[Penthode]

Hi All,

I have been away and I have been reviewing the documentration we have collected in recent months regarding CPA.

Upon comparing Nick's prototype and the RCA patent for the CPA switching, there are striking similarilties in the components. The CPA transformer and components around it in the prototype follows the patent even down to some components.

I am trying to derive a reliable field switcher using only four triodes (the two 12AU7's). This is proving a challenge because one pair of triodes would provide the bi-stable device for alternate fields, one triode for field recognition, one triode for alternate field bi-stable switching and an extra triode as the field pulse buffer (as shown in the figure 13 Hazeltine drawing). The field pulse buffer is the fifth triode so the question to ask is: is there a spare unused trode on the chassis besides the two 12AU7's? Or is there possible a spare low impedance winding on the vertical output transformer which could produce the pulse?

Also the transformer on the plate circuit of the 6U8 3.89MHz oscillator buffer appears redundant. The CPA transformer is all that would be required and I wonder if it originally attached to the buffer plate directly? I find the separate subchassis for the oscillator curious and wonder if it was later fitted to fulfil subsequent modifications?

In the timeline of development of CPA with symmetrical vestigial sideband chrominance channels thru to the development of the IQ asymmetrical sideband solution, I suspect there were intermediate but "unofficial" modifications to the system. If I were in their shoes in 1952 when CPA with the 3.89MHz subcarrier operation was proving difficult to tame and growing unpopular, I would have been inclined to experiment with CPA back at 3.58MHz. The benefit of 3.58MHz CPA would be to make each chrominance channel sidebands more symmetrical to help reduce the CPA flicker.

Introducing the IQ solution would mean different bandwidth I Q chrominance channels which would mean a second delay line for the I channel. At this stage they may or may not have thought about the design of a "simplified" narrowband chrominance channel decoder. Without seeing any evidence means that discussion of this is pure conjecture. Making the set a switchable 3.58MHz CPA/non-CPA receiver would not be much of a problem but full IQ demodulation I feel would be out of the question.

I plan to complete my diagram in Visio which I will hopefully tonight. Any thoughts, comments or criticisms are most welcome.

Terry