Quote:
Originally Posted by TuningIndicator
I read somewhere that Walker saw no point in changing from valve to transistor amps until there was good reason to do so.
I'm pretty sure that Quad were the last of what I call the 'big three' (The others, of course, being Leak and Radford) to put transistor amps into production.
|
The valve-to-transistor transition period was an interesting time. Crossover distortion was viewed as the big bugbear with transistor amplifiers, and was usually ascribed as being an inevitable outcome of Class B operation that needed to be smoothed out as best could be done, usually with copious negative feedback. But asymmetry (both in average slope and tightness of curve at the x-axis) in Lin-type quasi-complementary output stages was a more subtle aspect that needed attention. Quad may have been one of the first to address this in a production amplifier with its output triples as used in the 303. In 1969 Shaw then Baxandall separately developed progressively simpler circuits using a diode in the conjugate pair, and Baxandall wrote quite a bit on the topic, initially in Wireless World September, 1969. That said, I can’t actually think of a commercial amplifier that I know for sure used the Baxandall circuit, but I suspect that it was quite common.
Leak had been an early mover with the germanium quasi-complementary Stereo 30 in 1963. It moved up to silicon transistors in 1968 with the Stereo 70, and then reworked the smaller amplifier to use silicon transistors as the Stereo 30 Plus. But as I recall, it stayed with the basic Lin circuit, with lots of NFB.
Whereas Quad had worked at ironing out the shortcomings of the quasi-complementary circuit, J.E. Sugden saw Class A as the way to go to avoid “transistor sound”, and I think its early models, initially under the Richard Allan name, just predated the Quad 303. I suspect that the Quad view of the Class A option would have been that it was too inefficient, hence the need to pursue Class B improvements.
Radford went the fully-complementary route, maybe a year or so later than the Quad 303, although I am not sure of the timing. This was seen as overcoming the asymmetry problem, although there is a viewpoint that at higher audio frequencies some asymmetry creeps in. Drawbacks at the time were the scarcity, cost and fragility of PNP power transistors. Possibly when the Quad 303 was designed, PNPs that could do 45 watts per channel were either not available or prohibitively costly. I think that the original Radford transistor amplifier was something like 30 watts per channel. The SPA50, 50 watts fully complementary dated from circa 1971. New entrant Cambridge at the end of the 1960s also opted for the fully complementary output.
Odd-man-out was Rogers, who opted for transformer-drive for its initial solid-state Ravensbourne amplifier, a contemporary of the Quad 303. I haven’t figured out whether this configuration avoided the quasi-complementary asymmetry.
Cheers,