Non-linear tuning capacitor drive.

[G6Tanuki]

We all know the issue with traditional receiver tuning-capacitor drives - that cramped-at-the-HF-end effect which means that the entire 28-30MHz amateur band occupies about 2mm of the tuning scale, and accurate tuning to Luxembourg on a cheap 'trannie' was fiddly to say the least.

Many years ago I saw a rather posh US-manufactured receiver (from the 1930s - it had the old pre-octal 6C6 / 6D6 style valves and covered MW, SW and up to about 60MHz for the pre-war US Police- and FM-bands and TV-sound) which addressed this rather elegantly.

The 'drum' for the tuning-string on the capacitor was not circular; instead it was in two parts each shaped in a flat spiral* of opposite-handedness, with some sort of gear/spring-tensioning device between the two. As it rotated, string unrolled from one spiral and was taken up by the other one.

This gave the effect that depending on where in the rotation the capacitor was, a given length of string-movement gave different changes in the angular position of the capacitor, so practically linearising the tuning-scale (pointer attached to the string) and giving a much more constant KHz-change-per-revolution of the tuning knob.

I'm surprised this approach wasn't used more - it wouldn't have been expensive to implement and could have been a big selling-point for general-coverage LW/MW/SW radios.

* Something a bit like a Logarithmic Spiral: http://mathworld.wolfram.com/LogarithmicSpiral.html

[turretslug]

I'd wondered why more devices of this nature hadn't appeared over the years- particularly in an era like the mid/late '30s when the sophisticated/prestige superhet was the subject of considerable ingenuity and neuron-time in every aspect of design. The shaping of gang-capacitor vanes reduced the non-linear tuning scale tendency- but it needed to be combined with a variable-ratio drive to get anywhere near a linear scale without unimaginably peculiar vane-shaping. That device you describe sounds interestingly elegant- there's an Eddystone 750 here which uses a peg on the tuning drive's final gear to drive an arm on the gang-capacitor shaft to similar effect, the one sliding along the other to achieve scale linearising. In typically elegant Eddystone fashion, the peg has two little brass rollers that run down the U-channel arm- I like the mechanical detailing of many Eddystones, even if the electronics is a little pedestrian. This ingenious and simple mechanism is used on a handful of subsequent models with wide span per band tuning but not on more "pro" market sets like the 910 and 830 where the less ambitious tuning ratio meant that there was less scale cramping to compensate. The effect on an otherwise fairly ordinary, lower-rung comms receiver is a pleasing "precision" look to the tuning scale, with just a little stretching in the middle relative to either end.

I wonder how many other mechanisms were dreamt up to achieve scale linearising- the mechanism in the US receiver mentioned sounds as if it might have been inspired by the fusee drive in quality clocks, adapted for bi-directionality.

Edit: The Eddystone drive mentioned has three brass rollers on the tuning drive output peg, the outer two running along the U-channel limbs, the inner one rotating in the opposite direction along the gang-capacitor arm's tensioner spring- almost Swiss-level thoroughness!

[G6Tanuki]

Yes it puzzles me that it wasn't more commonplace. As you say, it could be considered as a sort-of single-turn version of the Fusee.

I wasn't familiar with the Eddystone 750's approach - it looks elegantly cunning!

Without delving into maths of a complexity that's inappropriate for a Monday lunchtime, I suspect that having complicatedly-sculpted vanes would not necessarily linearise all wavebands uniformly: the L:C ratio variation with respect to frequency-change gets inelegant and interesting as you go from LW to shortwave and beyond, which is why 'proper' wideband comms receivers like the AR88 have tuning-gangs with what amounts to two capacitors per tuned-circuit, selecting the low-value variable for HF bands and the larger-value one (sometimes paralleling the small value one too) for the LF bands.

[turretslug]

In a way, the narrow-angle scale window of the likes of the AR88, BC348 and half-a-million others helps to conceal scale cramping- though I doubt that this was a primary concern!- it's certainly obvious with the whole scale exposed. I think one of the most impressive examples of tuning scale linearity has to be the permeability-tuned "kHz" setting oscillator used in the likes of the Collins R388 et al, where the tuning scale rotates 10 times successively with something like less than a kHz deviation achievable overall. We are talking exceptional radios with well-beyond-mass-market architecture from an era when spending loadsamoney/unit was almost a requirement, though!