Perry Lind made the comment that the old Westinghouse Electric transmitter sounded better than the modern transmitters, I’ve got mixed feelings. The old high-level plate modulated transmitters did sound good if they were well engineered and maintained within the limits of their capabilities. Modulation transformers and reactors resulted in bounce and tilt and limited the modulation that could be obtained to lower levels than it otherwise could be without overmodulating negative peaks.
Some of the other systems like the RCA AmpliPhase or CCA’s Doherty modulation sounded ok if everything was perfectly tuned and tweaked and in good order but were difficult to maintain in that state. The ampliphase system was kind of icky in that the relationship between phase angle and power output wasn’t linear and I don’t think predistorting the audio to compensate did so adequately.
One problem with any of the old tube transmitters is that tubes don’t just run perfect until they fail, they gradually lose their ability to emit current from the cathodes efficiently and high power tubes are expensive so there was a reluctance to replace them until they were really sick. This would cause the RF finals to be incapable of handling positive peaks linearly or the modulator tubes would get weak and become incapable of creating enough power to modulate the transmitter completely.
I know the CCA’s were cable of high modulation. I don’t know what KHHO (formerly KTAC) is using these days, but back in the last 70′s they had a CCA beast of some sort and I remember being down there and looking at the modulation monitor, a unit with a couple of LEDS when peaks exceeded the threshold set on a couple of calibrated dials. The positive dial was calibrated up to 133% but the knob actually turned physically past that point (and was), and the LED was for the most part continuously on. And actually it didn’t sound that bad given the degree of audio crunching that was taking place.
The Harris pulse width modulation (class D) transmitters and later the digital modulation transmitters (which essentially is equivalent to pulse with modulation using the RF stage as the pulse width modulator stage at the carrier frequency) also sounded good and both were capable of heavy modulation.
The tendency of transistors to work until they don’t and the fact that they’re operating in a switched mode making linearity a non-issue, means that the audio quality and modulation capability doesn’t gradually degrade as the tubes age. The efficiencies are also high, which perhaps if you’re in a cold region of the country in the winter and not paying the power bills is not such a good thing.
I think what sounds not so good with many of todays stations is the audio processing and audio sources that are used. Take bad audio, run it through an audio chain that stomps it entirely flat, and no transmitter is going to sound good.