RF sweeper schematic, sweep voltage section better alternative?

Hello I would like to make a sweeper for the lab for the shortwave frequencies.
I have found many vcos but not many voltage drivers.
I have found this one http://www.aoc.nrao.edu/~pharden/hobby/TS2.jpg
Is it considered good?
have you got any better alternative to this?

How about a DDS chip instead? The AD5930 can take a 50MHz master clock, so is good up to about 25MHz output. It's very easy to drive from any MCU and can generate a programmable sweep and burst output.

I would suggst a simple oscillator, maybe an NE555, feeding an 8-bit binary counter then an 8-bit DAC. It has the advantage of a very linear ramp , adjustable sweep speed and it can easily be 'frozen' by stopping the counter or oscillator. For example, if you were scanning for a station you could connect the AGC line to the counter so it stops when a signal is found. The DDS soluton is good but you also need a micro and software to make it work.

Brian.

The "frozen" approach is a good feature but, I am worrying about that the dac will make steps and not smoothly cover the whole frequency.
as far as concern amplitude stability, I am thinking of using this one as a vco, I have tested the vco section and it works very amplitude steady through the whole range. http://neazoi.com/singlespanosc/index.htm
Maybe I will leave the fll out for this application. The vco has not good phase noise and on a shortwave receiver the beat tone is fluctuates a bit when using a varicap, but for filter characterizing should be ok.
The advantage would be that a single span DC-30MHz sweeper can be made. by adjusting the varicap driver voltage and/or switching different coils, lower range sweeper can be made with almost the same hardware.
What do you think?

The problem you will have is that if you want a linear frequency vs time sweep, you need a pretty non-linear VCO Vtune voltage vs time. You can get that with a DAC driven by a prom or micro that corrects the Vtune sweep to be a linear frequency sweep.

A PLL automatically corrects for the VCO non-linearity, and the DDS of course has no VCO so you do not have to linearize it.

If you want to stay fully analog, there are "break point linearizers' that use an op amp with diodes in the feedback path to make a piece-wise linear frequency sweep. i.e. they take a linear input ramp and turn it into a non-linear Vtune that you can personally set to match a particular VCOs characteristics.

Why a linear frequency VS time is so important. For filter characterization and other light use, a spectrum analyzer , should be able to display the filter response, even if the freq/time linearity of the sweeper is not good. It was to be really bad to be useless I think.
For more precise measurements (group delay etc) it could make a point though.