Leak "Trough-line 3" LO woes

[John_BS]

This tuner would not operate reliably above about 100MHz. The LO stopped oscillating at higher frequencies: it appears to be a Hartley oscillator with the cathode and grid of V3b tapped onto the transmission line resonator.
After much investigation to try to eliminate component problems (the resistors and many caps had been previously changed in this tuner) I concluded that the oscillator was becoming progressively weaker as the frequency rose, until eventually dying. R9, the heater resistor, sits alongside C11, and if the latter was removed the LO would tune above the top of the band (120.5MHz). However, if the orange preset C8 was then adjusted to bring it back down, the oscillation stopped.
The HT on the anode of V3b was always low, irrespective of whether the valve was oscillating or not: the main HT was 10v down due to the mains tap being on 250V (no 240V tap) .
If C8 was disconnected the oscillator would operate to much higher frequencies, but any attempt to add capacitance back seemed to kill the oscillation. All this suggested that the Q of the tuned circuit was marginal and/or the valve was not providing sufficient loop gain. The (Mullard) valve checked out 100%, so I was left to conclude that either

a) The cathode tap of V3b had been attached to the resonator too high up (the wire emerging from the cavity appears to be well off-centre, and/or
b) C8 had become lossier than when new, or
c) There was excessive loading via C10 (which was 10pf on this tuner), or
d) a combination of all three above.

If C10 was removed the oscillator again tuned to the top of the band and beyond. It was replaced with 8p2, plus a 3p3 to ground from the same tapping point.
Finally, care was taken to separate anything likely to induce loss from the “ hot “ nodes of this oscillator: R9 was moved a few millimetres away from C11.
The end result is that the LO now operates to about 104 MHz reliably, but I do wonder if this is a problem which has been seen before? It feels like a Friday Afternoon radio.
As part of this investigation I calculated the parameters of the transmission line (attached), which agree quite well with a separate assessment of the effective inductance calculated from adding capacitance and measuring delta F.

[G6Tanuki]

Those orange ceramic trimmers are known to go wobbly with age! The silver-plating on the ceramic plates seems to develop high resistance which kills the "Q" of the associated tuned-circuit.

(I've had this issue in WWII "GEE" RF26/27 units and now instinctively replace them with 'beehive' trimmers)

[John_BS]

Thanks for the feedback. That capacitor is well integrated into the assembly, and I was doubly-reluctant to try swapping it out as I suspect it was chosen to have a negative temp-co.; possibly -330ppm if orange?

John