[Synchrodyne]

In his final letter in the WW correspondence series, 1973 April, p.193, HPW made the comment:

“With the shunt feedback arrangement I have often found that amplifier noise is audible above tape noise and surface noise on discs, and, with the new generation of "noise-reduced" recordings and low-output heads, I do not consider this a satisfactory situation as I believe audible background noise to be a source of listening fatigue.”

In his 1972 May article on low-noise amplifiers (p.233ff), HPW ran through the noise calculations for the shunt- and series-feedback cases. With perfect amplifiers, the shunt feedback arrangement gave an S/N ratio of 58.5 dB against a 2 mV input, the series-feedback arrangement 72 dB. He measured the 1969 JLH shunt-feedback design at 58 dB, and his own series-feedback design at 70 dB.

For its P40, Cambridge claimed 60 dB against 3 mV input, which would be 56.5 dB against 2 mV. Interestingly, the same 60 dB against 3 mV was claimed for the P50. This had a Darlington emitter follower at the input, which evidently did not provide any noise benefit, although presumably it allowed the use of lower impedances in the following virtual earth stage.

One might say that if you are going to use a shunt-feedback RIAA stage, then the JLH design was very good of its kind, getting close to the theoretical best signal-to-noise ratio. But that S/N was just a little short of what was desirable in practice.


Cheers,