View Single Post
Old 23rd Nov 2017, 10:59 pm   #11
Pieter H
Tetrode
 
Join Date: Sep 2017
Location: Waalre, Netherlands
Posts: 67
Default Re: Introduction PCC189 in Mullard tuners?

Hi, thanks for all feedback and inputs.

To go back to the original question in post #5: in principle all Philips/Valvo/Mullard valves were optimized valves for the application. On my site on the early Philips TV development you can read all about the close co-operation between the set development and radio valve development labs. After the initial struggles with generic valves used in the tuners (UF42, EF42, EF80 and the ECC81 as a very first effort to RF optimization), from the PCC84 and PCF80 onwards all subsequent valves were designed primarily for their role in VHF cascode tuners with pentode mixer and triode oscillator. This architecture was used unchanged from 1954 to 1964, when the cascode was replaced by a single PC900 triode. Pure RF triodes like PC86 and PC88 were only used in UHF tuners, with one as pre-amp and ons as self-oscillating mixer. This architecture did not change until the introduction of transistors.

The result of the close co-operation between set/tuner development and valve development is illustrated by the extreme stability of the Philips tuner circuit. After two generations the circuit had almost completely stabilised, with minimal changes from then. New valves were what we call today "drop-in replaceable", requiring no changes to the application. The benefit of such stability in terms of minimal (re)-design effort and manufacturing streamlining should not be under-estimated, and has probably been one of the factors contributing to the dominant position of the Philips tuners over time. Jumping to completely different types of valves like the Nuvister, requiring serious concept changes, was then almost certainly considered too risky and not offering sufficient performance improvement to justify such a re-design.

In UHF the European approach of using a triode pre-amp was clearly superior to the American concept of a passive diode mixer with no pre-amp. Noise figure must have been pretty bad as any theoretical analysis of such a set-up will show. At the same time UHF remained a challenge for volume consumer modules, where probably the best solution was to use the minimal number of valves. The interconnect between the valves or valve sections probably gave more degradation than performance gain. Therefore just one triode as pre-amp instead of a cascode, and a single self-oscillating triode instead of separate oscillator and mixer. This approach didn't fundamentally change with the introduction of transistors, although a third transistor was used as oscillator.

All that in the next chapter of the tuner history!

Cheers, Pieter
Pieter H is offline