09-12-2016, 11:32 AM
I've now tracked down the original - it was in the wrong folder! - and updated the schematic which is attached.
I note that I also 'moved' a component value in the original - the missing resistor has the value of 220Ω that I assigned to the audio input series resistor (which probably isn't very important, anyway).
As for varicap diodes - which are only special versions which exploit the capacity effect in 'normal' diodes - as the reverse bias decreases and the capacitance increases, the effect becomes very non-linear (hence the 'special law' multi-turn pots used with varicap UHF tuners: using a linear law pot would result in the channels at the bottom of Band IV being crammed together with a subsequent loss of tuning accuracy).
It is for this reason that the diode is biased from the supply to move the centre point to make it more linear.
The change in capacity needed for ±75kHz deviation at 100MHz is very small - hence the 3ρ3 capacitor in series, to make the change more manageable. From memory, my prototype in the 60s needed an audio input of around 1V rms for a decent modulation level - although this would have been based on comparative listening tests as I had no way of measuring the actual deviation.
As Andy says, the value of this capacitor can be varied as required. If you can't obtain a reasonable modulation depth, try increasing the value of this cap although if you have a selection of diodes, I suggest you try swapping that first - even specifically designed varicaps can have a wide production spread* so different samples of the same type of diode of any kind can be expected to behave differently.
* When the BB105 was introduced for varicap TV tuners, Mullard made them available as matched sets to ensure that the tuners would perform correctly. Individual unmatched samples could well have made it impossible to align many tuners!
I note that I also 'moved' a component value in the original - the missing resistor has the value of 220Ω that I assigned to the audio input series resistor (which probably isn't very important, anyway).
As for varicap diodes - which are only special versions which exploit the capacity effect in 'normal' diodes - as the reverse bias decreases and the capacitance increases, the effect becomes very non-linear (hence the 'special law' multi-turn pots used with varicap UHF tuners: using a linear law pot would result in the channels at the bottom of Band IV being crammed together with a subsequent loss of tuning accuracy).
It is for this reason that the diode is biased from the supply to move the centre point to make it more linear.
The change in capacity needed for ±75kHz deviation at 100MHz is very small - hence the 3ρ3 capacitor in series, to make the change more manageable. From memory, my prototype in the 60s needed an audio input of around 1V rms for a decent modulation level - although this would have been based on comparative listening tests as I had no way of measuring the actual deviation.
As Andy says, the value of this capacitor can be varied as required. If you can't obtain a reasonable modulation depth, try increasing the value of this cap although if you have a selection of diodes, I suggest you try swapping that first - even specifically designed varicaps can have a wide production spread* so different samples of the same type of diode of any kind can be expected to behave differently.
* When the BB105 was introduced for varicap TV tuners, Mullard made them available as matched sets to ensure that the tuners would perform correctly. Individual unmatched samples could well have made it impossible to align many tuners!
