anyway, as I was saying before I sidetracked myself: even though visual signalling as it used to be is pretty much gone, ships still have signal searchlights, and even the Quartermasters have to be able to use them.  When we go on sea trials after a ship's been repaired, if there's another ship nearby, I'll go up to the signal bridge and use the searchlight.  And the "sonic interface" is still there on a visual medium.

I don't think you're weird at all!

Class A amplifiers are biased at the most linear portion of the tube's characteristic curve.  Class B amplifiers are biased right at cutoff.

AB1 and AB2 are biased somewhere between the Class A and Class B points, and as you said, AB1 indicates that the tube doesn't draw any grid current, whereas AB2 does draw grid current at some point.

And of course, Class C amplifiers are biased well beyond cutoff.