[time-nuts] Speaking of varicap diodes

[ed breya]

The recent discussion of tweaking up the tune range of an OCXO has 
reminded me of another of my ongoing projects, and I thought I'd mention 
some of it here. The varicap diode is near and dear to hearts - whether 
you know it or not - for tuning RF oscillators and other functions. One 
of my projects is a complete redo and upgrade of an old-school C-V 
meter, for looking at varicaps and other junction devices. It's a 
Princeton Applied Research (PAR) Model 410, from the late 1960s -1970s 
or so. In its original form, it was for looking at the gate to source or 
drain capacitance of the newly developed MOSFET technologies, but it's 
basically a 1 MHz capacitance meter, using lock-in methods to measure C 
and G, with applied variable DC bias.

I got this unit a while back for cheap, and managed to fix a number of 
problems (mostly power supply related) and get it working well enough to 
experiment with it. I never did find any details like manuals or 
schematics, but it's mostly simple enough to figure out, with plenty of 
reverse engineering and circuit tracing. Among the first things to go 
were all the low-speed plotting functions, to be replaced by "fast" 60 
Hz sweeps, like in a I-V curve tracer, for display on a scope in X-Y 
mode. It took a lot of redesign to make it go this fast, and other 
things to be useful in its new life, but it worked out pretty well.

The really cool part is this puff-o-meter engine, which I would keep no 
matter what happens to the rest of the unit, but it turned out that I 
could use the carcass and PS, with mostly new controls up front, and 
lots of new circuitry

So, it's set up to sweep any device in reverse bias, up to 100V, and 
display the C from 0.05 pf/div to 200 pF/div on a scope, versus bias 
voltage, or show static C or G and leakage current on its original built 
in meter, at manually set DC bias.

Ed