I have never tried to use the "5-lead" IC audio "power" amplifiers, so it is with great sorrow that I feel that I have missed the last sailing of that boat. All that seems left is a discrete component class a-b design, using either power transistors or beam-power vacuum tubes, the latter with audio output transformers in a
Williamson design configuration.
I am particularly interested in maybe reproducing, as accurately as possible, the "
100 watt ultra-linear amplifier" with separate plate and screen grid power supplies for the 6146 beam-power pentodes, and using rugged and reliable 6SN7 dual triodes as the pre-amp and drivers. These tubes may be impossible to procure at reasonable prices as NOS, but I think suitable replacements are being made in foreign countries. I could also "roll my own" preamps and driver using whatever tubes are currently available.
I think the output transformer may have to be custom made, but fortunately there is a small company in the USA that specializes in this service for amateur radio customers. I recall seeing a schematic many years ago for a similar circuit that used taps on the output transformer, instead of two separate windings, to apply the screen grid voltages to the pair of output tubes from a single B+ supply connected to the primary center-tap, whereas the linked reference above uses two separate windings and separate power supplies for the screen and plate supplies. Both circuits were identified as "Williamson" derivatives, which at that time (circa 1950 - 1960) meant very linear, low harmonic distortion, state-of-the-art audio power amplifiers. Today they would be called boat anchors, but I thought the ones I heard sounded pretty decent, although they were expensive and I could not afford one, much less two for a stereo "hi-fi" setup.
This is not a construction project for the inexperienced or faint of heart. LETHAL high voltages are required. Electrolytic capacitors with appropriate voltage ratings will be almost impossible to acquire. NOS is not suitable for this, but there are work-arounds using lower voltage capacitors connected in series, with voltage dividers to distribute the high voltage equally across each capacitor. Hardly anyone knows what a "swinging choke" is anymore, and they are probably unavailable, but there is probably a work-around for that too... my Novice rig didn't use a swinging choke, but it did have a voltage-regulated B+ supply with a series pass-tube and negative feedback to control the output voltage. That worked "perty gud" and everyone I worked using CW on the 80m Novice band said my signal sounded great... no chirps, no hum, just pure CW RF.
Fortunately, my first amateur Novice transmitter was a homebrew rig built sometime between 1965 and 1966 while I was serving in the U.S. Air Force. It used a single 6146 power pentode as a Class C RF amplifier. I grew up experimenting with 6SN7 vacuum tubes while learning a little bit of electronics along the way, so those are familiar to me. Playing around with electronics was always a lot of fun and I learned a lot too. This is important to you younger whipper-snappers coming of age here in the 21st Century. If it ain't FUN, why are you wasting your time doing it?
