EMRFD Message Archive 1849
Message Date From Subject 1849 2008-07-30 18:12:54 bkopski AN HB 100 mHz RF SWEEP GENERATOR UPDATE Hi Everyone,
A few weeks ago in post 1819 I described what was / is for me a long-
sought accomplishment that of realizing a good working HB ~1 to 100
mHz RF sweep generator.
In that post I described using a Mini Circuits POS-400 VCO along with
a fixed oscillator in a heterodyne configuration to produce the
frequency sweep. I also described using a separate linear voltage
ramp generator to drive the VCO over the necessary range. The POS-
400 specification describes the linearity of its voltage vs.
frequency performance and hence that of the generator - and roughly
speaking, it is more or less "linear" i.e. not "bad" but not
quite "good" either, as below.
The actual measured transfer behavior of the generator is shown both
in the red curve in the Excel sheet (see new File pdf) and in a new
Photo. The latter shows a 100 mHz frequency sweep with 10 mHz
harmonic markers in place and it is clear these begin to "bunch" a
bit past about 30 mHz. The photo display and the graphic are
As a challenge and for the fun of it, I wanted to "straighten things
out" and I found enough room in the RF box assembly to include some
add-on ramp-bending circuitry. Two new Photos show the "before"
and "after" of this addition.
The added small circuit accepts the same (but scaled) linear input
ramp and then slightly compresses it past about the 30 mHz point. In
actuality, two "break points" or three segments total are used in
this particular case. This piecewise-synthesized non-linear
approximation "unbends" the slight VCO nonlinearity previously shown
in the Excel graph and in the harmonic markers Photo to make
everything come out "more linear" in the end. This is seen in a
second similar Photo that replays the harmonic marker presentation
over the same 100 mHz sweep and it is clear that the 10 mHz lines
fall much closer to where they should. Yet another Photo shows the
two voltage sweeps captured "live" - as applied to the VCO the
original linear ramp, and the new curved voltage drive shape.
Compare this latter display with the Excel graph where I've added a
straight (blue) line extrapolation of the transfer curve beginning at
about 30 mHz. In summary: this works well.
My Thanks to those who have commented previously, and to those who
also requested more info both on-line and off. I've had a lot of fun
and learned a lot with this project, got a good working instrument
for the effort, and I did it all "my way". Boy that feels good!
And hey - isn't that what "Experimental Methods" is largely about?
Cordially and 73,